FR2698332A1 - Braking control method for road vehicle - applying control signals using pressure sensors, and detection of sliding of wheel to determine adjustment required to create optimum braking pressure - Google Patents

Abstract

The method is applied to a vehicle comprising means (11) for developing intermediate braking pressures (pi) from signals received from sensors (13, 16, 16 ', 17, 18, 19, 20). For each wheel, a signal representative of the actual slip rate (Sxi) is produced, which is compared with a signal representative of a setpoint slip rate (Sxc i) to deliver a deviation signal (epsiloni) which , after processing, corrects the intermediate signals (pi) to derive control signals (pi) for brake actuators, so that the actual slip of each wheel (Sxi) is substantially equal to the setpoint slip (Sxc i) of said wheel. According to the invention, the quantities (Sxi), (Sxc i) and (epsiloni) are sampled and they are processed in digital filtering blocks (R), (T), (S), respectively, established using a dynamic behavior model of each wheel during braking, and to take into account the pure delay of the actuator.

Description

 The invention relates to a braking deceleration control method of a road vehicle, where each wheel is equipped with a hydraulic pressure brake actuator and electrically controlled to certain signals emitted by pressure sensors and / or movement of the brake pedal to develop brake setpoint pressures. Such a system is described in document FR-A-2,620,097.

 During braking, it turns out that one can not simultaneously optimize the longitudinal deceleration and the transverse deceleration of the vehicle, as is clear from the variations in the longitudinal braking force, which goes through a maximum for a a significant value of the sliding of the wheel, while the transverse braking force only decreases as soon as the appearance of this same sliding.

For the record, it is recalled that the pseudo-slip or "slip" (Sx) is the difference, brought back to the speed of the vehicle (V), between the vehicle speed (V) and the tangential velocity (Vr) of the wheel, c that is to say,
Sx = (VV,) / V
To achieve the best braking conditions while respecting a good stability of the vehicle, it was proposed in the French patent application No. 92 11029 filed September 16, 1992 by the applicant, a control method according to which one develops, for each wheel , a signal representative of the actual sliding ratio (Sx), this signal is compared with a signal representative of a set slip ratio (Sxc) in order to deliver a difference signal which, after processing, provides a correction of the electric brake actuator control such that the actual slip of each wheel (Sx) is substantially equal to the target slip (Sxc) of said wheel.On choosing, or by determining a target slip (Sx,) according to a strategy braking, we then achieve an effective and optimal compromise depending on the type of braking between the longitudinal and transverse braking forces, ensuring both a value as reduced As far as possible from the braking distance, all of course without locking the wheels, and maintaining a good stability of the vehicle. According to a preferred form of implementation of the method described in this patent application, the effective braking pressure signals result from the product of the intermediate setpoint pressure signals by a factor which is a function of the actual slip (Sx) and the slip of setpoint (Sx,), the difference signal between these two slides subjected to a type of treatment
PID.

 Although the control method described in the aforementioned patent application generally gives satisfaction, it was realized that the PID processing of the deviation signal does not allow to take into account the pure delay of the brake actuator, c that is, the time between the application of a step control signal and the response of the actuator to that control step. Thus, in the context of PID regulation, the pure delay must be less than the sampling period of the calculator, since the PID can only process the systems having a pure delay less than the sampling period. . In addition, if the pure delay is greater than one quarter of the time constant of the device to be controlled, the PID processing results in slower closed-loop responses than in open-loop.

 A certain instability of the PID control of the deviation signal has also been noted in the unstable zone of the longitudinal / sliding force characteristic of the tire fitted to a wheel, that is to say in the zone at beyond the maximum of the longitudinal force exerted by the tire on the road, in response to the torque applied by the brake actuator to this wheel, maximum currently reached for a pseudo-slip of 5 to 15 *.

 The object of the present invention is therefore to provide a method for controlling the deceleration of a vehicle which makes it possible to achieve a value as small as possible of the braking distance, without locking the wheels, taking into account the pure delay of the vehicle. a wheel brake actuator and maintaining the stability of the control of this actuator including in the instability zone of the characteristic of the tire equipping the braked wheel.

 Another object of the present invention is to provide such a method which allows easier control of the control than that of a process involving PID processing of the deviation signal.

 The present invention also aims to provide a vehicle implementing the method according to the invention.

 These aims of the invention, as well as others which will become apparent in the following description, will be achieved with a braking deceleration control method of a road vehicle, in which each wheel is equipped with an actuator. electrically actuated hydraulic pressure brake servo-controlled to signals such as those emitted by brake fluid pressure sensors and / or brake pedal displacement, for developing intermediate braking pressures (Pi), according to which method for each wheel, a signal representative of the actual sliding ratio (Sx), which is compared to a signal representative of a set slip ratio (Sxci), is produced for outputting a difference signal (Ei) which, after processing, corrects the intermediate signals (Pi) to derive control signals (Pi) from brake actuators, so that the actual slip of each wheel (Sxi) is substantially equal to the slip of reference (Sxci) of said wheel, the method being remarkable in that the quantities (Sx), (Sxj) and (Ei) are sampled and processed in digital filter blocks (R), (T), ( S), respectively, established using a model of dynamic behavior of the wheel in the braking phase.

 By adopting a control of the "RST" type, it is possible, according to an advantageous characteristic of the method according to the invention, to adapt the functions of the digital filtering blocks to the ratio of the pure delay of the brake actuator associated with each wheel, at the sampling period. This avoids one of the disadvantages of a PID control when this pure delay can not be neglected.

The adoption of an "R-S-T" type control also makes it possible to improve the stability of the control in the zone of instability of the longitudinal force / slip characteristic of the tire equipping the wheel concerned. It still allows to take into account the bandwidth of the actuator concerned.

où
A est un gain reliant le couple de freinage à la pression de freinage,
I, l'inertie de la roue considérée, r, le rayon de la roue, t,, le retard pur de l'actionneur, t, la constante de temps estimée de l'actionneur,
K, la pente de la courbe caractéristique du pneumatique, s, la variable de Laplace.According to another characteristic of the present invention, in order to establish the functions of the digital filtering blocks R, S, T, a model of dynamic behavior of the wheel defined by a transfer function is used.
H (s) as

or
A is a gain connecting the braking torque to the braking pressure,
I, the inertia of the wheel considered, r, the radius of the wheel, t ,, the pure delay of the actuator, t, the estimated time constant of the actuator,
K, the slope of the characteristic curve of the tire, s, the Laplace variable.

 The present invention also provides a road vehicle, of the wheel type each equipped with a hydraulic pressure brake actuator and electrically controlled, comprising means for generating a signal representative of a braking force required by the driver. and linked to a depressing position of a brake pedal, means for evaluating braking forces on each wheel, a sensor for measuring the deceleration of the vehicle, means for processing said signals to form signaling signals. intermediate control (Pi)) brake actuators specific to each wheel, means for determining the slip of each wheel, means for producing, for each wheel (i), a difference signal (Ei) between the actual slip (Sxi) and a set slip value (SXci), means for processing said difference signal (fi) to form intermediate brake control signal correction signals (Pi) so as to establishing effective control signals (Pi) of the brake actuators, and digital filtering means established using a model of dynamic behavior of each wheel in the braking phase, these filtering means respectively processing the slips ( Sx), (Sxci)) and the difference (Ei) so as to form multiplicative correction signals (y) of the signals (Pi) for establishing the effective control signals (pi) of the brake actuators.

Other characteristics and advantages of the present invention will appear on reading the description which follows and on examining the appended drawing in which
FIG. 1 is a partial schematic view of a road vehicle implementing the deceleration control method according to the invention,
FIG. 2 is a functional diagram of a device for implementing the method according to the invention.

 The method according to the invention is applied to a passenger vehicle shown schematically in FIG. 1, each front wheel 1,2 and rear 3,4 of which is provided with a brake 23,23 ', 24,24' respectively, comprising a receiver cylinder in which a piston moves which applies the brake linings with the necessary force against a disk 10, 10 ', 11, 11' respectively, rotating with the wheel.

 At the brake pedal 12 is associated a measurement sensor 13 which transmits a signal representative of the pedal stroke and / or the force applied to the pedal by the driver of the vehicle.

 A control device 5 comprising a calculating section and a power section takes into account the signal delivered by the sensor 13 to develop magnitudes (voltage, current or frequency) for controlling AC or DC electric actuators 25, 25 15,15 'for adjusting or adjusting the pressures of a brake fluid in the brake cylinders 23,23', 24,24 '.

 The device 5 also takes into account signals correlated with the braking force of each wheel delivered, for example, by sensors 17 to 20 associating strain gauges with a deforming membrane, constituting braking pressure sensors. . Similarly, sensors 17 'to 20', each associated with a wheel, deliver signals representative of wheel speeds.

 The sensors 17, 18, 19, 20 thus deliver control signal generation signals, the deviations of which from setpoint values serve to establish the control signal of the actuators, so as to enslave the effort of braking each wheel to a set value stored in the control device 5.

 In addition, the device 5 receives a signal from an accelerometric sensor 16, representative of the actual deceleration of the vehicle, and a signal from a position sensor 16 'measuring the travel between the rear axle and the body .

 The determination of the reference speed V of the vehicle is made from the four wheel speeds, according to a conventional method.

In the functional diagram of FIG. 2, it appears that the device 5 for implementing the control method according to the invention comprises a block 11 developing four intermediate setpoint pressures (Pi), i varying from 1 to 4 to identify the corresponding wheel, according to the method described in the patent application No. 9115491 from brake fluid pressure signals provided by sensors 17,18,19,20, the degree of depression of the pedal 12, issued by the sensor 13, the displacement measured by the sensor 16 'and a signal from the sensor 16 representative of the actual deceleration of the vehicle.Une block 21 generates effective control signals (Pi) of the four brakes by producing the product of the pressures of intermediary setpoints (Pi) by a weighting factor y, resulting from the output of a block 5 which will be described later, the signals (Pi) then taking the form
Pi = Pi
A block 31 calculates the slip ratios Sxi = (V-Vri) / V of each of the four wheels i (i = 1 to 4) from the information from the sensors 17 'to 20' measuring the tangential speeds (Vri) of the wheels and the speed (V) of the vehicle, which can also be established using the signals delivered by these four sensors.

 According to the present invention, slip rates or "slip" are sampled and are operated in a digital control system known to the skilled as "R-S-T". Reference can be made to the book entitled "Identification and control systems" author Ioan Doré Landau, Hermes Edition, Paris 1988, pages 100 to 111, for the description and calculation of such a system by the method of "placement of poles. "

The actual slip rate (Sx) of a wheel feeds an R block that performs a digital filtering function to calculate a magnitude Sxri such that
Sxri = king # Sxi (t) + r1i # Sxi (t-Te) + r2i # Sxi (t-2Te) where Te is the measurement sampling period, t the current time. In the "automotive" application envisaged here, Te may be of the order of 10 ms, for example.

The set slip rate (Sxci) of a wheel feeds a block T. This block executes a numerical filtering function to calculate a magnitude (SXcti) such that Sxcti = you # Sxci (t) + t1i # Sxci (t- Te) + t2i # Sxci (t-2Te)
The difference (Ei) between the slip rate filtered by the block R (suri) and the set slip rate filtered by the block T (Sxcti) leaves a comparator 41.

A block S performs a digital filter function to calculate the weighting factor (Yi), from the difference (#i), according to the following law:
yi (t) = - (s1i-1) #yi (t-Te)
- (s2i-s1i) #yi (t-2Te)
- (s3i-s2i) #yi (t-3Te)
- (O-s3i) #yi (t-4Te)
+ E1 (t)
It will be noted that the R and S blocks define the control dynamics of the device while the T block defines the tracking dynamics of the sliding setpoint. Block S, in particular, acts as a high-frequency signal integrator or, according to another analogy, as a low-pass filter.

According to the invention, the coefficients r1, r1i and r2i introduced into the block R, the coefficients t'1, t1i and t2i introduced in the block T as well as the coefficients following s2i and s3i introduced in the block S are calculated, for each wheel, from a model of dynamic behavior of the wheel in braking phase, which takes into account the bandwidth of the actuator and its pure delay. This dynamic model depends on the physical characteristics of the vehicle which are
- the inertia I of the wheel,
- the speed V of the vehicle,
the radius r of the wheel,
the cutoff frequency f @ of the actuator of
braking,
the pure delay tr of the actuator,
- the estimated slope K of the characteristic curve of the
pneumatic (which represents the longitudinal force F1
at the pneumatic / ground contact, depending on the slip
(Sx)), slope taken at optimum operating point
longed for,
the gain A connecting the braking torque to the pressure
setpoint brake Pi.

expression qui prend la forme d'une fonction de transfert
H(s), dans laquelle s est la variable de Laplace et z la constante de temps estimée de l'actionneur, soit : z = l/2itfa
Ce modèle dynamique reliant le pseudo-glissement Sxi de la roue (i = 1 à 4), au coefficient pondérateur y, est ensuite numérisé à la période d'échantillonnage Te, qui est la période d'acquisition des données, ce qui conduit à un modèle numérique reliant la séquence d'acquisition {Sxi} des valeurs du pseudo-glissement de la roue à la séquence {yi} des coefficients pondérateurs.On pourra se reporter à cet égard aux méthodes de discrétisation-numérisation décrites dans l'ouvrage intitulé : "Automatique des systèmes linéaires" (Partie 1 - Signaux et systèmes) auteurs
Philippe de Larminat et Yves Thomas, éditions Flammarion
Sciences (Paris) pages 296-305 et, plus particulièrement, 303 et suivantes.This dynamic model expresses the relation between the real sliding rate Sxi of a wheel and the weighting factor yi of the intermediate setpoint pressure pi according to the following relation

expression that takes the form of a transfer function
H (s), where s is the Laplace variable and z is the estimated time constant of the actuator, ie: z = l / 2itfa
This dynamic model connecting the pseudo-slip Sxi of the wheel (i = 1 to 4), to the weighting coefficient y, is then digitized at the sampling period Te, which is the period of data acquisition, which leads to a numerical model linking the acquisition sequence {Sxi} of the pseudo-slip values of the wheel to the sequence {yi} of the weighting coefficients. In this respect, reference may be made to the discretization-digitization methods described in the book entitled : "Automatic linear systems" (Part 1 - Signals and systems)
Philippe de Larminat and Yves Thomas, Flammarion editions
Sciences (Paris) pages 296-305 and, more particularly, 303 et seq.

In the present application, the digitization of the transfer function H (s), provides a digitized or digital transfer function Hn (z-1) connecting the acquisition sequence of the measured slips {Sxi} to the sequence {yi}, following the following expression
Pi [b1 * # z-1 + b2 * # z-2] {Sxi}
Hnz-1 = =
1 + a1 * # z-1 + a2 * # z-2 {yi} where z-1 is the delay operator, b1, b * 2, a * 1 and has 2 coefficients determined as follows.

A number of intermediate variables are defined that are
## EQU1 ##
t # I # V t # I # V t I # V
@ ~ -Te @ ~ r2 # K # Te
T t
# 1 # e # 2- # 2 # e # 1 e # 1-e # 2
α =, ss =
# 1- # 2 # 1- # 2
d1 = a
d2 = p.Te
d3 = -a1ssTe
d4 = a - a2pTe
b1 = b / a1 (- α; + 1), b2 = b # ss # Te
The mathematical expressions for calculating the coefficients b * 1, b * 2, a1 * and a * 2 of the digitized transfer function Hn (z-1) are as follows
b1 * = b1
b * 2 = b2.d2 - b1.d4
a * 1 = - d1 - d4
a2 * = d1.d4 - d3.d2
As indicated in the aforementioned Landau work, the digital transfer function can be noted
(Z - @) # B (z @)
Hn (z-1) =
A (z-1)
tr where d = ~, ratio of the pure delay of the actuator to the
The sampling period, rounded to the nearest integer part.

B (z-1) = (b1 * # z-1 + b2 * # z-2) #pi,
A (z-1) = 1 + a1 * # z-1 + a2 * # z-2,
The coefficients r0, r1i and r2i introduced into the block
R, the coefficients you, t1i and t2i introduced into the block T as well as the coefficients s1i, s2i and s3i introduced into the block S are calculated for each wheel (i) from the digital transfer function Hn (z-1) , according to a procedure of the type "placement of poles" described in the book of
Landau, for example.

P1 = - 2.kl.k2 p2 = k12 où k1 = exp(-###n)#Te

où Un est choisi tel que 0,25 < #n#Te < 1,5 et # est choisi tel que 0,7 < # < 1
Ces choix correspondent au respect du théorème de
Shannon, quant à l'échantillonnage des signaux à traiter.As an example of implementation of the present invention, the coefficients r1, r1i, r2i, you, t1i, t2i, s1i, s2i and s3i are determined in the case where d = 1, that is to say in the case where the pure delay of the braking actuator is equal to Te, sampling period and in the presence of a digital integrator in the filter block '5, so as to eliminate static errors in the control phase. We ask first
m1 = b1 * # pi
m2 = b2 * # pi
l1 = a1 * - 1
12 = a 2 - a1 *
13 = - a 2
v1 = P1 - a1 * + 1
v2 = P2 - a * 2 + a * 1
v3 = a2 *
It will be noted that p1 and p2 characterize the desired dynamics in closed loop, according to the work cited above. They can be calculated, for example, using the following expressions

P1 = - 2.kl.k2 p2 = k12 where k1 = exp (- ### n) #Te

where A is chosen such that 0.25 <# n # Te <1.5 and # is chosen such that 0.7 <# <1
These choices correspond to the respect of the theorem of
Shannon, as to the sampling of the signals to be processed.

These ranges indicate that the choice of the sampling frequency for the control-regulation systems is based on the desired bandwidth for the closed-loop system. Reference can be made in this regard to the above-mentioned work by Doré Doré Landau, pages 42 to 49.

one is an evolution coefficient function of the real slip Sxi of a wheel and of the setpoint slip Sxci, so as to accelerate the convergences of the sliding towards the setpoint, during adhesion transitions of the weak adhesion to high adhesion type, in regulation on stable area of the tire. #n can be calculated as follows
If Sxi # Sxci,
(#n max- # n min) (#n max- # n min) #n = # Sxi2-2 # - # Sxi + #n max
Sxci Sxci
If Sxi> Sxci, #n = #n min
#n min represents the desired nominal dynamic which ensures, under all circumstances, a good overall stability of the looped system. xi evils represents the fastest dynamics that one wishes to give to the closed-loop system, in accordance with the range of imposed tolerances (#n max <1,5 / Te).

 Such a pace makes it possible not to disturb the regulation phases around the setpoint, and to accelerate the convergences when the slip Sxi is much lower than the setpoint Sxci (this case occurs during the adhesion transition mentioned above). ).

So as not to act in an unstable zone of the tire (where the convergence problems do not arise), we adapt #n max according to the requested set point Sxci and the optimum slip estimated on strong adhesion, which we note
Sxcpti.

The adaptation of in max is such that
#n min - #max
If Sxci # Sopti, #n max = @ Sxci + #max
Sxopti If Sxci> Sopti, #n max = #n min
#max represents the maximum natural pulsation that can be given to the closed-loop system without destabilization.

 This choice of evolution of on max makes it possible to adapt the response dynamics as a function of the regulation zone.

S3i = O
v2 l1v1 s2i
roi=
m1 m1 m1
l3m1 s2i l3
r1i=( -l2) - # v1
m2 m2 m2
-l3
r2i= # s2i
m2
toi = 1/(m1 + m2)
t1i = p1/(m1+m2)
t2i = P2/(m1+m2)
Bien entendu, l'invention n'est pas limitée au mode de réalisation décrit, qui n'a été donné qu'à titre d'exemple.The coefficients of the filtering functions executed in the R, S and T blocks can be calculated using the following expressions, in the case where tr = Te (ie d = 1) Sli = V1

S3i = O
v2 l1v1 s2i
King =
m1 m1 m1
l3m1 s2i l3
r1i = (-l2) - # v1
m2 m2 m2
-L3
r2i = # s2i
m2
you = 1 / (m1 + m2)
t1i = p1 / (m1 + m2)
t2i = P2 / (m1 + m2)
Of course, the invention is not limited to the embodiment described, which has been given by way of example.

Thus, in the case where the pure delay of the actuator could be assimilated to a multiple of the sampling time Te greater than 1, the coefficients of the functions executed in the blocks R, S and T and the number terms of these functions, could be adapted to such a pure delay, by a calculation procedure similar to that described above.

Claims (9)

 A method of controlling deceleration by braking a road vehicle, wherein each wheel (1, 2, 3, 4) is equipped with a brake actuator (25, 25 ', 15, 15') with hydraulic pressure and electrically controlled to signals such as those emitted by brake fluid pressure sensors (17,18,19,20) and / or displacement of the brake pedal (16), to develop intermediate braking pressures ( pi), according to which a signal representative of the actual sliding ratio (Sxi) is produced for each wheel (i), which is compared with a signal representative of a set slip ratio (Sxci) for outputting a difference signal (Ei) which, after processing, corrects the intermediate signals (pi) to derive control signals (Pi) from brake actuators, so that the actual slip of each wheel (Sx) is substantially equal to the set slip (Sxci) of said wheel, characterized in that sampler maps the magnitudes (Sx), (Sxi) and (Ei) and processes them in digital filter blocks (R), (T), (S), respectively, established using a dynamic behavior model of the wheel in the braking phase.  2. Method according to claim 1, characterized in that adapts the digital filter blocks (R, S, T) to the ratio (d) of the pure delay (tr) of the brake actuator associated with each wheel, at the sampling period (Te).  3. Method according to any one of claims 1 and 2, characterized in that draws the control signals (Di) braking actuators of the relationship  P = Yi x where yi is a multiplicative correction weighting factor delivered by the sampled output of the filter block (S).  4. Method according to claim 3, characterized in that, to establish the functions of the digital filter blocks (R, S, T), a dynamic behavior model defined by a transfer function H (s) such that

 or A is a gain connecting the braking torque to the braking pressure, I, the inertia of the wheel considered, r, the radius of the wheel, sorting the pure delay of the actuator, t, the estimated time constant of the actuator, K the slope of the characteristic curve of the tire, s, the Laplace variable.  5. Method according to claim 4, characterized in that the block (R) delivers a filtered real slip (Sxri) such that  Sxri = king # Sxi (t) + r1iSxi (t-Te) + r2i # Sxi (t-2te) where t is the current moment, Te the sampling period of the measure (Sx) of the slip, king, r1 , r2i constants.  6. Process according to claim 4, characterized in that the block T delivers a filtered set point slip (Sxcti) such that:  Sxcti = you # Sxci (t) + t1i # Sxci (t-Te) + t2i # Sxci (t-2Te) where t is the current moment, Te the sampling period, t ,,, t1i, t2i constants .  7. Process according to claim 4, characterized in that the block (S) delivers the weighting factor (yi) such that  yi (t) = #i (t) - (s1i-1) #yi (t-Te) - (S2i-S1i) #yi (t-2Te) - (S3i-S2i) #yi (t-3Te) - (O-S3i) #yi (t-4Te) t being the current moment, Te the sampling period, s1i, s2i, s3i, constants.  8. A method according to all of claims 5 to 7, characterized in that said constants are a function of a predetermined dynamic imposed on the system, itself a function of the actual slip rate (Sxi) and setpoint (Sxci).  9. Road vehicle designed for the implementation of the method according to any one of claims 1 to 8, of the wheel type (1,2,3,4) each equipped with a brake actuator (25,25 '). 15,15 ') with hydraulic pressure and electrically controlled, comprising means (13) for generating a signal representative of a braking force requested by the driver and linked to a driving position of a pedal brake means (12), evaluation means (17, 18, 19, 20) of braking forces on each wheel, a sensor for measuring the deceleration of the vehicle, means for processing (5) said signals to formulate intermediate control signals (pi) of the brake actuators (25,25 ', 15,15') specific to each wheel (1,2,3,4), means (31) for determining the slip of each wheel , means for producing (41), for each wheel, a difference signal (Ei) between the actual slip (Sx) and a set slip value (Sxci), means s of processing said deviation signal (Ei) to form correction signals (Yi)) of the intermediate braking control signals (Pi) so as to establish effective control signals (Pi) of the brake actuators (25, 25 '15, 15'), characterized in that it comprises digital filtering means (R, S, T) established using a model of dynamic behavior of each wheel in the braking phase, these means ( R, S, T) respectively processing the slips (Sxi), (Sxi) and the gap (Ei) so as to form multiplicative correction signals (yi) of the signals (Pi)) to establish the actual control signals ( Pi) brake actuators.