 Open Access
 Total Downloads : 1540
 Authors : Pradeep Rohilla, Jitender, Amit, Akshay Dhingra
 Paper ID : IJERTV5IS041074
 Volume & Issue : Volume 05, Issue 04 (April 2016)
 DOI : http://dx.doi.org/10.17577/IJERTV5IS041074
 Published (First Online): 26042016
 ISSN (Online) : 22780181
 Publisher Name : IJERT
 License: This work is licensed under a Creative Commons Attribution 4.0 International License
Design and Analysis of Controller for Antilock Braking System in Matlab/Simulation
Pradeep Rohilla
Department of Mechanical Engineering The NorthCap University
Gurgaon, India
Amit
Department of Mechanical Engineering The NorthCap University
Gurgaon, India
Jitender
Department of Mechanical Engineering The NorthCap University
Gurgaon, India
Akshay Dhingra
Department of Mechanical Engineering The NorthCap University
Gurgaon, India
Abstract Antilock braking mechanism enhances the vehicle steadiness and steering ability to stop a vehicle wheel without locking and minimizing stopping distance. A scientific model of electronically monitored slowing mechanism (ABS) of quarter auto model has been produced. The different sorts of controllers such as P, PI, PD and PID have been executed and results are analyzed in Matlab/Simulink environment. The controllers parameters are streamlined to control wheel slip and stopping distance. A Genetic algorithmic optimization technique has been used to obtain gain parameters of controllers. The simulated results of an ABS system are compared with and without controller and further between the distinctive sorts of controllers. The output response of PID controller is better as compared to different controllers.
Keywords Antilock Braking System (ABS), Proportional controller (P), Proportional Integrative (PI), Proportional Derivative (PD), PID, modeling quarter car model, genetic algorithm. Slip ratio, steering ability.

INTRODUCTION
The target of ABS (Antilock braking system) is to produce the biggest conceivable braking power progressively while keeping the vehicle stable and avoiding excessive wheel slippage. ABS works when the braking force is more than the force of adhesion [1]. The ABS screens the pace of every wheel to identify locking. When it recognizes sudden breaking, it will discharge breaking pressure for a moment and then continue optimum braking pressure to each wheel [2]. By repeating this procedure in brief time frame, it upgrades steering control amid sudden stops. Thus, it will likewise enhance the soundness of halting the vehicle. Accordingly, ABS advantages in two ways: You will stop prior, and you will have the capacity to direct while you stop. Coefficient of friction between tire and road, the tire slip proportion, and the vertical force on the wheel are the essential procedure parameters influencing the control quality. The estimation of slip ratio between wheel and road surface is highly uncertain. The reason for this instability for the most part comprises of vertical contact force between tire and the road surface, slip ratio and sometimes rapid variation of the road conditions with its related large variations of friction coefficients [3]. The
plant to be controlled incorporates elastically suspended wheel, braking servo system and actuator. For the configuration of ABS controllers, various methodologies have been proposed. A prescient methodology to design a non direct modelbased controller for the wheel slips is put forward by [49]. A staticstate feedback control calculation for Antilock braking system is proposed by [10]. The model of a quartervehicle and an ABS has been described by [11]. ABS that is updated for one kind of surface can't be trusted upon to work honorably on a different kind of surface. To make ABS work successfully for various road conditions, we need to perceive the perfect wheel slips on a given surface. The goal of the present paper is to outline and examine an efficient controller for antilock braking system.

PROBLEM FORMULATION
ABS system helps to achieve shorter stopping distances as compared to those of locked up wheels, furthermore to maintain vehicles steadiness and steering capability which is explained by the relation between wheel slip ratio ( ) and the coefficient of friction ( ). The friction coefficient relies upon a lot of factors, and thus fluctuates in a wide range. The variables are recorded beneath:

The slip proportion between the road and the tire,

Tire brand (seasonal tires),

Condition of road surface i.e. wet or dry,

Slip angle of tire, and

Vehicle speed.
The wheel slip ratio likewise changes as per the kind of road [13]. Considering all sorts of road surfaces, the frictional coefficient of almost all the road surfaces is worst when the wheel is locked and the slip proportion of wheel is 1 and is optimum when the wheel slip ratio is 0.2 approximately. So, in order to maximize the coefficient of friction for any road surface, we need to control the wheel slip ratio to an estimation of around 0.2. This is the objective of ABS controller.


MODELLING OF VEHICLE BRAKING SYSTEM In this section, a mathematical model of quarter vehicle
dynamic motion has been obtained by using the physics law. The improved model of quarter vehicle as shown in Fig. 1 was derived [1415]. The nonlinear dynamics can be depicted as follows.
The force balance in the longitudinal direction:
max = rFN (1) The slip ratio is defined by:
= Vx R
Vx
Summing torques about the wheel centre,
(2)
Fig. 1: Quarter vehicle dynamic motion
TABLE 1. Input Parameters
Vehicle Parameters
Values
Radius of the wheel (R)
0.33 m
Mass of the vehicle (m)
342 kg
Moment of Inertia ()
1.13 kg2
Gravitational Constant(g)
9.81 m/2
Maximum Braking Torque (u)
1200 Nm
Linear velocity of vehicle ()
27.78 m/s
Rotational speed of wheel ()
27.780.33 = 84.18
rad/s
Desired slip ( )
0.2
Vehicle Parameters
Values
Radius of the wheel (R)
0.33 m
Mass of the vehicle (m)
342 kg
Moment of Inertia ()
1.13 kg2
Gravitational Constant(g)
9.81 m/2
Maximum Braking Torque (u)
1200 Nm
Linear velocity of vehicle ()
27.78 m/s
Rotational speed of wheel ()
27.780.33 = 84.18
rad/s
Desired slip ( )
0.2
JWW = u + rRFN (3)
Using equations 1 & 2, and rearranging for yields,
= (1 + 2) +
(4)
The above equations 1, 2, 3 & 4 are used to develop a
sim
ulink model of ABS in Matlab as shown in Fig. 2. In this s ,
imulink model the tire torque and the desired slip of value
0.2 is taken as input; whereas the vehicle speed, wheel speed, stopping distance, and slip are the output received. To calculate the value ofr, a standard graph between r and is used, where the value of is calculated from the model. The input parameters that have been used are given in Table 1.

CONTROL SYSTEM
Fig. 2: ABS Model in Matlab/Simulink
further fed into the system as shown in Fig. 3, which controls the brake pressure modulator so as to maintain the desired slip
The conventional controller such as P, PI, PD, PIDhas been used in control system, the gain parameters play a vital role in controlling action; so in the present work these parameters are optimized by genetic algorithmic in matlab. A closed loop system is implemented where the output (slip ratio) is compared with the desired value and an error is generated which is controlled by the controllers. The controller output is
ratio. The various plots of ABS system without controller that are slip ratio vs. time, stopping distance vs. time and vehicle & wheel velocity vs. time are shown in Figs. [46].
Fig. 3: Feedback Control System
1
Where is the proportional gain parameter of the controller, whose value is optimized using GA technique (Fig. 7). The optimized value of is 1.817 and the same is applied in the model. The various plots that are slip ratio vs. time, stopping distance vs. time and vehicle & wheel velocity vs. time are shown in Figs. [810]. The stopping distance obtained by implementing Proportional control is 46.9494 meters, and the stopping time is 3.1229 seconds.
Best: 0.095556 Mean: 0.0979158
0.8
Best fitness
Slip
Slip
Fitness value
Fitness value
0.6
0.5
0.4
Mean fitness
0
0 1 2 3 4
Time (s)
Fig. 4: Slip vs. Time without controller
Stopping Distance (m)
Stopping Distance (m)
60
0.2
0
0 20 40 60 80 100
Generation
Current best individual
Current best individual
Current Best Individual
Kp = 1.817
Kp = 1.817
2
1
Sd = 52.3364
T = 3.3284
Sd = 52.3364
T = 3.3284
40
20
0
0 1 2 3 4
Time (s)
0
1
0.8
1
Number of variables (1)
Fig. 7: GA optimized value of
Fig. 5: Stopping distance vs. time without controller
vehicle speed
vehicle speed
100
Wheel Speed
Wheel Speed
80
Speed
Speed
60
40
20
0
0.6
Slip
Slip
0.4
0.2
0
50
0 1 2 3 4
Time (s)
Fig. 8: Slip ratio vs. time for P Controller
0 1 2 3 4
Time(s)
Fig. 6: Vehicle speed and wheel speed vs. time without controller

Proportional Control ( P Controller )
P controller endeavors to control the output by applying input to the system which is proportional to the measured error (e) between the output and the reference point. Here control torque is defined as,
u = e (5)
40
Stopping Distance(m)
Stopping Distance(m)
30 Sd = 46.9494
T = 3.1229
20
10
0
0 1 2 3 4
Time (s)
Fig. 9: Stopping distance vs. time for P Controller
100 1
Vehicle speed
Slip
Slip
80 Wheel speed
Speed
Speed
60
40
20
0
0 1 2 3 4
Time(s)
Fig. 10: Vehicle & wheel speed vs. time for P Controller

Proportional Derivative Control ( PD Controller )
This controller feeds the error with constant gain () as well
0.5
0
0 1 2 3
Time (s)
Fig. 12: Slip ratio vs. time for PD Controller
50
Stopping Distance(m)
Stopping Distance(m)
40
as the differentiation of error with constant gain () to the 30
system so as to maintain the output of system at the reference
point. 20
Sd = 43.3268
T = 2.9779
u = e +
(6) 10
The GA optimized values of and are 4.052 and 0.187 (Fig.11). After implementing the same value, various plots that are slip ratio vs. time, stopping distance vs. time and vehicle & wheel velocity vs. time are drawn as shown in Figs. [1214]. The stopping distance obtained by implementing PD control is 43.3268 meters, and the stopping time is 2.9779 seconds.
Best: 0.00711449 Mean: 0.00771639
0
0 1 2 3
Time (s)
Fig. 13: Stopping distance vs. time for PD Controller
Vehicle speed
Wheel speed
Vehicle speed
Wheel speed
100
80
Best fitness Mean fitness
Best fitness Mean fitness
1
Fitness value
Fitness value
Speed
Speed
60
0.5 40
0
0 20 40 60 80 100
Generation
Current Best Individual
Current best individual
Current best individual
6
20
0
0 1 2 3
Time(s)
Kp = 4.052
Kp = 4.052
4
Kd = 0.187
Kd = 0.187
2
0
1 2
Number of variables (2)
Fig. 11: GA optimized value of Kp and Kd
Fig.14: Vehicle & wheel speed vs. time for PD Controller

Proportional Integral Control ( PI Controller )
In this controller, input to the system is the error with constant gain () in addition to the integral of error with constant gain () to control the system output.
u = e + e (7)
The GA optimized value of and are 2.503 and 0.01 (fig. 15). Obtained graph of slip ratio vs. time, stopping distance vs. time and vehicle & wheel velocity vs. time are shown in Figs. [1618]. The stopping distance obtained by implementing PI control is 45.1382 meters, and the stopping time is 3.0582 seconds.
Best: 0.105758 Mean: 0.106128
Best fitness Mean fitness
Best fitness Mean fitness
Fitness value
Fitness value
4
2
100
Vehicle speed Wheel speed
Vehicle speed Wheel speed
80
Speed
Speed
60
0
0 20 40 60 80 100
Generation
Current Best Individual
40
20
0
0 1 2 3 4
Current best individual
Current best individual
3 kp = 2.503
2
Ki = 0.01
Ki = 0.01
1
0
1 2
Time(s)
Fig. 18: Vehicle & wheel speed vs. time for PI Controller

Proportional Integral Derivative Control (PID Controller )
In PID controller, system input is the summation of error with constant gain (), integral of error with constant gain (Ki),
and differential of error with constant gain ().
Number of variables (2)
u = e + + (8)
Fig.15: GA optimized value of Kp and Ki
1
0.8
Slip
Slip
0.6
The G The GA optimized value of , and is 5.009, 0.001 and 0.189 (Fig. 19). Obtained graphs of slip ratio vs. time, stopping distance vs. time and vehicle & wheel velocity vs. time are shown in Figs. [2022]. The stopping distance obtained by implementing PID control is 42.6188 meters, and the stopping time is 2.9514 seconds.
0.4
0.2
Stopping Distance(m)
Stopping Distance(m)
0
0 1 2 3 4
Time (s)
Fig.16: Slip ratio vs. time for PI Controller
50
40
Sd = 45.1382
T = 3.0582
Sd = 45.1382
T = 3.0582
30
20
10
0
0 1 2 3 4
Time (s)
Fig. 17: Stopping distance vs. time for PI Controller
1
Fitness value
Fitness value
0.5
0
Current best individual
Current best individual
6
4
2
0
Best: 0.0049855 Mean: 0.0053636
Best fitness Mean fitness
Best fitness Mean fitness
0 20 40 60 80 100
Generation
Current Best Individual
Kp= 5.009
Ki= 0.001
Kd= 0.189
Kp= 5.009
Ki= 0.001
Kd= 0.189
3
3
1 2
Number of variables (3)
Fig.19: GA optimized value ofKp, Ki and Kd
1
0.8
Slip
Slip
0.6
0.4
0.2
0
0 1 2 3
Time (s)
Fig. 20: Slip ratio vs. time for PID Controller
50
Stopping distance(m)
Stopping distance(m)
40
Cotroller type
Stopping Distance (m)
Stopping time (s)
Without Controller
52.3364
3.3284
P Controller
46.9494
3.1229
PI Controller
45.1382
3.0582
PD Controller
43.3268
2.9779
PID Controller
42.6188
2.9514
Controller type
Stopping Distance (m)
Stopping time (s)
Without Controller
52.3364
3.3284
P Controller
46.9494
3.1229
PI Controller
45.1382
3.0582
PD Controller
43.3268
2.9779
PID Controller
42.6188
2.9514
30
as 2.9779 seconds which is superior to PI controller where the stopping distance is 45.1382 meters and the stopping time is 3.0582 seconds. After that, we have actualized PID controller in which stopping distance results out to be 42.6188 meters and stopping time turns out as 2.9514 seconds which is superior to PD controller where the stopping distance is 43.3268 meters and the stopping time is 2.9779 seconds. So, after observing all these results, it is found out that PID controller yields the best result in terms of least stopping distance and least stopping time. By applying PID controller, the stopping distance is reduced by 9.7176 meters and stopping time is decreased by 0.377 seconds as compared to braking without controller. These results are shown in Table 2.
TABLE 2. Comparison of different types of controllers
Sd = 42.6188
20 Time= 2.9514
10
0
0 1 2 3
Time (s)
Fig. 21: Stopping distance vs. time for PID Controller
Vehicle speed
Wheel speed
Vehicle speed
Wheel speed
100
80
Speed
Speed
60
40
20
0
0 1 2 3
Time(s)
Fig. 22: Vehicle & wheel speed vs. time for PID Controller


RESULTS
The P, PD, PI and PID controllers have been effectively executed and consequences of ABS framework with and without controller are analyzed. The output response of P controller in which stopping distance results out to be 46.9494 meters and stopping time comes out as 3.1229 seconds which is superior to braking without controller where the stopping distance is 52.3364 meters and the stopping time is 3.3284 seconds. Also, we have executed PI controller in which stopping distance results out to be 45.1382 meters and stopping time results out to be 3.0582 seconds which is better than P controller where the stopping distance is 46.9494 meters and the stopping time is 3.1229 seconds. Thirdly, we have executed PD controller in which stopping distance results out to be 43.3268 meters and stopping time comes out

CONCLUSION
A mathematical model of antilock braking system is made in matlab/simulink. We actualized different controllers, for example; P, PI, PD and PID. By comparing the outcomes that are shown in table 2, PID is the best amongst every one of them. The stopping distance obtained in PID controller is 42.6188 meters and stopping time is 2.9984 seconds.
REFERENCES

Bosch Automotive Handbook (2004), 6th Edition, Robert Bosch GmbH. Wu. and Shih. simulated and experimental study of Hydraulic antilock braking systems using sliding mode PWM Control, Mechatronics (2003), vol. 13, pp. 331351.
Available : http://dx.doi.org/10.1016/S09574158(01)000496

Wei, zhang and Gua Xuexun. An ABS control strategy for commercial vehicle, IEEE/ASME Trans Mechatronics, 2015, vol. 20, pp. 384

Huang and Shih (2010), Design of a hydraulic Antilock braking system (ABS) for motorcycle, J. Mech. Sci. Technol, pp. 11411149,[Online].

S. Anwar and B. Zheng. (2007 May). An antilock braking algorithm For an eddycurrentbased brakebywire system. IEEETrans. Veh. Technol. 56(3), pp. 11001107.
 [6]Hoseinnezhad and BabHadiashar, Efficient antilock braking by

Direct maximization of tireroad frictions, IEEE Trans. Ind. Electron,2011, vol. 58, pp. 35933600.

T. A. Johansen, I. Petersen, J. Kalkkuhl, and J. LÂ¨udemann. (2003),Gainscheduled wheel slip control in automotive brake systems. IEEE Trans. Control Syst. Technol. 11, pp. 799811.

Pedro, Dahunsi, and Nyandoro. (2012),Direct Adaptive neural control of antilock braking systems incorporated with Passive suspension dynamics. J. Mech. Sci. Technol. 26,pp. 4115 4130.

Mirzaeinejad and Mirzaei, A novel method for nonlinear control of wheel slip in antilock braking systems, Control Engineering Practice 2010 vol. 18, pp. 918926

Baslamisli, KÃ¶se and Anlas, Robust control of antilock brake system, Vehicle System Dynamics,2007, vol. 45, pp. 217232,

K.Z. Rangelov, SIMULINK model of a quartervehicle with an antilock braking system, Masters Thesis Eindhoven: Stan Ackermans Institute, 2004.

Raesian, Khajehpour, and Yaghoobi, A New Approach in Antilock Braking System (ABS) Based on Adaptive Neuro Fuzzy Selftuning PID Controller, in IEEE (ICCIA) 2011, pp.530535

Assadin, F., Mixed HN and Fuzzy Logic controllers for the automobile ABS, PSA PeugeotCitroen, SAE 2001 World congress, March 58, Detroit, Michigan, USA, 2001.

Nouilllant, C., et al., 2002. A cooperative control for car suspension And brake systems, Int. J. Automotive. Technol. 3,pp. 147155.