Algorithm of anti-lock braking system for two-axle vehicles with one driving axle with adaptive redistribution of braking forces

Cover Page


Cite item

Abstract

The main purpose of active vehicle safety systems is to prevent an emergency situation. If such a situation arises, the system independently (without the participation of the driver) assesses the probable danger and, if necessary, prevents it by actively intervening in the driving process.

One of the ways to increase the active safety of vehicles when braking is the use of anti-lock braking systems (ABS). The main problems in ensuring the operation of the ABS, built on different control principles and with different control parameters, are the impossibility of directly determining the vehicle speed and, as a result, the slip coefficient, as well as the inability to effectively respond to changing road conditions during braking. For example, when braking on a slippery supporting surface and trying to avoid an obstacle in front, there is a risk of losing traction and skidding. The algorithms of the ABS operation developed at present do not ensure the prevention of the occurrence and development of skidding under the conditions indicated above.

The aim of the work is to increase the stability and controllability of two-axle vehicles with one driving axle during braking due to the adaptive redistribution of braking forces on the wheels. An algorithm for the operation of an anti-lock braking system with adaptive redistribution of braking forces on the wheels of a vehicle is proposed. Thanks to this algorithm, when braking on a slippery surface of a two-axle vehicle with one driving axle, the absence of wheel blocking and also skid resistance are ensured. The efficiency and effectiveness of the proposed algorithm when braking a two-axle vehicle with one driving axle on a slippery supporting surface were proved by the methods of simulation.

Full Text

Introduction

The anti-lock braking system (ABS) is one of the solutions to the problem of increasing vehicle active safety during braking. Recognizing this fact, the legislators of several countries are encouraging vehicle manufacturers to implement the ABS. As a result, in Russia, all M2 buses with more than 8 passenger seats are required to have an ABS (in the European Economic Community, since 2004, every new vehicle has been equipped with an ABS). Simultaneously, the algorithms for controlling ABS operations are being improved, resulting in a higher level of control over vehicle movement parameters during braking.

Based on the control parameters, the ABS is categorized by the following [1‒6]:

  • the value of the wheel slip coefficient corresponding to the maximum wheel adhesion (s-regulation);
  • the maximum interaction coefficient value (μ-regulation); and
  • the value and sign of the dμ/ds parameter, which characterizes the degree of approach to the maximum adhesion (gradient regulation).

When using s-regulation, the following basic algorithms and their combinations are typically used [7−13]:

  • the equality mode of angular wheel and linear decelerations of the vehicle;
  • the wheel slip coefficient and its further maintenance within the specified limits; and
  • the threshold deceleration of the braking wheel.

Most of the disadvantages of s-regulation are since neither wheel slip nor wheel deceleration provides sufficient information to determine the optimal braking force control. The impossibility of directly determining vehicle speed, and thus the slip coefficient, and the impossibility of effectively responding to changes in road conditions during braking are the main problems in ensuring ABS operation based on different principles and with different control parameters.

This work is aimed at increasing the stability and controllability of two-axle vehicles during braking owing to the adaptive redistribution of braking forces on wheels.

Algorithm for estimating vehicle movement parameters during braking

Wheels are known to slow down with an increased braking torque during braking. At a certain point, the wheel deceleration exceeds the value that the vehicle deceleration cannot physically exceed. As the braking torque increases, the wheel declaration (not the vehicle) also increases. The physical vehicle declaration determines wheel deceleration threshold , and can be approximately calculated as follows:

ω˙n=aOXTrs,

where aOXT is the current linear vector of acceleration projection aO of the wheel center O (Fig. 1) on the plane of its rotation; rs is the static radius of the wheel.

 

Fig. 1. Acceleration plan for the center of the wheel during curvelinear motion of vehicle

 

To determine aOXT, we consider the acceleration plan for the wheel center during curvilinear vehicle movement and assume that the rolling plane of the wheel is perpendicular to the flat support base.

The acceleration aO (Fig. 1) of point O (wheel center) during plane motion is equal to the vector sum of acceleration aC of the center of mass of the vehicle (point C) and acceleration aOC of point O during rotational motion around pole C:

aO=aC+aOC. (1)

In Figure 1, C is the center of mass of the vehicle; О is the center of the vehicle wheel; CXY represents axes of the coordinate system associated with the center of mass of the vehicle; ОXTYТ represents the coordinate system axes associated with the center of the vehicle wheel; aC represents the vector of acceleration of the vehicle mass center; aO represents the vector of acceleration of the vehicle wheel center; aOCτ represents the vector of tangential acceleration; aOCn is the vector of normal acceleration; aOXT represents the current linear vector of acceleration projection, aO of the center O of the wheel on the XТ axis; Θ is the angle of rotation of the controlled wheel; ωZ is the angular speed of the vehicle rotation about the vertical axis.

In the associated coordinate system, we take into account that the transfer velocity vector VOC of point O relative to the pole C is as follows:

VOC=ω×OC, (2)

where ω=ωx,  ωy,  ωz is the vector of the angular velocity of point O relative to point C and OC=xo,  yo,  zo is the radius vector from point O to point C in the axis of the associated coordinate system CXY projections.

Thus,

aOC=ε×OC+ω×ω×OC, ε=dωdt,  (3)

where ε is the vector of the angular acceleration of the vehicle.

It is noteworthy that the acceleration vector aOC consists of tangent and normal components:

aOCτ=ε×OC, aOCn=ω×ω×OC.  (4)

The vector of tangential acceleration aOCτ is directed perpendicular to the CO ray. The normal acceleration vector aOCn is directed from the center of the wheel О to the center of mass С of the vehicle.

Thus, the vector modulus aOXT=aOXT can be defined as follows:

aOXT=aOXcosΘ+aOYsinΘ, (5)

where aOX, aOY are the projections of the center О of the wheel acceleration vector aO on the X and Y axes of the coordinate system associated with the center of mass of the vehicle.

The intended purpose of braking torques on wheels

The braking torque МТi on the i-th wheel can be determined as follows, taking into account the ABS operation:

МТi=hbrakehABSihfbiTmax, i=1,,N, (6)

where hbrake=01 is the degree to which the driver presses the brake pedal; hABSi=01 is the reduction degree of the effective braking torque on the i-th wheel due to the ABS; hfbi=01 is the redistribution degree of the braking torque on the i-th wheel when braking on a straight line (taking into account the normal reaction redistributions between the front and rear axles); Tmax is the maximum braking torque developed by the wheel brake mechanism; N is the number of wheels on the vehicle.

The value hABSi can be defined as follows:

hABSi=ω˙порω˙iωiωmax, i=1,, N, ωmax=maxωi, i=1,, N, (7)

where ωi is the current angular speed of rotation of the i-th wheel.

The cofactor ω˙nорω˙i in Eq. (7) allows the braking torque on the i-th wheel to be reduced when its angular deceleration ω˙i exceeds the threshold value ω˙пор. Using the fastest wheel of the vehicle as a reference, cofactor 2 ωiωmax allows for an adjustment in braking torque reduction.

An adaptive algorithm for braking force redistribution on the vehicle wheels

When a vehicle brakes on a straight-line section of motion, the vehicle “bounces” forward, the rear wheels are relieved from normal loads, and the front wheels take on additional load owing to inertial forces. Therefore, the dynamic normal load R1d on the wheels of the front axle and R2d on the wheels of the rear axle can be determined as follows for a two-axle vehicle:

R1d=R1s+ΔR1, R2d=R2sΔR2, R1s=Ml1L, R2s=Ml2L,

where R1s, R2s are the normal reactions on the wheels of the front and rear axles, respectively in a static position; ΔR1, ΔR2 represent an increment of normal responses to the front and rear axles, respectively, during braking; M is the weight of the vehicle sprung parts; l1, l2 are the distances from the center of the vehicle mass to the front and rear axles, respectively; and L=l1+l2 is the vehicle wheelbase.

On the assumption that the stiffness of the suspensions of all wheels is approximately equal, the increment of normal reactions to the front and rear axles ΔR1 and ΔR2, is defined as follows:

ΔR1=MaCxghcl1l12+l22, ΔR1=MaCxghcl2l12+l22,   (8)

where aCx is the projection module of the center of mass acceleration onto the X-axis of the associated coordinate system and hc is the height of the vehicle center of mass.

We defined the value hfbi=RidRis. Finally, using Eq. (8), we obtained the following for braking a vehicle in a straight-line section of motion (Θ3°):

hfb1,3=1+aCxghcLl12+l22 − for the front axle wheels,

hfb2,4=1aCxghcLl12+l22 − for the rear axle wheels.  (9)

If Θ3°, and hfbi=1.

Testing the performance and efficiency of the ABS algorithm

Theoretical vehicle braking studies were performed using simulation mathematical modeling. The aspects of the mathematical model of motion have been considered in previous studies [14–19].

Using simulation modeling methods in testing the performance and efficiency of the proposed algorithm, it was discovered that emergency braking on a slippery road (coefficient of adhesion at full slip 0.35) of a passenger vehicle with a gross weight of 6000 kg at an initial speed of 60 km/h with a simultaneous turn of the steering wheel (the driver’s attempt to bypass the obstacle) causes front axle drift. The trajectory of the vehicle’s motion during braking is presented in Figure 2.

 

Fig. 2. Trajectory of movement of a vehicle with a gross weight of 6000 kg when braking with ABS without anti-skid function of the front axle

 

To avoid this drift in the front axle, it is required to first recognize the occurrence and development of this process. For this purpose, we used previous data [20], where a parameter δV=VC1VC2 represents the difference in the estimate of the linear velocities of the center of the vehicle mass, first using the linear speed of the center of the front axle (vector VC1), and subsequently using the linear speed of the center of the rear axle (vector VC2), as a diagnostic sign of the onset of front axle drift or rear axle skidding. Figure 3 presents a graph of the change in time of the diagnostic sign δV while the vehicle is braking.

 

Fig. 3. The graph of the change in time of the diagnostic characteristics δV when braking the vehicle

 

The graph in Figure 3 shows a diagnostic sign that appears during braking δV>0, indicating the front axle drift occurrence.

A counter-rotation moment for skid resistance at the front axle is required owing to increased braking of the rear wheel inner concerning the direction of rotation. However, because more braking can cause the wheel to become stuck, it is necessary to release the brakes of all wheels, except for the rear wheel inner, concerning the rotational direction. Thus, Eq. (6) for determining the braking torque on each wheel is as follows:

Мтi=hbrakehABSihfbihESPiTmax, i=1,,N, (10)

where hESPi=01 is the degree of reduction of the effective braking torque on the i-th wheel due to the skid resistance algorithm at the front axle during braking (anti-skid function of the front axle).

Thus, considering the rule of signs adopted in the simulation, the algorithm for determining the value hESPi, i=1,,N should be as follows.

If Θ1>0° (turn left) and δV>0 (front axle drift), then hESP1=hESP3=hESP4=1CuδV; hESP2=1.

If Θ1<0° (turn to the right) and δV>0 (front axle drift), then hESP1=hESP2=hESP3=1CuδV; hESP4=1.

In the above equations, Cu is the controller’s gain which is adjusted individually for each vehicle.

Using simulation modeling methods, the motion of a two-axle vehicle with a total mass of 6000 kg was simulated under the same conditions as described earlier to access the efficiency and performance of the proposed ABS operation during braking. Figure 4 presents the trajectory of the vehicle when braking with the ABS and the anti-skid function of the front axle, Figure 5 demonstrates the dependence of vehicle speed on time, Figure 6 presents graphs of changes in the angular velocities of the wheels on time, and Figure 7 presents a graph of the change in time of the diagnostic sign δV during braking.

 

Fig. 4. Vehicle trajectory when braking with ABS and anti-skid function of the front axle

 

Fig. 5. Dependence of vehicle speed on time

 

Fig. 6. Graphs of changes in angular speeds of wheels from time to time

 

Fig. 7. The graph of the change in time of the diagnostic characteristics δV when braking a vehicle with ABS and with the function of countering the drift of the front axle

 

Figures 4 to 7 illustrate that when braking with ABS and the anti-skid function of the front axle, the wheels do not lock, and the maximum value of the diagnostic sign δV decreases by 40%, indicating that the proposed algorithm for the operation of an ABS with an anti-skid function of the front axle is operable and efficient.

Conclusions

When braking a vehicle on a slippery supporting surface with a simultaneous steering wheel rotation, an algorithm for the operation of an ABS with an anti-skid function of the front axle for two-axle vehicles is proposed and characterized by not only the absence of wheel blocking but also an increase in vehicle controllability.

The operability and efficiency of the proposed algorithm for the operation of an ABS with an anti-skid function of the front axle have been proved using simulation modeling methods of braking a vehicle on a slippery supporting surface with a simultaneous steering wheel movement.

×

About the authors

M. M. Zhileykin

Bauman Moscow State Technical University

Author for correspondence.
Email: jileykin_m@mail.ru

DSc in Engineering

Russian Federation, Moscow

D. S. Chugunov

Bauman Moscow State Technical University

Email: dan0634@mail.ru
Russian Federation, Moscow

References

  1. Ergin, A.A., Kolomejtseva, M.B., Kotiev, G.O. Antiblocking control system of the brake drive of automobile wheel (2004) Pribory i Sistemy Upravleniya, (9), pp. 11–13.
  2. Aref M.A. Soliman, Mina M.S. Kaldas. An Investigation of Anti-lock Braking System for Automobiles. SAE In-ternationalby Warwick University, Thursday, May 05, 2016.
  3. Chendi Sun and Xiaofei Pei. Development of ABS ECU with Hard ware-inthe-Loop Simulation Based on Labcar System. SAE International by Warwick University, Thursday, May 05, 2016.
  4. Edoardo Sabbioni, Federico Cheli and Vincenzo d’Alessandro. Politecnico di Milano Analysis of ABS/ESP Control Logics Using a HIL Test Bench. SAE International by Warwick University, Thursday, May 05, 2016.
  5. Farthad Assadian. Mixed

Supplementary files

Supplementary Files
Action
1. JATS XML
2. Fig. 1. Acceleration plan for the center of the wheel during curvelinear motion of vehicle

Download (112KB)
3. Fig. 2. Trajectory of movement of a vehicle with a gross weight of 6000 kg when braking with ABS without anti-skid function of the front axle

Download (41KB)
4. Fig. 3. The graph of the change in time of the diagnostic characteristics δV when braking the vehicle

Download (98KB)
5. Fig. 4. Vehicle trajectory when braking with ABS and anti-skid function of the front axle

Download (40KB)
6. Fig. 5. Dependence of vehicle speed on time

Download (110KB)
7. Fig. 6. Graphs of changes in angular speeds of wheels from time to time

Download (131KB)
8. Fig. 7. The graph of the change in time of the diagnostic characteristics δV when braking a vehicle with ABS and with the function of countering the drift of the front axle

Download (96KB)

Copyright (c) 2021 Zhileykin M.M., Chugunov D.S.

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

This website uses cookies

You consent to our cookies if you continue to use our website.

About Cookies