International Journal of Engineering Trends and Technology- Volume3Issue5- 2012

Simulation and Analysis of Passive and Active Suspension System Using Quarter Car Model for Different Road Profile
Abdolvahab Agharkakli#1, Ghobad Shafiei Sabet*2, Armin Barouz#3
Department of Mechanical Engineering, Shahrood Branch, Islamic Azad University, Shahrood, Iran * Department of Mechanical Engineering, Shahrood Branch, Islamic Azad University, Shahrood, Iran # Ms.C student, Sari branch, Islamic Azad University, Sari, Iran 1 [email protected] 2 [email protected] 3 [email protected]
#

Abstract- The objectives of this study are to obtain a mathematical model for the passive and active suspensions systems for quarter car model. Current automobile suspension systems using passive components only by utilizing spring and damping coefficient with fixed rates. Vehicle suspensions systems typically rated by its ability to provide good road handling and improve passenger comfort. Passive suspensions only offer compromise between these two conflicting criteria. Active suspension poses the ability to reduce the traditional design as a compromise between handling and comfort by directly controlling the suspensions force actuators. In this study, the Linear Quadratic Control (LQR) technique implemented to the active suspensions system for a quarter car model. Comparison between passive and active suspensions system are performed by using different types of road profiles. The performance of the controller is compared with the LQR controller and the passive suspension system. Keywords- Quarter Car-model, Active Suspension system, LQR Control Design, Road Profile

I.

INTRODUCTION

A car suspension system is the mechanism that physically separates the car body from the wheels of the car. The performance of the suspension system has been greatly increased due to increasing vehicle capabilities. Apple yard and Well stead (1995) have proposed several performance characteristics to be considered in order to achieve a good suspension system. Suspension consists of the system of springs, shock absorbers and linkages that connects a vehicle to its wheels. In other meaning, suspension system is a mechanism that physically separates the car body from the car wheel. The main function of vehicle suspension system is to minimize the vertical acceleration transmitted to the passenger which directly provides road comfort. Traditionally automotive Suspension designs have been compromise between the three conflicting criterias namely road handling, load carrying, and passenger comfort. The suspension system must support the vehicle, provide directional control using

handling manoeuvres and provide effective isolation of passengers and load disturbance. A passive suspension has the ability to store energy via a spring and to dissipate it via a damper. Its Parameters are generally fixed, being chosen to achieve a certain level of compromise between road handling, load carrying and ride comfort. An active suspension system has the ability to store, dissipate and to introduce energy to the system. It may vary its parameters depending upon operating conditions. There are three Types of suspension system; passive, semi-active and active suspension system. Traditional suspension consists springs and dampers are referred to as passive suspension, then if the suspension is externally controlled it is known as a semi active or active suspension. The passive suspension system is an open loop control system. It only designs to achieve certain condition only. The characteristic of passive suspension fix and cannot be adjusted by any mechanical part. The problem of passive suspension is if it designs heavily damped or too hard suspension it will transfer a lot of road input or throwing the car on unevenness of the road. Then, if it lightly damped or soft suspension it will give reduce the stability of vehicle in turns or Change lane or it will swing the car. Therefore, the performance of the passive suspension depends on the road profile. In other way, active suspension can gave better performance of suspension by having force actuator, which is a close loop control system. The force actuator is a mechanical part that added inside the system that control by the controller. Controller will calculate either add or dissipate energy from the system, from the help of sensors as an input. Sensors will give the data of road profile to the controller. Therefore, an active suspension system shown is Figure1 is needed where there is an active element inside the system to give both conditions so that it can improve the performance of the suspension system. In this study the main objective is to observe the performance of active by using LQR controller and passive suspension only and representing in ref. [10].

ISSN: 2231-5381

http://www.internationaljournalssrg.org

Page 636

International Journal of Engineering Trends and Technology- Volume3Issue5- 2012

representation of the quarter car model. Suspension system is modelled as a linear suspension system. The state variable can be represented as a vertical movement of the car body and a vertical movement of the wheels. Figure3 shows a basic two-degree-of freedom system representing the model of a quarter-car and representing in ref. [1] . the model consists of the sprung mass M2 and the unsprung mass M1. The tire is modelled as a linear spring with stiffness Kt. The suspension system consists of a passive spring Ka and a damper Ca in parallel with an active control force u. The passive elements will guarantee a minimal level of performance and safety, while the active element will be designed to further improve the performance. This combination will provide some degree of reliability.
Figure1: Active Suspension System

II.

MATHEMATICAL MODELLING OF ACTIVE SUSPENSION FOR QUARTER CAR MODEL

Quarter-car model in Figure2 is very often used for suspension analysis; because it simple and can capture important characteristics of full model. The equation for the model motions are found by adding vertical forces on the sprung and unsparing masses. Most of the quarter-car model suspension will represent the M as the sprung mass, while tire and axles are illustrated by the unsparing mass m. The spring, shock absorber and a variable force-generating element placed between the sprung and unsparing masses constitutes suspension. From the quarter car model, the design can be expend into full car model and representing in ref. [12].

Figure3: Active Suspension for Quarter car Model

From the Figure3 and Newton's law, we can obtain the dynamic equations as the following: For M1 , F = Ma K t Xw r Ka Xs Xw Ca Xs Xw Ua = M1 Xw =

(1)

For M2 , F = Ma Ka Xs Xw Ca Xs Xw + Ua = M2 Xs
Figure2: Quarter Car Model

The main focus is to provide background for mathematical model of a quarter car model. The dynamic model, which can describes the relationship between the input and output, enables ones to understand the behaviour of the system. The purpose of mathematical modelling is to obtain a state space

Xs =

Ka Xs Xw Ca Xs Xw + Ua M2

(2)

Let the state variables are;

ISSN: 2231-5381

http://www.internationaljournalssrg.org

Page 637

International Journal of Engineering Trends and Technology- Volume3Issue5- 2012

X1 = X s X w X2 = Xs X3 = Xw r X4 = Xw Where, Xs Xw = Suspension travel Xs = Car Body Velocity Xs = Car Body Acceleration Xw r = Wheel Deflection Xw = Wheel Velocity Therefore in state space equation, the state variables are established in equation (3). Therefore, equations (1) and (2) can be written as below X t = Ax t + Bu t + Where, X1 = X s X w X 2 X 4 X2 = Xs X3 = Xw r X4 r X4 = Xw Rewrite equation (4) into the matrix form yield X1 X2 = X3 X4
Ka M2 Ka M1

III. (3)

CONTROLLER DESIGN USING LINEAR QUADRATIC REGULATOR (LQR) STATE FEEDBACK DESIGN

The main objective of this section is to design LQR controller for the active suspension system. In optimal control, the attempts to find controller that can provide the best possible performance. The LQR approach of vehicle suspension control is widely used in background of many studies in vehicle suspension control. . The strength of LQR approach is that in using it the factors of the performance index can be weighted according to the designers desires or other constraints. In this study, the LQR method is used to improve the road handling and the ride comfort for a quarter car model. LQR control approach in controlling a linear active suspension system and representing in ref. [21]. A system can be expressed in state variable form as X t = Ax t + Bu t + (7)

With x (t) R n, u (t) R m, the initial condition is x (0). We assume here that all the states are Measurable and seek to find a state-variable feedback (SVFB) control u t = Kx t + v(t) (8)

(4)

That gives desirable closed-loop properties. The closed-loop system using this control becomes (5) X t = A BK X t + ( ) (9)

0 0

Ca M2 Ca M1

1 0

0 0
K t M1

1
Ca M2 Ca M1

0 X1 0 1 X2 M2 0 + Ua + r X3 0 1 1 X4 0 (6) M1

With Ac the closed-loop plant matrix and v (t) the new command input. Note that the output matrices C and D are not used in SVFB design. Since many naturally occurring systems are optimal, it makes sense to design man-made controllers to be optimal as well. To design a SVFB that is optimal, we may define the performance index (PI) J=
2 0 1

x t T Qx t + u t T Ru t

dt

(10)

Parameters of Quarter Car Model for Simple Passenger Car; M1 =59kg M2 =290kg Ka=16812(N/m) K t =190000(N/m) Ca=1000(Ns/m)
Table1: Parameter for Quarter Car Model

Substituting the SVFB control into this yields J=2

1 x 0

t T (Q + K T RK)x dt

(11)

We assume that input v (t) is equal to zero since our only concern here are the internal stability Properties of the closedloop system. The objective in optimal design is to select the SVFB K that minimizes the performance Index J. The performance index J can be interpreted as an energy function, so that making it small keeps small the total energy of the closed-loop system. Note that both the state x (t) and the Control input u (t) are weighted in J, so that if J is small, then neither x (t) nor u (t) can be too large. Note that if J is

ISSN: 2231-5381

http://www.internationaljournalssrg.org

Page 638

International Journal of Engineering Trends and Technology- Volume3Issue5- 2012

minimized, then it is certainly finite, and since it is an infinite integral of x (t), this implies that x (t) goes to zero as t goes to infinity, This in turn guarantees that the closed loop system will be stable. The two matrices Q (an n n matrix) and R (an m m matrix) are selected by the design engineer. Depending on how these design parameters are selected, the closed-loop system will exhibit a different response. Generally speaking, selecting Q large means that, to keep J small, the state x (t) must be smaller. On the other hand selecting R large means that the control input u (t) must be smaller to keep J small. This means that larger values of Q generally result in the poles of the closed-loop system matrix Ac = A BK being further left in the s-plane so that the state decays faster to zero. on the other hand, larger R means that less control effort is used, so that the poles are generally slower, resulting in larger values of the state x(t). One should select Q to be positive semidefinite and R to be positive definite. This means that the scalar quantity x t T Qx t is always positive or zero at each time t for all functions x (t), and the scalar quantity u t T Ru t is always positive at each time t for all values of u (t). This guarantees that J is well-defined. In terms of eigenvalues, the eigenvalues of Q should be non-negative, while those of R should be positive. If both matrices are selected diagonal, this means that all the entries of R must be positive while those of Q should be positive, with possibly some zeros on its diagonal. Note that then R is invertible. Since the plant is linear and the PI is quadratic, the problem of determining the SVFB K to minimize J is called the Linear Quadratic Regulator (LQR). The word 'regulator' refers to the fact that the function of this feedback is to regulate the states to zero. This is in contrast to Tracker problems, where the objective is to make the output follow a prescribed (usually nonzero) reference command. To find the optimal feedback K we proceed as follows. Suppose there exists a constant matrix P such that
d dt

It has been assumed that the external control v (t) is equal to zero. Now note that the last equation has to hold for every x (t). Therefore, the term in brackets must be identically equal to zero. Thus, proceeding one sees that A BK T P + A BK + Q + K T RK = 0 AT P + PA + Q + K T RK K T B T P PBK = 0 (15)

This is a matrix quadratic equation. Exactly as for the scalar case, one may complete the squares. Though this procedure is a bit complicated for matrices, suppose we select K = R1 B T P Then, there results AT P + PA + Q + R1 B T P R(R1 BT P) (R1 B T P)T B T P PB(R1 B T P) = 0 (17) AT P + PA + Q PBR1 B T P = 0 This result is of extreme importance in modern control theory. Equation (17) is known as the algebraic Riccati equation (ARE). It is a matrix quadratic Equation that can be solved for the auxiliary matrix P given (A, B, Q, and R). Then, the optimal SVFB gain is given by (16). The minimal value of the PI using this gain is given by (15), which only depends on the initial condition. This mean that the cost of using the SVFB (16) can be computed from the initial conditions before the control is ever applied to the system. The design procedure for finding the LQR feedback K is: Select design parameter matrices Q and R Solve the algebraic Riccati equation for P Find the SVFB using K = R 1BT P There are very good numerical procedures for solving ARE. The MATLAB routine that performs this is named lqr (A, B, Q, R). (16)

(x Px) = x (Q + K RK)x

(12)

Then, substituting into equation (11) yields J=2

1 d 0 dt

x T Px dt = 2 x T 0 Px 0

(13)

Where we assumed that the closed-loop system is stable so that x (t) goes to zero as time t goes to infinity. Equation (13) means that J is now independent of K. It is a constant that depends only on the auxiliary matrix P and the initial conditions. Now, we can find a SVFB K so that assumption (13) does indeed hold. To accomplish this, differentiate (13) and then substitute from the closed-loop state equation (9) to see that (13) is Equivalent to x T Px + x T Px + x T Qx + x T K T RKx = 0 x T Ac T Px + x T PAcx + x T Qx + x T K T RKx = 0 x T Ac T P + PAc + Q + K T RK x = 0
Figure4. Linear Quadratic Regulator

(14)

ISSN: 2231-5381

http://www.internationaljournalssrg.org

Page 639

International Journal of Engineering Trends and Technology- Volume3Issue5- 2012

V. RESULT AND DISCUSSION

IV.

SYSTEM MODELLING

Designing an automatic suspension system for a car turns out to be an interesting control problem. When the suspension system is designed, a 1/4 car model (one of the four wheels) is used to simplify the problem to a one dimensional springdamper system.

Simulation based on the mathematical model for quarter car by using MATLAB/SIMULINK software will be performed. Performances of the suspension system in term of ride quality and car handling will be observed, where road disturbance is assumed as the input for the system. Parameters that will be observed are the suspension travel, wheel deflection and the car body acceleration for quarter car. The aim is to achieve small amplitude value for suspension travel, wheel deflection and the car body acceleration. The steady state for each part also should be fast. Two type of road disturbance is assumed as the input for the system. The road profile 1 is assumed to be a single bump taken from [2]. The disturbance input representing in [2] is shown below where a denotes the bump amplitude. The sinusoidal bump with frequency of 8 HZ has been characterized by,
a 1 cos 8 t

Figure5: active suspension system

r(t) =

0,

, 0.5 t 0.75 otherwisw

(19)

Where * Body mass (M1) = 290 kg, * Suspension mass (M2) = 59 kg, * spring constant of Suspension system (K1) = 16182 N/m, * spring constant of wheel and tire (K2) = 190000 N/m, * damping constant of suspension system (b1) = 1000 Ns/m. * Control force (u) = force from the controller we are going to design. Equations of motion: From the picture above and Newton's law, we can obtain the dynamic equations as the following: MX1 = b1 X1 X2 K1 X1 X2 + U (18)

Where,( a= 0.05 (road bump hieght 10 cm ).

Figure7: Road Profile 1

MX2 = b1 X1 X2 + K1 X1 X2 + K2 W X2 U Quarter car simulation model:

The road profile2 is assumed have 3 bumps as below where a denotes the bump amplitude. The sinusoidal bump with frequency of 8 HZ, 4 HZ and 8 HZ has been characterized by,
a 1 cos 8 t 2

r(t) = 0,

0.5 t 0.75, 3 t 3.25, 5 t 5.25 otherwisw

(20)

Figure 6: simulation modelling for Quarter car Model

ISSN: 2231-5381

http://www.internationaljournalssrg.org

Page 640

International Journal of Engineering Trends and Technology- Volume3Issue5- 2012

Where, a= 0.05 (road bump hieght 10, 5 cm and 10 cm).

Figure9: Excitation Force and Force Generated (Road profile1) Figure8: Road Profile 2

Comparison between Passive and Active Suspension for Quarter Car Model: Computer simulation work is based on the equation (4) has been performed. Comparison between passive and active suspension for quarter car model is observed. For the LQR controller, the waiting matrix Q and waiting matrix R is set to be as below. First we can choose any Q and R. and then, if we getting more control input then we increase the R, if we want faster performance then we need more control so, Q should be more and R is less. If we want slow performance then Q should be less and R high. 50 0 0 0 0 50 0 0 Q= 0 0 200 0 0 0 0 250 And R = 0.01 Therefore, the value of gain k K = 949630 66830 764050 1230 (23) (22)
Figure10: Excitation Force and Force Generated (Road profile2)

(21)

Effect of the Suspension Performance on Various Road Profiles: In this section, the performance of the quarter car active suspension system is evaluated under different road profile between the passive and LQR controller. It shows the LQR with disturbance observer and the proposed observer is robustness to overcome and performance better than passive one. Figures (11-16) shows the performance for road profile 1, the Figures (17-22) shows the performance for road profile 2. Excitation Force and Generate Force by the actuator for Two Road Profile is shown in Figure9, Figure10, the Excitation Force gets cancelled by the Generated Force. This indicates the efficiency of the LQR controller.

Figure11: Car Body Displacement with Road Profile 1

Figure12: Car Body Velocity with Road Profile 1

ISSN: 2231-5381

http://www.internationaljournalssrg.org

Page 641

International Journal of Engineering Trends and Technology- Volume3Issue5- 2012

Figure13: Car Body Acceleration with Road Profile1

Figure17: Body Displacement with Road Profile 2

Figure18: Body Acceleration with Road Profile 2 Figure14: Wheel Displacement with Road Profile1

Figure19: Body Velocity with Road Profile 2 Figure15: Wheel Deflection with Road Profile1

Figure16: Suspension Travel with Road Profile1

Figure20: Wheel Displacement with Road Profile 2

ISSN: 2231-5381

http://www.internationaljournalssrg.org

Page 642

International Journal of Engineering Trends and Technology- Volume3Issue5- 2012

Figure21: Wheel Deflection with Road Profile1

Figure22: Suspension Travel with Road Profile

By comparing the performance of the passive and active suspension system using LQR control technique it is clearly shows that active suspension can give lower amplitude and faster settling time. Suspension Travel for two types of Road Profile can reduce the amplitude and settling time compare to passive suspension system. Body Displacement also improve even the amplitude is slightly higher compare with passive suspension system but the settling time is very fast. Body Displacement is used to represent ride quality. LQR controller design approach has been examined for the active system. Suspension travel in active case has been found reduced to more than half of their value in passive system. By including an active element in the suspension, it is possible to reach a better compromise than is possible using purely passive elements. The potential for improved ride comfort and better road handling using LQR controller design is examined. MATLAB software programs have been developed to handle the control design and simulation for passive and active systems. VI. CONCLUSIONS The methodology was developed to design an active suspension for a passenger car by designing a controller, which improves performance of the system with respect to design goals compared to passive suspension system. Mathematical modelling has been performed using a two degree-of-freedom model of the quarter car model for passive and active suspension system considering only bounce motion to evaluate the performance of suspension with respect to various contradicting design goals. LQR controller design approach has been examined for the active system. Suspension travel in active case has been found reduced to more than half of their value in passive system. By including an active element in the suspension, it is possible to reach a better compromise than is possible using purely passive elements. The potential for improved ride comfort and better road handling using LQR controller design is examined.

The objectives of this project have been achieved. Dynamic model for linear quarter car suspensions systems has been formulated and derived. Only one type of controller is used to test the systems performance which is LQR.

REFERENCES
[1] Robert L.W. & Kent L.L., Modeling and Simulation of Dynamic System, Second Edition, Prentice-Hall, 1997 [2] Sam Y.M., O sman H.S.O., Ghani M.R.A., A Class of Proportional-Integral sliding Mode Control with Application to Active Suspension System,System and Control Letters, 2004. 51:217-223 [3] Kim C., Ro P.I, An Accurate Full Car Ride Model Using Model Reducing Techniques, Journal of Mechanical Design. 2002. 124:697-705 [4] Hespanhna J.P., Undergraduate Lecture Note on LQG/LGR controller Design, University of California Santa Barbara; 2007 [5] Sam Y.M., Ghani M.R.A and Ahmad, N. LQR Controller for Active Car Suspension. IEEE Control System 2000 I441-I444 [6] Wu S.J., Chiang H.H., Chen J.H., & Lee T.T., Optimal Fuzzy Control Design for Half-Car Active Suspension Systems IEEE Proceeding of the International Conference on Networking, Sensing and Control. March, Taipei, Taiwan: IEEE. 2004. 21-23 [7] Astrom K.J. & Wittenmark B., Adaptive Control, Second Editon, Addison- Wesley Pub, 1995. [8] Chen H.Y. & Huang S.J., Adaptive Control for Active Suspension System International Conference on Control and Automatic. June. Budapest, Hungary: 2005. [9] Shariati A., Taghirad H.D. & Fatehi A. Decentralized Robust H- Controller Design for a Full Car Active Suspension System Control, University of Bath, United Kingdom, 2004 [10] William D.L., Hadad W.M., Active Suspension Control to Improve Vehicle Ride and Handling, Vehicle System Dynamic. 1997. 28: 1-24 70 [11] Sam Y.M. Proportional Integral Sliding Mode Control of an Active Suspension System, Malaysia University of Technology, PHD Dissertation, Malaysia University of Technology; 2004 [12] kenaga S., Lewis F.L., Campos J., & Davis L., Active Suspension Control of Ground Vehicle Based on a Full Car Model Proceeding of America Control Conference, June, Chicago, Illinois: 2000 [13] Joo D.S. & Al-Holou N., Development & Evamation of Fuzzy Logic Controller for Vehicle Suspension System, (1995) MI 48219; pp. 295-299 [14] Hung V.V. & Esfandiari S.R., Dynamic Systems: Modeling and Analysis, Second Edition, McGraw-Hill, 1997

ISSN: 2231-5381

http://www.internationaljournalssrg.org

Page 643

International Journal of Engineering Trends and Technology- Volume3Issue5- 2012

[15] Hyvarinen J.P., The Improvement of Full Vehicle SemiActive Suspension through Kenimatical Model, Master Science Disertation. Oulu of University Findland; 2004 [16] Lowen J. Shearer & Kulakowski B.T., Dynamic Modeling and Control of Engineering Systems Second Edition, Macmillan Publishing, 1990 [17] Gallestey P., Undergraduate Lecture Note on Applied Nonlinear Control Design, University of Califonia Santa Barbara; 2000 [18] Yoshimura T., Kume A., Kurimoto M., Hino J., Construction of an Active Suspension System of a Quarter-Car Model Using the Concept of Sliding Mode Control, Journal of Sound and Vibration 2000, 239:187-199 [19] Winfred K.N., Dion R.T., Scott C.J., Jon B., Dale G., George W.A., Ron R., Jixiang S., Scott G., Li C. Development and Control of a Prototype Pneumatic Active Suspension System, IEEE Transaction on Education. 2002. Vol. 45; No.1; pp. 43-49. [20] Hrovat, D. & Tseng, T., Some Characteristics of Optimal Vehicle Suspensions Based on Quarter Car Model IEEE Proceedings ofthe 29th Conferences on Decision and Control, December, Honolulu, Hawaii: IEEE. 1990. 2232-2237 [21] E. Esmailzadeh and H.D. Taghrirad Active Vehicle Suspensions with Optimal State-Feedback Control, International Journal of Mechanical Science, pg 1- 18, 1996

ISSN: 2231-5381

http://www.internationaljournalssrg.org

Page 644