U.S. patent application number 10/786051 was filed with the patent office on 2004-12-09 for autonomous control method for a small, unmanned helicopter.
Invention is credited to Fujiwara, Daigo, Hazawa, Kensaku, Matsusaka, Keitaro, Nonami, Kenzo, Shin, Jin Ok.
Application Number | 20040245378 10/786051 |
Document ID | / |
Family ID | 33115256 |
Filed Date | 2004-12-09 |
United States Patent
Application |
20040245378 |
Kind Code |
A1 |
Nonami, Kenzo ; et
al. |
December 9, 2004 |
Autonomous control method for a small, unmanned helicopter
Abstract
An objective of the present invention is to provide an
autonomous control method that autonomously controls a small
unmanned helicopter toward target values, such as a set position
and velocity, by deriving model formulas that are well suited for
the autonomous control of small unmanned helicopters, by designing
an autonomous control algorithm based on the model formulas, and by
calculating the autonomous control algorithm. The autonomous
control system for a small unmanned helicopter of the present
invention comprises: sensors that detect the current position, the
attitude angle, the altitude relative to the ground, and the
absolute azimuth of the nose of the aforementioned small unmanned
helicopter; a primary computational unit that calculates optimal
control reference values for driving the servo motors that move
five rudders on the helicopter from target position or velocity
values that are set by the ground station and the aforementioned
current position and attitude angle of the small unmanned
helicopter that are detected by the aforementioned sensors; an
autonomous control system equipped with a secondary computational
unit that converts the data collected by said sensors and the
computational results as numeric values that are output by said
primary computational unit into pulse signals that can be accepted
by the servo motors, such that these components are assembled into
a small frame box, thereby achieving both size and weight
reductions; a ground station host computer that can also be used as
the aforementioned computational unit for the aforementioned
autonomous control system; if the aforementioned ground station
host computer is used as the aforementioned computational unit for
the aforementioned autonomous control system, in the process of
directing the computational results that are output from said
ground station host computer to said servo motors through a manual
operation transmitter, a radio control generator that converts said
computational results as numerical values into pulse signals that
said manual operation transmitter can accept; a servo pulse
mixing/switching apparatus, on all said servo motors for said small
unmanned helicopter, that permits the switching of manual operation
signals and said control signals that are output from said
autonomous control system or mixing thereof in any ratio; an
autonomous control algorithm wherein the mathematical model for
transfer function representation encompassing pitching operation
input through pitch axis attitude angles in the tri-axis
orientation control for said small unmanned helicopter is defined
as 1 G ( s ) = - Ls K n s 2 ( s 2 + 2 ??? s s s + n s 2 ) ( T s + 1
) s such that the aforementioned small unmanned helicopter is
controlled autonomously based on the aforementioned mathematical
model;
Inventors: |
Nonami, Kenzo; (Tokyo,
JP) ; Shin, Jin Ok; (Seoul, KR) ; Fujiwara,
Daigo; (Kasukabe-shi, JP) ; Hazawa, Kensaku;
(Hitachi-shi, JP) ; Matsusaka, Keitaro;
(Fuchu-shi, JP) |
Correspondence
Address: |
HAUPTMAN KANESAKA BERNER PATENT AGENTS, LLP
Suite 310
1700 Diagonal Road
Alexandria
VA
22314
US
|
Family ID: |
33115256 |
Appl. No.: |
10/786051 |
Filed: |
February 26, 2004 |
Current U.S.
Class: |
244/17.13 |
Current CPC
Class: |
A63H 30/04 20130101;
A63H 27/12 20130101; G05D 1/0858 20130101; G05D 1/0033
20130101 |
Class at
Publication: |
244/017.13 |
International
Class: |
B64C 015/02 |
Foreign Application Data
Date |
Code |
Application Number |
Feb 26, 2003 |
JP |
2003-049570 |
Claims
1. An autonomous control program for a small unmanned helicopter
wherein the program detects the current position, the attitude
angle, the altitude relative to the ground, and the absolute
azimuth of the nose of the small unmanned helicopter; establishes
position or velocity reference values from the ground station;
determines optimal control reference values for driving the servo
motors that move a number of helicopter rudders from the current
position and attitude angle of said small unmanned helicopter
detected by said sensors; based upon said computational processing
results, effects translational motion control and tri-axial
orientation control on the small unmanned helicopter; or defines
the mathematical model for transfer function representation
including pitching operation input and pitch axis attitude angles
in the tri-axial orientation control for said small unmanned
helicopter as 16 G ( s ) = - Ls K n s 2 ( s 2 + 2 ??? s s s + n s 2
) ( T s + 1 ) s ( 13 ) and based upon said model equations, causes
said primary computational unit to calculate optimal control
reference values for driving the servo motors.
2. An autonomous control program for a small unmanned helicopter
wherein the program detects the current position, the attitude
angle, the altitude relative to the ground, and the absolute
azimuth of the nose of the small unmanned helicopter; establishes
position or velocity reference values from the ground station;
determines optimal control reference values for driving the servo
motors that move a number of helicopter rudders from the current
position and attitude angle of said small unmanned helicopter
detected by said sensors; based upon said computational processing
results, effects translational motion control and tri-axial
orientation control on the small unmanned helicopter; or defines
the mathematical model for a transfer function representation
including the rolling input and roll axis attitude angles in the
tri-axial orientation control for said small unmanned helicopter as
17 G ( s ) = - Ls K n s 2 ( s 2 + 2 ??? s n s s + n s 2 ) ( T s + 1
) s ( 14 ) and based upon said model equations, causes said primary
computational unit to calculate optimal control reference values
for driving the servo motors.
3. An autonomous control method for a small unmanned helicopter
wherein the method detects the current position, the attitude
angle, the altitude relative to the ground, and the absolute
azimuth of the nose of the small unmanned helicopter; establishes
position or velocity reference values from the ground station;
determines optimal control reference values for driving the servo
motors that move a number of helicopter rudders from the current
position and attitude angle of said small unmanned helicopter
detected by said sensors; based upon said computational processing
results, effects translational motion control and tri-axial
orientation control on the small unmanned helicopter; or defines
the mathematical model for transfer function representation
including yawing input and yawing axis attitude angles in the
tri-axial orientation control for said small unmanned helicopter as
18 G ( s ) = - Ls K n s 2 ( s 2 + 2 ??? s n s s + n s 2 ) s ( 15 )
and based upon said model equations, causes said primary
computational unit to calculate optimal control reference values
for driving the servo motors.
4. An autonomous control method for a small unmanned helicopter
wherein the method detects the current position, the attitude
angle, the altitude relative to the ground, and the absolute
azimuth of the nose of the small unmanned helicopter; establishes
position or velocity reference values from the ground station;
determines optimal control reference values for driving the servo
motors that move a number of helicopter rudders from the current
position and attitude angle of said small unmanned helicopter
detected by said sensors; based upon said computational processing
results, effects translational motion control and tri-axial
orientation control on the small unmanned helicopter; or defines
the mathematical model for transfer function representation
including pitching axis attitude angles and the longitudinal speed
in the tri-axial orientation control for said small unmanned
helicopter as 19 Vx = g T s + T a s - a ( - ) ( 16 ) and based upon
said model equations, causes said primary computational unit to
calculate optimal control reference values for driving the servo
motors.
5. An autonomous control method for a small unmanned helicopter
wherein the method detects the current position, the attitude
angle, the altitude relative to the ground, and the absolute
azimuth of the nose of a small unmanned helicopter; establishes
position or velocity reference values from the ground station;
determines optimal control reference values for driving the servo
motors that move a number of helicopter rudders from the current
position and attitude angle of said small unmanned helicopter
detected by said sensors; based upon said computational processing
results, effects translational motion control and tri-axial
orientation control on the small unmanned helicopter; or defines
the mathematical model for transfer function representation
including rolling axis attitude angles and the lateral speed in the
translational motion control for said small unmanned helicopter as
20 Vy = g T s + T a s - a ( 17 ) and based upon said model
equations, causes said primary computational unit to calculate
optimal control reference values for driving the servo motors.
6. An autonomous control method for a small unmanned helicopter
wherein the method detects the current position, the attitude
angle, the altitude relative to the ground, and the absolute
azimuth of the nose of the small unmanned helicopter; establishes
position or velocity reference values from the ground station;
determines optimal control reference values for driving the servo
motors that move a number of helicopter rudders from the current
position and attitude angle of said small unmanned helicopter
detected by said sensors; based upon said computational processing
results, effects translational motion control and tri-axial
orientation control on the small unmanned helicopter; or defines
the mathematical model for transfer function representation of the
vertical speed in the translational motion control for said small
unmanned helicopter as 21 Vz = k s i ( 18 ) and based upon said
model equations, causes said primary computational unit to
calculate optimal control reference values for driving the servo
motors.
7. The autonomous control method for a small unmanned helicopter of
claim 1, wherein, in determining optimal control reference values
for driving the servo motors that drive the rudders for the small
unmanned helicopter, said primary computational unit for said
autonomous control system autonomously controls said small unmanned
helicopter by executing independent autonomous control algorithms
on the six physical quantities of said small unmanned helicopter:
pitch axis attitude angle, roll axis attitude angle, yaw axis
attitude angle, longitudinal speed, lateral speed, and vertical
speed.
8. The autonomous control method for a small unmanned helicopter of
claim 1, wherein said small unmanned helicopter is autonomously
controlled by constructing the respective autonomous control
algorithms as a type 1 servo system so that for the respective
physical quantities of said small unmanned helicopter, the
steady-state deviation from any reference value will be zero.
9. The autonomous control method for a small unmanned helicopter of
claim 7, wherein said small unmanned helicopter is autonomously
controlled by applying either the linear quadratic Gaussian (LQG)
theory or the linear quadratic integral (LQI) theory to the
autonomous control algorithms that are constituted as a type 1
servo system, by treating the respective autonomous control
algorithms as uncoupled transfer function representation
mathematical models.
10. The autonomous control method for a small unmanned helicopter
of claim 1, wherein, in determining optimal control reference
values for driving the servo motors that drive the rudders for the
small unmanned helicopter, dynamic characteristics consisting of
longitudinal speeds and lateral speeds are represented as
mathematical models for which pitch axis attitude angles and roll
axis attitude angles are input quantities, and said small unmanned
helicopter is controlled autonomously by calculating the respective
attitude angles that are necessary for effecting arbitrary
longitudinal and lateral speeds.
11. The autonomous control method for a small unmanned helicopter
of claim 1, wherein, in determining optimal control reference
values for driving the servo motors that drive the rudders for the
small unmanned helicopter, in order to move said small unmanned
helicopter to an arbitrary position, longitudinal, lateral, and
vertical speed reference values are defined as follows:
Vxref=.alpha.(Pxref-Px) (19) for a longitudinal reference value,
Vyref=.alpha.(Pyref-Py) (20) for a lateral value, and
Vzref=.beta.(Pzref-Pz) (21) for a vertical reference value, thereby
effecting the autonomous control of said small unmanned
helicopter.
12. The small unmanned helicopter autonomous control method of
claim 1, wherein, in determining optimal control reference values
for driving the servo motors that drive the rudders for the small
unmanned helicopter, and further, the transfer function
representation mathematical model including the rolling force and
rolling axis attitude angles in the tri-axial orientation control
for said small unmanned helicopter is defined as 22 G ( s ) = - L s
K n s 2 ( s 2 + 2 s n s s + n s 2 ) ( T s + 1 ) s ( 22 ) and said
small unmanned helicopter is autonomously controlled based on said
mathematical model.
13. The small unmanned helicopter autonomous control algorithm of
claim 1, wherein, in determining optimal control reference values
for driving the servo motors that drive the rudders for the small
unmanned helicopter, and further, the transfer function
representation mathematical model including the yawing force and
yawing axis attitude angles in the tri-axial orientation control
for said small unmanned helicopter is defined as 23 G ( s ) = - L s
K n s 2 ( s 2 + 2 s n s s + n s 2 ) s ( 23 ) and said small
unmanned helicopter is autonomously controlled based on said
mathematical model.
14. The small unmanned helicopter autonomous control method of
claim 1, wherein, in determining optimal control reference values
for driving the servo motors that drive the rudders for the small
unmanned helicopter, and further, the transfer function
representation mathematical model including the pitching axis
attitude angle and the longitudinal speed in the tri-axial
orientation control for said small unmanned helicopter is defined
as 24 Vx = g T s + T a s - a ( - ) ( 24 ) and said small unmanned
helicopter is autonomously controlled based on said mathematical
model.
15. The small unmanned helicopter autonomous control method of
claim 1, wherein, in determining optimal control reference values
for driving the servo motors that drive the rudders for the small
unmanned helicopter, and further, the transfer function
representation mathematical model including the rolling axis
attitude angle and the lateral speed in the translational motion
control for said small unmanned helicopter is defined as 25 Vy = g
T s + T a s - a ( 25 ) and said small unmanned helicopter is
autonomously controlled based on said mathematical model.
16. The small unmanned helicopter autonomous control method of
claim 1, wherein, in determining optimal control reference values
for driving the servo motors that drive the rudders for the small
unmanned helicopter, and further, the transfer function
representation mathematical model for the vertical speed in the
translational motion control for said small unmanned helicopter is
defined as 26 Vz = k s t ( 26 ) and said small unmanned helicopter
is autonomously controlled based on said mathematical model.
17. The autonomous control method for a small unmanned helicopter
of claim 1, wherein, in determining optimal control reference
values for driving the servo motors that drive the rudders for the
small unmanned helicopter, the transfer function representation
mathematical model that describes the dynamic characteristics of
said servo motors included in said Eqs. 13 and 14 is defined as 27
G s ( s ) = n s 2 ( s 2 + 2 s n s s + n s 2 ) ( 27 ) wherein an
autonomous control algorithm is designed by entering M-series
signals (pseudo-white signals) into said servo motors, by applying
a partial space identification method based on input/output
relationships, by determining the unknown parameters .omega..sub.ns
and .zeta..sub.s in Eq. 15, and by using the resulting values,
thereby autonomously controlling said small unmanned
helicopter.
18. The autonomous control method for a small unmanned helicopter
of claim 1 wherein, in determining optimal control reference values
for driving the servo motors that drive the rudders for the small
unmanned helicopter, the parameters that are included in said Eqs.
13 through 18, specifically, parameters K.sub..theta., and
T.sub..theta., contained in Eq. 13; parameters K.sub..phi., and
T.sub..phi., contained in Eq. 14; parameters K.sub..psi. and
T.sub..psi. contained in Eq. 15; parameters T and a contained in
Eq. 17; and parameter k contained in Eq. 18 are adjusted so that
their experimental results and simulation results will agree,
autonomous control algorithms are designed according to the
resulting values, and said small unmanned helicopter is thereby
autonomously controlled.
19. A small unmanned helicopter autonomous control method wherein,
in determining optimal control reference values for driving the
servo motors that drive the rudders for the small unmanned
helicopter, sine wave signals representing arbitrary frequency,
amplitude, and time values are entered into said servo motors by
manual operation, and during this process, said sine wave signals
are obtained by using a function that outputs manual operation
signals that are input into said servo pulse mixing/switching unit
of a servo pulse mixing/switching unit that is an external device
to said autonomous control system; wherein, simultaneously, the
flight status of said small unmanned helicopter is obtained by
means of said sensors for said autonomous control system, and
wherein the parameters contained in Eqs. 13 through 18 in claim 1
are determined by analyzing the interrelationship between the
various pieces of data thus obtained, and wherein autonomous
control algorithms are created based on the mathematical models
thus obtained.
Description
BACKGROUND OF THE INVENTION
[0001] 1. Field of the Invention
[0002] This invention relates to an autonomous control system for
small unmanned helicopters, and autonomous control algorithms that
control the rudders for said small unmanned helicopter based on the
aforementioned mathematical models.
[0003] 2. Description of the Prior Art
[0004] Helicopters are flying bodies that have operating ranges
such as longitudinal motions, lateral motions, vertical motions,
and hovering, which are not exhibited by an aircraft; as such, they
have the advantage of being able to flexibly respond to various
situations.
[0005] Based on this advantage, there are hopes for constructing
unmanned, small helicopters for use in places that are difficult or
dangerous for manned operations, for example, in high-altitude
work, such as the inspection of power transmission lines, or in
emergency rescue operations or the detection of land mines.
Previously, an autonomous control system using an unmanned
helicopter for agricultural chemical spray applications was
described in Patent Reference 1 listed below:
[0006] [Patent Reference 1]
[0007] Patent Disclosure 2000-118498
[0008] First, let us provide a logical explanation of the mechanism
of autonomous control for helicopters. Helicopters are the objects
of control; they are flying bodies that are capable of changing
their orientation by means of servo motor actions and are capable
of three-dimensional motions. The purpose of flying by autonomous
control is to move the helicopter according to positional and speed
target values. The required maneuvering follows the computational
results generated by calculation computer. In order to delegate the
piloting of a helicopter to a computer, the calculation computer
must have sensing and actuation functions. Devices that have the
function of sensing the various flight conditions of a helicopter
are called sensors. Actuators that move the helicopter's rudders by
receiving autonomous control signals that are generated by
determining control reference values based on computational results
from the computer and by converting these results into signals are
referred to as servo motors. The helicopter can be autonomously
controlled toward a given target value by means of a feedback
control loop that links "(sensor)--(calculation computer)--(servo
motor)--(helicopter), Reference FIG. 1".
[0009] Following is a description of autonomous control for a small
unmanned helicopter described in Patent Reference 1, with reference
FIG. 2 to drawings. This system can be divided into a mobile
station, which includes the helicopter and a ground station.
Mounted on the ground station are a helicopter body 101; a sensor
102 that detects the current position and attitude angle of the
helicopter body 101; servo motors 103 that move the rudders for the
helicopter body 101; a backup receiver 104 that receives manual
maneuver signals from a backup transmitter 110; and a wireless
modem that communicates with the ground station.
[0010] Employed in the sensor 1021 are the GPS (not shown in the
figure) that detects the current position of the helicopter body
101 and tri-axis orientation sensors (not shown in the figure) that
detect the tri-axial attitude angles of the helicopter body 101.
Installed on the ground station are a computer CPU 108 for the
input of reference values on which speed reference values are
entered; a computer CPU 109 for internal computation purposes; and
a backup transmitter 110 that permits the operator to perform
manual operations in the event of the occurrence of a dangerous
situation. The CPU 109 calculates position and attitude angle
reference values from target speed values, compares the results
with the current position, speed, and attitude angle obtained from
the sensors 102, and based on these results, calculates control
instruction values that bring the helicopter to the reference
values. By forming a feedback control loop by linking "(sensors
102)--(CPU 109)--(servo motors 103)--(helicopter body 101)"
(intervening components omitted), it is possible to effect the
autonomous control of the helicopter toward its reference
values.
[0011] Following is a description of the operation of the system.
The operator sets four speed reference values (Vx*, Vy*, Vz*,
.omega.*) consisting of longitudinal, lateral, vertical, and
rotational speeds, on the CPU 58. The CPU 59 integrates these speed
reference values with respect to time, obtaining a longitudinal
target position X*, a lateral target position Y*, a vertical target
position Z*, and a rotational target position (yawing angle)
.psi.*. Similarly, the CPU 59 differentiates the four speed
reference values (Vx*, Vy*, Vz*, .omega.*) and multiplies the
results by coefficients to calculate a target pitching angle
.theta.* and a target rolling angle .phi.*. The differences between
the target values that are set in this manner and the detected
values (.theta., .phi., .psi., .omega.) for the body attitude,
speed (X, Y, Z, Vx, Vy, Vz) that are detected by the sensors 52
consisting of the GPS and tri-axis attitude sensors that are
installed on the helicopter body 51 are calculated as follows:
.DELTA.X=X*-X
.DELTA.Y=Y*-Y
.DELTA.Z=Z*-Z
.DELTA.Vx=Vx*-Vx
.DELTA.Vy=Vy*-Vy
.DELTA.Vz=Vz*-Vz
.DELTA..theta.=.theta.*-.theta.
.DELTA..phi.=.phi.*-.phi.
.DELTA..psi.=.psi.*-.psi.
.DELTA..omega.=.omega.*-.omega.
[0012] Based upon these differences (errors), the CPU 109
calculates control reference values for the servo motors 103 that
move the rudders for the helicopter body 101. Four types of control
reference values are computed: elevator servo (longitudinal)
instruction, aileron servo (lateral) instruction, corrective servo
(vertical) instruction, and rudder servo (rotational) instruction.
After computing these four types of control reference values, the
CPU 109 supplies them to the aforementioned servo motors 103, and
performs feedback control on these operations until the differences
become zero (0).
[0013] Helicopters that are used in the aforementioned conventional
autonomous control system were originally intended for the spraying
of agricultural chemicals, with a maximum weight of approximately
30 kg. Although the aforementioned sensors and computational unit
for the aforementioned conventional autonomous control system are
large and heavy, the helicopter can adequately fly even when
carrying these items. However, the unloaded helicopter used in the
aforementioned conventional autonomous control system weights
approximately 60 kg empty, and approximately 90 kg when fully
loaded. Therefore, such a helicopter cannot easily be carried. In
addition, in order to use the system, the helicopter must have a
flight range sufficiently larger than the actual helicopter, which
limits the range over which the helicopter can be deployed. In some
cases, manned helicopter operations involve a narrow space in which
the aforementioned conventional helicopter cannot negotiate. On the
other hand, the small unmanned helicopter addressed in the present
invention refers to a helicopter comparable in size and weight to,
and compatible with, a commercially available hobby-type
small-scale radio-controlled helicopter. Although the above problem
can be solved by effecting autonomous control in such a helicopter,
the smaller the weight of a helicopter, the more difficult it is to
control. In other words, the autonomous control system is subject
to stringent constraints in terms of size and weight, and small
helicopters tend to be unstable in terms of dynamic properties.
Therefore, with the aforementioned conventional autonomous control
system, it is impossible to mount the autonomous control system on
the aforementioned small unmanned helicopter as is. Further,
applying the autonomous control algorithms for the aforementioned
conventional autonomous control system to the aforementioned small
unmanned helicopter as is does not guarantee adequate control
performance. Further, the calculation of a control instruction
value is a time-consuming process due to a large number of
computational steps involved in the determination of servo motor
control instruction values; consequently, when one attempts to
achieve size reductions in the autonomous control system, one must
contend with the conflicting requirements of accommodating a large
number of computational steps and the stringent constraints imposed
on the capabilities of the computational equipment and the size of
the control program. Further, beyond the computational equipment,
the sensors are also subject to stringent constraints on size and
weight, which clearly adds to the difficulty of construction of
small and lightweight autonomous control systems. An autonomous
control system that can be mounted and flown on the type of small
unmanned helicopter for which the present invention is intended has
not been successfully developed. Manual operation being independent
from autonomous control, can be thought of as being completely
unrelated to autonomous control. However, the process of designing
an autonomous control algorithm may require the measurement of
manual operation signals. In creating mathematical models in the
present invention, we used a technique called system
characterization, wherein a mathematical model is obtained by
associating input signals into the aforementioned servo motors for
the aforementioned small unmanned helicopter with output signals,
which represent the flying conditions of the aforementioned small
unmanned helicopter as measured by sensors installed on the
aforementioned autonomous control system, and by analyzing the
signals. The system characterization requires the collection of
input/output data under the condition in which the aforementioned
small unmanned helicopter is flying, and this process is referred
to as a characterization experiment. Conducting a characterization
experiment requires that the aforementioned servo motors be driven
using characterization input signals that are appropriate for
system characterization. In such a case, however, characterization
input by itself can cause a substantial tilt in the attitude of the
aforementioned small unmanned helicopter and sudden acceleration,
with a potential risk of accidents. To prevent such a problem, it
is necessary to stabilize the motion of the aforementioned small
unmanned helicopter by controlling the correction rudder by means
of manual operation. However, because the correction rudder is
considered to be part of characterization input, for system
characterization, the aforementioned manual operation signals must
also be obtained as measurement data.
[0014] As stated above, the present invention also takes into
consideration the use of the aforementioned autonomous control unit
as an auxiliary unit for manual operation, i.e., as an operator
assist unit. In terms of the objective and the technique of
operation assistance, manual operation signals, for example, could
be associated with the target value input signals for the
aforementioned autonomous control algorithm, and the drive signals
that are actually output to the aforementioned servo motors could
all be treated as autonomous control signals that are computational
results of the aforementioned autonomous control algorithm. In
other words, the method may be described as follows: even though
the human operator may have the illusion of operating the
helicopter himself, in actuality, he merely provides motion
commands, which are target values, to the aforementioned autonomous
control small unmanned helicopter; upon receipt of the target
values, the aforementioned autonomous control algorithm is
calculated, and autonomous control is effected. The method can
provide the aforementioned autonomous control algorithm with target
values without using the aforementioned ground station computer,
with the advantage that people not versed in computer operation can
safely fly the aforementioned small unmanned helicopter, enjoying
the benefits of the aforementioned autonomous control algorithm.
This approach, however, requires a new technique for associating
the aforementioned manual operation signals with target values.
[0015] The development of autonomous control algorithms for the
autonomous control of the aforementioned small unmanned helicopter
requires mathematical models that describe the dynamic
characteristics of the helicopter. The use of mathematical models
permits the application of various control theories, which have
made strides in recent years and whose effectiveness has amply been
recognized, to the development of autonomous control algorithms,
which should improve the flight performance in situations in which
the aforementioned small unmanned helicopter is controlled
autonomously.
[0016] However, the dynamic characteristics of a helicopter are
subject to a complex interplay of dynamical action and fluid
dynamic action, which makes analysis an extremely difficult task.
Although a detailed analysis of the dynamic characteristics of
manned helicopters has been pursued aggressively, little detailed
analysis has been performed with regard to helicopters of the size
addressed in the present invention. In addition, there have been no
reports on the dynamic characteristics of the aforementioned servo
motors.
[0017] Even if a mathematical model that describes the dynamic
characteristics of the aforementioned small unmanned helicopter in
detail exists, if the mathematical model is a highly complex one,
the development of an autonomous control algorithm will also be
difficult. Autonomous control algorithms that are developed and
based on complex mathematical models are generally complex, and may
not necessarily be appropriate for execution by a computer that is
subject to stringent restrictions on its computational capabilities
due to weight limitations.
[0018] There is a system identification method that avoids
theoretical analyses and draws inferences on the dynamic
characteristics of a given physical system based on its
input/output relationships. System identification requires the
input of signals containing frequency components encompassing a
broad bandwidth into the physical system. However, entering such
signals into the aforementioned small unmanned helicopter involves
a risk. In addition, such a system will also require devices for
the measurement of the aforementioned input/output signals and an
instrumentation system. There have been no cases where system
identification is run on the type of small unmanned helicopter
addressed in the present invention and where the soundness of the
identification model thus obtained is validated.
SUMMARY OF THE INVENTION
[0019] An objective of the present invention is to provide an
autonomous control method that autonomously controls a small
unmanned helicopter toward target values, such as a set position
and velocity, by deriving model formulas that are well suited for
the autonomous control of small unmanned helicopters, by designing
an autonomous control algorithm based on the model formulas, and by
calculating the autonomous control algorithm.
[0020] The autonomous control system for a small unmanned
helicopter of the present invention comprises: sensors that detect
the current position, the attitude angle, the altitude relative to
the ground, and the absolute azimuth of the nose of the
aforementioned small unmanned helicopter; a primary computational
unit that calculates optimal control reference values for driving
the servo motors that move five rudders on the helicopter from
target position or velocity values that are set by the ground
station and the aforementioned current position and attitude angle
of the small unmanned helicopter that are detected by the
aforementioned sensors; an autonomous control system equipped with
a secondary computational unit that converts the data collected by
said sensors and the computational results as numeric values that
are output by said primary computational unit into pulse signals
that can be accepted by the servo motors, such that these
components are assembled into a small frame box, thereby achieving
both size and weight reductions;
[0021] a ground station host computer that can also be used as the
aforementioned computational unit for the aforementioned autonomous
control system;
[0022] if the aforementioned ground station host computer is used
as the aforementioned computational unit for the aforementioned
autonomous control system,
[0023] in the process of directing the computational results that
are output from said ground station host computer to said servo
motors through a manual operation transmitter, a radio control
generator that converts said computational results as numerical
values into pulse signals that said manual operation transmitter
can accept;
[0024] a servo pulse mixing/switching apparatus, on all said servo
motors for said small unmanned helicopter, that permits the
switching of manual operation signals and said control signals that
are output from said autonomous control system or mixing thereof in
any ratio;
[0025] an autonomous control algorithm wherein the mathematical
model for transfer function representation encompassing pitching
operation input through pitch axis attitude angles in the tri-axis
orientation control for said small unmanned helicopter is defined
as 2 G ( s ) = - Ls K n s 2 ( s 2 + 2 ??? s s s + n s 2 ) ( T s + 1
) s
[0026] such that the aforementioned small unmanned helicopter is
controlled autonomously based on the aforementioned mathematical
model;
[0027] an autonomous control algorithm wherein the mathematical
model for transfer function representation encompassing rolling
input through roll axis attitude angles for the aforementioned
tri-axis orientation control is defined as 3 G ( s ) = - Ls K n s 2
( s 2 + 2 ??? n s s + n s 2 ) ( T s + 1 ) s
[0028] such that the aforementioned small unmanned helicopter is
controlled autonomously based on the aforementioned mathematical
model;
[0029] an autonomous control algorithm wherein the mathematical
model for transfer function representation encompassing yawing
input through yaw axis attitude angles for the aforementioned
tri-axis orientation control is defined as 4 G ( s ) = - Ls K n s 2
( s 2 + 2 ??? s n s s + n s 2 ) s
[0030] such that the aforementioned small unmanned helicopter is
controlled autonomously based on the aforementioned mathematical
model;
[0031] an autonomous control algorithm wherein the mathematical
model for transfer function representation encompassing pitch axis
attitude angles through longitudinal speeds in the translational
motion control for the aforementioned small unmanned helicopter is
defined as 5 Vx = g T s + T a s - a ( - )
[0032] such that the aforementioned small unmanned helicopter is
controlled autonomously based on the aforementioned mathematical
model;
[0033] an autonomous control algorithm wherein the mathematical
model for transfer function representation encompassing roll axis
attitude angles through lateral speeds in the translational motion
control for the aforementioned small unmanned helicopter is defined
as 6 Vy = g T s + T a s - a
[0034] such that the aforementioned small unmanned helicopter is
controlled autonomously based on the aforementioned mathematical
model;
[0035] an autonomous control algorithm wherein the mathematical
model for the transfer function representation of vertical speeds
in the translational motion control for the aforementioned small
unmanned helicopter is defined as 7 Vz = k s t
[0036] such that the aforementioned small unmanned helicopter is
controlled autonomously based on the aforementioned mathematical
model;
[0037] The autonomous control method of the present invention
permits the autonomous control of small unmanned helicopters
according to position and velocity target values that are set by
the human operator. In addition to implementing an autonomous
control unit by mounting the autonomous control device used in this
method on the body of a helicopter, we have developed a servo pulse
mixing/switching unit that either mixes or switches autonomous
control signals and manual operation signals, a pulse generator
unit that converts these signals into pulse signals that the manual
operation signal transmitter can accept, and an autonomous control
algorithm that is well suited for the autonomous control of small
unmanned helicopters; as a result, small unmanned helicopters, in
the same class as hobby-oriented small radio-controlled
helicopters, can be autonomously controlled according to target
values. This permits complete autonomous control with
hobby-oriented small radio-controlled helicopters; consequently,
such helicopters become easily portable and can be adapted to small
spaces in which it is impossible to perform operations using the
conventional unmanned helicopter, and in which it is difficult to
perform manned operations, and in this manner, the application of
such helicopters can be broadened.
BRIEF DESCRIPTION OF THE DRAWING
[0038] The features and advantages of the present invention will be
apparent from the following description of an exemplary embodiment
of the invention and the accompanying drawing, in which:
[0039] FIG. 1 is a feedback control loop system
[0040] FIG. 2 is a configuration of autonomous control system using
conventional unmanned helicopter system
[0041] FIG. 3 is a configuration of autonomous control system using
autonomous control device of small, unmanned helicopter of the
invention
[0042] FIG. 4 is a coordinate system of small, unmanned helicopter
of the invention
[0043] FIG. 5 is a simulated and experimental results of pitch axis
of the invention
[0044] FIG. 6 is a simulated and experimental results of roll axis
of the invention
[0045] FIG. 7 is a simulated and experimental results of yaw axis
of the invention
[0046] FIG. 8 is an experimental results of pitch angle, roll angle
and yaw angle with control of the invention
[0047] FIG. 9 is a relation between attitude angle and body
acceleration
[0048] FIG. 10 is a simulated and experimental results
[0049] FIG. 11 is a configuration of position control loop in the X
and the Y directions
[0050] FIG. 12 is an experimental result of trajectory following
control by control algorithm
[0051] FIG. 13 is an experimental result of fixed hovering
control
[0052] FIG. 14 is an experimental results of lateral and
longitudinal velocity control
[0053] FIG. 15 is an experimental results of altitude control
[0054] FIG. 16 is an internal structure of servo pulse
mixing/switching device
[0055] FIG. 17 is a configuration of control system using manual
piloting signal transmitter as reference input device
[0056] FIG. 18 is a Configuration of convert system to the control
signal from control referenece signal using pulse generator
device
DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0057] The inventor of the present invention has consistently
pursued autonomous control research from the design of the
requisite hardware through the development of autonomous control
algorithms. The inventor developed the aforementioned autonomous
control system, the aforementioned radio-controlled pulse
generator, the hardware for the aforementioned servo pulse
mixing/switching unit, and the software on a novel and independent
basis. In addition, the inventor conducted flight experiments for
the characterization of the aforementioned mathematical models,
derived mathematical models from the experimental results, and
developed autonomous control algorithms for the small unmanned
helicopter using the mathematical models. By using the autonomous
control system, the inventor successfully achieved the complete
autonomous control of a small helicopter body, with an unloaded
weight of approximately 9 kg, according to target values.
[0058] In what follows, we provide a logical explanation of the
hardware for the autonomous control of the small unmanned
helicopter, and then describe research accomplishments on the
tri-axial attitude control of the helicopter body and autonomous
control algorithms for cruise speed/position control.
[0059] The helicopter body used in this invention is a helicopter
with a size and weight comparable to those of small
radio-controlled helicopters that are commercially available for
hobby purposes. Its empty weight is approximately 9 kg. Using this
small unmanned helicopter, the present invention can be applied to
operations that are performed at a high altitude, such as the
inspection of power transmission lines, to emergency rescue
operations, to the detection of land mines, or to operations
performed in narrow spaces that are too difficult or dangerous for
human operators to perform. The small unmanned helicopter, which is
the object of control, is a flying object whose orientation can be
changed by the operation of five servo motors and which can be
moved three-dimensionally.
[0060] The purpose of flight based on autonomous control, as in the
aforementioned prior technologies, is to move the helicopter
according to position and speed target values, and the maneuvering
of the helicopter obeys the autonomous control algorithms based on
mathematical models that have been developed by the inventor. The
aforementioned autonomous control algorithms are installed on the
calculation computer for the autonomous control system developed by
the inventor, wherein the results of computations generated by the
aforementioned calculation computer provide control signals for the
respective servo motors.
[0061] In order to delegate the piloting of a helicopter to a
computer, the calculation computer must be endowed with sensing and
actuation functions. Devices that have the function of sensing the
various flight conditions of a helicopter are called sensors.
Actuators that move the helicopter's rudders by receiving
autonomous control signals that are generated by determining
control reference values based on computational results from the
computer and by converting them into signals are referred to as
servo motors. By building an autonomous control system comprised of
these components and mounting it on the helicopter body, it is
possible to form a feedback control loop that links
"(sensor)--(calculation computer)--(servo motor)--(helicopter)", as
shown in FIG. 1.
[0062] The inventor of the present invention has developed an
autonomous control system, which is a piece of hardware, as a
platform that permits the autonomous control of the aforementioned
small unmanned helicopter. Because the aforementioned servo motors
among the pieces of hardware that make up the autonomous control
system are incorporated into the body of the aforementioned small
unmanned helicopter, the aforementioned autonomous control system
contains built-in sensors and computational units. Sensors that are
incorporated into a conventional autonomous control system include
tri-axis attitude sensors that detect the orientation of the
helicopter and the Global Positioning System (GPS) that detects the
current position of the helicopter. In addition, the autonomous
control system of the present invention incorporates a ground
altitude sensor that detects the distance between the body of the
aforementioned small unmanned helicopter and the ground surface,
and this permits the use of an autonomous control algorithm that
utilizes the altitude with respect to the ground as sensor
information. The autonomous control system also incorporates the
primary computational unit, which is a computation CPU used for the
implementation and calculation of autonomous control algorithms, as
well as a secondary computational unit, which is a CPU that
collects and processes data from the aforementioned sensors, and
converts the aforementioned computational results into control
signals. Efforts to reduce the size and weight of the autonomous
control system create a problem of stringent restrictions on the
functionality and processing speeds of the computational unit. We
solved this problem by diving the computational unit into two
ptechnologies, the primary computational unit and a secondary
computational unit, wherein the secondary computational unit is
assigned the role of all computations with the exception of
autonomous control algorithms, and the primary computational unit
is allowed to concentrate on the computation of autonomous control
algorithms. In addition, the autonomous control system incorporates
a power supply unit that outputs power with a voltage required by
the aforementioned sensors and the aforementioned computational
units, as well as a wireless modem that communicates with the
ground station. The aforementioned power supply unit takes a DC
power supply of a positive voltage as input and can supply power at
three voltages, including a negative voltage. The aforementioned
wireless modem comprises two sets, wherein one set is used for the
exchange of control information with the ground station host
computer, and the other set for the exchange of supplementary
information that improves the accuracy of GPS current position
information. The autonomous control system of the present
invention, which is designed under the assumption that it would be
installed on the aforementioned small unmanned helicopter, must
necessarily be small and lightweight. The aforementioned small
unmanned helicopter of the present invention has an approximate
maximum payload of only 5 kg. Therefore, in constructing the
autonomous control system, the inventor of the present invention
excluded any extraneous components that would lead to an increase
in weight or size. Based upon this consideration, the inventor
assembled the aforementioned sensors, the aforementioned
computational units, the aforementioned power supply unit, and the
aforementioned wireless modem into a small frame box of W 190
mm.times.D 290 mm.times.H 110 mm, and thus succeeded in the
development of an autonomous control system with an approximate
weight of 2.9 kg. For batteries, we adopted a commercially
available high-capacity lithium ion battery, approximately 700 g in
weight, for notebook personal computers, which assured a continuous
operating time of approximately 1 hour when fully charged.
Including stays and other accessories for installing the unit of
the aforementioned small unmanned helicopter, a weight reduction
extensive enough to fall within the aforementioned maximum payload
was accomplished for this autonomous control system.
[0063] Although the aforementioned autonomous control system of the
present invention has an adequate autonomous control algorithm
computational capacity, to provide for the future possibility of
loading more complex autonomous control algorithms that cannot be
handled by the aforementioned primary computational unit, the
inventor of the present invention made provisions for the use of a
ground station host computer for the computation of autonomous
control algorithms so that a computer significantly more powerful
than the aforementioned primary computational unit can be employed.
This arrangement permits various operating modes, such as a mode in
which, for example, attitude control is loaded on the
aforementioned primary computational unit of the aforementioned
autonomous control system and everything else is loaded on the
aforementioned ground station host computer, and a mode in which
only the autonomous control algorithm related to some of the servo
motors is loaded on the aforementioned primary computational unit
of the aforementioned autonomous control system and autonomous
control algorithms for any other servo motors are loaded on the
aforementioned ground station host computer, in addition to the
operating mode in which all autonomous control algorithms are
loaded on the aforementioned ground station host computer, thereby
substantially improving the expansion potential of the
aforementioned autonomous control system.
[0064] The manual operation transmitter for hobby purposes is
provided with a function of receiving external operation signals
that permits beginning operators to practice piloting. When a
ground station computer is used for the computation of autonomous
control algorithms, it is possible to drive the aforementioned
servo motors for the aforementioned small unmanned helicopter by
providing autonomous control computation results as external
operation signals to the aforementioned manual operation
transmitter. Further, by using the aforementioned servo pulse
mixing/switching unit, it is possible to drive the aforementioned
servo motors by first transmitting the aforementioned external
operation signals to the aforementioned autonomous control system
to generate another computational processing, instead of driving
the aforementioned servo motors directly by means of the
aforementioned external operation signals.
[0065] The aforementioned manual operation transmitter can only
accept the aforementioned external operation signals that are
encoded in pulse format. Therefore, the aforementioned autonomous
control computational results need to be converted into the pulse
format. However, no conversion equipment that can be directly
connected to the aforementioned ground station computer has been
developed.
[0066] In view of this fact, the inventor of the present invention
developed a radio control pulse generator unit that converts
computational results, in the form of numerical values that are
output from the aforementioned ground station host computer, into
the aforementioned pulse signal format that can be accepted by the
aforementioned manual operation signal transmitter. This unit uses
the RS-232C type serial communication method on the input side,
which is a common data communication technique, and thus can be
used on almost any existing computer. For the power supply, the
unit uses the power that is supplied by the aforementioned manual
operation transmitter, which eliminates the need for providing a
special power supply for the operation of the unit, thus
facilitating the handling of the unit. In addition, the unit is
endowed with functions that feed the number of maneuvering
operations and switching states of the aforementioned manual
operation transmitter to the aforementioned computer.
[0067] For the present invention, we developed a servo pulse
mixing/switching unit that permits a switch over to manual
operation in the event of an emergency situation during autonomous
control so that the aforementioned small unmanned helicopter can
fly safely. The following is a description of how the
aforementioned servo pulse mixing/switching unit works, with
reference to FIG. 16. The servo pulse mixing/switching unit 7, by
means of a pulse-processing computation CPU71 that is built in the
unit, processes the pulse signals that are transmitted from the
manual operation receiver 6 and the secondary computational unit 42
of the autonomous control unit, and generates servo pulse signals
to be output to the servo motors 3. The pulse signals that are
input from the manual operation signal receiver 6 into the servo
pulse mixing/switching unit 7 are composed of manual operation
signals, switching instruction signals between the manual operation
and autonomous control modes, and mixing ratio instruction signals
between manual operation and autonomous control signals. On the
other hand, the pulse signals that are input from the secondary
computing unit 42 of the autonomous control system into the servo
pulse mixing/switching unit 7 are autonomous control signals. The
servo pulse mixing/switching unit 7 recognizes the switching
instruction signals and the mixing ratio instruction signals. Based
on these signals, the servo pulse mixing/switching unit mixes
manual operation signals and the autonomous control signals that
are generated in the secondary computing unit 42 in an arbitrary
ratio, generates servo pulse signals based on the ratio, and can
output the results to the aforementioned servo motors 3.
Construction of the servo pulse mixing/switching unit 3 in this
manner has made it possible to switch between manual operation
signals and autonomous control signals dynamically at any time
during the flight. In view of the essential requirements for size
and weight reductions for this unit, as in the case of the
aforementioned autonomous control system, we achieved a weight
reduction to a final weight of only 250 g.
[0068] In the present invention, the aforementioned servo pulse
mixing/switching unit 7 is provided with the function of mixing
manual operation signals and autonomous control signals in an
arbitrary ratio. This will be described with reference to FIG. 16.
To mix manual operation signals and autonomous control signals, a
mixing signal instruction signal that provides a mixing ratio is
also input into the unit. As may be understood from the figure, the
mixing ratio instruction signal can also be supplied by the manual
operation signal transmitter. The pulse-processing computational
CPU71 recognizes the mixing ratio instruction signal, determines
the mixing ratio, and based on this ratio, performs calculations
for mixing manual operation signals and autonomous control signals,
generates the aforementioned serve pulse signals, and can drive
each of the servo motors 3. In this manner, the present invention
permits the use of the aforementioned autonomous control system as
a manual operation assist unit, endowed with a function for the
step-wise development of manual operation skills. Because the
mixing can be accomplished in any ratio, when the operator is not
versed in the operation of the radio-controlled helicopter, the
relative weight of autonomous control signals and manual operation
signals can be adjusted in favor of autonomous control so that
autonomous control signals can take recovery action if the operator
performs unskilled maneuvering to prevent the helicopter from
crashing, and as the operator becomes used to the operation of the
radio-controlled helicopter, the relative weight can be adjusted in
favor of manual operation. In this manner, the invention permits
gradual improvements in the manual operation of the
radio-controlled helicopter.
[0069] The present invention is designed so that the aforementioned
servo pulse mixing/switching unit is permitted to draw its power on
a shared basis from the power supply for the manual operation
system. In FIG. 16, the power supply for the manual operation
system is the battery 61. The power is supplied to the
aforementioned servo pulse mixing/switching unit 7 through the
signal lines for the manual operation signal, the switching
instruction, or the mixing ratio instruction that is input from the
manual operation signal transmitter 8 into the aforementioned servo
pulse mixing/switching unit 7. In addition, the power is supplied
to the aforementioned servo motors 3 through the inside of the body
of the aforementioned servo pulse mixing/switching unit 7. This
eliminates the need for excessively increasing the number of
batteries, prevents the problem of missing a dead battery by
oversight, forgetting to turn the power switch on, or
mis-connecting the wires due to increased complexity of wiring and
other types of human error, and thus enhances the safety of the
autonomous control system for the aforementioned small unmanned
helicopter at the practical level. With regard to the sharing of
the power supply, there is a basis for sharing the power supply for
the aforementioned manual operation system rather than the
aforementioned autonomous control unit. The reason is that if the
power supply for the aforementioned manual operation system runs
down, the aforementioned servo motors 3 will also cease to operate,
with the result that even when the power is being supplied to the
autonomous control unit, the aforementioned small unmanned
helicopter will become inoperable. Also, to maintain compatibility
between the small unmanned helicopter and the hobby-oriented
radio-controlled helicopter, it is necessary to keep manual
operation enabled even when the power for the autonomous control
unit is cut off.
[0070] In addition, in order to maintain the reliability of the
manual operation system also through the introduction of the
aforementioned servo pulse mixing/switching unit into the manual
operation system, the present invention incorporates into the
aforementioned servo pulse mixing/switching unit 7 a mechanism that
automatically detects any malfunction of the pulse-processing
computational CPU71, which is the key to the aforementioned servo
pulse mixing/switching unit that automatically restores the CPU,
and immediately returns it to the original operating condition.
Built into the aforementioned pulse-processing computational CPU71
is a timer unit 72 with the function of resetting the
aforementioned pulse-processing computational CPU71, and this timer
is used actively. The pulse-processing computational program, which
is written into the aforementioned servo pulse mixing/switching
unit 7 inputs at fixed time intervals a normal operation
notification signal indicating the state of normal operation into
the aforementioned timer unit 72. Malfunctioning of the
aforementioned pulse-processing computational CPU71 disables the
input of the aforementioned normal operation notification signal
into the aforementioned timer unit 72. If a certain length of time
elapses in this condition, the aforementioned timer unit 72
automatically inputs a reset signal to the aforementioned
pulse-processing computational CPU71, and effects a recovery
through a process similar to the power-on process. The reliable
operation of this function required the development of the
processing structure for the aforementioned pulse-processing
computational program under the assumption that the function will
be used. Specifically, we developed a program having a structure
wherein the processing component that transmits normal operation
notification signals to the aforementioned timer unit is completely
separated from the pulse input/output data unit which has a high
priority level and which has the potential of being called even in
the midst of a malfunction of the pulse-processing computational
CPU71. This permits the reliable detection of abnormal operating
conditions of the pulse-processing computational CPU71. It should
be noted that the length of time between the occurrence of a
malfunction to the issuance of a reset signal can be specified in
the aforementioned pulse-processing computational program. All
operations from the time a reset signal is input to the time the
system is recovered are performed automatically and
instantaneously. In this manner, the aforementioned
pulse-processing computational CPU71 can be restored in a matter of
dozens or hundreds of milliseconds from the time a malfunction
occurred. The assurance of continuously normal operations in terms
of software ensures the safety and reliability of the
aforementioned servo pulse mixing/switching unit.
[0071] In this invention, we developed for the aforementioned servo
pulse mixing/switching unit a function that automatically
recognizes the line connection status of the manual operation
signals, switching instruction signals, mixing ratio instruction
signals, and autonomous control signals for the signal lines that
are connected to the aforementioned servo pulse mixing/switching
unit 7, and the presence or absence of pulse signals that are input
into the aforementioned servo pulse mixing/switching unit through
these signal lines, as well as a function that generates
appropriate servo pulse signals according to the conditions that
are detected. The pulse-processing computational CPU71 continuously
measures the manual operation signals, switching instruction
signals, mixing ratio instruction signals, and autonomous control
signals of the pulse signals that are input. If measured values on
these pulse signals indicate an abnormal value, the
pulse-processing computational software determines whether the
signal line associated with the pulse signal is loose or the pulse
signal has become extinct. Specifically, if switching instruction
signals cease, all servo motors are forced to turn to the manual
operation mode; if autonomous control signals cease, the affected
servo motor 3 is forced to turn to the manual operation mode; if
mixing ratio instruction signals cease, no arbitrary-ratio mixing
is performed and the servo pulse mixing/switching unit operates
solely in the switching function; and if manual operation signals
cease, the affected serve motor is forced to turn to the autonomous
control mode. Moreover, if all signals cease to be supplied, the
software is programmed so that the servo motors will maintain the
last operating status that was in effect. In this manner, even when
the aforementioned autonomous control system is in a power-off
condition, i.e., in a condition where no autonomous control signals
are being supplied, the helicopter can be operated manually,
similar to a hobby helicopter, provided that the power is supplied
to the manual operation system. In addition, in the event of
failure to supply autonomous control signals due to a dead power
supply battery for the aforementioned autonomous control system in
the midst of autonomous control by the aforementioned autonomous
control system, the extinction of the autonomous control signals is
instantaneously recognized and the system is switched over to
manual operation. In this manner, helicopter crashing accidents can
be prevented by means of manual operation by the operator. In other
words, the manual operation system does not depend on the presence
or absence of the autonomous control system, and the autonomous
control system does not affect the manual operation system at all.
Thus, the use of this functionality has permitted the
implementation of the separation of the two signal systems in a
safe manner.
[0072] In this invention, the aforementioned servo pulse
mixing/switching unit is provided with a function that converts
pulse measurement data on the manual operation, switching
instructions, and mixing ratio instruction signals of the signals
that are input from the manual operation signal receiver 6 to the
aforementioned servo pulse mixing/switching unit into numerical
data and outputs the results to the aforementioned autonomous
control unit (in the figure, numerical data of manual operation
signals (rudders). The pulse-processing computational CPU71 of the
aforementioned servo pulse mixing/switching unit 7 constantly
measures manual operation, switching instructions, and mixing ratio
instruction signals. The servo pulse mixing/switching unit is
provided with a mechanism that outputs these measured data as
numerical data to the aforementioned autonomous control unit
through RS-232C serial communications, which are a general-purpose
interface. The data that can be output in the form of numerical
data is by no means limited to the types of data mentioned above;
any input signals. and output signals to and from the
aforementioned servo pulse mixing/switching unit 7 can be output.
The aforementioned autonomous control unit treats this data in a
manner similar to handling data from other sensors. This function
permits the recording and storage of the aforementioned manual
operation signals similar to other sensor data, thereby allowing
the implementation of safe, system characterization experiments
using correction rudders for manual operation, and system
characterization based in the input/output data that is obtained
from system characterization experiments.
[0073] In addition, in this present invention, we devised a
function capable of providing target values from the manual
operation signal transmitter 8 to the aforementioned autonomous
control algorithm implemented in the aforementioned primary
computer of the aforementioned autonomous control unit, and this
function was implemented in the aforementioned primary computer.
This is described with reference to FIG. 17. Using the function,
incorporated in the aforementioned servo pulse mixing/switching
unit 7, that outputs manual operation signals as numerical
measurement data to the aforementioned autonomous control unit,
target values are generated by multiplying the manual operation
signals by a constant of proportionality, and in this manner, the
autonomous control of the aforementioned small unmanned helicopter
was achieved. It should be noted that the aforementioned constant
of proportionality need not be a fixed number; it can be any
quantity that associates manual operation signals with target
values. In this manner, it is possible to use the manual operation
signal transmitter 8 as a target value signal input unit for the
autonomous control algorithm. Using the manual operation signal
transmitter 8, the operator can intuitively provide target values,
such as the direction of motion and velocity, to the aforementioned
small unmanned helicopter, and can accomplish given objectives by
safely maneuvering the helicopter as if manually operated. In this
method, even though the operator may be under the illusion of
performing manual operations, autonomous control signals that are
provided to the aforementioned servo motors 3 are actually computed
by the autonomous control algorithm. The operation of the
<autonomous control algorithm><autonomous control
signal><servo pulse mixing/switching unit 7><servo
motor><small unmanned helicopter 1><sensor
data><autonomous control algorithm> feedback loop ensures
the safety of the aforementioned small unmanned helicopter at all
times, and in this manner, extremely safe flying can be
realized.
[0074] Following is a description of the results of research on
tri-axial attitude control for the body of the helicopter and
autonomous control algorithms for cruising speed/position
control.
[0075] (Tri-Axial Attitude Control)
[0076] First, we explain tri-axial attitude control using the
control system for the small unmanned helicopter.
[0077] We define a coordinates system as in FIG. 4. The attitude
angles .theta., .phi., and .psi. represent the pitching angle, the
rolling angle, and the yawing angle, respectively. In this
coordinates system, the positive X direction is the forward
direction from the body, the positive Y direction is the direction
to the right of the body, and the positive Z direction is the
direction downward from the body
[0078] (Modeling of Servo Motors)
[0079] Servo motors similar to those used in a small hobby-type
radio-controlled helicopter are used as actuators for the body of
the helicopter, wherein the input pulse width is equal to the
rotational angle of a servo motor. Parameters .zeta..sub.s and
.omega..sub.ns were determined by the partial space identification
method by assuming the characteristics of a servo motor as in Eq.
(1) using a transfer function and by entering M-series of signals
as pulse widths into the servo motors. 8 G s ( s ) = n s 2 s 2 + 2
??? s n s s + n s 2 ( 1 )
[0080] (Pitching Rolling Model)
[0081] The autonomous control system used in the present invention
includes sensor delay time and delay time in the wireless space.
Such delay time represents three times the sampling time.
Therefore, the transfer function representation from an elevator
servo (longitudinal direction instruction to the pitch axis
attitude angle .theta., as in Eq. (2), takes the form of the
following transfer function equation (3) from an aileron servo
(lateral direction) instruction to the roll axis attitude angle
.phi.: 9 G ( s ) = - Ls K n s 2 ( s 2 + 2 ??? s n s s + n s 2 ) ( T
s + 1 ) s ( 2 ) G ( s ) = - Ls K n s 2 ( s 2 + 2 ??? n s s + n s 2
) ( T s + 1 ) s ( 3 )
[0082] (Yawing Model)
[0083] The system incorporates a yaw axis speed stabilization rate
gyro unit similar to that used in a small hobby-type
radio-controlled helicopter. The unit performs controls by treating
the input data as a yaw axis rotational angle speed. The helicopter
body used in the present mode of embodiment also incorporates an
angular velocity servo controller that uses an angular velocity
gyro sensor. This sensor is assumed to be a second-order delay
system. Therefore, the yaw axis rotational motion model takes the
form of a system that includes delay time, a second-order delay
system, and one integrator, as shown in Eq. (4): 10 G ( s ) = - Ls
K n s 2 ( s 2 + 2 ??? s n s s + n s 2 ) s ( 4 )
[0084] where the model gains (K.sub..phi., K.sub..theta.,
K.sub..psi.) and the time constants (T.sub..phi., T.sub..theta.)
are adjusted and determined by comparing experimental and
simulation data so that the two sets of numerical data will
agree.
[0085] In the present mode of embodiment, the LQG optimal control
theory is applied to the above three mathematical models to design
an autonomous control algorithm. In addition, a first-order servo
system was constructed to eliminate steady-state deviations. An
optimal feedback gain is determined by assuming that the pitching
angle .theta., the rolling angle .phi., and the yawing angle .psi.
are either single-input, single-output (SISO) systems or uncoupled
systems, respectively. FIG. 8 shows experimental results that were
produced by the controller thus obtained.
[0086] In this section, we explain translational motion controls
using the control system for a small unmanned helicopter.
[0087] (Longitudinal/Lateral Motion Model)
[0088] From a simple dynamical analysis, the following equation on
the lateral speed Y of the body can be obtained: 11 Vy = g 1 s ( 5
)
[0089] where g denotes the acceleration due to gravity; and .phi.,
the rolling attitude angle. The derivation of Eq. (5) is based on
the assumption that the helicopter body is at a fixed altitude and
that .phi.<<1. The experimental data shown in FIG. 9 suggests
the existence of some motion properties between the attitude angle
of the body and the acceleration. By approximating these properties
in terms of a first-order delay and by comparing experimental
results with simulations, we added one instability pole to Eq. (5),
and used the following transfer function representation
mathematical model in the final design of the control system: 12 Vy
= g T s + T a s - a ( 6 )
[0090] A similar mathematical model was also used for the
longitudinal speed of the helicopter body: 13 Vx = g T s + T a s -
a ( - ) ( 7 )
[0091] where .THETA. denotes the pitch attitude angle of the
body.
[0092] (Vertical Motion Model)
[0093] The helicopter performs vertical motions by varying the
blade corrective pitch, thus varying the rotor lift. According to
blade wing element theory, the lift due to a rotor can be
represented as follows: 14 T = b 4 a 2 R 3 ( t + t ) c ( 8 )
[0094] where b denotes the number of rotor blades; .rho., the
density of air; a, the slope of the two-dimensional lift; .OMEGA.,
the rotor rpm; R, the radius of the rotor blade; .theta..sub.t the
corrective pitch; .phi..sub.t, the inflow angle; and c, the chord
length. Assuming that all of these variables with the exception of
.theta..sub.t, are constants, the following transfer function
representation mathematical model for the velocity in the Z
direction can be formulated: 15 Z = k x t ( 9 )
[0095] (X-Y Direction Autonomous Control Algorithms)
[0096] For the models represented by Eqs. (6) and (7), we designed
independent velocity control algorithms for the X, Y, and Z
directions based on LQI control theory. FIG. 11 shows an overall
image of the X and Y position control loops. The velocity control
algorithms for the X and Y directions are computational algorithms
that calculate the attitude angle of the body to yield the required
vehicle speeds when certain, respective, target speeds for the
vehicle are given.
[0097] To move the body to an arbitrary target position, the
respective target velocities are calculated according to the
following equations:
V.sub.xref=.alpha.(P.sub.xref-P.sub.x) (10)
V.sub.yref=.alpha.(P.sub.yref-P.sub.y) (11)
[0098] where Vxref and Vyref denote target velocities in the X and
Y directions, respectively; Pxref and Pyref denote target positions
in the X and Y directions, respectively; Px and Pydenote the X- and
Y-coordinates of the body, respectively; and .alpha. is an
arbitrary constant.
[0099] The control loop has a structure in which the attitude
controller, the velocity controller, and the position controller
are arranged serially. Compared with a single controller, this
structure has the following advantages: (1) the ability to limit
the attitude angle within a safe range; (2) an improved position
control overshooting through the application of a velocity limiter;
and (3) a controller internal state that is not dependent on
positional coordinates.
[0100] The third advantage avoids the use of a complex algorithm
for coordinate transformations when the yawing angle .psi. is
varied in an arbitrary coordinate system.
[0101] In view of the fact that the position and velocity for each
axis can be observed by means of high-precision RTK-DGPS, one state
quantity in the 3D model shown in Eqs. (6) and (7) is estimated by
means of a minimum-dimension observer. Because the dynamics for the
orientation servo systems are assumed to be sufficiently faster
than the translational motion, their properties are not taken into
considerations in the design of the control system.
[0102] (Z-Direction Autonomous Control Algorithm)
[0103] The velocity control algorithm for the Z direction was
designed by comparing experimental results with simulation results,
by adjusting and determining the value of k in Eq. (9) so that the
two sets of numerical data agree, and by using the resulting value
and applying LQI control theory. The velocity control algorithm
with respect to the Z direction is a computational algorithm that
calculates the corrective pitch necessary to implement the required
speed of the vehicle when a target speed value in the Z direction
is given.
[0104] To move the body to an arbitrary target position, the target
velocity in the Z direction is calculated according to the
following equation:
V.sub.zref=.beta.(P.sub.zref-P.sub.z) (12)
[0105] where Vzref denotes a target velocity the Z direction;
Pzref, a target position value in the Z direction; Pz the
Z-coordinate of the vehicle; and .beta., an arbitrary constant.
[0106] FIG. 12 shows the reference trajectory following control
results, and FIG. 13 shows the experimental results of the hovering
control. FIG. 14 shows the lateral and longitudinal velocity
control results and FIG. 15 shows the altitude control results.
Following is a description of a mode of embodiment of the
autonomous control algorithm based on the aforementioned servo
pulse mixing/switching unit, the aforementioned radio-controlled
pulse generator unit, and the aforementioned mathematical models,
such that the autonomous control algorithm is implemented on the
aforementioned small unmanned helicopter. The configuration of the
system can be divided broadly into a mobile station, wherein the
aforementioned autonomous control system and the aforementioned
servo pulse mixing/switching unit are installed on the
aforementioned small unmanned helicopter; and a ground station
wherein the state of autonomous control is monitored and target
values are input. Installed on the mobile station are a helicopter
body 1; sensors 2 that detect the current position and the attitude
angle of the helicopter body 1; servo motors 3 that operate the
five rudders for the helicopter body 1; a CPU4 that independently
calculates control reference values on optimal motion and direction
for each of the servo motors 3 by using the aforementioned
autonomous control algorithm, from the current flight state of the
helicopter obtained from the sensor 2, and the target values that
are set by the ground station; a wireless modem 5 that communicates
with the ground station; a radio-controlled receiver 6 that
receives manual operation signals from the radio-controlled
transmitter 8; and a servo pulse mixing/switching unit 7. The
aforementioned autonomous control system is built upon the sensor 2
and the CPU 4. In addition to its computation function, the CPU 4
is also endowed with the function of exchanging signals with the
wireless modem 5 to permit monitoring of the sensor information
obtained from the sensor 2 on the ground station and entering
target values that are established by the ground station. Included
in the sensor 2 are the GPS (not shown in the figure) that detects
the position of the helicopter body 1; a tri-axial orientation
sensor (not shown in the figure) that detects the orientation
states of the three axes; a ground altimeter (not shown in the
figure) that determines the altitude of the helicopter body; and a
magnetic azimuth meter (not shown in the figure) that measures the
azimuth. The servo pulse mixing/switching unit is a unit that can
switch between manual operation signals and autonomous control
signals, and can mix manual operation signals and autonomous
control signals in any proportion. For example, if the autonomous
control system for the helicopter fails during flight for some
reason or the CPU 4 fails, disabling the autonomous control of the
helicopter body, the aforementioned servo pulse mixing/switching
unit will automatically switch to the manual operation mode, thus
preventing the helicopter from crashing. In addition, the provision
of the aforementioned servo pulse mixing/switching unit 7 permits
the use of the autonomously controlled aforementioned small
unmanned helicopter for training in the use of a radio-controlled
helicopter. When the operator is not proficient in the operation of
a radio-controlled helicopter, the ratio between autonomous control
and manual control can be set so that it is heavily skewed in favor
of autonomous control, such as 9:1 and 8:2, so that even if the
operator maneuvers the helicopter unskillfully, the autonomous
control signals transmitted by the CPU 4 will perform recovery
operations to prevent crashing. As the operator gains familiarity
with the operation of the radio-controlled helicopter, the ratio
can be changed in favor of manual control, thereby improving the
operator's skill in performing manual operations on the
radio-controlled helicopter.
[0107] The ground station is provided with a CPU 91 for the input
of position or velocity target values; a monitor CPU 92 that
monitors sensor information on the helicopter body 1 that is
transmitted from the mobile station; and a wireless modem 10 that
performs communications with the mobile station. Also provided is a
radio control transmitter 8 that permits the operator to perform
manual operations.
[0108] Following is a description of the operation of the system.
The operator enters position or velocity target values from the CPU
9 of the ground station. The target values, through the wireless
modems 11 and 5, are entered into the CPU 4 that performs
calculations. Based on the target values that are entered and the
current flight status of the helicopter body 1 obtained from the
sensors 2, the CPU 4 performs computations of the autonomous
control algorithms that are designed based on aforementioned
mathematical models (2), (3), (4), (6), (7), and (9). Based on the
results of these computations, five types of control instructions
involving optimal motion and direction for each of the servo motors
3 are determined. If full autonomous control is on, the servo pulse
mixing/switching unit 7, not under the influence of manual
operation signals, controls each of the servo motors 3 based on the
controls signals having computed control reference values, thus
achieving the fully autonomous control of the motions of the
rudders for the helicopter body 1 according to target values. If
manual operation signals are to be mixed by the servo pulse
mixing/switching unit 7, control signals, mixed according to a set
proportion, are transmitted to the servo motors 3.
[0109] As described above, according to the autonomous control
system for the small unmanned helicopter based on this mode of
embodiment, the implementation and installation of the autonomous
control algorithms designed based on the above-described
mathematical models (2), (3), (4), (6), (7), and (9), as well as
the CPU 4, the sensors 2, and the servo motors 3, on a hobby
helicopter body permit the calculation of control reference values
that specify the optimal motion and direction for the respective,
corresponding servo motors 3 from the current flight state obtained
from the sensors 2 and from given position and velocity target
values, by using the aforementioned autonomous control algorithms,
thus achieving complete control of the helicopter body according to
target values.
[0110] It should be noted that, as another mode of embodiment, the
autonomous control of the aforementioned small unmanned helicopter
can be similarly achieved by installing a computer similar in
computational capabilities to the CPU 4, on the ground station.
However, in this case, the feedback control loop that is formed
would contain two wireless sections in the wireless modem and in
the manual operation transmitter/receiver. In such a case, if
wireless communications break down, the logical structure of the
control system will collapse, thus posing a risk of instability of
the flight of the aforementioned small unmanned helicopter. From a
flight safety standpoint, it is not desirable to implement all of
the aforementioned autonomous control algorithms on a host computer
installed on the ground station. In such a case, the CPU 4 and the
aforementioned ground station host computer can be used in
combination for control computation purposes by combining the
aforementioned servo pulse mixing/switching unit and the
aforementioned radio control pulse generator unit. Many widely
different embodiments of the present invention may be constructed
without departing from the spirit and scope of the present
invention. It should be understood that the present invention is
not limited to the specific embodiments described in the
specification, except as defined in the appended claims
* * * * *