BACKGROUND AND SUMMARY OF THE INVENTION
The present invention relates to a method of controlling the braking
of a vehicle, as more specifically defined in the preamble of Claim 1.
Document DE-A-37 22 107 is illustrative of such a prior electronic
braking system. This document shows a brake system and method directed to the generation
and distribution of hydraulic brake pressure to various brake mechanisms. The method
is directed to the generation and distribution of hydraulic brake pressure to various
brake mechanisms, by fencing wheel skid, generating a signal indicative of brake
effort and monitoring the change in the velocity of wheel speed.
In conventional braking systems, a pressure is applied to each brake
unit; the braking torque generated from the pressure acting on a piston (or pistons)
forces some friction material to act on a rotating element (usually a drum or a
disk). The same pressure is normally applied to both wheels of an axle. Frequently
to prevent vehicle instability due to locking of the rear wheels before lockup
of the front wheels, a proportioning valve(or valves) modulates to rear wheel brake
pressure. Sometimes an adjustable proportioning valve, sensitive to vehicle spring
deflection, is used which attempts to maintain proper brake balance. The adjustment
rebalances the brakes (front to rear) as a result of changes in vehicle loading.
The usual (conventional) approach has some shortcomings. The coefficient
of friction between the brake pads/shoes and the rotating element is not uniform.
In high-volume production, the coefficient of friction may vary significantly batch
to batch. Allowances are made for this variation, to insure that the rear wheels
do not lock prematurely, causing vehicle instability. This "safety margin" results
in less than best brake balance. Furthermore, the usual brake balancing schemes
assume a nominal friction coefficient of friction of the brake pads and linings
and also assume that the reaction torque at the tire/road interface is consistently
related to brake torque. This is not always true since if the tire size (rolling
radius) changes appreciably, the lever arm, through which the torque acts, will
These shortcomings result in compromises in brake effectiveness,
and can cause uneven wear of tires, brake linings, etc. A cost penalty, as well
as some degradation in reliability, results from the addition of a load-sensing
or a deceleration sensing proportioning valve to adjust the rear braking pressure
as a function of the front braking pressure.
Although not a functional element, the subjective reaction of drivers
to the perception of force required versus pedal travel, is limited by the need
to provide enough fluid displacement to place the brake shoes/pads in contact
with the drum/rotor. Pedal travel is dependent on the compliance of a brake and
is also affected by any air entrained in the brake hydraulic fluid. The above factors
determine the stroke of the master cylinder and therefore the stroke brake pedal.
Geometry of the pedal/master cylinder combination can be varied within limits,
but for ease of use, excessive travel of the pedal cannot be accommodated. Further,
to develop the pedal force required to produce a pressure adequate to stop the
vehicle under worst-case conditions, there are other constraints based on physical
limitations of the operator.
The present invention is directed to a substantially improved method
of regulating the braking effort at the individual wheels, which obviates most
or all of the deficiencies noted above.
Specifically, the invention describes a method of regulating the
rate of change of velocity (deceleration) of each wheel, based on a deceleration
command generated by the operator. The invention envisions an electrically controlled
braking system, where the input is an electrical signal derived by any of several
sensing processes. The input signal is then examined by a microcontroller, and
the appropriate braking activity at each wheel produced.
It is an object of the present invention to provide a braking system
in which an allowance is made for differing wheel speeds when negotiating a bend
or a corner. A further object of the present invention is to provide a braking
system which accommodates the use of different sized tires such as resulting from
the use of a temporary spare tire, the use of a severely deflated tire, or the
replacement of a previously worn tire with a new one. A further object of the
present invention is to provide for antilock braking control with no additional
hardware and as such, to provide a cost effective braking system.
Accordingly, the invention comprises: a method of controlling the
braking behavior of various wheels of a vehicle during a normal braking mode of
operation and during an antilock braking mode of operation. The method uses a
control unit or microcontroller which calculates various parameters over a known
time or sampling increment. The method including the steps of:
- 1.1 obtaining a value of the actual rotational velocity of each wheel (W&sub1;,W&sub2;);
- 1.2 generating a first signal indicative of braking effort P(in) or desired
- 1.3 generating a per wheel velocity command signal the slope or deceleration
of which is proportional to the desired vehicle deceleration p(in) and a multiplicative
scale factor (ABSGi+Gi) for adjusting such deceleration during
the normal braking mode and antilock braking mode,
- 1.4 generating a per wheel error signal Ei as a difference between the wheel
velocity command and actual rotational velocity of each wheel,
- 1.5 operating upon the error signal Ei to generate a brake activity command
- 1.6 regulating brake force in response to the brake activity command signal
Many other objects and purposes of the invention will be clear from
the following detailed description of the drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
In the drawings:
DETAILED DESCRIPTION OF THE DRAWINGS
- FIGURE 1 diagramatically illustrates a brake control system incorporating the
- FIGURE 2 diagramatically illustrates a four wheel vehicle.
- FIGURE 3 illustrates a deceleration command curve.
- FIGURE 4 shows a time history of wheel velocity versus time.
As mentioned above, the present invention utilizes an input signal
which is examined by a microcontroller to generate a signal indicative of braking
activity. Reference is briefly made to FIGURE 1 which diagramatically illustrates
the braking system 10 incorporating the teachings of the present invention. There
is shown a pedal force sensor 12 for measuring the operator applied braking force
as the operator presses upon a brake pedal 14. The output of the sensor 12 diagramatically
shown on line 16 is a pedal force input signal p(in). As can be appreciated, the
units of the p(in) signal may be in volts per unit of applied braking force. This
input signal is received by the microcontroller 20 which also receives as input
signals additional system parameters and constants to generate a plurality of
braking activity signals generally shown as Bi. In the present invention
the pedal force input signal is sealed within the microcontroller 20 to yield a
nominal deceleration command signal which is also referred to as p(in). This deceleration
command signal is appropriately adjusted for various scale factors to generate
a plurality of wheel deceleration command signals as described below.
If, for example, the brake mechanism 30 is a hydraulic brake, the
brake activity signal Bi may be used to control the level of hydraulic
pressure within the brake 30. FIGURE 1 diagramatically illustrates one method
of achieving this control wherein the actual brake pressure is measured by a pressure
sensor 32, the output of which is combined with the brake activity signal such
as B&sub1; which would be a pressure command signal, to generate a pressure error
signal which is communicated to a motor 34, which in turn rotates a pump 36 to
pressurize the brake line. Pressure decay within the brake 30 may be achieved by
opening an electrically responsive valve 38 in response to a valve activation
signal Vi (i=1,2--n) to return brake fluid to the sump 40 of the pump 36. The
brake 30 operates directly on a wheel 42, wheel speed Wi is sensed by
a wheel speed sensor 44, the output of which is communicated to the microcontroller
20. FIGURE 1 also illustrates another method of brake activation utilizing an
electrically controlled brake 30 wherein the brake activation signal, such as for
example, B&sub2; is combined with a sensor 50 to generate an error signal causing
a motor 52 to move a jack screw 54 thereby urging the friction material against
the disk or drum. The applied braking force on the brake 30 is sensed by sensor
50 which can be implemented in many ways such as a torque sensor, a position sensor
measuring the number of turns of the motor, or a current sensor measuring motor
current which is indicative of developed motor torque. As before, a wheel speed
sensor such as 44 is used to generate a signal indicative of the speed of the wheel
42 which is communicated to the microcontroller 20.
In view of the above, the invention contemplates obtaining a value
indicative of the actual rotational velocity of each wheel of a vehicle. This is
accomplished as the microcontroller 20 interrogates the various wheel speed sensor
input signals Wi.
Reference is briefly made to FIGURE 2 which diagramatically illustrates
a vehicle having four wheels. As will be seen from the description below, the
present invention operates on sets of wheel speed information. In the preferred
embodiment of the invention, the microcontroller 20 defines the front wheel rotational
velocities (WFR, WFL) as a first set of input parameters
and the rear wheel velocities (WRR, WRL) as a second independent
set of input parameters. It should be appreciated that the sets of input parameters
can also be defined in a split manner, that is, one set utilizing the front left
wheel velocity WRR
and rear right velocity WFL, while the other
set using the front right velocity WFR and rear left velocity WRL.
The microcontroller operates on each set of velocities identically. In the preferred
embodiment of the invention, the wheel velocities for an exemplary set of such
wheel velocities will be described as W&sub1; and W&sub2;.
The microcontroller identifies a command wheel velocity signal for
each wheel in the set. If, for example, the microcontroller is implemented to keep
track of the time variable t, each wheel velocity command signal Wci
would be defined by equation 1.
Wci (t) is the wheel velocity signal at time t for
the ith wheel, i = 1,2,
Wi(t) is the actual rotational velocity of the ith
p(in) is a parameter or signal indicative of braking effort
scaled approximately as a deceleration. This parameter may be viewed as a nominal
deceleration command signal.
The combination scale factor (ABSGi+Gi) scales
the nominal deceleration command signal to achieve a per wheel deceleration command
p(in)(ABSGi+Gi) for each wheel. ABSGi is a variable
gain factor increment used to adjust the rate of change or deceleration of the
wheel velocity command signal during the antilock braking mode of operation. The
parameter ABSGi is defined as zero when the braking system 10 is in
its normal braking mode of operation. The parameter Gi is an additional
gain factor used to adjust the rate of change or deceleration of the vehicle command
signal during the normal braking mode of operation (straight ahead braking or
braking in a turn).
In the preferred embodiment of the invention the microcontroller
20 is a sampled data device which samples various input parameters at sampling
increments n, n+1, n+2, etc. and generates a corresponding plurality of signals.
The wheel velocity command signals shown in FIGURE 1 can be implemented in such
microcontroller 20 by utilizing equations 2 and 3.
Wci(n)=+p(in)*(ABSGi+Gi)+ Wci(n-1)   (3)
With regard to equation 2, the variable Wci(0) is
the initialized value of the ith wheel velocity command signal which is initialized
to the value of the initial measured velocity of the ith wheel, i.e., Wi(0)
each time the pedal is depressed. Thereafter, the wheel velocity command signal
at any sample period n is given by equation 3 which may be generated within a
register within the micro- controller 20 in a known manner. Thereafter and as diagramatically
illustrated in FIGURE 1, an error signal E is generated for each wheel (i=1,2)
for each set of wheel velocity parameters in accordance with equation 4.
Ei(n)=Wci(n) - Wi(n)   (4)
A braking activity signal Bi as mentioned above, is generated for each
wheel. This signal may be thought of as a brake force command, brake torque command,
pressure command, etc. Some appropriate scaling may be required. In the preferred
embodiment of the invention, the braking activity signal Bi is obtained
by operating on the error signal associated with each wheel, by a proportional,
integral, differential controller (PID) which is diagramatically shown by equation
Bi=[PID CONTROLLER] x Ei(n)   (5)
More specifically, the actual brake activity signal Bi
as a result of
the use of the PID controller is implemented utilizing the scheme shown in equations
6a and 6b wherein PK, DK and IK are constants of proportionality respectively
associated with a term proportional to the error signal, its derivative and summation
or integral value.
Bi=Ei(n)*PK + [Ei(n)-Ei(n-1)]*DK
+ Ei(n)*IK   (6a)
Bi=[Wci(n)-Wi(n)]*PK + [Wci(n-1)
- Wi(n-1)]*DK + [E(n)]*IK   (6b)
Reference is made to FIGURE 3 which shows a stop action moment
in time and illustrates a situation where the wheel velocities of one of the sets
of wheels (W&sub1;, W&sub2;) are different. This figure will be useful in illustrating
how the gain factor Gi is obtained for straight line braking. It also
illustrates how the present invention allows for the differing wheel speeds resulting
from different sized tires and will also be useful in illustrating in how the
present invention allows for differing wheel speeds when negotiating a bend or
corner during normal braking operation. For whatever reason, at time T0, the wheel
speeds W&sub1; and W&sub2; are of differing rates. The present invention envisions
that the microcontroller 20, upon receipt of a brake activation signal (at time
T0), will interrogate the wheel speed velocities W&sub1; and W&sub2;. Thereafter
the microcontroller will generate the gain factor Gi associated with
each wheel to appropriately adjust the deceleration (slope of the wheel velocity
command) of each wheel such that each wheel will reach zero speed simultaneously.
Returning to equation 3, and assuming for the moment that the gain factor ABSGi
is zero, that is none of the wheels are in the antilock mode of operation, and
recalling that the command signal p(in) is a braking effort signal scaled in terms
of deceleration which would yield a desired nominal deceleration of the vehicle.
In view of the above the microcontroller 20 will extrapolate the wheel speeds
from the initial stored velocities Wi(0) to zero speed in the following
manner. With regard to the wheel exhibiting the slowest rotational wheel velocity
at time T0 the microcontroller 20 will extrapolate or decrement the initial wheel
speed velocity of the slowest wheel W&sub1;(0) to zero at a rate which is generally
equal to p(in) x G1 wherein G1 for the slowest wheel is equal to a constant K1.
In the preferred embodiment of the invention the constant K1 is equal to 1. This
extrapolation to zero at the above constant deceleration rate of p(in)xG1 is shown
in FIGURE 3. Thereafter the microcontroller determines a preferred deceleration
rate for the higher rotating wheel, i.e., W&sub2;, such that its velocity would
reach zero speed simultaneously with that of the extrapolated value for W&sub1;.
The adjusted deceleration rate for the higher speed wheel W&sub2; is also diagramatically
shown in FIGURE 3 as p(in) G&sub2;. The above adjustment in the deceleration can
be accomplished in a straightforward manner by permitting the gain factor G2 to
be equal to
These gain factors are recalculated each time the pedal is depressed. In view of
the above, the microcontroller has now established the command rate at which any
particular wheel should decelerate during the normal braking mode of operation,
this value being equal to p(in)*Gi.
Consider the dynamics of braking which occurs in a conventional hydraulically
braked vehicle as the vehicle is negotiating a turn. When a vehicle is in a turn,
the outside wheel, for example the left front wheel, of FIGURE 2, will rotate faster
than the right front wheel. In addition because of the vehicle's weight transfer
during the turn, the left front brake will typically be capable of generating a
greater braking force at the tire/road interface than will the right front brake.
This imbalance in braking forces may ultimately tend to destabilize the vehicle
during an aggressive turn or accident avoidance maneuver. It is a design goal
to achieve a brake system that yields a balance between the right and left hand
braking forces while the vehicle is in a turn. This goal, however, is not often
realized. This deficiency is addressed and solved by the present invention in the
following manner. The microcontroller 20 first determines an index relating to
the brake balance between for example the front left and front right wheels. If
this index is greater than a threshold value, corrective action is taken to correct
the brake balance. This is accomplished by determining an index value which is
equal to the absolute difference between the right and left side brake activity
commands in accordance with equation 8 below.
B&sub1; - B&sub2; > BT   (8)
wherein B&sub1; is the brake activity signal associated with one wheel, B&sub2;
the brake activity associated with a second wheel and BT a threshold
value. While the brake activity command signals Bi have been used in
equation 8, it should be apparent that a measure of the brake activity, brake
force, brake application, etc., can also be achieved by measuring the actual developed
pressure in a hydraulic system or alternatively if an electrically braked system
were used, the measurement of motor current, brake torque, etc., alternatively
the position of the jack screw 34, can be used.
If the absolute value of the differences in the brake activity signals
as defined in equation 8 is less than the threshold value BT, then one
can assume the wheels are essentially operating in a balanced brake mode of operation.
Further, if one is attempting to maintain a left/right brake balance in the vehicle
during a turning maneuver, then one needs information to indicate to the microcontroller
20 that the vehicle has in fact begun such a turning maneuver. It is not desirable
to monitor the position of the steering wheel or the tires themselves since this
requires additional sensors and electronics, resulting in an increased cost of
the system. In the present invention, however, the initiation of a turn is obtained
implicitly by monitoring the brake activity command signals Bi (or alternatively
any of the above mentioned feedback signals: pressure, position, torque, current,
etc.). If the brake activity command such as B&sub1; of one wheel is increasing
and if the brake activity command of the other wheel such as B&sub2; is decreasing
as it normally will during a turning maneuver, then the gain factor Gi
associated with each wheel will be modified to adjust the commanded deceleration
(p(in)Gi) of that wheel. More specifically, the gain factor Gi
associated with the wheel having a decreasing brake actuator command Bi
is incremented while the gain factor associated with the wheel exhibiting an increasing
brake activity command is decremented. This process is shown below in equation
Gi(n)=Gi(n-1)*K + Gi(n-1)   (9)
wherein Gi(n) is the current gain factor, Gi(n-1) is the
old value of the gain factor and K is a constant of proportionality or an incremental
index which will be a positive and negative value as the case may be. If K is
a positive value then the appropriate gain factor Gi will increase or
be incremented and if K is a negative quantity then the gain factor Gi
will be decremented. As can be appreciated, by modifying the gain factor Gi
of the appropriate wheel, its desired rate of change or deceleration p(in)*Gi
will appropriately increase or decrease, thereby resulting in a relative increase
or decrease in the brake activity signal which will ultimately yield a more balanced
braking condition while negotiating a turn.
Reference is made to FIGURE 4 which illustrates the wheel velocity
command signal WC1 (in solid line) and actual wheel speed (in fragmented line).
As can be seen from this FIGURE, the slope or commanded deceleration of the wheel
velocity reference signal is p(in)G1. During normal brake operation it is anticipated
that the actual wheel speed will only deviate marginally from the commanded velocity.
Having determined the gain factor indices Gi which essentially
define the desired or recommended rate of change or decleration of each wheel during
straightline braking and during a turning maneuver, it is now desirable to determine
whether or not any particular wheel should be under antilock control. This is
done by determining whether the deceleration ai of any wheel has exceeded
its desired commanded deceleration by a given increment or threshold ABSTi
as shown in equation 10.
Ai>Gi*p(in)*k&sub2; = GixK&sub2; = ABSTi   (10)
wherein k&sub2; is a scale factor greater than 1 and K&sub2; equals the product
of the brake effort or deceleration command p(in) x constant k&sub2;. k&sub2; is
typically in the range of 1.05 to 1.1.
In a microcontroller which calculates the various parameters over
a known sampling increment, a value of the actual wheel acceleration ai
can be obtained from a comparison of wheel velocity signals as generated by the
wheel speed sensor 44 as shown in equation 11.
ai=Wi(n) - Wi(n-1)   (11
Reference is again made to FIGURE 4 which shows one of the wheels entering a skid
condition. On inspection of the wheel velocity curve W&sub1;, one can see that
the actual deceleration of this curve exceeds the desired deceleration p(in)*G&sub1;
by an amount for the purpose of illustration, which is greater than the deceleration
threshold value ABST&sub1;. This situation is indicative of the fact that the
commanded brake activity command for this particular wheel is greater than the
forces which can be generated at the tire/road interface causing an increased
deceleration (negative acceleration) of the wheel in excess of the desired or commanded
value. In this condition the wheel is tending toward lock-up. To bring this wheel
under control it is desirable to decrease the brake actuation or activity force.
This is done by commanding a less aggressive deceleration command for that particular
wheel. This is accomplished in concert with equation 3 by modifying the gain factor
ABSGi to decrease the slope of the commanded deceleration. This is achieved
by generating a recurring index within the microcontroller in accordance with
equation 12 as shown below:
ABSGi(n) = ai + ABSGi(n-1)   (12)
wherein ABSGi(n) is the new value of this parameter and ABSGi(n-1)
is the prior or old value of this parameter. As can be seen from FIGURE 4 and equation
12 with the wheel decelerating rapidly, the new value of the gain factor ABSGi(n)
will be a quantity which when added to the gain factor term Gi will
decrease the slope of the commanded deceleration as shown in the interval subsequent
to time t2 or sampling event n2. As such it can be seen that the commanded change
in velocity of the wheel or its commanded deceleration has decreased (the slope
of the curve has decreased) which in turn will cause for a smaller brake activity
command Bi or reduced the brake activity or brake force applied to
the brake 30. Subsequently, the wheel will begin to accelerate at some point in
time such as t3 after the braking forces on the wheel have been sufficiently reduced.
When the indexed value ABSGi(n) is again zero the normal braking mode
of operation is entered and the factor ABSGi is maintained at zero
until it is determined that the actual wheel deceleration exceeds the threshold
value as described above whereupon the antilock lock mode of operations is again
As can be seen, the antilock behavior of any particular wheel has
been altered by modifying the reference deceleration command to a value less than
that which would be commanded by the operator. Subsequently, the brake activity
force is modified to allow the previously skidding wheel to reaccelerate to a
speed nearly synchronous with that of the vehicle speed.
In view of the above, it can bee seen that the present invention
provides for an automatic proportioning, that is each wheel will automatically
do its share of braking. The proportioning achieved by the present invention is
idependent of vehicle loading or load distribution. Control of any of the wheels
in an antilock braking mode of application is implemented with minimal additional
Many changes and modifications in the above described embodiment
of the invention can, of course, be carried out without departing from the scope
thereof. Accordingly, that scope is intended to be limited only by the scope of
the appended claims.