Modeling Simulation and Control of active

suspension system in Matlab Simulink

environment
Mohd. Avesh and Rajeev Srivastava

Suspension system is the most important factor responsible in

Abstract- This paper proposes the study of automobile active order to provide good riding comfort and stability to a vehicle
suspension system for the purpose of improving ride comfort to [4, 6].
passengers and simultaneously improving the stability of vehicle In early day’s simple coils and leaf springs were used as a
by reducing vibration effects on suspension system. With the suspension system. The first steel spring put on a vehicle was
increased demand of fine control of parameters, the mechatronic a single flat plate installed on carriages by the French in the
controller is being used for increasing the comfort of vehicle with
18th century. In recent years as the electronic and computer
their application to the suspension system of the vehicle. This is
the prime location for the consideration of inclusion of controller.
technology had a tremendous growth [7, 8]. Effect of this
The modeling and simulation is carried out in Simulink growing technology can also be seen in the field of
environment and the PID (Proportional Integral derivative) automobile industry. Modern Suspension system is the
controller is introduced for improving the performance of the integration of mechatronic controllers as well the mechanical
suspension system. The active control of the suspension system is system [3].
being carried out by proper tuning of the controller. There are three basic approaches to control a suspension
The dynamic performance of the integrated system is system as Passive suspension, semi active suspension and
improved significantly by using the controller. active suspension [4, 8]. There are some factors, such as
suspension design, quick response against the jerk,
Keywords—Active suspension, Modeling and Simulation of intelligence, and ability to work in real time operating
half car suspension, PID control of suspension.
conditions etc., have a very important role for providing a
good ride comfort, road holding, and improving the stability
I. INTRODUCTION of the vehicle [3, 6].

T he basic purpose of suspension system is to isolate a

vehicle body from road irregularities in order to improve
the driving comfort, safety and retain continuous road wheel
Specifically, passive suspension is an older conventional
system have non-controlled springs and shock-absorbing
dampers with fixed parameters [8, 9]. Passive system designs
contact in order to provide road holding [1, 2, 3]. for a specific operating condition and haven’t the ability to
Suspension system connects the wheel and the vehicle body adjust its parameter according to the operating conditions [10].
by springs, dampers and some linkages, the spring carries the While in real world application operating conditions change
body mass by storing energy and helps to isolate the body i.e. the load on the vehicle and road profiles etc. Beyond this a
from road disturbances, while damper dissipates this energy new technology required by which Suspension system should
and helps to damp the oscillations [4, 5]. be able to retain the comfort level with variable operating
Suspension system is one of the important concerns of the condition [3]. A variable or adjustable, damper or spring
researchers in automobiles from the very beginning of considerably improved the driving comfort and vehicle
automobiles era. So far the human expectations also increased stability in comparison with fixed parameters, this strategy
with time and a lot of researches have been done for adapted in the basic suspension system in the early 1970’s
improving the performance of automobile suspension system [11] and give the name semi active control. Semi-active
for getting a comfortable riding of the vehicle [3]. The main suspension systems offer a considerable improvement in the
factors always attracted the researchers i.e. durability or dynamic performance of the vehicle. Electro Rheological (ER)
stability, road holding and comfortable riding for the various [12, 13] and Magneto Rheological (MR) [6, 14] fluid dampers
operating conditions. are often preferred in semi-active suspension system, such
dampers have the ability to change damping coefficient
accordingly [5].
Mohd. Avesh, PG student, Department of Mechanical Engineering,
Motilal Nehru National Institute of Technology, Allahabad - 211004, India Beyond this, a controlled action of force introduced against
(e-mail: [email protected]). the suspension travel by means of the external actuator. These
Rajeev Srivastava, Associate Professor, Department of Mechanical additional forces are controlled by various control algorithms
Engineering, Motilal Nehru National Institute of Technology, Allahabad- using the data from sensor attached to the vehicle body [8]. In
211004, India (e-mail: [email protected]).

978-1-4673-0455-9/12/$31.00 ©2012 IEEE

active suspension system passive components are augmented Modelling of suspension system has taking into account the
by actuators that supply additional forces while pulling down following observations.
or pushing up the body masses for gathering the desired level • The suspension system modelled here is considered four
of comfort in order to suppress the vibrations due to the road degree of freedom system and assumed to be a linear or
irregularities [4]. The systems with active control can improve approximately linear system for a half car. However, as the
its qualities through the addition of an active shock absorber, model is practically denoted as a system including non-
which generates instantaneous forces making it possible to negligible non-linearity and uncertainty, the modelling of
support the loads and to ensure the safety and comfort against the active control becomes relatively complicated [25].
the constraints [15]. Main drawback of Active suspension is • Some minor forces (including flex in the vehicle body,
that actuation energy fed to the system from the external movement and backlash in various linkages, joints and gear
source leading to increase the system cost [16]. So far various systems) are neglected for reducing the complexity of the
control approach introduced for active control in suspension model because effect of these forces is minimal due to low
system application such as back stepping control [18], H∞ intensity. Hence these are left out for the system model.
control [19], sliding control [20], fuzzy control [1, 7], adaptive • Tyre material is considered as having damping property as
sliding fuzzy control [21] and intelligent control [22, 23] etc. well as stiffness.
Various control approaches of such type were invented and
applied to control the active suspension system. In this paper,
the PID controller is considered for providing the fine control A. Modelling of passive system
to vehicle suspension system. The proposed approach has The passive system consists of an axel which connects rear
more advantages as compared with the conventional passive and front wheel together through the Passive springs and
approach [3, 5, 11, 15]. dampers, where each tyre modelled as springs & dampers
It has been observed from the comprehensive literature [26].
study that active suspension system can maintain good It is represented as a mass connected by the front and rear
comfort and durability of vehicle for various operating suspension in fig. 1.
conditions [4, 15]. It is attracted the attention of numerous
researchers and automotive industries in the field of active
suspension system because its modern features which can
overcome the limitation of passive suspension system [8, 16].
The modeling and simulation has been carried out in
Simulink environment and further sophisticated controller
implemented i.e. PID. Simulink is a versatile interface can
handle various types of controllers easily [5, 6, 24]. It is
requires, for dynamic performance analysis, constructing a
basic suspension model and implements a controller as an
integrated system.
The present work tries to analyse the performance of active
control in the application of a four degree of freedom vehicle
suspension. Dynamic response with road uncertainties
simulated with fixed parameter of the system. Further
implements the PID controller and the performance
improvement analyses are carried out for such system. The
whole system works as a close loop control system. A sensor
measures the physical parameter (vertical displacement,
acceleration and velocity) and decode into electrical signal.
These acquired signals proceed by the PID and generate a
Fig. 1. Model of a four degree of freedom half car suspension system.
controlled signal which controls the action of the actuator for .
fine response in real time [8].
The responses are recorded and it can be observed from the Equations of motion are derived considering different
graph of system dynamic responses in fig. 7, an active parameters for this proposed system. These differential
suspension system with PID controller provides a high level of equations are obtained by utilizing physical laws (Newton’s
comfort and suppresses the body vibration at the large scale. laws of motion) and it will be linear time invariant for
considered linear system. Considered parameters are tabulated
II. SYSTEM MODELING in Table 1.
The mathematical modelling of a four degree of freedom
half car body for a passive and active suspension system is
being carried out by using basic law of mechanics.
 TABLE 1
PARAMETERS USED IN THE SYSTEM SIMULATION
After equating it is possible to get reduce (1) and (2) as
S.N. Parameter Symbol Unit following equations (3) & (4).
1 Mass of vehicle’s body ms kg (3)
•• A+ B +C + D
ys = −[ ]
2 Total Length of the body l m ms
3 Length of the body front from l m •• l ( A + C) − l2 (B + D) (4)
centre of gravity
1 θ s = −[ 1 ]
Is
4 Length of the body rear part from l m Bottom section: The second approach involved bottom
2
centre of gravity section of the system; in which unsprung mass is considered
5 Stiffness of the Tyre material kt1 kt2 N/m and derived the following two equations by balancing the
upward and downward forces.
6 Coefficient of springs k s1 k s2 N/m • • • •
•• ( ys1 − yu1 ).ks1 + ( y s1 − yu1 ).cs1 − ( yu1 − yt1 ).kt1 − ( yu1 − yt1 ).ct1 (5)
7 Damping coefficients of the c c N-s/m yu1 =
dampers
s1 s2 mu1
8 Damping coefficients of the Tyre c c N-s/m • • • •
t1 t 2 •• ( ys2 − yu2 ).ks2 + ( y s2 − yu2 ).cs2 − ( yu2 − yt2 ).kt2 − ( yu2 − yt2 ).ct2 (6)
material yu2 =
9 Moment of inertia of the body kg m 2 mu2
I s
So far the mathematical modelling for the proposed passive
10 Angular acceleration of the body •• Rad/sec2
θs suspension system is done as in above equations (1), (2), (3),
(4), (5) & (6).
11 Unsprung masses of the suspension m m kg
u1 u2
B. Modelling of active system
12 Body vertical Displacements y y m
(Sprung mass)
s1 s2 Active suspension system consists of actuators with springs
13 Suspension Travel(unsprung mass y y m and dampers. Actuators feed the desired force to the
vertical displacement)
u1 u2 suspension system in order to suppress the unnecessary
14 Tyre vertical travel due to road y y m vibrations as well as the vertical travel of the vehicle sprung
disturbance (Jerk Height)
t1 t2 mass.
For providing the controlled action (u) an external energy
Note- Suffix 1&2 shows the front and rear suspension parameters source also needed through which this actuator works. A
respectively. Proportional Integral Derivative (PID) controller proposed
here for fine controlling the actuating force in real time
In this section whole suspension system modeled by top operating.
and bottom section considered individually as shown in Fig. 1 The actuating force considered as ‘u’. It provides equal and
and derived mathematical equations for both section as opposite force in both upper and lower direction between the
follows sprung and unsprung masses as shown in fig. 2.
Top section: Initially by considering only top section of the
suspension systems following two mathematical equations are
derived.
• • • •
•• ( ys1 − yu1 ).ks1 + ( ys1 − yu1 ).cs1 + ( ys2 − yu2 ).ks2 + ( ys2 − yu2 ).cs2
ys = − (1)
ms

• • • •
•• l1[( ys1 − yu1 ).ks1 + ( ys1 − yu1 ).cs1 ] −l2[( ys2 − yu2 ).ks2 + ( ys2 − yu2 ).cs2 ]
θ s =− (2)
Is
Equation (1) relates vehicle body mass (ms), vertical
acceleration at the centre of gravity with their respective
displacements and velocities. Whereas (2) relates the moment
of inertia (Is) and the angular acceleration with the vehicle
moments at front and rear position.
For feasibility in modelling it have consider the following Fig. 2. 1/4th Model of active suspension system with actuating force (u)
between sprung and unsprung mass.
terms in A,B ,C and D as shown below
A = ( y s1 − y u 1 ) . k s1 Further, Mathematical modelling is done for half car active
B = ( y s 2 − y u 2 ) .k s 2 suspension system model through balancing the vertical forces
• • of the body, and following two equations are derived.
C = (y s1 − y u 1 ) . c s1
• •
D = (y s2 − y u 2 ) .c s 2
 • • • • closed loop an error signal fed to adjust the input in order to
•• ( ys1 − yu1 ).ks1 + ( ys1 − yu1 ).cs1 − ( yu1 − yt1 ).kt1 −( yu1 − yt1 ).ct1 u reach the output to desired setppoint as shown in fig. 5. This
yu1 = − (7)
mu1 mu1 error signal acquired from the vehicle body as the difference
• • • •
between desired and actual valuues.
•• ( ys2 − yu2 ).ks2 + ( ys2 − yu2 ).cs2 − ( yu2 − yt2 ).kt2 − ( yu2 − yt2 ).ct2 u (8)
yu2 = −
mu2 mu2
Modelling of entire system is completed in both ways as
passive and active suspension and total eight
e mathematical
equations derived.

III. SIMULATION IN MATLAB SIM

MULINK
Fig. 5. PID as a feedback loop control.
In this approach system needs to simullate for getting the
dynamic response. The entire suspension system is simulated There are three gains as Prooportional Kp, Integral Ki and
with the application of Matlab Simulinkk. In last section derivative gain Kd can easily tunne or adjust in order to provide
mathematical modeling of proposed system m is done and eight fine control for the applicatioon [3]. In Table 2 shows the
mathematical equations are derived as (1)), (2), (3), (4), (5), variation in gain with some resppective factor for fine tuning of
(6), (7) & (8). the controller.
The Simulink library is utilized and the logic is developed
TABLLE 2
according to the mathematical model, furtheer the entire system GAIN VARIATION WITH SO
OME RESPECTIVE FACTOR
is simulated in the Simulink as shown in figg. 3 and fig. 4.
CL Rise Overrshoot Settling S-S
Response time time Error
Kp Decrease Increease Small Decrease
Change
Ki Decrease Increease Increase Eliminate
Kd Small Decrrease Decrease Small
Change Change

For fine tuning of the contrroller in order to reducing the

overshoot and settling time the following
f gain values are taken
Fig. 3. Road disturbance (tr ) simulated by unit steep input with feedback into consideration.
controller (PID). Kp=1000, Kd =4000 & Ki =50000

The above selected values of o gains are taken into account

by adjusted it manually in Simulink
S where the minimum
overshoot and as well as thhe minimum settling time is
possible, while Simulink platfo form also have a provision by
which it can auto tune the gainss. The manual tuning is carried
out taking in account of the criteeria mentioned in Table 2.

Actually when vehicle runns on different type of road

profiles, different irregularities come into play such as plane
and smooth road, rough road and a bumpy road etc, therefore
this work is to predict the vehicle dynamic response for such
type of road irregularities. In order
o to overcome such effects
on the durability and ride qualityq the active control is
implemented.
A speed breaker is considereed as a jerk for which average
height is considered 100 mm and vehicle runs slowly at 10
km/hr at which it can easily annalyze the dynamic response of
rear and front suspension for thee jerk.
Fig. 4. Simulation of the basic suspension system in Matlab
M Simulink.
The following parameters are a taken into account for the
simulation of suspension system m under analysis:
IV. CONTROL METHODOLOGY AND
D TUNING
Bump height = 100mm.
It is designed a Proportional Integral derrivative (PID) as a Speed of Vehicle = 10 km/h.
feedback loop controller for the proposeed system. In this Time delay* = 0.84 sec.
*The chassis length of the vehicle is L=2.34m and vehicle run at 10km/h so
the time taken by the rear wheel is 0.84 sec for getting the earlier position of
front wheel.
So far the system dynamic response recorded for 50
seconds which shown in Fig. 6 and 7. In order to investigate
the effect of uncertainties, the numerical data is considered for
simulation are mentioned in table 3 [17].

TABLE 3
NUMERICAL VALUES FOR THE MODEL PARAMETERS.

Parameter Value Unit

ms 430 kg
mu 30 kg
1
mu 25 kg
2
Fig. 7. Time response plots of vehicle body displacement of active
l1 0.871 m suspension system.
l2 1.469 m TABLE 4
kt 152 N/m HIGHEST VALUES OF DISPLACEMENT, VELOCITY AND SETTLING TIMES FOR
1 BOTH ACTIVE AND PASSIVE SYSTEMS, AND THE PERFORMANCE EVALUATION.
kt 152 N/m
2 Passive Active Improvement
ks 6666.67 N/m Displacement -6.6468 -0.72941 89.0262%
1
ks 10000 N/m (mm)
2 Velocity -12.0724 5.0408 58.2433%
cs 500 N-s/m (mm/sec)
1
cs 400 N-s/m Settling 100 8 92%
2 time(sec)
ct 24 N-s/m
1
ct 24 N-s/m
2 V. CONCLUSION
Is 600 Kg m2 The simulated model is analyzed after running it to predefine
time and different observations are drawn in light of the
The time response plot which is obtained during simulation performance and behavior of suspension system.
while considering passive and active system is shown in fig. 6 • The comparison between active and passive suspension
and fig.7 respectively. system is made and dynamic characteristics are compared.
It has been observed that performance is improved in the
form of vertical displacement and settling time by 89% and
92% respectively which will increase the passenger
comfort and stability of vehicle.
• The proposed PID control gives percentage reduction in
body vertical displacement, velocity and settling time.
• It is apparent that, the displacement and velocity signal
amplitude of active control reduced by 9 times and 2.5
times of the passive signal respectively for the same
operating conditions.
• The observation drawn from the analysis is carried out
with the help of software, the practical simulation may
differ which can be further supported by control logics.

VI. REFERENCES
[1] Chen-sheng Ting, Tzuu-Hseng S. LI and Fan chu Kung, “Design
Fig. 6. Time response plots of vehicle body displacement of passive
suspension system. of fuzzy controller for active suspension system” Journal of
Mechatronics, Vol.5, No. 4, pp. 365-383, January 1995.
[2] Kalyamoy deb and Vikas saxena, “Car suspension design for [21] Jeen Lin, Ruey-Jing Lian, Chung-Neng Huang and Wun-Tong Sie,
comfort using genetic algorithms” unpublished. “Enhanced fuzzy sliding mode controller for active suspension
[3] Mouleeswaran Senthil kumar, Member, IAENG, “Development of systems” Mechatronics, pp. 1178–1190, March 2009.
Active Suspension System for Automobiles using PID Controller” [22] A. Aldair and W. J. Wang, “Design an intelligent controller for full
World Congress on Engineering London, U.K Vol II WCE 2008, vehicle nonlinear active suspension systems” International journal
July 2008. on smart sensing and intelligent systems vol. 4, no. 2, pp.224-243,
[4] Mohammad Ali Nekoui , Parisa Hadavi, “Optimal control of an June 2011.
active suspension system” 14th International Power Electronics [23] Jeen Lin and Ruey-Jing Lian, “Intelligent control of active
and Motion Control Conference, EPE-PEMC 2010. suspension systems” IEEE Transactions on industrial electronics,
[5] Dirman Hanafi, “PID controller design for semi-active car vol. 58, no. 2, pp.618-628, February 2011.
suspension based on model from intelligent system identification” [24] Chih-Hsuan Lee, Chiu-Ling Chen, Shun-Hsu Tu, Wes S. Jeng,
IEEE Computer society, Second International conference on Jeng-Fu Shen and Feng Tyan, “Modeling and simulation of half
computer engineering and applications, pp. 60-63, 2010. car suspension system with a MR damper using RecurDyn and
[6] Luis C. Felix-Herran, Jose de Jesus Rodriguez-Ortiz, Rogelio Soto Simulink” CSME-1254.
and Ricardo Ramirez-Mendoza, “Modeling and Control for a [25] Y. Şahin and G. Husi , “Design of a modified linear quadratic
Semi-active Suspension with a Magnetorheological Damper regulator for vibration control of suspension systems of military
including the actuator dynamics” IEEE Computer society, and civil vehicle” Int. Rev. Appl. Sci. Eng. 1,pp.55-60, 2010.
Electronics, Robotics and Automotive Mechanics Conference, pp. [26] W. Gao, N. Zhang and H. P. Du, “A half-car model for dynamic
338-343, 2008. analysis of vehicles with random parameters” 5th Australasian
[7] Paul L Ro, Chul Kim and Hoyong Kim, “An Active Suspension Congress on Applied Mechanics, Brisbane, Australia, December
System using Fuzzy Logic Control” American control conference, 2007.
California, pp. 2252-2253, July 1983.
[8] M. Appleyard, P.E. Wellstead, “Active suspensions: some
background” IEE Proc.-Control theory Appl., Vol. 142, No. 2,
March 1995.
[9] Jianmin Sun and Qingmei Yang, “Compare and analysis of passive
and active suspensions under random road excitation” IEEE,
International Conference on Automation and Logistics Shenyang,
China, pp. 1577-1580, August 2009.
[10] Jianmin Sun and Yi Sun, “Comparative study on control strategy
of active suspension system” IEEE Computer society, Third
international conference on measuring technology and
mechatronics automation, pp.729-732, 2011.
[11] Marius-Constantin O.S. Popescu and Nikos E. Mastorakis
“Testing and Simulation of a Motor Vehicle Suspension”
International Journal of systems applications, engineering &
development, Issue 2, Vol. 3, pp. 74-83,2009.
[12] Reiza Zakia Mukhis, Nik abdullah Nik Mohamed & Mohd Jailani
Mohd Nor, “The design and simulation of flow mode
electrorheological damper” Jurnal Teknologi, 44(A), pp. 115–125
June 2006.
[13] Nik Abdullah Nik Mohamed, Mohd Jailani Mohd Nor, Reiza
Zakia Mukhlis, “Compromising vehicle handling and passenger
ride comfort using ER-damper” MECHANIKA, pp.50-54.Feb
2005.
[14] Liu Haibo and Yang Jianwei, “Study on Semi-Active Suspension
System Simulation Based on Magnetorheological Damper” IEEE
Computer society, Second International Conference on Intelligent
Computation Technology and Automation, 2009.
[15] M. Khemliche, I. DIF, S. Latreche and B. Ould Bouamama,
“Modelling and analysis of an active suspension ¼ of vehicle with
bond graph” IEEE, pp. 811-814, 2004.
[16] Massimo Canale, Mario Milanese and Carlo Novara IEEE
Transactions on control systems technology, Vol.14, No.6,
November 2006.
[17] Yahaya Md. Sam and Johari Halim Shah Bin Osman, “Modeling
and control of the active suspension system using proportional
integral sliding mode approach ” Asian Journal of Control, Vol. 7,
No. 2, pp. 91-98, June 2005.
[18] J. Campos, F. L. Lewis, L. Davis, and S. Ikenaga, “Backstepping
based fuzzy logic controller of active vehicle suspension system”
American Control Conference Chicago, Illinois, pp.4030-4035,
June 2000.
[19] Monika Zuscikova, Cyril Belavy, “H∞ Controler Design for
Active Suspension System” 18th International Conference on
Process Control, Tatranska Lomnica, Slovakia, pp.394–399, June
2011.
[20] T. Yoshimura , A. Kume, M. Kurimoto and J. Hino, “Construction
of an active suspension system of a quarter car model using the
concept of sliding mode control” Journal of Sound and
Vibration,239(2),pp. 187-199, 2001.