U.S. patent application number 15/090566 was filed with the patent office on 2016-07-28 for centripetal acceleration determination, centripetal acceleration based velocity tracking system and methods.
The applicant listed for this patent is Merlin Technology, Inc.. Invention is credited to John E. Mercer.
Application Number | 20160216379 15/090566 |
Document ID | / |
Family ID | 50339662 |
Filed Date | 2016-07-28 |
United States Patent
Application |
20160216379 |
Kind Code |
A1 |
Mercer; John E. |
July 28, 2016 |
CENTRIPETAL ACCELERATION DETERMINATION, CENTRIPETAL ACCELERATION
BASED VELOCITY TRACKING SYSTEM AND METHODS
Abstract
The flight of an aircraft is characterized in terms of
Earth-based reference system data for a first position and a second
position of the aircraft that are separated by a time increment. A
centripetal acceleration of the aircraft is determined based on the
Earth-based reference system data for the positions in conjunction
with the time increment. A rate of rotation is established
corresponding to the time increment in an aircraft-based reference
system. Aircraft airspeed is determined based on the centripetal
acceleration and the rate of rotation. A turn can be detected as
characterized by a change in track angle from a prior GPS packet
compared to a new packet. Aircraft turns can be detected based on a
change in track angle in GPS packets. Compensation can be applied
to a gyro based on centripetal acceleration such that the gyro more
accurately responds to the Earth gravity axis.
Inventors: |
Mercer; John E.; (Gig
Harbor, WA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Merlin Technology, Inc. |
Kent |
WA |
US |
|
|
Family ID: |
50339662 |
Appl. No.: |
15/090566 |
Filed: |
April 4, 2016 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14032375 |
Sep 20, 2013 |
9309004 |
|
|
15090566 |
|
|
|
|
61704338 |
Sep 21, 2012 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G01P 5/00 20130101; G01P
7/00 20130101; G01P 15/14 20130101; B64D 43/02 20130101; G01S 19/39
20130101 |
International
Class: |
G01S 19/39 20060101
G01S019/39; G01P 15/14 20060101 G01P015/14 |
Claims
1. An apparatus, comprising: an instrument for characterizing the
flight of an aircraft during a turn in terms of Earth-based
reference system data for a first position and a second position of
the aircraft that are separated by a time increment; a gyro that is
supported by the aircraft and provides gyro outputs that are
intended to be responsive to an Earth gravity axis; and a processor
configured for determining a centripetal acceleration of the
aircraft responsive to said turn based on the Earth-based reference
system data for the first and second positions in conjunction with
the time increment and for applying compensation to said gyro based
on the centripetal acceleration such that the gyro more accurately
responds to the Earth gravity axis.
2. The apparatus of claim 1 further comprising: a GPS system that
is carried by the aircraft to produce the Earth-based reference
system data as GPS data and the processor uses the GPS data and the
time increment to determine the centripetal acceleration.
3. The apparatus of claim 2 wherein the GPS data comprises a first
GPS packet for the first position and a second GPS packet for the
second position.
4. The apparatus of claim 2 wherein the processor determines the
centripetal acceleration using no more than the GPS data and the
time increment.
5. The apparatus of claim 3 wherein the GPS data for each of the
first and second GPS packets is characterized as a groundspeed
vector including a magnitude and a track angle.
6. The apparatus of claim 3 wherein the GPS data for each of the
first and second GPS packets is characterized by a north component
and an east component of groundspeed.
7. The apparatus of claim 1, further comprising: said processor
further configured to resolve a difference between the Earth-based
reference system data for a first groundspeed vector at the first
position and a second groundspeed vector at the second position
into northing components and easting components and to determine
the centripetal acceleration based on a change in the northing
component from the first position to the second position, a change
in the easting component from the first position to the second
position and the time increment.
8. The apparatus of claim 7 wherein said processor is further
configured to represent the northing component change as
.DELTA.V.sub.N, the easting component change as .DELTA.V.sub.E, the
time difference as .DELTA.t, and to determine the centripetal
acceleration, .alpha., as .alpha. = ( .DELTA. V N .DELTA. t ) 2 + (
.DELTA. V E .DELTA. t ) 2 . ##EQU00003##
9. The apparatus of claim 1 wherein the centripetal acceleration
combines with the Earth's gravitational acceleration to produce an
apparent acceleration and the compensation shifts the axis of
reference of the accelerometer from the apparent acceleration
toward the Earth gravity axis.
10. The apparatus of claim 1, further comprising: a GPS system that
is carried by the aircraft to serve as said instrument, which GPS
system outputs GPS data as a series of GPS packets for selective
use as the Earth-based reference system data such that a first
packet is selected from the series in association with the first
position and a second packet is selected from the series subsequent
to the first packet and the second packet is associated with the
second position.
11. The apparatus of claim 10 wherein said processor saves at least
a portion of the series of GPS packets in a rolling buffer for
selection of at least the first packet therefrom based on a
comparison to a current packet that terminates the series of GPS
packets.
12. The apparatus of claim 10 wherein said processor is further
configured to apply a packet selection interval to the series of
GPS packets such that the first selected packet and the second
selected packet are spaced apart by a predetermined number of
unused packets.
13. The apparatus of claim 10 wherein said processor is further
configured to apply a packet selection interval to the series of
GPS packets such that the first selected packet and the second
selected packet are spaced apart by at least a predetermined amount
of time.
14. The apparatus of claim 10 wherein said processor is further
configured to apply a packet selection interval to the series of
GPS packets having a duration that is based on a rounding error in
the GPS data such that the first selected packet and the second
selected packet are spaced apart in time by at least a
predetermined amount of time.
15. A method, comprising: characterizing the flight of an aircraft
during a turn in terms of Earth-based reference system data for a
first position and a second position of the aircraft that are
separated by a time increment; determining a centripetal
acceleration of the aircraft responsive to said turn based on the
Earth-based reference system data for the first and second
positions in conjunction with the time increment; and applying
compensation to a gyro that is intended to slave to the Earth
gravity axis based on the centripetal acceleration such that the
gyro more accurately responds to the Earth gravity axis.
16. The method of claim 15 wherein the Earth-based reference system
data includes GPS data comprising a first GPS packet for the first
position and a second GPS packet for the second position and the
centripetal acceleration is determined using no more than the GPS
data and the time increment.
17. The method of claim 16 wherein the GPS data for each of the
first and second GPS packets is characterized by a north component
and an east component of groundspeed and determining the
centripetal acceleration includes resolving a difference between
the Earth-based reference system data for a first groundspeed
vector at the first position and a second groundspeed vector at the
second position into northing components and easting components and
establishing the centripetal acceleration based on a change in the
northing component from the first position to the second position,
a change in the easting component from the first position to the
second position and the time increment.
18. The method of claim 17 including representing the northing
component change as .DELTA.V.sub.N, the easting component change as
.DELTA.V.sub.E, the time difference as .DELTA.t, and determining
the centripetal acceleration, .alpha., as .alpha. = ( .DELTA. V N
.DELTA. t ) 2 + ( .DELTA. V E .DELTA. t ) 2 . ##EQU00004##
19. The method of claim 15 wherein the centripetal acceleration
combines with the Earth's gravitational acceleration to produce an
apparent acceleration and applying the compensation shifts the axis
of reference of the accelerometer from the apparent acceleration
toward the Earth gravity axis.
Description
RELATED APPLICATION
[0001] This application is a divisional application of copending
U.S. patent application Ser. No. 14/032,375 filed on Sep. 20, 2013,
which claims priority from U.S. Provisional Patent Application Ser.
No. 61/704,338 filed on Sep. 21, 2012, each of which are hereby
incorporated by reference in their entirety.
BACKGROUND
[0002] The present application is generally related to a system,
instruments and methods for monitoring the flight of an aircraft
and, more particularly, to an advanced system, apparatus and
methods for monitoring the flight of the aircraft based on a
determination of centripetal acceleration.
[0003] In one aspect of the prior art, modern aircraft and other
vehicles such as boats and ships often include one or more GPS
(Global Positioning System) receivers for purposes of monitoring
the progress and movement of a craft. Applicant recognizes,
however, that a GPS receiver is ineffective in terms of
characterizing movement of an ambient that supports the moving
craft such as, for example, air or water. Using the example of an
aircraft, it can be critical to know the airspeed of the aircraft,
especially in windy conditions, for purposes of avoiding operation
of the aircraft beyond its design limits. As will be further
discussed below, attempting to maintain a constant GPS groundspeed
of an aircraft subject to a tail wind can result in exceeding the
airspeed design limits of the aircraft when turning from downwind
to upwind.
[0004] One approach for avoiding airspeed design limitations
resides in providing an airspeed sensor such as, for example, a
pitot tube. While such sensors are generally effective, it should
be appreciated that the cost for the addition of such a sensor is
not trivial. Further, airspeed sensors can be subject to failure.
For example, a pitot tube can be subject to icing over or becoming
clogged with other foreign material. Thus, even in an aircraft that
includes an airspeed sensor, Applicants recognize that an
additional system for independently determining airspeed can be
valuable.
[0005] One recent approach for determining airspeed without using
an airspeed sensor is described in a posting entitled Wind
estimation without an airspeed sensor
(http://diydrones.com/forum/topics/wind-estimation-without-an,
dated Jan. 29, 2010). This approach, however, must utilize a pair
of GPS readings in conjunction with a direction cosine matrix for
purposes of estimating the airspeed. It should be appreciated that
the formation of the cosine matrix is not trivial and requires
measurements along all three axes of rotation. In a typical
implementation, triaxial rate gyros or triaxial attitude gyros and
triaxial accelerometers, and triaxial magnetometers are needed in
conjunction with the GPS for slaving the rate gyros. Triaxial rate
gyros and triaxial accelerometers are not normally included as
standard equipment in an aircraft's equipment.
[0006] In another aspect of the prior art, a gyro is sometimes used
to augment a magnetic compass to form what can be referred to as a
slaved gyro. As will be further discussed, the reference for this
slaved gyro should be the Earth's gravitational vector, so that a
turn is represented by the slaved gyro in the Earth's gravitational
axis system. Unfortunately, however, centripetal acceleration
during a turn produces an acceleration that adds to the
gravitational acceleration when an aircraft makes a coordinated
turn such that the total acceleration is towards the floor of the
aircraft. Applicant recognizes that slaving a gyro to this apparent
acceleration can be problematic since the gyro will tend to shift
to an axis as a reference which is perpendicular to the floor of
the aircraft in a turn, if the aircraft executes a turn of
sufficiently long time duration unless the centripetal acceleration
factor is compensated.
[0007] The foregoing examples of the related art and limitations
related therewith are intended to be illustrative and not
exclusive. Other limitations of the related art will become
apparent to those of skill in the art upon a reading of the
specification and a study of the drawings.
SUMMARY
[0008] The following embodiments and aspects thereof are described
and illustrated in conjunction with systems, tools and methods
which are meant to be exemplary and illustrative, not limiting in
scope. In various embodiments, one or more of the above-described
problems have been reduced or eliminated, while other embodiments
are directed to other improvements.
[0009] In one aspect of the present disclosure, a method and
associated apparatus are described for determining an airspeed of
an aircraft in flight during a turn. The flight of the aircraft is
characterized in terms of Earth-based reference system data for a
first position and a second position of the aircraft that are
separated by a time increment. A centripetal acceleration of the
aircraft is determined based on the Earth-based reference system
data for the first and second positions in conjunction with the
time increment. A rate of rotation is established corresponding to
the time increment in an aircraft-based reference system. The
airspeed is determined based on the centripetal acceleration and
the rate of rotation.
[0010] In another aspect of the present disclosure, a method and
associated apparatus are described for monitoring the flight of an
aircraft, the aircraft carrying a GPS system which outputs GPS data
as a series of GPS packets for selective use as Earth-based
reference system data. At least selected ones of the series of GPS
packets are periodically monitored to serve as the Earth-based
reference system data for detecting that the aircraft is engaged in
a turn based on a track angle that forms part of each GPS packet. A
turn is detected as characterized by a change in the track angle
from a prior packet corresponding to a prior position of the
aircraft to a new packet corresponding to a new position of the
aircraft. A centripetal acceleration is determined based on the
prior packet, the new packet and a time increment that separates
the prior packet and the new packet. A rate of rotation
corresponding to the time increment in an aircraft-based reference
system is established. The airspeed is determined based on the
centripetal acceleration and the rate of rotation.
[0011] In still another aspect of the present disclosure, an
instrument is described including a GPS receiver for characterizing
the movement of a craft in terms of Earth-based reference system
data for a first position and a second position of the craft that
are separated by a time increment. At least a selected one of (i) a
heading gyro provides a heading output for tracking a heading of
the craft in a craft-based reference system and (ii) a rate sensor
monitors a rate of change of a yaw orientation of the craft in the
craft-based reference system. A processor is configured for
determining a centripetal acceleration based on the Earth-based
reference system data for the first and second positions in
conjunction with the time increment, establishing a rate of
rotation corresponding to the time increment in the craft-based
reference system at least based on the heading output if the
heading gyro is selected or at least based on the rate of change of
the yaw orientation if the rate sensor is selected, and a velocity
of the craft in a turn is determined based on the centripetal
acceleration and the rate of rotation.
[0012] In yet another aspect of the present disclosure, a method
and associated apparatus are described for monitoring the flight of
an aircraft, the aircraft carrying a GPS system which outputs GPS
data as a series of GPS packets. Selected ones of the series of GPS
packets are at least periodically monitored for detecting that the
aircraft is engaged in a turn based on a track angle that forms
part of each GPS packet as characterized by a change in the track
angle from a prior packet corresponding to a prior position of the
aircraft to a subsequent packet corresponding to a subsequent
position of the aircraft. Responsive to detecting a turn, a
centripetal acceleration of the aircraft is determined based on the
prior packet, the subsequent packet and a time increment that
separates the prior packet and the subsequent packet. A rate of
rotation of the aircraft is established corresponding to the time
increment in an aircraft-based reference system. An airspeed of the
aircraft is determined based on the centripetal acceleration and
the rate of rotation.
[0013] In a continuing aspect of the present disclosure, a method
and associated apparatus are described in which the flight of an
aircraft is characterized during a turn in terms of Earth-based
reference system data for a first position and a second position of
the aircraft that are separated by a time increment. A centripetal
acceleration of the aircraft is determined responsive to the turn
based on the Earth-based reference system data for the first and
second positions in conjunction with the time increment.
Compensation is applied to a gyro that is intended to slave to the
Earth gravity axis based on the centripetal acceleration such that
the gyro more accurately responds to the Earth gravity axis.
BRIEF DESCRIPTION OF THE DRAWINGS
[0014] Embodiments are illustrated in referenced figures of the
drawings. It is intended that the embodiments and figures disclosed
herein are illustrative rather than limiting.
[0015] FIG. 1 is a diagrammatic perspective, partial view of a
helicopter including components of an autopilot system for purposes
of serving as a framework for the discussion of centripetal
acceleration determination according to the present disclosure as
well as its various applications.
[0016] FIG. 2 is an overhead diagrammatic perspective, partial view
of the helicopter of FIG. 1, shown here to illustrate further
details with respect to components of the system of the present
disclosure.
[0017] FIG. 3 is a diagrammatic, perspective partial view of an
embodiment of an actuator and an embodiment of a force limited link
that can serve as components of the system of the present
disclosure.
[0018] FIG. 4 is a diagrammatic, perspective view of an embodiment
of a gear drive arrangement that can form part of the actuator of
FIG. 3 along with a redundant pair of actuator drive motors.
[0019] FIG. 5 is a block diagram that illustrates an embodiment of
the autopilot system of the present disclosure.
[0020] FIG. 6 is a schematic diagram of an embodiment of a voting
section that receives votes which are cast by a set of triplex
processors.
[0021] FIG. 7 is a flow diagram that illustrates an embodiment of a
method for the operation of an inner control loop and an outer
control loop of the system of the present disclosure.
[0022] FIG. 8 is a schematic diagram of an embodiment of a dynamic
braking system that can form part of the system of the present
disclosure.
[0023] FIG. 9 is a schematic diagram of another embodiment of a
dynamic braking system that can form part of the system of the
present disclosure.
[0024] FIG. 10 is a vector diagram illustrating a ground speed
vector for each one of two different positions of an aircraft.
[0025] FIG. 11 is a flow diagram illustrating an embodiment of a
method for executing centripetal acceleration velocity system
determinations of airspeed.
[0026] FIG. 12 is another flow diagram for executing centripetal
acceleration velocity system determinations of airspeed including a
turn detection procedure.
DETAILED DESCRIPTION
[0027] The following description is presented to enable one of
ordinary skill in the art to make and use the invention and is
provided in the context of a patent application and its
requirements. Various modifications to the described embodiments
will be readily apparent to those skilled in the art and the
generic principles taught herein may be applied to other
embodiments. Thus, the present invention is not intended to be
limited to the embodiments shown, but is to be accorded the widest
scope consistent with the principles and features described herein
including modifications and equivalents. It is noted that the
drawings may not be to scale and may be diagrammatic in nature in a
way that is thought to best illustrate features of interest.
Descriptive terminology may be adopted for purposes of enhancing
the reader's understanding, with respect to the various views
provided in the figures, and is in no way intended as being
limiting.
[0028] FIG. 1 is a perspective, partial view of a helicopter 10,
shown here for purposes of illustrating various components of an
embodiment of an autopilot system 12 in relation to the helicopter.
It should be appreciated that much of the physical structure of the
helicopter itself has been rendered as invisible in FIG. 1 for
purposes of illustrative clarity, however, it is understood that
this structure is present. The autopilot of the present disclosure
is electromechanical and can provide flight control of a helicopter
without requiring a hydraulic flight control system. The helicopter
can be, by way of non-limiting example, a Robinson R22 helicopter.
The teachings that are brought to light herein, however, can
readily be adapted for use with any suitable helicopter, either
currently available or yet to be developed. For example, the
autopilot of the present disclosure can be used with helicopters
having hydraulic cyclic assistance, as will be further described
below.
[0029] Helicopter 10 includes a stick or cyclic 14 having a control
handle or grip 18 that is configured for engagement with the hand
of a pilot. As will be appreciated by one of ordinary skill in the
art, stick 14 can be moved fore and aft (toward and away from an
instrument console 20) to control pitch of the helicopter and
transversely for purposes of controlling roll of the helicopter in
a coordinated manner to produce controlled flight. Additional
control inputs are provided by the pilot via a pair of pedals in
order to control the yaw orientation of the helicopter by changing
the pitch of a tail rotor. It is noted that these yaw orientation
control components have not been shown since they are not
controlled by the autopilot of the present disclosure. Further, the
pilot also remains in control of the collective of the helicopter
as well as the throttle settings. The autopilot of the present
disclosure, however, can exert full control authority over stick 14
by moving the stick in any direction to the limits of its travel
under appropriate circumstances. Stick 14 passes below a deck 24 of
the helicopter and engages pitch and roll linkages of the
helicopter in a manner that is familiar to one of ordinary skill in
the art so as to control cyclic actuation of the main rotor of the
helicopter. The term "cyclic" refers to the variation in pitch of
the rotor blades of the helicopter on a per revolution basis. In
this regard, cyclic control can refer to manipulation of the stick
or the stick itself can be referred to as the cyclic. An autopilot
display processor unit (ADPU) 28 can be mounted in instrument
console 20 to provide indications to the pilot as well as to
provide processing capability and other capabilities, as will be
further described.
[0030] The cyclic, in particular, handle 18 includes a Switch
Module Assembly 26 that can be mounted as shown. Details of handle
18 are shown in a further enlarged inset view. The switch module
can contain switches including an engage/disengage switch 29a and a
trim/mode "top-hat" switch 29b (4-way with a center contact). The
top-hat switch allows the pilot to trim the course, speed and
altitude. In another embodiment 5 independent switches can replace
the top-hat switch. There can be a time-out feature in the
autopilot processor which prevents switch faults or wiring faults
from causing continuous trimming. The mode switch can select and
deselect altitude, speed, hover or position hold modes based on
current flight condition.
[0031] Still referring to FIG. 1, autopilot 12 implements cyclic
control through a number of component assemblies that are
appropriately located on the helicopter. A main autopilot unit 30
is located below the main deck of the helicopter. In the present
embodiment, main unit 30 includes an L-shaped enclosure 31 that
supports electronics as well as a pitch control linkage 32a and a
roll control actuator linkage 32b, which may be referred to
generally or collectively by the reference number 32. Each of these
linkages includes an actuator that is located within the main unit
enclosure, as will be further described. A distal end of each of
the linkages engages the lowermost end of stick 14 to implement
what is known as a parallel control system. In this regard, it
should be appreciated that the original cyclic control linkages of
helicopter 10 between stick 14 and the rotor remain intact. That
is, inputs from the helicopter pilot as well as the autopilot are
input directly to the stick. Details with respect to the pitch and
roll control linkages provide for a parallel control input
arrangement. A series type autopilot control system, in contrast,
requires breaking the original cyclic control linkages of the
helicopter between the stick and rotor such that the autopilot
actuators can be inserted into the break.
[0032] Turning to FIG. 2, components of the helicopter and
autopilot are shown in an overhead perspective view. In this view,
a pitch actuator 60a and a roll actuator 60b (which may be referred
to generally or collectively by the reference number 60) can be
seen within L-shaped enclosure 31 with the lid of the enclosure
rendered transparent. Main unit electronics 66 are located within
the enclosure and are suitably electrically interfaced (not shown)
both externally and to the actuators.
[0033] Referring to FIG. 3, an embodiment of actuator 60 that can
be used for the pitch and roll actuators throughout this disclosure
is seen in a perspective view installed within enclosure 31 and
connected to a control linkage 32. Each actuator includes a housing
82 having a gear arrangement, yet to be illustrated, within the
housing, dual motors Motor A and Motor B, and a clutch arrangement
84 for selectively engaging and disengaging the motors to rotate an
output shaft which is not visible on the opposite side of housing
82. As will be seen, the gear arrangement allows motors A and B to
simultaneously drive the output shaft or either one of the motors
to individually drive the output shaft. In the present embodiment,
motors A and B are brushless DC motors having a Y stator winding
configuration which requires coordinated inputs to drive the motor
phases in a particular sequence. As such, the motors cannot runaway
under their own power. The motors include Hall effect sensors that
are used for purposes of timing electrical drive pulses to the
stator of the motor. Further details with respect to the motors and
related drive considerations are provided at one or more
appropriate points hereinafter.
[0034] FIG. 4 illustrates an embodiment of a gear drive arrangement
100 that can be used in the actuator of FIG. 3. Initially, it is
noted that the gear drive arrangement is a multi-stage reduction
drive, for example, on the order of about 1750:1. Also, teeth have
not been illustrated on a number of the gears to be described, but
are understood to be present. Other embodiments may not require
gears with teeth. Motors A and B have output shafts supporting
gears that engage a gear 102 on a first shaft 104. An opposing end
of shaft 104 supports a smaller gear 106 that drives a gear 110
that is supported on a second shaft 112 which also supports a
smaller gear 114 (partially hidden in the view of the figure). It
is noted that shaft 112 can comprise a clutch shaft that can move
laterally to selectively engage or disengage the actuator motors
from the remaining gears of the gear drive. A suitable clutch
arrangement is described, for example, in U.S. Pat. No. 7,954,614
which is incorporated by reference. The clutch arrangement relies
upon movement of the clutch shaft along its elongation axis by
using a permanent magnet that is mounted on a distal end of the
shaft. A clutch actuator 113 (FIG. 3) can selectively move (for
example, rotate) another permanent magnet in relation to the clutch
shaft mounted permanent magnet such that the clutch shaft is
magnetically biased to move between an engaged position and a
disengaged position. The clutch shaft remains in a current
operational position despite a power failure. Gear 114, in turn,
selectively drives a gear 120 that is supported on a third shaft
122. The latter also supports a smaller gear 124 that drives a gear
130 that is supported on a forth shaft 132. The forth shaft, in
turn, supports a smaller gear 134 which is arranged to rotate an
output gear 140 that is supported on an output shaft 142 of the
actuator. The output gear is configured to provide sufficient
rotation to move stick 14 through its full range of motion. In an
embodiment, the actuators of the present disclosure are
sufficiently robust, in terms of the generated level of actuation
force, so as to be capable of controlling the cyclic of a
hydraulically equipped helicopter through a failed hydraulic
system. For example, actuator 60 is capable of applying forces of
at least 100 pounds to the cyclic. While the present embodiment has
been designed to provide actuation forces at this level, it should
be appreciated that in another embodiment, significantly higher
force can be provided since the geared motor drive of the actuator
arrangement that has been brought to light herein is capable of
providing forces at the output shaft of the actuator as high as 300
pounds. As seen in FIGS. 1 and 2, the actuator forces are applied
to the bottom of the cyclic whereas pilot forces are applied to the
top of the cyclic. Accordingly, the pilot is provided with a
mechanical advantage due to the different lever-arm lengths. On the
R22 helicopter, the mechanical advantage that the pilot has at the
top of the stick compared to the bottom of the stick where the
actuators are attached is roughly 7:1. In such a case, an actuator
applied force of 100 pounds is equivalent to about 14 pounds of
pilot applied force. Similarly, while the actuator can generate
very large forces, the force-limited-link that is described below
generally will not transmit forces of such a magnitude through to
the base of the cyclic, unless a much stiffer force-limited link is
installed.
[0035] Having described the mechanical components of the autopilot
in detail above, it is now appropriate to describe the autopilot in
terms of the relationship between the aforedescribed components and
related control electronics. In particular, FIG. 5 is an embodiment
of a block diagram of autopilot 12. In this regard, main unit 30
comprising enclosure 31, the pitch and roll actuators 60 and
electronics 66 may be referred to hereinafter as the Motor Control
Processor Unit (MCPU) or main autopilot unit 30. The MCPU includes
three microprocessors, each of which may be referred to as a Motor
Control Processor (MCP). There are three MCPs, individually
designated as MCP A, MCP B and MCP C. These processor units each
access a sensor suite of tri-axial MEMS rate sensors and tri-axial
MEMS accelerometers. The MCPs are used to provide an inner loop of
an overall control system having an inner loop and an outer loop.
The MCPs provide commands to brushless DC motors, Motor A and Motor
B of pitch actuator 60a and roll actuator 60b, driving the control
system for the helicopter. All inter-processor communication can be
through a serial bus that is natively supplied on each of the
processors. Data integrity can be protected, for example, through
the use of a cyclic redundancy check (CRC) incorporated into the
data stream.
[0036] Each one of the three MCPs can execute identical level "A"
software to constitute a triple-redundant system. The motor control
processors are interconnected so that they can share data. Each
processor reads its sensor suite and compares its data with sensor
data coming from the other two processors for purposes of
consistency and each motor control processor computes averages of
all the corresponding sensors to use for further processing. Sensor
data determined to be erroneous is eliminated from the averaging. A
warning signal of sound and light can be sent to autopilot display
processor unit (ADPU) 28 on instrument panel 20 (FIG. 1). In an
embodiment, status lights, best seen in the enlarged inset view of
the ADPU in FIG. 1, include green (normal), amber (caution) and red
(failure), as well as dual warning horns to provide system status
indications. The warning horns also provide system status
notifications and alarms along with the status lights. Both the
status lights and horns interface directly to the MCPs. In some
embodiments, sounds and/or warnings can be transmitted over the
helicopter audio system such that notifications can be heard in the
pilot's headset as well as being issued from the ADPU.
Complementing the status lights and horns is a display which
provides current autopilot system settings such as engagement
status, course, magnetically slaved gyroscopic heading, speed over
ground and any warning messages. Also on the panel is a testing
button which initiates an Initiated Built-In Test (IBIT).
[0037] The MCPs also read Hall sensor data from the actuator
motors, which can be used to indicate the current position of each
actuator, and a command signal coming from an autopilot display
processor (ADP) which forms part of the ADPU. In this regard, the
ADPU serves as the outer control loop to provide command signals to
the inner loop. Using all these data, each MCP calculates a control
signal for the motors in terms of a PWM (Pulse Width Modulation)
and direction of rotation. Each processor also uses the Hall sensor
data to control the power connections to the armature of the
brushless motors assigned to it. Each MCP compares its PWM command
signal and rotation direction for the pitch and roll actuators with
commands generated by the other two MCPs for agreement. Since all
processors are using the same data to compute motor commands, they
should produce identical output signals. Signals for
agreement/disagreement with the other two processors are sent to a
Hardware Voting circuit 200 that will disable control input
capability of any MCP that is in disagreement with the other two
MCPs.
[0038] It should be appreciated that for a gear ratio of 1750:1,
one revolution of the motor rotates the actuator output shaft by
only about 0.2 degrees. In and by itself, this resolution can be
sufficient for monitoring the actuator output position. For
example, rotation of the motor shaft can be detected using a magnet
that is mounted on the shaft, as is familiar to one having ordinary
skill in the art. In an embodiment, however, Hall sensor data from
the motors can be used to determine the incremental position of the
actuator output shaft of each actuator. In this regard, each
actuator motor includes 3 Hall sensors. The Hall sensor pulses can
act like an incremental up/down counter. The position of the
arm/output shaft relative to a reference location can be tracked
constantly. For example, a zero reference position of the actuator
output shaft can be defined when the actuator is engaged via clutch
84. Such zero reference position tracking can be used for certain
failures wherein the best approach resides in restoring the
actuator shafts to their averaged positions prior to the failure.
Since each motor includes 3 Hall sensors and 4 poles, there are 12
Hall state changes per revolution of each motor. Remarkably, by
monitoring the Hall state changes, resolution can be increased by a
factor of 12 such that a resolution of about 0.017 degrees is
provided at the output shaft of the actuator. In an embodiment, a
corresponding movement at the top of the stick in FIG. 1 can be
about 0.0039 inch.
[0039] As described above, each actuator includes motor A and motor
B. Each individual motor is controlled by one MCP. Thus only MCP A
and MCP B control motors. In particular, MCP A controls motor A in
each of pitch actuator 60a and roll actuator 60b, while MCP B
controls motor B in each of pitch actuator 60a and roll actuator
60b. MCP C (the third processor) does not control a motor but
performs all calculations to generate stick commands as if it were
controlling a motor. In this regard, a third motor can readily be
added to each actuator (see FIG. 4) that would engage gear 102 in
the same manner as motor A and motor B, but responsive to MCP C.
The latter, however, votes in a manner that is identical to the
other two processors. For example, if MCP A and MCP C agree on the
control of the pitch motor, but MCP B does not, then MCP B will be
voted out from control of its pitch motor, MCP B will still control
its roll motor unless MCP A and MCP C also vote out control of that
motor. On the other hand, if MCP C is voted out, no actuator motors
will be affected, but a warning light and horn can be actuated as
would be the case for the MCPs which control motors. Further
details with respect to this architecture are provided
hereinafter.
[0040] The actuators are designed such that either one of motor A
or motor B is independently capable of driving the actuator to
control the helicopter. The output shaft of a failed motor will be
rotated by the remaining motor. If one of MCP A or MCP B is voted
out, the autopilot can continue to function despite the fact that
each of these MCPs controls motors. As stated, there can be a
warning light and a brief sounding of the horn to notify the pilot
that there has been a non-critical autopilot malfunction.
[0041] The MCPs have full authority over the controls and are not
rate limited. The MCP control section is the only portion of the
autopilot that can create a critical or major hazard malfunction.
Accordingly, the MCPU is designed as triple-redundant with level
"A" designated software for purposes of operating the inner loop of
the autopilot. These factors greatly reduce the probability of a
critical failure. Applicants recognize, however, that the software
corresponding to the outer loop can be partitioned from the inner
loop software in a way that allows the outer loop software to be
designated at a lower level "C" certification. The outer loop
software is handled by the ADP (Autopilot Display Processor) in
ADPU 28. The MCPs convert requested autopilot commands from the ADP
into actuator control signals that can drive the actuator motors
within defined operational limits. In this regard, it should be
appreciated that level A software is handled by the triple
redundant MCPs while level C, outer loop software is handled by a
completely different processor. By way of still further
explanation, a single executable runs on each MCP. The MCPs, which
may be referred to as the triplex processors, can execute identical
software. Thus, the autopilot control laws are partitioned between
the ADP and triplex processors. The ADP processes the outer loop
dynamics and autopilot modes while the triplex MCPs process the
inner loop dynamics. The ADP further provides the pilot's graphical
and testing interface to the autopilot and executes the autopilot
control laws to determine actuator commands based on sensor and GPS
data. A GPS 160 can provide GPS data 162 to ADPU 28. The GPS data
is typically provided in packets, indicated as P.sub.x that can be
spaced apart in time by an interval I. The GPS data, for example,
can include altitude, groundspeed, commands, track angle, latitude
and longitude. Generally, the GPS unit will output velocity data to
the nearest knot and heading data to the nearest degree. Thus,
there can be some amount of rounding error implicated with the use
of this data, as will be further discussed. GPS updates, by way of
the illustrated packets P.sub.x, are received on the basis of a
given GPS unit that is in use. Packets can be received relatively
rapidly, for example, as often as five per second, however, a
longer update interval can be involved. In some embodiments, GPS
packets are received at one second or three second intervals.
Accordingly, the processor interfaces with the GPS, magnetometer,
accelerometer and yaw rate gyro which provide the position,
altitude, groundspeed, course and heading data. The ADP monitors
the health of these sensors but does not check the validity of the
data. The IBIT test switch also interfaces to the ADP. In another
embodiment, the ADP can be designed in the same manner as the MCPU
with triple redundancy. With both the MCPU and ADP in a triple
redundancy configuration, the autopilot can tolerate a single
failure in either or both units and still remain fully
functional.
[0042] The MCPs accept data from the ADP which can include commands
as well as data from external GPS 160. The data can be screened by
each MCP to detect errors or malfunctions. The control is
rate-displacement limited by the MCPs. The MCPs will not allow a
command from the ADP to create a hazardous response from the
helicopter. GPS and magnetometer data are used in both the MCP and
ADP. The GPS and magnetometer data are used in the MCPs to remove
drift errors associated with the rate sensors of each sensor suite.
The GPS data can also be checked for errors.
[0043] The MCPs constantly monitor for both internal and external
faults. In the event of an ADP failure, any one MCP can immediately
recognize the situation based on update rate and control signal
conformity. In response, the MCPU, in one embodiment, will then
cause the inner loop to hold the helicopter straight and level. In
another embodiment, the MCPU can act in the manner of a SAS
(Stability Augmentation System) or a dead reckoning system and
control the helicopter based on internal rate signals. The MCPs
will attempt to hold zero rates and/or heading and also actuate a
horn and light to indicate a failure. It has been empirically
demonstrated that the helicopter can maintain prolonged flight with
only MCP control, providing more than ample time for the pilot to
take control and disengage the autopilot. The ability to detect
excessive autopilot response resides in the triplex motor
controllers as detailed herein. The triplex processors monitor
sensors and also check to confirm that calculated responses are
within limits. Pitch and roll commands from the ADP are limited
based on such command filtering by each of the triplex processors.
Each triplex processor can detect whether a limit has been exceeded
and can initiate safe shut down of the autopilot. Pitch and roll
axis commands can be monitored identically but with different limit
values. The monitors are dynamic; that is, the limit values can be
frequency/rate dependent. Redundancy management features for each
axis can include stick rate limiting and body rate monitoring.
[0044] Each MCP processor can be provided with an independent power
supply. A total power failure of the helicopter's electrical power
system can cause the actuators to lock in position for about five
seconds using a dynamic braking feature that is described in detail
below. This five second time period is sufficient for the pilot to
take over control. In this regard, the autopilot does not let the
cyclic stick flop over by releasing control responsive to a power
failure. Even though the actuators are locked, however, the pilot
can still perform control over the helicopter since there are
override or force limited links 300a (pitch, seen in FIG. 1) and
300b (roll, seen in FIGS. 1 and 2) between each actuator and the
cyclic stick. These links are rigid for forces below an unseating
value and compliant at higher forces to allow the pilot to safely
maneuver and land the helicopter even if disengagement of the
system cannot be achieved. It has been empirically demonstrated
that a pilot can control the helicopter, including hovering and
landing, with both actuators in what is referred to as a "locked"
state. The locked state is provided by shorting all windings of the
actuator motors and is used in a dynamic braking embodiment
described below. The override links are described in detail in a
separate patent application. In a helicopter that does not utilize
a hydraulic interface to the cyclic, cyclic vibration isolators
302a (pitch) and 302b (roll) can be located on the output shaft of
each actuator. The vibration isolators may be optional for use with
a helicopter having hydraulic cyclic control since the hydraulic
system generally provides damping of cyclic oscillations. The
vibration isolators reduce the two per revolution oscillating
motion, that is present in the R22 rotorcraft control linkage and
other light helicopters, to prevent vibratory loads on the
rotorcraft control and to increase the fatigue life of the actuator
components. The cyclic vibration isolators are described in detail
in a separate patent application.
[0045] The sensor suite of each MCP can also include memory such
as, for example, EEPROM or other suitable memory. If there is an
error detected by an MCP during operation, the error code can be
stored in the EEPROM of the sensor suite associated with the MCP.
The EEPROM can later be read in the context of determining the
cause of failure. The EEPROMs can also contain parameters specific
to the model of the helicopter in which the autopilot is installed
such as, for example, control loop constants and sensor offsets and
gains.
[0046] FIG. 6 is a schematic representation of an embodiment of
hardware voting section 200 of FIG. 5. It should be appreciated
that one having ordinary skill in the art may readily implement a
software version based on the hardware configuration that is shown.
Main unit electronics 66 (FIGS. 2 and 5) includes an individual
driver for Motor A and Motor B of each actuator. In particular, a
first driver 600 drives Motor B of roll actuator 60b, a second
driver 602 drives Motor B of pitch actuator 60a, a third motor
driver 604 drives Motor A of roll actuator 60b and a fourth motor
driver 606 drives Motor A of pitch actuator 60a. In this regard,
each MCP generates separate commands for pitch and roll that are
targeted for pitch and roll actuators 60a and 60b, respectively.
For example, MCP A delivers pitch actuations to Motor A of actuator
60a and delivers roll actuations to Motor A of actuator 60b. For
purposes of the present description, a logic high signal on disable
inputs 610 of each driver (individually designated as 610a-610d)
will result in disabling that driver, although any suitable logic
scheme can be employed. During normal operation, these drivers
operate in a manner that will be familiar to those of ordinary
skill in the art with respect to driving the armature coils of
brushless DC motors in timed coordination. As will be seen, the
status for a given motor is determined independently, based on
independent pitch and roll vote indications that are cast by the
MCPs that do not control the given motor.
[0047] Still referring to FIG. 6, each motor driver disable input
610a-610d is electrically connected to a respective output of one
of a set of two-input AND gates 614a-614d. Further, each AND gate
614 receives vote indications from the two MCPs that are not
associated with the particular motor driver to which each AND gate
is connected. For example, AND gate 614a, which can disable driver
600 for Motor B of roll actuator 60b, receives a first roll vote
indication from MCP A that is designated as "MCP A vs. B roll vote"
to indicate that the vote is cast by MCP A for or against the
command generated by MCP B. Similarly, AND gate 614a receives a
second roll vote indication from MCP C that is designated as "MCP C
vs. B roll vote" to indicate that the vote is cast by MCP C for or
against the command generated by MCP B. Thus, roll votes cast by
MCP A and MCP C are individual indications by these two MCPs as to
whether a current roll stick movement command being generated by
each of MCP A and MCP C agrees or disagrees with the current roll
stick movement command being generated by MCP B. In the present
implementation, a vote by MCP A or MCP C against or in disagreement
with the MCP B roll command is characterized as a high logic level.
If only one of MCP A or MCP C casts a roll control vote against MCP
B, only one input of AND gate 614a is logic high such that the
output of AND gate 614a remains at logic low, which does not
disable driver 600 to maintain Motor B of actuator 60b in a normal
operational status. On the other hand, if both MCP A and MCP C cast
a vote against roll control by MCP B, AND gate 614a will output a
logic high level that disables motor driver 600 such that Motor B
of roll actuator 60b is deactivated. Control of each of the
remaining three motors is implemented in a manner that is
consistent with the foregoing descriptions, as illustrated by FIG.
6.
[0048] Attention is now directed to further details with respect to
the inner and outer control loops of the present disclosure. In an
embodiment, the inner loop can be configured for providing control
of one or more selected orientation parameters of the helicopter
such as, for example, attitude hold including a given level of
redundancy and/or software certification applied to the inner loop.
The outer, autopilot loop can be configured for providing at least
one navigation function with respect to the flight of the aircraft
including a different level of redundancy and/or software
certification as compared to the inner loop. The redundancy and/or
certification level applied to the inner loop can be greater than
the redundancy and/or certification level applied to the outer
loop.
[0049] FIG. 7 is a flow diagram, generally indicated by the
reference number 700, which illustrates an embodiment of a method
for operating an inner loop 702 and an outer loop 704, as well as
interaction between these loops. The method starts at 710 and
proceeds to 712 which reads an ADP command that is passed from the
outer loop, as will be further described. For the moment, it is
sufficient to note that an ADP command is obtained for each
iteration though the inner loop. An ADP command filtering decision
is made at 713 as to whether the ADP command is within acceptable
limits, for example, as described above. If the command is
acceptable, operation proceeds to 714. On the other hand, if the
command is not acceptable, operation proceeds to failure handling
716 which can initiate the issuance of warnings and/or shut down
the autopilot. At 714, each MCP reads the sensors of its sensor
suite (FIG. 5) while the ADP reads ADP sensors 718 and GPS 719. At
720, the ADP sensor data is shared with the MCPs. At 722, the MCPs
share MCP sensor suite data (FIG. 5) with one another to form an
average set of sensor data that is used by each MCP and which is
shared with the ADP. Further, the MCPs determine an attitude for
the helicopter which is also shared with the ADP as indicated by a
connection 724. At 726, each MCP determines actuator motor
commands. At 728, voting is performed based on the commands, for
example, using the hardware implementation of FIG. 6 or a software
equivalent. At 729, the results of voting are compared. In the
event that there is a processor dispute, operation transfers to
failure handling 716. Any appropriate action can be taken as a
failure handling depending on the voting results. For example, if
control has been voted out for one motor of a particular actuator,
that motor can be deactivated, as discussed above. Appropriate
warnings can be issued. If step 730 does not identify a voting
dispute, operation proceeds to 730, wherein the motors are actuated
based on the voting.
[0050] Still referring to FIG. 7, attention is now directed to
further details with regard to the operation of outer loop 702. It
is noted that inner loop 702 and outer loop 704 execute in parallel
in a cooperative manner. In this regard, at 740, the outer loop
determines an ADP command that is based on the current flight mode
and control laws for the particular rotorcraft in which the
autopilot is installed. The control laws and related parameters can
be customized on a per rotorcraft basis. The determination is
based, at least in part, on rate data from the MCPs as well as an
attitude for the helicopter that is generated by step 722 of the
inner loop, taken in conjunction with data from ADP sensors 718 and
GPS 719. At 760, command filtering is applied which serves to limit
ADP commands for subsequent use by the inner loop. The current ADP
command, subject to filtering, is the read by step 712. In this
regard, it should be appreciated that step 726 applies command
limiting, as described above.
[0051] FIG. 8 is a schematic diagram of an embodiment of a dynamic
braking system, generally indicated by the reference number 800,
that can be used, for example, with actuator 60 of FIG. 3. As
described above, each motor can include a Y-connected stator. In
particular, each motor includes three stator coils, designated as
A1-A3 for Motor A and as B1-B3 for Motor B. For purposes of the
present discussion, it should be noted that the motors are selected
for characteristically exhibiting a resistance to rotation of the
drive shaft of the motor in response to shorting or grounding the
drive coils. Sets of motor driver lines 802 and 804 are connected
to appropriate motor drivers, for example, as seen in FIG. 6. Each
stator coil is also electrically connected to the drain terminal D
of one of a group of six n-channel enhancement mode MOSFETs,
individually designated as 806a-f and which can be referred to
collectively as MOSFETs 806. The source terminal S of each of these
transistors is connected to a ground 810. Accordingly, a positive
voltage on a gate terminal G of these MOSFETs turns on each MOSFET
such that the drain to source channel is essentially shorted, so as
to function as a switch to connect or short the associated stator
coil to ground. A drive circuit 820 receives input power from the
helicopter, which is designated as V.sub.in, and can comprise
battery power from the helicopter. For purposes of drive circuit
820, the input power for proper operation can range from 9-32 volts
DC. When power is present, during normal operation of the
helicopter, a zener diode D2 regulates to 9 volts which biases the
gate terminal of a p channel depletion mode MOSFET 830 to 9 volts.
Current flows through a diode D1 and a 1K ohm resistor to another
zener diode D3 which biases a source terminal of MOSFET 830 to 7
volts, also charging a capacitor C1 to 7 volts. Thus, V.sub.GS of
MOSFET 830 is 2 volts DC such that the transistor is biased into an
off state. Since MOSFET 830 is off, the drain terminal of this
MOSFET is at zero volts which biases the gate of each of MOSFETs
806 to zero volts such that each of these transistors is also
off.
[0052] Still referring to FIG. 8, responsive to a power failure in
which Vin goes to zero volts, the gate voltage of MOSFET 830 drops
to zero volts which turns the transistor on. Once MOSFET 830 is on,
it provides a discharge path for capacitor C1 to a 100K ohm
resistor R3. This discharge current results in the application of a
positive gate voltage to each of MOSFETs 806 such that these
transistors turn on, thereby connecting the stator coils of motors
A and B to ground as capacitor C1 discharges through R3.
Accordingly, MOSFETs 806 will remain on based on an RC time
constant that is determined primarily by capacitor C1 and resistor
R3. In the present example, the time constant is approximately 4.7
seconds. In practice, MOSFETs 806 will remain on for about 4
seconds. While this time period can be varied through the selection
of component values, it should be selected to provide for a period
of time that is sufficient for the pilot to take over manual
control from the autopilot. Even during the time period during
which braking is applied, the pilot is able to take over and
maintain control of the helicopter due to the presence of force
limited links 300a and 300b, as described above and shown in FIGS.
1 and 2. One of ordinary skill in the art will recognize that the
circuit of FIG. 8 can readily be modified and adapted in view of a
particular installation.
[0053] Referring to FIG. 5 in conjunction with FIG. 8, it should be
appreciated that two instantiations of the circuit of FIG. 8 are
utilized. That is, one instance of the circuit of FIG. 8 is
connected to the motors of each of the roll actuator and the pitch
actuator. By temporarily shorting the motor coils to ground based
on the time constant described above, resistance to rotation of the
output shaft of each motor is produced. The degree of resistance is
amplified by the gear arrangement of the actuator such that
significant force is needed to move the stick from the position at
which the power failure occurred. Empirical results have
demonstrated that the stick will not flop over as a result of a
power failure while allowing ample time for the pilot to take over
control of the helicopter from the autopilot. It should be
appreciated that dynamic braking, as taught herein, can be used
with any motor that exhibits resistance in response to shorting at
least selected drive coils.
[0054] FIG. 9 illustrates another embodiment of a dynamic braking
system, generally indicated by the reference number 900, that can
be used, for example, with actuator 60 of FIG. 3. In this
embodiment, each motor winding is connected to one terminal of a
normally closed contact, individually designated as NC.sub.1,
NC.sub.2 and NC.sub.3. The opposite terminal of each NC contact is
connected to ground 810. Each of a first relay 902 and a second
relay 904 includes a relay coil that is driven by Vin. While the
present example illustrates the use of a 3 pole, single throw relay
(having only normally closed contacts) in association with each
motor, it should be appreciated that any suitable type of relay can
be used. During normal operation, Vin is applied to each relay coil
such that the normally closed contacts are in an open condition. If
Vin is lost, however, the normally closed contacts close to connect
each stator coil to ground thereby applying dynamic braking, as
described above. Since the motors remain in a braked condition
responsive to power failure, the pilot is able to take over and
operate the helicopter due to the presence of force limited links
300a and 300b, as described above and shown in FIGS. 1 and 2.
[0055] Referring again to FIG. 5, it should be appreciated that the
MCPs do not require an input from an airspeed sensor such as, for
example, a pitot tube. In this regard, Applicant has discovered a
heretofore unseen approach for determining airspeed based on the
configuration of the system that has been described above and as
will be described immediately hereinafter.
[0056] Attention is now directed to FIG. 10 which is a vector
diagram, generally indicated by the reference number 1000, having a
North compass direction oriented toward the top of the view. A
first groundspeed vector 1004 (V.sub.GPS.sub.t) that is measured at
a first time is shown in conjunction with a second groundspeed
vector 1008 (V.sub.GPS.sub.t+.DELTA.t) that is taken at a second
time, .DELTA.t, after the first groundspeed vector. As shown, a
ground track is associated with these vectors. Based on the
difference in track orientation between vectors 1004 and 1008, it
should be appreciated that the aircraft is engaged in a turn. Thus,
a rate of turn .omega. is associated with this turn and is
indicated by a circular arc. Points A and B terminate the
respective first and second groundspeed vectors. The difference
between these two groundspeed vectors, between points A and B, can
be reduced to a northing component 1012 and an easting component
1016. The northing component can be mathematically designated as
.DELTA.V.sub.N while the easting component can be mathematically
designated as .DELTA.V.sub.E as shown in FIG. 10. Applicant
recognizes that dividing these incremental components by .DELTA.t,
the time increment between the times at which the vectors are
measured, can determine a centripetal acceleration 1020, as
indicated by an arrow. If ascent or descent of the aircraft is
involved, the incremental change vector can be three-dimensional.
In most cases, however, the vertical component of movement is much
smaller than the in-plane components shown in FIG. 1, such that the
vertical component can be ignored, although this is not required.
Thus, the magnitude of centripetal acceleration vector, .alpha.,
can be expressed as:
.alpha. = ( .DELTA. V N .DELTA. t ) 2 + ( .DELTA. V E .DELTA. t ) 2
( 1 ) ##EQU00001##
[0057] The magnitude of the centripetal acceleration, like other
acceleration magnitudes, is invariant of the inertial system that
it is measured from. Therefore, if the centripetal acceleration is
measured in an Earth-fixed reference system, as is case with GPS
measurements, the magnitude is the same as that which would be
obtained from a system moving with the prevailing wind that carries
an aircraft along.
[0058] With the centripetal acceleration vector denoted as a:
.alpha.=.omega..times.V (2)
[0059] where rotation speed vector .omega. is shown in FIG. 10 and
is measured in the selected reference system along with a velocity
vector, V. If it is assumed that rotation vector, .omega., is
normal to velocity vector, V, which is typically a reasonably
accurate, if not better, assumption for aircraft flight, then an
expression can be framed in terms of the magnitudes of the vectors
according to:
V=.alpha./.omega. (3)
If .omega. is measured in the Earth-referenced system, for example,
based on GPS data, then V is the groundspeed (GPS speed). On the
other hand, if .omega. is measured in a system that is moving with
the aircraft system as rate of heading change, then V is the
airspeed. Based on this recognition, two simulations were
performed, as described immediately hereinafter.
[0060] In a first simulation, an aircraft was flying on a north
track (Earth system) at an airspeed of 80 knots. A 30 knot wind is
assumed coming from the north-east. The aircraft started a constant
rate roll into a standard rate turn (3 degrees/sec.). The roll rate
was such that it took two seconds to enter the bank angle to
produce this standard rate turn. The track turn rate was measured
in the Earth-fixed coordinate system and not constant due to the
wind. The aircraft then rolled out at the same rate as it rolled
into the standard rate turn. V was determined from Equation 3 using
Earth-based measurements for centripetal acceleration and track
turn rate and found to be equal, at least to a reasonable
approximation, to the groundspeed as produced by a GPS receiver, as
expected.
[0061] In a second simulation, again using the same model including
the constant heading turning rate, V was again determined. This
time, however, an aircraft-fixed coordinate system was used with
the heading taken from a gyro on-board the aircraft to determine
turn rate .omega. in Equation 3. The determined value for speed was
an aircraft frame of reference airspeed, that was at least
approximately equal to the actual, input airspeed of the aircraft,
as expected. Accordingly, Applicant has discovered that by using
only GPS data and a rate gyro, airspeed can be determined during
turns such that an airspeed measuring system is not necessarily
required in an embodiment of an autopilot. In another embodiment,
compensation for a failed airspeed detection system can be provided
based on the approach that has been brought to light herein. For
example, upon detection of a failed or malfunctioning airspeed
detector, a partial turn can be performed by an autopilot or pilot
to obtain airspeed based on the described approach. In this regard,
it should be appreciated that airspeed data can be important with
respect to preventing an aircraft from overspeeding, for example,
when flying in an autopilot controlled constant speed mode, that
might cause structural failure. That is, if an aircraft turns from
a downwind direction into the wind and attempts to maintain
groundspeed at a constant groundspeed, the now opposing wind can
result in exceeding the airspeed design limits of the aircraft to
maintain a constant groundspeed. For example, consider an aircraft
flying at 80 knots with a direct tailwind of 30 knots. The GPS
groundspeed being maintained is 110 knots. Now, if the aircraft
performs a 180 degree turn while maintaining groundspeed, the
aircraft will roll out of the turn with a 110 knot groundspeed into
a 30 knot head wind. This means the aircraft will be flying at an
airspeed of 140 knots, which could be well beyond the allowable
airspeed of the aircraft.
[0062] For fixed wing aircraft, stall is an important
consideration. Embodiments of the method of the present disclosure
can provide a back-up to an airspeed indicator in the instance of a
malfunction due to ice covering the pitot tube of the airspeed
sensor or if insects or debris block the pitot tube. Applicant
recognizes that both fixed wing and rotary wing aircraft normally
include the necessary panel-installed instruments and/or sensors
for purposes of implementing the airspeed determination method of
the present disclosure. For example, the heading rate can be
provided by measuring the change in heading indicated by a heading
gyro. In such an embodiment, an electrical output from the heading
gyro is interfaced to the new airspeed system. In this regard, such
heading gyros, with an electrical output, are common in the
aircraft market.
[0063] In view of the foregoing, systems and methods based on the
recognitions that have been brought to light herein may generally
be referred to hereinafter as CAVS (Centripetal Acceleration
Velocity System). A CAVS autopilot design, according to the
recognitions herein, can hold the GPS groundspeed until the start
of a turn, which is readily detectable, and then switch to holding
a centripetal force based airspeed until the end of the turn. At
the end of the turn, as detected, the autopilot can hold a new
groundspeed associated with the groundspeed that is detected as the
aircraft rolled out of the turn. In the above example, the aircraft
would then maintain a 50 knot groundspeed with an 80 knot
airspeed.
[0064] Attention is now directed to FIG. 11 which is a flow diagram
that illustrates an embodiment of a CAVS method, generally
indicated by the reference number 1200, for determining airspeed
based on centripetal force in a manner that is consistent with the
foregoing descriptions. Initially, it is noted that that the only
input requirements for the system are: (1) heading from a heading
gyro, (2) altitude, (3) track angle and (4) groundspeed, wherein
the latter three items can be provided from a GPS receiver. In this
regard, some GPS units supply the north and east components of the
groundspeed. For those units, the track angle on the ground can be
determined from the two groundspeed components at a given position;
conversely, the two speed components can be determined from the
track angle at the given position when only the magnitude of the
groundspeed is provided. The altitude can be used to determine the
vertical velocity (rate of climb) such as between points A and B of
FIG. 10, which would only have a noticeable impact if the vertical
velocity acceleration became a significant percentage of the
groundspeed-heading change acceleration such as, for example,
greater than 10 percent. The vertical component of acceleration can
be measured by a rate gyro or multiple axis rate gyros, resolved
into the proper component of rate. If the vertical velocity
acceleration (i.e., pitch rate) becomes significant, the
acceleration vector can be taken as three-dimensional. The method
begins at 1204 and moves to 1208. This step monitors for the
availability of GPS data and captures new data whenever such data
is available. The data, as is the case with any data described in
conjunction with the embodiments herein, can be directed to a
rolling buffer, for example, in ADPU 28 (FIG. 5) which can store
data corresponding to a time interval that extends from present
time and back in time by a selected number of seconds. Of course,
such a rolling buffer and/or other memory can be provided at any
suitable location within the system. The time interval can be
selected based on the particular update rate and data resolution
that is in use. For a typical light aircraft having data
resolutions of 1 degree for heading and 1 knot for speed
(hereinafter, standard light aircraft resolution), a buffer length
of 4 seconds can be used, although any suitable value can be used.
Having captured a packet of GPS data at 1208, operation can move to
1210 which imposes a packet selection interval. This selection
interval can be used for purposes of customizing the time
difference between GPS packets that are fed as inputs to the
process in view of the data resolution of a given GPS that is in
use. For example, Applicant has demonstrated useful results for
standard light aircraft resolution when input GPS packets are
spaced apart in time by three seconds. If the given GPS provides a
packet every second, step 1210 will result in acceptance of every
third GPS packet, skipping two intermediate packets. In FIG. 5,
packets P.sub.1, P.sub.4, P.sub.7 and so on, are accepted, while
intervening packets are not used. For packets that are not
accepted, operation returns to 1208. Of course, if the GPS provides
packet updates every three seconds, every packet is accepted and
step 1210 is not needed. In the instance of a GPS providing 5
updates per second, 14 update packets can be skipped such that
every 15.sup.th packet is accepted. Accordingly, this customization
provides for acceptable or better accuracy in determinations made
by the process since the change in track angle of the aircraft over
the selection interval that is chosen can be large in comparison to
the rounding errors that are present in the GPS data. For example,
in a standard turn of 3 degrees per second, the change in track
angle is 9 degrees which is large in comparison to a one-half
degree rounding error. Stated in another way, the difference in
track angle should be sufficiently large such that rounding induced
errors are small compared to the track angle change.
[0065] Once a GPS packet has been accepted, operation moves to 1214
and can separately store the GPS data for future reference in
addition to its availability via a rolling buffer. At least two GPS
packets are needed, of course, for purposes of performing CAVS
determinations, however, one or more previous GPS packets can be
accessed, for example, from the rolling buffer. In an embodiment,
whenever a packet is accepted, heading data can be collected, for
example, based on a heading gyro in ADPU 28 of FIG. 5 and stored
with reference to corresponding GPS data. Using heading, the method
only requires determining the difference between the current
heading and the heading read at the last time interval/packet
divided by the time difference .DELTA.t to produce a heading rate.
This represents the average rate of change in heading during the
time interval. In another embodiment, a heading rate sensor can be
monitored with a number of heading rates acquired over the interval
and averaged to produce an average rate over the interval which
should be equal to the difference in the heading over the entire
time interval.
[0066] At step 1216, the track angle data can be tested to insure
that the aircraft is actually in a turn. It should be appreciated
that the decision to accept a particular packet as being
representative of a sufficient amount of turning can be configured
according to a wide range of suitable embodiments such as, for
example, based on heading change. Step 1216 can monitor turns for
purposes of initiation of the CAVS airspeed determination method
when a sufficient amount of turning is detected, continuing the
method when ongoing turning is detected and/or terminating the
method when an insufficient amount of turning is detected. In an
embodiment, the decision can be made based on the change in track
(angle over the ground) from one packet to the next. The threshold
for this latter decision, for example, can be 9 degrees, although
any suitable value can be used. In another embodiment, the
packet-to-packet change in track can be monitored for more than two
consecutive packets. For example, if two consecutive packets
exhibit a 2 or 3 degree difference, the first packet can be held
for another two or three packets and, if at any time during that
process, a 9 degree difference is detected then operation can
proceed to step 1220. When a turn is detected, an autopilot can
switch from maintaining GPS-based groundspeed to holding CAVS
airspeed. If an insufficient amount of track change is detected,
operation can route to step 1224 which can output an exit
groundspeed value for display and/or autopilot use based, for
example, on whether a turn was already in progress. Further, the
autopilot mode can switch from holding CAVS determined airspeed
based on centripetal force to holding groundspeed to a value that
corresponds to the exit groundspeed value at the end of the turn.
Speed can then be displayed based on GPS groundspeed. If the track
angle data remains indicative of a turn, operation can proceed to
1220 which determines the northing and easting components
.DELTA.V.sub.N and .DELTA.V.sub.E based on the associated headings
and in a manner that is consistent with FIG. 10, for example, by
determining the differences between the northing and easting
groundspeeds for the current time period versus the prior time
period. Step 1220 can also compare rate of climb to determine if a
vector analysis including a vertical dimension is necessary. At
1228, the magnitude of centripetal acceleration .alpha. can be
determined, according to equation (2) by squaring the northing and
easting components, dividing each component by .DELTA.t, adding the
resulting values together and then taking the square root of the
sum. At 1232, the magnitude of the centripetal acceleration is
divided by the heading rate value .omega., in accordance with
equation 3, to produce the airspeed V.
[0067] At 1236, the known information including airspeed can be
used to determine the northing and easting airspeed velocities, the
wind speed and wind direction. In this regard the GPS-based
groundspeed and track are known as well as the airspeed and
aircraft-based heading. Based on ground-based speed and ground
track, northing and easting ground velocity components can be
determined. Similarly, northing and easting aircraft based
velocities can be determined based on airspeed and aircraft-based
heading. Subtracting the aircraft-based velocity components from
the corresponding track velocity components yields the wind
velocity components. A northing aircraft velocity V.sub.AN and an
easting aircraft velocity V.sub.AE can be expressed as:
V.sub.AN=V cos H (4)
V.sub.AE=V sin H (5)
[0068] Where heading H is available from any suitable source such
as, for example, from the MCPs. The northing wind speed W.sub.N and
the easting wind speed W.sub.E can be determined using:
W.sub.m=V.sub.GN-V.sub.AN (6)
W.sub.E=V.sub.GE-V.sub.AE (7)
[0069] Where V.sub.GN and V.sub.GE can be the GPS-based northing
and easting velocity components, respectively for the velocity
vector at the current location such as represented by vector 1008
of FIG. 10. The magnitude of the wind is given as:
|W|= {square root over (W.sub.N.sup.2+W.sub.E.sup.2)} (8)
[0070] The wind direction is given by:
W .cndot. = tan - 1 ( W N W E ) ( 9 ) ##EQU00002##
[0071] The wind speed quadrant can be determined based on the
mathematical signs of the northing and easting wind velocity
components. The wind direction is the wind velocity vector rotated
by 180 degrees since the wind direction is specified as the
direction from which the wind is coming.
[0072] Subsequent to step 1236, the method proceeds to 1224 at
which the results can be displayed directly and/or used for other
equipment such as an autopilot. Using the described centripetal
acceleration based method, it is not necessary to have an airspeed
indicator in order to hold airspeed in a turn. When not turning,
the groundspeed such as, for example, GPS speed can be used to hold
speed. In a turn, an autopilot can hold airspeed using the above
procedure so that excessive airspeed is not be produced as might
occur when holding groundspeed in a turn from a strong tailwind
into a strong headwind. Following 1224, the method can proceed to
step 1240 which can return the method to start 1204 or end the
method. While excessive airspeed has been used as an example, it
should be appreciated that too low an airspeed can also be a
concern. Turning from a head wind into a tail wind can create this
condition. Low airspeed can place the aircraft on the backside of
the power curve such that more power is required to go slower. If
the altitude is being held by an autopilot via controlling the
pitch angle, then the process will become unstable. For example, if
the aircraft loses altitude, the autopilot will pitch the aircraft
upward in an attempt to regain the altitude. This will slow the
aircraft down which will require even more power to hold altitude
thereby causing the aircraft to descend even faster. For a fixed
wing aircraft, this can induce a stall.
[0073] Referring to FIG. 12, a flow diagram illustrates another
embodiment of a CAVS based method, generally indicated by the
reference number 1400, for determining airspeed based on
centripetal acceleration in a manner that is consistent with the
descriptions relating to embodiment 1200 beginning at step 1216.
Accordingly, descriptions of like steps have not been repeated for
purposes of brevity. Embodiment 1400, however, includes a turn
detection procedure 1404 that is shown within a dashed rectangle
and which precedes step 1216. The method begins at start 1410 and
moves to step 1414 which monitors changes in aircraft yaw as
represented by a value .DELTA.Y. This latter value, in an
embodiment, can be based on the change in GPS track angle on the
ground in a manner that is consistent with the descriptions above.
For example, .DELTA.Y, can be determined as rapidly as packets are
available from the particular GPS that is in use. It should be
appreciated that two or more GPS packets are needed in order to
determine .DELTA.Y. Of course, any suitable number of GPS packets
and related data can be available from a rolling buffer, for
example, located in the ADP and/or any other suitable location. In
some embodiments, at least some GPS packets can be skipped on the
basis of a relatively rapid GPS packet update rate. For example,
for a five packet per second GPS, at least some of the packets can
be skipped insofar as their contributing to the determination of
.DELTA.Y, although this is not required. It should be appreciated
that the number of updates that is skipped can be customized, for
example, based on the track angle resolution of the GPS since the
effects of rounding errors decrease in proportion to increasing
track angle resolution.
[0074] Continuing with the description of step 1414, the value of
.DELTA.Y, in some embodiments, can be based on monitoring aircraft
heading H as determined by the MCPs. This determination, as
described above, can be based at least on MCP sensor data, GPS data
and magnetometer data from the ADP. It should be appreciated that
the value of .DELTA.Y, when determined in this manner, can be
updated very rapidly in what is, at least from a practical
standpoint, at least approximately on a real time basis. For
example, .DELTA.Y can be determined as rapidly as the MCP updates
are available or at least some of the updates can be skipped, for
example, on the basis of the angular resolution that is provided.
In an embodiment, the rate for periodic acceptance of MCP updates
can be determined based on the GPS packet update rate. For example,
.DELTA.Y can be updated at one second intervals when the GPS packet
update rate is one second. It should be appreciated that practical
embodiments can use any suitable interval for determining .DELTA.Y
irrespective of whether the value is based on heading H or GPS
track angle. Moreover, heading based determinations of .DELTA.Y,
like GPS based determinations can be available via a rolling buffer
and/or other suitable memory. In still another embodiment, the
value of .DELTA.Y can essentially correspond to the output of a
rate sensor which indicates a heading rate of change. At 1418, the
value of .DELTA.Y is compared to a window minimum value (WMV). The
latter will be described in further detail at an appropriate point
hereinafter. For the moment, it is sufficient to note that this WMV
for .DELTA.Y represents what would likely be considered as a
relatively steep and/or rapid turn, for example, as compared to a
standard rate turn such as a 3 degree per second turn. If .DELTA.Y
is less than the WMV, operation proceeds to 1422 which now compares
.DELTA.Y to a trigger threshold value that is indicative of a turn.
The trigger threshold value can be based, for example, on a
standard rate turn such as 3 degrees per second and is less than
the WMV that is used by step 1418. Useful embodiments can use a
value in a range from 1 degree to 3 degrees per second. A
particular embodiment can use 2 degrees per second. Of course, any
suitable value can be used based on a standard rate turn for a
given aircraft. By way of background, it is noted that, in
aviation, a standard rate turn is often defined as 2 minutes for
light aircraft and 4 minutes for jet aircraft. When .DELTA.Y is
less than the trigger value, operation returns to step 1414 for an
updated value.
[0075] On the other hand, when the value of .DELTA.Y is equal to or
exceeds the selected trigger value, a turn may be underway or the
aircraft may have experienced, for example, turbulence that induced
a momentary change in heading. In this regard, some amount of
side-to-side variation or deviation from an average/intended
heading is normal, especially under windy conditions. In this
regard, step 1422 initiates a determination that is intended to
distinguish between an actual turn and momentary heading
deviations. Accordingly, when .DELTA.Y is equal to or exceeds the
minimum trigger threshold value, operation proceeds to step 1426
which initiates additional yaw monitoring. In an embodiment, the
yaw monitoring can continue in view of the window minimum value,
WMV, that was briefly introduced above with respect to step 1418.
The WMV can specify a minimum amount of change in yaw angle over a
specified period of time. When this amount of change in yaw angle
is met or exceeded for the specified period of time, the aircraft
can be considered as being engaged in a turn. On the other hand,
when the minimum change in yaw angle is not met for the specified
period of time, the detected change in yaw orientation can be
attributed to random deviations from an intended heading and as not
reasonably being indicative of a turn. Thus, step 1426 can collect
further updates for the value of .DELTA.Y over the specified period
of time for subsequent comparative purposes. The WMV therefore can
comprise a combination of an angular value and the duration of the
time period. In this regard, any suitable combination of values can
be used; these values can be customized for a given aircraft. In an
embodiment, the values can be selected based on a standard rate
turn, for example, of 3 degrees per second. As noted above, the
time period can be selected on the basis of rounding error that is
present in the value of .DELTA.Y. In one embodiment, the minimum
angular change can be selected as 8 degrees while the time period
can be selected as 4 seconds. Thus, an average rate of turn of 2
degrees per second is implied. A series of yaw values
.DELTA.Y.sub.i can be determined over the specified time period and
can be based on either changes in heading H or GPS track angle. In
the instance of random deviation from an intended heading, some
values will be positive while others are negative such that these
random variations cancel one another. Changes that are responsive
to an actual turn, on the other hand, will be cumulative in one
angular direction. The total change in yaw angle at the end of the
time period, as compared to the value at the beginning of the
period, can be specified as .SIGMA..DELTA.Y.sub.i.
[0076] At 1430, the total yaw change is compared to the WMV. If the
total yaw change is less than the WMV, it is assumed that a turn is
not underway and operation is returned to step 1414. If, on the
other hand, the total yaw change is greater than or equal to the
WMV, it is assumed that a turn is underway and operation is routed
to step 1440. This step, for example, can initiate switching the
display of airspeed from GPS-based to CAVS centripetal force based
and/or switching the operational mode of an autopilot from a mode
that holds speed based on GPS groundspeed data to holding CAVS
airspeed. Of course, it should be understood that the airspeed
determinations, going forward from this point in the process, can
remain CAVS based until the aircraft is no longer in a turn. At
1442, the method can collect and store additional data such as, for
example, correlating GPS groundspeed with airspeed and updates of
GPS data. The process described above with regard to FIG. 11 can be
re-entered at step 1216 which relies on the additional collected
data to initially confirm that a turn remains in progress.
[0077] Attention is now returned to the determination that is made
by step 1418. If .DELTA.Y is greater than or equal to the WMV, for
example, of 9 degrees, it can be assumed that the aircraft is
engaged in a turn such that operation can proceed immediately to
step 1440 without the need to invoke steps 1422, 1426 and 1430, as
described above. It should be appreciated that in some embodiments
which enter step 1440 from either step 1418 or 1430 on the basis of
monitoring heading rate, for example, as provided by a rate sensor,
the next available GPS packet can be utilized by the CAVS
method.
[0078] In addition to the benefits of the CAVS centripetal
acceleration-based airspeed determination approach described above,
Applicant recognizes that a wide range of additional benefits can
be provided. For example, because the airspeed and groundspeed of
the aircraft are known over a range of heading angles during a turn
or partial turn, an autopilot system can determine the wind speed
and direction by using data from two or more angular positions
during a turn. Navigation and power settings can then be selected
to reduce the amount of fuel consumed during a given flight. The
described approach is applicable to aircraft that are capable of
any airspeed. For relatively fast aircraft, a compressibility (Mach
number) correction can be used. Of course, Applicants recognize
that a GPS unit itself can readily be configured to display CAVS
airspeed during a turn based on the teachings that have been
brought to light herein, for example, with the addition of a rate
or heading sensor to the GPS or taking advantage of an existing
rate or heading sensor in the aircraft's panel. Further, it should
be appreciated that any technique that relies on centripetal
acceleration as an input can benefit. In the prior art, the
centripetal acceleration is generally calculated based on the true
airspeed and the Earth's axis based turning rate. The true
airspeed, however, is not measured directly. First, a pitot-static
pressure is measured, then the outside air pressure and temperature
are measured. The latter two measurements are used to calculate the
air density which is used to correct indicated airspeed as
calculated from the pitot-static pressure measurement which may
require other corrections. Applicant submits that a significant
number of inaccuracies are introduced by all these measurements
whereas the inaccuracies associated with Applicant's technique are
more constrained. For example, the accuracy that is generally
associated with the GPS measurements can have a measurement
accuracy error that is at least as low as 0.025%.
[0079] Still another benefit of the described GPS-based centripetal
acceleration determination approach as exemplified, for example, by
FIG. 10 and associated descriptions relates to slaving of aircraft
gyros. When a gyro is used to augment a magnetic compass, the
reference for the gyro should be the Earth's gravitational vector,
so that a turn represented by the gyro represents a turn in the
Earth's axis system. When an aircraft makes a coordinated turn,
however, the centripetal acceleration produces an acceleration that
adds to the gravitational acceleration in such a way that the total
acceleration is towards the floor of the aircraft. Slaving a gyro
to this apparent acceleration can be problematic since the gyro
will tend to shift to an axis as a reference which is perpendicular
to the floor of the aircraft in a turn, if the aircraft executes a
turn of sufficiently long time duration. Applicant recognizes that
compensation can be provided via the described approach for
determining centripetal acceleration. This centripetal acceleration
determination can be used to correct the apparent vertical
acceleration in a turn and force a gyro to slave to the correct
Earth axis as a reference. This technique can be particularly
useful with respect to MEMS gyros which exhibit a high degree of
accuracy with respect to short term angular changes (for example,
in time intervals up to 10 to 30 seconds) but a significant amount
of drift with respect to long term angular changes. By way of
example, some MEMS gyros can exhibit a drift of 2.degree. per
minute. Accordingly, a slaving function can be applied to
compensate for this drift in a way that removes the effects of
centripetal acceleration during turns such that the gyro reference
corresponds over time more closely to the Earth's gravitational
axis irrespective of the aircraft performing turns.
[0080] Although the centripetal acceleration method that has been
brought to light herein has been described in terms of aircraft, it
enjoys broader applicability. While a boat is subject to movement
responsive to wind moving over the water, the boat is also subject
to the movement of the water over the ground. That is, a boat can
be subject to two moving ambients in relation to the ground. Based
on the movement of a boat or other such floating platform, the
approach of the present disclosure can be used, for example, to
compute tidal flow in saltwater or water flow in a river.
[0081] The foregoing description of the invention has been
presented for purposes of illustration and description. It is not
intended to be exhaustive or to limit the invention to the precise
form or forms disclosed, and other modifications and variations may
be possible in light of the above teachings wherein those of skill
in the art will recognize certain modifications, permutations,
additions and sub-combinations thereof.
* * * * *
References