U.S. patent application number 10/723291 was filed with the patent office on 2004-07-22 for method and system for determining tire pressure imbalances.
Invention is credited to Rosseau, James R..
Application Number | 20040143376 10/723291 |
Document ID | / |
Family ID | 30448050 |
Filed Date | 2004-07-22 |
United States Patent
Application |
20040143376 |
Kind Code |
A1 |
Rosseau, James R. |
July 22, 2004 |
Method and system for determining tire pressure imbalances
Abstract
A tire deflation warning system is based solely on the measured
distance traveled by each tire of a vehicle. The vehicle is
equipped with a sensor which generates pulses representative of the
distance traveled by each wheel such as generated by an ABS brake
system. The system accumulates the pulses in the controller to
determine whether the vehicle is stable. If the vehicle is stable,
it performs a test to cover individual tires, opposing diagonal
tires or any combination of three tires using an average of the
diagonal ratio. These tests are performed by utilizing the pulses
generated by each wheel representing the distance traveled by the
wheel.
Inventors: |
Rosseau, James R.;
(Birmingham, MI) |
Correspondence
Address: |
CHRISTOPHER DEVRIES
General Motors Corporation
Legal Staff, Mail Code 482-C23-B21
P.O. Box 300
Detroit
MI
48265-3000
US
|
Family ID: |
30448050 |
Appl. No.: |
10/723291 |
Filed: |
November 26, 2003 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
10723291 |
Nov 26, 2003 |
|
|
|
10347151 |
Jan 17, 2003 |
|
|
|
6684691 |
|
|
|
|
Current U.S.
Class: |
701/31.4 |
Current CPC
Class: |
B60T 2240/07 20130101;
B60C 23/061 20130101 |
Class at
Publication: |
701/029 |
International
Class: |
G06F 007/00 |
Claims
1. A system of determining tire pressure faults in a vehicle
comprising: determining a distance a first tire has traveled;
determining a distance a second tire has traveled; comparing the
first and second distances to determine if a pressure fault has
occurred in said first or second tire.
2. The system of claim 1 wherein the distances are calculated using
pulse generating sensors coupled to the first and second tires,
wherein a series of pulses equate to a distance.
3. The system of claim 2 wherein a ratio of pulses is used to
determine a tire pressure fault.
4. A system of determining tire pressure faults in a vehicle
comprising: determining distances a plurality of tires have
traveled; comparing the distances to determine if a pressure fault
has occurred in said plurality of tires.
5. The system of claim 4 wherein the distances are calculated using
pulse generating sensors coupled to the first and second tires,
wherein a series of pulses equate to a distance.
6. A system for detecting tire pressure imbalance comprising: a
vehicle; a plurality of wheels coupled to said vehicle; a plurality
of sensors operatively coupled to said plurality of wheels, each
said sensor sensing one of said plurality of wheels, said sensors
generating pulses indicative of distance traveled by each said
wheel; a controller for receiving said pulses generated by said
sensors; wherein a tire pressure fault is determined by analyzing
the distance traveled by each said wheel.
7. The system of claim 6 wherein said plurality of sensors are
coupled to an anti lock brake system and said anti-lock brake
system transmits said pulses to said controller.
Description
RELATED APPLICATION
[0001] This application is a continuation of U.S. Ser. No.
10/347,151 filed Jan. 17, 2003.
TECHNICAL FIELD
[0002] This invention relates to a method of and system for
determining imbalances in tire pressure on vehicles equipped with
wheel rotation sensors.
BACKGROUND OF THE INVENTION
[0003] Some suppliers of automotive chassis control systems offer
algorithms that use wheel speed for pneumatic tire inflation
monitoring. The typical wheel speed based tire inflation-monitoring
algorithm resides within an anti-lock brake controller. This system
architecture is required because of the need for high resolution
and high wheel speed data throughput. These systems rely on
accurate microprocessor timer information to perform the required
speed calculations. The measured time between speed ring teeth is
used to calculate wheel speed and wheel slip. Due to the nature of
these calculations, such systems are prone to noisy data at low and
high speeds, under some road conditions, due to imprecise machining
of speed ring teeth, and microprocessor timing limitations. A
typical wheel speed based system may operate well while traveling
in a straight line, at steady speeds, and on smooth surfaces, but
exhibit a high rate of false warnings when conditions are not
optimal.
SUMMARY OF THE INVENTION
[0004] While a vehicle is in motion, the present invention
determines when any combination of one, two, or three wheels are in
a state of rotational error through an adaptive measurement of the
distance that the four wheels have traveled and not through any
calculation of wheel speed. This invention has the inherent ability
to accurately discern a measure of pneumatic tire pressure
imbalance in a way not yet achieved.
[0005] More precisely, this invention provides a means to
accurately monitor tire pressure imbalance through the measurement
of distance by way of digital pulse devices included in existing
anti-lock braking systems and vehicle transmissions. Moreover, this
invention does not use nor care about the measured time between
digital pulses and will operate at the lowest possible speed at
which a digital pulse may be sensed. This invention will operate
and maintain accuracy equally as well at unrestricted high speeds.
This invention will operate expediently on any surface. Tire
pressure imbalance is determined while all four wheels are on a
similar surface. Furthermore, this invention is able to detect
certain surfaces such as gravel, snow, grass, etc.
[0006] The first step in this process is to compute a "perfect"
average of the diagonal ratio (the ratio of pulses accumulated on
one front and its opposite rear over the total pulses). This
computation must be performed when the vehicle is traversing a
non-deformable surface, has zero acceleration, and is going
straight. This "perfect" average is then used to generate a tire
deformation adjustment to compensate for acceleration,
deceleration, or turning during data collection. This adjustment
prevents the false triggering of a warning on tires that are lower
than the nominal pressure, but still above the warning threshold.
False triggering can be caused by the effect of weight transference
due to acceleration, deceleration or cornering on tire rolling
radius.
[0007] Three separate tests are performed to determine tire
pressure loss. The first test covers any individual tire, opposing
diagonal tires, or any combination of three low tires. The first
step in this process is to compute an average of the diagonal
ratio. A requirement of this test is that the diagonal ratio stays
within a narrow band. If the diagonal ratio goes outside of the
acceptable band, the data is deemed unreasonable, and is not
included in the average diagonal ratio. In the case of a rapid loss
of pressure, the data will be ignored initially, and then accepted
when it meets the criteria of repeatability. If the data is deemed
repeatable, the tire deformation adjustment is applied to it and a
running average is formed. If the running average is greater than
the calibration amount by more than the pressure loss threshold, a
warning occurs. If the running average is less than the trigger
point, a test is performed to determine if a warning is currently
set. If no warning is set, the routine proceeds to the multiple
wheel tests. If a warning is set, the routine checks for constant
speed and heading. If this criterion is met, the diagonal ratio is
tested for being within three PSI of the "perfect" average computed
previously. If this condition is met, the warning is cleared.
[0008] The two additional pressure loss tests cover the loss of
pressure from both of the front, rear, left, or right tires. These
multiple wheel tests require that the vehicle is not accelerating,
decelerating, or turning. The algorithm looks for repeatability
over several tenths of a mile to determine low tires. The distance
test must be set to accommodate the longest curve likely to be
encountered. To perform this test, lateral and longitudinal
distance ratios must be computed. The lateral ratio consists of the
ratio of the pulses accumulated on one side of the vehicle to the
total pulses. If the lateral or longitudinal ratios are greater
than a calibration, then the routine resets the respective test
odometer. When the longitudinal odometer counts down to zero, the
longitudinal ratio minus a speed related compensation is compared
to a pressure loss threshold. A warning occurs if the threshold is
exceeded, and is cleared if the longitudinal ratio is within three
PSI of the nominal calibration. When the lateral odometer counts
down to zero, the lateral ratio is compared to a pressure loss
threshold. A warning occurs if the threshold is exceeded, and is
cleared if the lateral ratio is within three PSI of the nominal
calibration.
[0009] The calibration mode is very strict about the data it will
accept. It will not accept data during acceleration, deceleration,
turning, or any non-repeatable data ratios. If the longitudinal
ratio varies outside of a tight calibration, the algorithm rejects
the data. If the vehicle is traversing a deformable surface, the
routine rejects the data. If the data passes both of these tests,
tight repeatability criteria insure that the measurement is valid.
The data from the calibration mode is used to compensate the
incoming data to eliminate the effects of variation in tire
loading, tire size, tire wear that are normally present.
[0010] Moreover, this invention will accommodate any steering
angle, longitudinal acceleration, or lateral acceleration. In
addition, this invention uses an adaptive method to eliminate false
warnings by tracking variations in tire pressure that are less than
the warning trigger point and adjusting the measurements based on
the tire deformation characteristics due to acceleration induced
weight transference and/or vehicle load.
[0011] Advantageously, this invention will calibrate all four
wheels during periods of zero longitudinal and lateral acceleration
and adaptively correct for speed variations of all wheels.
[0012] Advantageously, this invention will trigger a warning upon
reaching a predetermined pressure loss threshold and clear said
warning without recalibration upon a tire reaching a predetermined
recovery threshold. This is a self-clearing feature which clears
the warning when a tire or tires are refilled to a predetermined
level within the nominal calibration thresholds or when simply
clearing a false trigger precipitated by a juxtaposition of
unlikely data events.
[0013] Advantageously, this invention will trigger a warning upon
the rapid loss of tire pressure. This will allow detection of all
types of pressure leakage, from a pinhole leak to a catastrophic
failure.
[0014] Advantageously, this invention may reside in any vehicle
controller or custom built controller that may receive digital
pulses from an ABS system and/or a vehicle transmission at any
regularly scheduled time interval by any available form of
communication.
BRIEF DESCRIPTION OF THE DRAWINGS
[0015] This invention may be best understood by reference to the
preferred embodiment and to the drawings in which:
[0016] FIG. 1 is a hardware diagram of a vehicle equipped with an
anti-lock braking system having a rotation sensor on each wheel. In
addition, said anti-lock braking system controller having
communication with an additional controller in which (in this
embodiment) the invention will operate. The secondary controller
having communication with the vehicle to provide a means to inform
the vehicle operator through a driver information device of a loss
of pneumatic tire pressure balance.
[0017] FIGS. 2-5 illustrate a sequence of operations that are
executed to carry out the tire pressure warning method function
outlined in FIG. 1.
DESCRIPTION OF THE PREFERRED EMBODIMENT
[0018] A method and apparatus for detecting under-inflated tires
for a vehicle using distance traveled by each of the tires. For the
purposes of this invention, only the mechanical and electrical
components are shown. In the preferred embodiment, as shown in FIG.
1, a vehicle 18 is equipped with an Anti-lock Braking System (ABS)
and a means to sense all four wheel rotations. Alternatively, the
method does not require an ABS system. The method requires at a
minimum that wheel rotation sensors are included as well a means of
measuring the rotation sensor signals in a controller. The rotation
sensor generates a predetermined number of pulses for each
revolution of the wheel as in an ABS system.
[0019] Referring to FIG. 1, when the left front tire 1 and the
matching wheel bearing 5 rotate about an axis, a motion sensor 19
which includes an integrated machined tooth passes a mounted
proximity sensor 9 generates a rising edge digital pulse upon first
contact and a falling edge digital pulse when the tooth is no
longer within the sensor's proximity. An ABS controller 13 receives
the rising and/or falling edge digital pulses for each wheel and
accumulates the sum of these pulses for each wheel. The ABS system
communicates the sum of the digital pulses of each individual wheel
to the Body Control Module (BCM) 14 by way of an ordinary serial
data interface 15, at regular intervals. The same can be said of
the remaining wheels, where the combination of tire and wheel 2,
wheel bearing 6, and sensor 10, represent the right front assembly,
wheel 3, wheel bearing 7, and sensor 11, represent the left rear
assembly, and wheel 4, wheel bearing 8, and sensor 12, represent
the right rear assembly. When the sequence of operations as
outlined in FIG. 2 determine that a tire or tires are in a state of
rotational error because of tire pressure loss, the BCM controller
14 issues a message to the driver information device 16 by way of
an ordinary serial data interface 17.
[0020] Since this invention is based solely on the distance
traveled, a more detailed description of the properties of this
invention follows. The method is based solely on the measured
distance traveled by each tire. The quantity of machined teeth and
rolling circumference of each wheel determines the relationship
between sensor pulses and distance. It is well known that a
deflated tire has a smaller rolling radius. The basis for this
invention is that an under-inflated tire will rotate more than a
properly inflated tire. For example, if we have a vehicle having
four matching fifteen inch steel belted radial tires, with one tire
having a pressure of ten PSI less than the other three tires, then
the under-inflated tire will need to rotate the equivalent of 2.8
feet more than the three properly inflated tires will in one-tenth
of a mile. One rotation of this tire measures 6.3 feet and produces
forty-eight digital pulses in that distance. In one-tenth of a
mile, these tires rotate approximately eighty-four times with each
tire producing four thousand and thirty-two digital pulses. The
additional 2.8 feet of travel of the under inflated tire will
produce an extra twenty-one digital pulses in one-tenth of a mile
or one extra digital pulse every four wheel rotations.
[0021] The pulse counting method has the inherent ability to
function properly at any vehicle speed, on any surface, and
relatively noise free under adverse driving conditions. A
substantial advantage of this method is that it may be utilized in
any vehicle controller, as long as it receives the wheel pulse
count at regular intervals. This method does not require high
throughput and may operate on any real time interrupt time base.
Time between teeth is not used.
[0022] Referring to FIG. 2, a method of detecting under-inflated
tires by measuring the rotational variation of vehicle wheels is
diagrammed to illustrate the control processes of this embodiment.
Referring to FIG. 1, wheel rotation is measured by the ABS
controller 13 and all of the digital pulse transitions from
proximity sensors 9, 10, 11, and 12, are accumulated by the
controller 13 and the individual sums of each wheel transmitted to
the BCM 14 at a convenient time interval.
[0023] At every interval of one hundred milliseconds, for example,
a series of controller instructions generally illustrated by the
operations of FIG. 2 starting at step 20 and proceeding to step 25
are executed. At step 25, all of the digital pulses of every
individual wheel are received and stored into an individual account
bearing the names of: left front (LF), right front (RF), left rear
(LR), and right rear (RR) in a storage device. In step 30,
determination is made as to whether the calibration switch is
depressed. If it is determined that the calibration switch is on,
calibration mode is enabled by setting CalMode=True, and
KDAIG=KDAIG2=Kdcal, in step 35, as discussed more fully below.
[0024] At every interval of one hundred milliseconds, for example,
a series of controller instructions generally illustrated by the
operations of FIG. 2 starting at step 20 and proceeding to step 25
are executed. At step 25, all of the digital pulses of every
individual wheel are received and stored into an individual account
bearing the names of: left front (LF), right front (RF), left rear
(LR), and right rear (RR) in a storage device. In step 30,
determination is made as to whether the calibration switch is
depressed. If it is determined that the calibration switch is on,
calibration mode is enabled by setting CalMode=True, and
KDAIG=KDAIG2=Kdcal, in step 35, as discussed more fully below.
[0025] If step 45 determines that the wheels are not slipping, then
the system proceeds to step 50, where the new digital pulse
information is summed, LFT=LFT+LF, RFT=RFT+RF, LRT=LRT+LR,
RRT+RRT+RR and dist=LFT+RFT+LRT+RRT. This is followed by step 60,
which determines whether one or more complete revolutions of all
wheels--the short distance occurred a complete revolution has
occurred and if the total number of digital pulses are greater than
calibration KSHORTD, the routine proceeds to step 70 which
calculates several parameters to allow a determination to be made
of the acceptability of the data thus far collected. If not, the
system proceeds to step 640 to wait for the next time period.
[0026] The next steps, 80-110, determine the suitability of the
data for determining tire pressure imbalance. A series of steps are
performed to see if the vehicle is stable. Typical criteria used
are: lateral acceleration below a threshold (step 80), turning
radius above a limit (step 90). At step 110, the difference between
the diagonal ratio and the average value of this parameter
calculated at steps 220-260 is used to eliminate data that is
impossible to achieve due to pressure imbalances. The diagonal
ratio is the ratio of the distance traveled by the left front and
right rear tires over the distance traveled by all of the tires.
This is done by summing the pulses for the wheels. If any of the
required criteria are not met, the algorithm proceeds to step 140,
where the values are reset. If the data accumulated thus far is
acceptable, at step 150 this data is further aggregated into sums
LFA, RFA, LRA, and RRA. At step 160, a determination is made to
determine if a second total distance has been accumulated.
[0027] If the second required distance has been achieved (KMIDD),
then at step 170, additional calculations are made to allow the
following steps to determine the validity of the data. At step 180,
the longitudinal ratio, which is the ratio of the distance traveled
by the front wheels over the total distance traveled by all of the
wheels, is used to determine if the vehicle is on a loose surface.
Again, this is done by using the number of pulses. An additional
compensation is made in step 180 to account for the additional
distance the driven wheels of a vehicle will travel due to the
tractive slip required to maintain vehicle velocity. This
compensating factor is generally related to the speed of the
vehicle when the data is recorded. If the vehicle is on a loose
surface, the data is only accepted if the diagonal ratio is
repeatable within a tolerance (KDIAGREP) at step 120, and not
outside of a bound (KDAIG2) with respect to the average value of
the diagonal ratio in step 130.
[0028] If the data meets the previous criterion, a last pulse
accumulation is made to form the sums LFS, RFS, LRS, RRS at step
190. The acceleration, lateral acceleration and speed are also
accumulated to keep track of the vehicle conditions as each record
is added to the final sum. If at step 200, the distance LONGD
(typically 0.05-0.2 miles) has been achieved, then the routine
proceeds to step 210. At this point, the previous values of the
ratios are saved and the longitudinal, diagonal and lateral ratios
are calculated. The lateral ratio represents the ratio of the
distance traveled by the left side tires over the distance traveled
by all of the tires. At steps 220-260, an average value of the
diagonal ratio is created by using data only when the vehicle is
traveling in a very consistent manner. This is typically determined
by low acceleration (step 220), low lateral acceleration (step
230), and repeating values of the diagonal ratio (step 250). At
step 260, the new data is averaged into the old to form the value
AvgDiag.
[0029] At step 270, compensations are calculated to prevent false
pressure imbalance detection caused by acceleration or turning. The
basis for this compensation is the characteristic that a
low-pressure tire will deform more than a properly inflated one,
under loading from weight transfer caused by vehicle maneuvering.
Without this compensation, a tire that is low, but above the
detection threshold, could appear lower than the limit during
maneuvers, giving a false or premature warning. At step 280, if the
diagonal ratio is repeating within a narrow band, a running
diagonal ratio average is created using the current value of the
diagonal ratio. Additionally, in step 290, the pressure loss
variable is calculated using the running average and the calibrated
value of the diagonal ratio as well as the compensations calculated
previously. It should be noted that the calibrated values of the
diagonal ratio, lateral ratio, and the longitudinal ratio represent
the values of these factors when the tires are in a known nominal
state i.e. when the pressures are correct. The difference between
these calibrated values and the current value of these ratios are
generally proportional to the pressure imbalance being measured. If
the diagonal ratio is not repeating within the band Krepeat, then
in step 300, the running diagonal ratio is updated with the value
AvgDiag, which represents the current value of the diagonal ratio
when no maneuvers are occurring. In this way, the running average
is prevented from being corrupted by vehicle conditions that could
lead to false warnings. At step 310, the calculated pressure
imbalance is compared to the desired threshold, Ktrip. If the limit
is exceeded, an alert is indicated at step 320. If the pressure
loss is below the threshold, step 330 determines if a warning is
already in effect, if so, at steps 340-350, a determination is made
if the alert should be cancelled. The alert is set to OFF at step
360 if the running average is close to the value AvgDiag and the
calculated pressure loss is below the value Ktripoff. At step
370-380, a determination is made if the vehicle is changing speed
or turning, if so, no further pressure loss checks are performed
and the routine exits to step 570.
[0030] Steps 390-470 are used to detect the simultaneous loss of
pressure in two tires on the same side of the vehicle. This type of
pressure loss is difficult to determine using wheel rotation data
because the effect of this type of pressure imbalance is the same
as the effect of a vehicle performing a turn. At steps 390-410, the
lateral ratio is checked to determine if it is repeating within a
narrow band (Klat) at least Kcnt times in a row. If the value
repeats as required, the routine proceeds to step 420, otherwise it
exits to step 480, clearing the counter if appropriate. At step
420, a comparison is made to determine if the loss alert should be
set to ON. The comparison is based on the difference between the
calibrated value of the lateral ratio and the current filtered
value. If the pressure loss is above Ktrig, the repeat counter is
cleared at step 450 and the warning set at step 460. If the
calculated pressure loss is below the level to set the warning, at
step 430, the loss is compared to the calibration Ktripoff to
determine if the warning should be cleared. If required, the
warning is cleared at step 440.
[0031] Steps 480-560 are used to detect the simultaneous loss of
pressure in two tires on the same end of the vehicle. This type of
pressure loss is difficult to determine using wheel rotation data
because the effect of this type of pressure imbalance is the same
as the effect of a vehicle traveling in a situation where a large
amount of power is required to maintain speed. Some examples are
driving at high speed, climbing a long grade or towing. At steps
480-500, the compensated value of longitudinal ratio is checked to
determine if it is repeating within a narrow band (Klong) at least
Kcnt times in a row. If the value repeats as required, the routine
proceeds to step 520, otherwise it exits to step 570, clearing the
counter if appropriate. At step 520, a comparison is made to
determine if the loss alert should be set to ON. The comparison is
based on the difference between the calibrated value of the
longitudinal ratio and the current filtered value. If the pressure
loss is above Ktrig, the repeat counter is cleared at step 530 and
the warning set at step 540. If the calculated pressure loss is
below the level to set the warning, then at step 550, the loss is
compared to the calibration Ktripoff to determine if the warning
should be cleared. If required, the warning is cleared at step 560.
The routine then proceeds to step 570, where a check is made to see
if the calibration mode is in effect. If the calibration mode is
not active, the routine is complete at step 640.
[0032] In calibration mode, at steps 580-590, each of the ratios is
checked for repeatability within the band Kcal. Any ratio that
varies outside this band causes the process to proceed to step 640
to wait for the next time period. At step 595, calcnt is checked to
determine if the values of A1, A2, A3, need to be initialized, if
this is the first execution of the calibration routine, at step 600
the ratio averages are initialized. At step 610, the counter calcnt
is incremented and a running average created for each of the three
ratios: lateral, longitudinal and diagonal. The routine proceeds to
step 620, where the value of calcnt is compared to Kcalrep to
determine if the required number of repeating ratios has occurred
to allow the determination of a valid calibration. At step 630, the
running ratio averages are stored in the respective variables
LatCal, LongCal, and DiagCal. In addition, the several other
variables are initialized and the calibration mode is completed.
The routine then goes to step 640 where it remains dormant until
the next real-time interrupt, which will begin execution at
referenced step 10.
* * * * *