Documente Academic
Documente Profesional
Documente Cultură
By
Mochammad Rizky Diprasetya S
11111034
BACHELORS DEGREE
in
MECHANICAL ENGINEERING MECHATRONICS CONCENTRATION
FACULTY OF ENGINEERING AND INFORMATION TECHNOLOGY
June 2015
Page 2 of 125
I hereby declare that this submission is my own work and to the best of my knowledge,
it contains no material previously published or written by another person, nor material
which to a substantial extent has been accepted for the award of any other degree or
diploma at any educational institution, except where due acknowledgement is made in
the thesis.
Date
Approved by:
Date
Date
Date
Page 3 of 125
ABSTRACT
By
Mochammad Rizky Diprasetya S
Kirina Boediardjo, ST, M.Sc, Advisor
Yunita Umniyati, PhD, Co-Advisor
This thesis works purposes are to develop a mathematical model for quarter car passive
and active suspension, to simulate its behavior, and analyze the difference of both
suspension. The active suspension use PID controller to control the displacement of the
car in vertical direction caused by different road profile. The displacement has to be
small, these condition are achieved by manipulating the force actuator inside the
suspension. The force actuator manipulated according to the output of controller
system. V-REP simulator is used to visualize the physical simulation. The physical
model of suspension developed in V-REP. The result from V-REP compared with the
result from MATLAB Simulink.
Keywords: Quarter Car Suspension, MATLAB, Simulink, V-REP Simulator, Full Car
Suspension, PID, LQR.
Page 4 of 125
Copyright 2015
by Mochammad Rizky Diprasetya S
All rights reserved
Page 5 of 125
DEDICATION
Page 6 of 125
ACKNOWLEDGEMENTS
I thank All for His constant blessing and supports during this thesis development.
I would like to thank to my parents for their patience in these past four years.
I would like to thank Kirina Boediardjo, ST, M.Sc for her support throughout the
development of the thesis, time, and her guidance in building the system model
schematic during development of this thesis.
I would like to thank to Yunita Umniyati, PhD for her inputs, advice, and support for
this thesis.
Last but not least I thank all of my beloved friends who have supported me from the
very beginning until the very end of this thesis development.
Page 7 of 125
TABLE OF CONTENTS
Page
STATEMENT BY THE AUTHOR ................................................................................ 2
ABSTRACT ................................................................................................................... 3
DEDICATION ............................................................................................................... 5
ACKNOWLEDGEMENTS ........................................................................................... 6
TABLE OF CONTENTS ............................................................................................... 7
LIST OF FIGURES ..................................................................................................... 10
LIST OF TABLES ....................................................................................................... 13
CHAPTER 1 - INTRODUCTION ............................................................................... 14
1.1 Background ............................................................................................................ 14
1.2 Thesis Purpose ....................................................................................................... 16
1.3 Objective ................................................................................................................ 16
1.4 Hypothesis.............................................................................................................. 16
1.5 Thesis Scope .......................................................................................................... 17
1.6 Thesis Limitations .................................................................................................. 17
1.6 Thesis Organization ............................................................................................... 17
CHAPTER 2 - LITERATURE REVIEW .................................................................... 19
2.1 Theoretical Perspectives ........................................................................................ 19
2.1.1 Suspension System.............................................................................................. 19
2.1.2 Spring in Suspension System .............................................................................. 21
2.1.3 Damper in Suspension System............................................................................ 22
2.1.4 Force Actuator in Suspension System................................................................. 23
2.1.5 Hydraulic Actuator .............................................................................................. 23
2.1.6 Electromagnetic Actuator.................................................................................... 24
2.1.7 PID Control ......................................................................................................... 25
2.1.8 Linear Quadratic Regulation (LQR) ................................................................... 27
2.1.9 MATLAB Simulink ............................................................................................ 28
2.1.10 V-REP Simulator ............................................................................................... 29
2.1 Previous Studies ..................................................................................................... 31
2.1.1 Modeling, Simulating, and Analyzing an Overhead Crane Using MATLAB
Simulink and V-REP Simulator [5] .............................................................................. 31
CHAPTER 3 RESEARCH METHODS ................................................................... 33
Mochammad Rizky Diprasetya S
Page 8 of 125
Page 9 of 125
4.7 Comparison V-REP and MATLAB Simulink Rear Right Suspension of Car Result
with Impulse Input ....................................................................................................... 95
4.8 Comparison V-REP and MATLAB Simulink V-REP Front Right Suspension of
Car Result with Step Input ......................................................................................... 103
CHAPTER 5 CONCLUSIONS AND RECCOMENDATIONS ............................ 110
5.1 Conclusions .......................................................................................................... 110
5.2 Recommendations ................................................................................................ 111
GLOSSARY ............................................................................................................... 112
REFERENCES .......................................................................................................... 113
APPENDICES ........................................................................................................... 114
APPENDIX A - MATLAB CODE FOR LQR CONTROLLER OF QUARTER CAR
SUSPENSION ........................................................................................................... 115
APPENDIX B - MATLAB CODE FOR LQR CONTROLLER OF FULL CAR
SUSPENSION ........................................................................................................... 117
APPENDIX C STATE SPACE MATRIX OF FULL CAR PASSIVE SUSPENSION
.................................................................................................................................... 120
APPENDIX D STATE SPACE MATRIX OF FULL CAR ACTIVE SUSPENSION
.................................................................................................................................... 123
CURRICULUM VITAE ............................................................................................ 124
Page 10 of 125
LIST OF FIGURES
Figures
Page
Page 11 of 125
Page 12 of 125
Page 13 of 125
LIST OF TABLES
Table
Page
43
44
48
48
56
71
Table 4.2 Result Comparison V-REP Simulator and MATLAB Simulink for Rear
Right Suspension due to Impulse Input
102
Table 4.3 Result Comparison V-REP Simulator and MATLAB Simulink for Rear
Right Suspension due to Step Input
108
Page 14 of 125
CHAPTER 1 - INTRODUCTION
1.1 Background
The suspension system must upkeep the vehicle, deliver directional control using
handling manoeuvres and deliver actual isolation of person along for the ride and load
disturbance. The criteria of the suspension personally depend on the purpose of the
vehicle. For example, a normal car driver will need a quite soft ride for low to medium
speed handling and at ease drive.
Page 15 of 125
Page 16 of 125
1.4 Hypothesis
The main purpose of this thesis are to analyze quarter car passive and active suspension
system, and analyze the difference between them. Determine which suspension has
good performance to achieve comfort and safety while ride a vehicle.
These are the hypothesis:
Page 17 of 125
will add or dissipate energy according to the control scheme. Hence, this
suspension system will greatly isolated the body of vehicle from road
disturbance.
The actuators is not design in dynamic model. Because the mechanism of the actuator
cannot be developed in V-REP Simulator.
1.6 Thesis Organization
Chapter 1 Introduction
This chapter briefly explains about the purpose, scope, and limitation of the thesis
Page 18 of 125
Page 19 of 125
Page 20 of 125
Page 21 of 125
Page 22 of 125
Page 23 of 125
Page 24 of 125
Page 25 of 125
Inside the controller block, the error is processed with three coefficient which are:
Proportional, Integral, and Derivative, shown in figure 2.9.
Page 26 of 125
() = () + () +
0
()
Where:
u = control signal
e = error value
Kp = proportional gain
Ki = integral gain
Kd = derivative gain
Here below explained each functions for each coefficient:
Proportional Coefficient
Kp is the constant of proportionality, or simply the gain of the amplifier. The larger the
value of this coefficient, the faster the system output will respond to the error. If it is
too high, the system can no directly stop at the targeted reference input, and this
phenomenon is called overshoot.
Integral Coefficient
Ki is implemented when the steady state error still occurs in the system, so as long as
an error exists, the output of Ki will grow with time, until the reference value is reached.
The usage of integral control can also create overshoot phenomenon, especially when
the proportional coefficient is large.
Derivative Coefficient
Kd is a derivative control of the output of the process. In an ideal process, an error must
be corrected as quickly as possible without overshoot, and this can be achieved if there
Page 27 of 125
is maximum gain at the beginning and at the end reducing the gain as the error
approaches zero.
In real applications, there are several possible ways of how to tune PID control in the
system, such as: manual tuning with basic knowledge from characteristic of coefficient,
Ziegler-Nichols, and auto tuning software. Here below in figure is the example of the
system output based on several different usages of PID control possibilities.
Page 28 of 125
1
( + )
2 0
Page 29 of 125
The MATLAB has an add-on called Simulink, which is able to deliver a collaborating
graphical situation for modelling, simulating and analysing dynamic systems. In this
thesis, Simulink will be used as the main software to simulate the mathematical
modelling.
There are three types of joint in V-REP that can be used to simulate movements in real
life:
Page 30 of 125
Revolute: This type of joint can create rotational movement in cylindrical shap,
and the joint which is already connected to the center of the cylinder can be
driven as a motor or just acting as a passive joint. The speed can be adjusted
here.
Spherical: This type of joint has spherical movement, this provides rotational
movement in three axes.
Prismatic: This type of joint has linear movement that can generate back and
forth movement.
V-REP provides several calculation modules to support the operation of the elemental
object:
A forward and inverse kinematics module is used in solving kinematics calculation for
various mechanism. The module uses the damped least squares pseudo inverse method.
The inverse kinematics module is particularly helpful when dealing with manipulators.
Dynamic or physics module that allow the dynamic simulation of rigid bodies done
with the Bullet Physics Library.
A path planning module is based on random tree algorithm.
A collision detection module that can occur during one scene and this can allow the
user to calculate minimum distance between two points or bodies. Both these modules
use the method of Oriented Bounding Boxes.
To control the simulation, V-REP uses script methodology, and there two types: main
script and child script. Main script as the main scene and child script is responsible for
scene object.
Page 31 of 125
Identify the problem to determine the problem and the behavior of the system.
To know what is the real life condition.
Page 32 of 125
Establish the scenario for the simulation consider what input that will be given
and any certain limitation must be defined.
Page 33 of 125
Develop a model After finding the right mathematical equation, the next step
is to convert the equation into simulation.
Simulation Part
Establish simulation scenario Consider what input will be given into the
system and any certain limitation must be defined.
Page 34 of 125
()
=
The second law can also be stated in terms of an objects acceleration. Since Newtons
second law is only valid for constant-mass system, mass can be taken outside. Thus,
=
Page 35 of 125
Page 36 of 125
From the figure 3.1 obtain mathematical equation using the second Newton Law, where
positive sign for every upward direction and negative sign for every downward
direction.
For M1
= 1
( ) + ( ) + ( ) = 1
=
( ) + ( ) + ( )
1
(3.1)
For M2
= 2
( ) ( ) = 2
=
( ) ( )
2
(3.2)
From the equations 3.1 and 3.2 above, let the state variable are:
1 =
2 =
3 =
4 =
where
= Suspension Travel (m)
Page 37 of 125
Therefore in state space equation, equations 3.1 and 3.2 can be written as:
() = () + ()
Where
1 = 2 4
2 =
3 = 4
4 =
Rewrite equations 3.1 and 3.2 into matrix form
0
1
2
2
=
0
3
[4 ]
[ 1
2
0
0
0
0
1
1
0
2 2
0
[ ] + [ ]
3
1
1
4
0
]
1
(3.3)
State-space matrix equation 3.3 above is developed because the matrix will be given
into the LQR function. The LQR function use state-space matrix to calculate the gain
for the LQR controller output.
Page 38 of 125
Page 39 of 125
From the figure 3.2 obtain mathematical equation, where positive sign for every upward
direction and negative sign for every downward direction.
For M1
= 1
( ) + ( ) + ( ) = 1
=
( ) + ( ) + ( )
1
(3.4)
For M2
= 2
( ) ( ) = 2
=
( ) ( ) +
2
(3.5)
From the equations 3.4 and 3.5 above, let the state variable are:
1 =
2 =
3 =
4 =
where
= Suspension Travel (m)
= Car Body Velocity (m/s)
Mochammad Rizky Diprasetya S
Page 40 of 125
Therefore in state space equation, equations 3.4 and 3.5 can be written as:
() = () + () + ()
Rewrite the equations 3.4 and 3.5 into matrix form
0
1
2
2
=
0
3
[4 ]
[ 1
2
0
0
0
0
1
0
1
1
0
2 2
2
0
[ ]+
+ [ ]
1
0
1
3
4
1
0
[1 ]
1 ]
(3.6)
State-space matrix equation 3.6 above is developed because the matrix will be given
into the LQR function. The LQR function use state-space matrix to calculate the gain
for the LQR controller output.
Page 41 of 125
Page 42 of 125
Page 43 of 125
where
ms = mass of the car body (Kg)
muf and mur = front and rear mass of the wheel (Kg)
Ip and Ir = pitch and roll of moment inertia (Kg m2)
Zs = car body displacement (m)
Zs1, Zs2, Zs3, and Zs4 = car body displacement for each corner (m)
Zu1, Zu2, Zu3, and Zu4 = wheel displacement for each corner (m)
a = distance from center of the car body to front wheel (m)
b = distance from center of the car body to rear wheel (m)
Cf and Cr = front and rear damping (Nm/s)
Kf and Kr = stiffness of front and rear car body spring (N/m)
Ktf and Ktr = stiffness tire (N/m)
The state variables of the system are shown in Table 3.1 and the definition of each state
variable is given in Table 3.2.
8 =
2 =
9 =
3 =
10 =
4 = 1
11 = 1
5 = 2
12 = 2
6 = 3
13 = 3
7 = 4
14 = 4
Page 44 of 125
Definition
Roll angle
Roll rate
Pitch angle
Pitch rate
Vertical displacement
Vertical velocity
Page 45 of 125
Page 46 of 125
Page 47 of 125
2 = + +
For Zs3
3 = + +
3 = + +
For Zs4
4 = + +
4 = + +
where
ms = mass of the car body (Kg)
muf and mur = front and rear mass of the wheel (Kg)
Ip and Ir = pitch and roll of moment inertia (Kg m2)
Zs = car body displacement (m)
Zs1, Zs2, Zs3, and Zs4 = car body displacement for each corner (m)
Zu1, Zu2, Zu3, and Zu4 = wheel displacement for each corner (m)
a = distance from center of the car body to front wheel (m)
b = distance from center of the car body to rear wheel (m)
Cf and Cr = front and rear damping (Nm/s)
Kf and Kr = stiffness of front and rear car body spring (N/m)
Ktf and Ktr = stiffness tire (N/m)
U1 and U2 = front right and left force actuator
U3 and U4 = rear right and left force actuator
Mochammad Rizky Diprasetya S
Page 48 of 125
The state variables of the system are shown in Table 3.3 and the definition of each state
variable is given in Table 3.4
8 =
2 =
9 =
3 =
10 =
4 = 1
11 = 1
5 = 2
12 = 2
6 = 3
13 = 3
7 = 4
14 = 4
Definition
Roll angle
Roll rate
Pitch angle
Pitch rate
Vertical displacement
Vertical velocity
Page 49 of 125
Page 50 of 125
Page 51 of 125
0
1000
0
0
0
0
1000
0
0
0
]
0
1000
= 0.03
Page 52 of 125
State space formed by A, B, C, and D matrix. Rewrite these matrix and their relationship
as:
() = () + ()
() = () + ()
Matrix A is the main suspension system matrix, and relates how the current state affects
the state change . Matrix B is the control matrix, and determines how the system input
affects the state change. Matrix C is the output matrix, and determines the relationship
between the system state and the system output. Matrix D is the feed-forward matrix,
in this thesis the system is not use feed-forward controller which means matrix D is
zero matrix.
After determine the Matrix A, B, C, and D, MATLAB will calculate the gain of matrix
K using lqr() function. Then the matrix K will produce new value of matrix A:
= ( )
Matrix B, C, and D will be the same as previous value. Then the new state space will
be constructed with the new value of matrix.
3.4 Mechanical Part
Type of suspension system that is taken as main design on this thesis is double wishbone
suspension which commonly used by every car and it is more stable then single
wishbone. The design is completely designed using V-REP Simulator.
The main frame is created by cuboid shape, the wheel is created by cylindrical shape,
the joint is created by revolute joint which can generate force in rotational direction,
and the suspension is created by prismatic joint which can generate force in translational
direction.
Page 53 of 125
Page 54 of 125
Page 55 of 125
In figure 3.10 shown that the lower arm connected to the chassis by suspension. The
stiffness suspension (K) and damper coefficient (C) can be adjust in the joint dynamic
properties. The joint can act like a suspension which contain spring and damper.
Page 56 of 125
Value
0.3104 m
0.3104 m
36927 N/m
36927 N/m
4000 N/ms
4000 N/ms
500 Kg
60 Kg
0.1527 m
0.1527 m
Page 57 of 125
Page 58 of 125
Page 59 of 125
Page 60 of 125
( ) + ( ) + ( )
1
( ) ( )
2
(4.1)
(4.2)
Page 61 of 125
Page 62 of 125
Page 63 of 125
The first experiment is to give an impulse road input 0.3 m at t = 1 s, can be seen in
figure 4.4.
Page 64 of 125
The system response shown in figure 4.5 is showing the movement of the car body due
to impulse input. It occurs oscillating movement to the car body and create
uncomfortable movement for the passenger.
Page 65 of 125
In the figure 4.6, it can be seen the suspension travel due to impulse input. The
suspension travel calculate the travel of the car body minus by wheel displacement. The
negative value means that the wheel displacement act first due to the input, and after
that the car body move because of the suspension create force due to the displacement
of the wheel to the car body.
Page 66 of 125
In figure 4.7, it can be seen that the wheel displacement occur due to impulse input.
Page 67 of 125
Page 68 of 125
In figure 4.9, it can be seen that the suspension travel occur due to impulse input. The
suspension travel calculate the travel of the car body minus by wheel displacement. The
negative value means that the wheel displacement act first due to the input, and after
that the car body move because of the suspension create force due to the displacement
of the wheel to the car body.
Page 69 of 125
In figure 4.10, it can be seen that the car body displacement occur due to impulse input.
The car body displacement is oscilating
Page 70 of 125
In figure 4.11, it can be seen that the wheel displacement occur due to impulse input.
In figure 4.9, 4.10, and 4.11 shown that the suspension is not stable because the
suspension produced oscillating output due to the road disturbance which are step input
and impulse input. The suspension need controller to make it more stable.
Page 71 of 125
Value
0.1527 m
4000 Ns/m
4000 Ns/m
36927 N/m
36927 N/m
190000 N/m
190000 N/m
0.3104 m
0.3104 m
60 Kg
60 Kg
500 Kg
Page 72 of 125
The full car suspension consist of three type of motion, pitching, rolling and bouncing.
For each motion has its own block diagram in Simulink.
Page 73 of 125
Page 74 of 125
Page 75 of 125
Page 76 of 125
Page 77 of 125
Figure 4.18 Car Body Displacement of Front Right Suspension due to Impulse Input
Page 78 of 125
In figure 4.18 above shown that the car body displacement of front right suspension due
to impulse input. The car body displacement oscillates which makes uncomfortable
driving.
4.1.2.1.2 Step Input
The experiment is to give an step road input 0.3 m to the front right suspension, can be
seen in figure 4.19
Page 79 of 125
Figure 4.20 Car Body Displacement of Front Right Suspension due to Step Input
In figure 4.20 above shown that the car body displacement of front right suspension due
to step input. The car body displacement oscillates which makes uncomfortable driving.
4.3 Control Analysis
To prevent the oscillation of the car body due to road disturbance, controller is required.
There are two different types of controllers as previously discussed in chapter 3, PID
controller and linear quadratic regulation. Comparison is done in this chapter to see
which method is more effective.
4.3.1 PID Controller for Quarter Car Suspension
The system is needed to be controlled. The system of the suspension system input is
reference distance between car body and wheel, and the output is force which will be
produced by the force generator. The PID controller for suspension system transfer
function can be formed into block diagram below in figure 4.21:
Page 80 of 125
The PID controller control the suspension travel. The output from the PID controller is
in meter unit. In the controller there is a converter from displacement to force. The
displacement derevatived by twice convert it into acceleration. The acceleration
multiply by the mass of car and wheel and make it into force unit. The force fed into
the suspension. The closed loop system can be shown below in figure 4.22.
Page 81 of 125
oscillation. The car body displacement also achieve its better performance than passive
suspension as suspension travel.
After the controller is developed, this controller is integrated with the system.
4.3.2 PID Controller for Full Car Suspension
The difference between PID controller for full car suspension and quarter car
suspension is, full car suspension have three motion as mentioned in the previous
chapter which are bouncing, rolling and pitching. Each motion affected by the
suspension of each corner of the car. The force that produced by the PID controller fed
to the bouncing, rolling and pitching system. The reference is the same which is the
suspension travel of each suspension. Each suspension produced its own controlled
force.
Page 82 of 125
In figure 4.23 above shown the control system of full car suspension. Inside of the
control system box can be shown in figure. Each suspension has its own PID controller
which produce force with reference of its suspension travel.
Page 83 of 125
Page 84 of 125
Figure 4.26 Comparison Suspension Travel with PID Controller due to Step Input
Figure 4.27 Comparison Car Body Displacement with PID Controller due to Step Input
Page 85 of 125
Figure 4.28 Comparison Wheel Displacement with PID Controller due to Step Input
Overall from figure 4.26, 4.27, and 4.28 shown that the PID controller makes the
suspension travel, wheel displacement and car body displacement smooth movement
than the passive suspension which produced oscillating movement due to step input.
The comfortable driving achieved with this controller.
Page 86 of 125
Figure 4.29 Comparison Wheel Displacement with PID Controller due to Impulse Input
Figure 4.30 Comparison Car Body Displacement with PID Controller due to Impulse
Input
Page 87 of 125
Figure 4.31 Comparison Suspension Travel with PID Controller due to Impulse Input
Overall from figure 4.29, 4.30, and 4.31 shown that the PID controller makes the
suspension travel, wheel displacement and car body displacement smooth movement
than the passive suspension which produced oscillating movement due to impulse input.
The comfortable driving achieved with this controller.
Page 88 of 125
In figure 4.32 shown that the car body displacement that use active suspension has
smooth travel displacement. In real life, passive suspension with oscillating car body
displacement make the passenger feel uncomfortable while driving the car.
Page 89 of 125
In figure 4.33 shown that there is no huge difference between passive suspension and
active suspension on the wheel displacement.
Page 90 of 125
Page 91 of 125
Figure 4.35 Comparison Car Body Displacement of Rear Right Suspension with PID
Controller due to Step Input
In figure 4.35 above shown that the PID controller makes the car body displacement
smooth movement than the passive suspension which produced oscillating movement
due to step input. The comfortable driving achieved with this controller.
Page 92 of 125
Figure 4.36 Comparison Car Body Displacement of Rear Right Suspension with PID
Controller due to Impulse Input
In figure 4.36 above shown that the PID controller makes the car body displacement
smooth movement than the passive suspension which produced oscillating movement
due to impulse input. The comfortable driving achieved with this controller.
4.6 Full Car Suspension Simulink Model and Full Car Suspension V-REP
Comparison
The parameter for this experiment shown in Table 3.5. The parameter is used for both
V-REP and MATLAB Simulink. The car is placed 0.2 m above the ground and the car
fall to the ground.
Page 93 of 125
In figure 4.37 shown the result of car body displacement in MATLAB Simulink. The
graph shown that the car body displacement travel to -0.046 m from the reference which
is -0.071 m. After that because of the suspension, the car body travel back to its
reference point. From that data, the car body travel -0.03 m from its reference point.
Page 94 of 125
Page 95 of 125
4.7 Comparison V-REP and MATLAB Simulink Rear Right Suspension of Car
Result with Impulse Input
In figure 4.39 below is the overall result using V-REP Simulator rear right suspension
with impulse input. The impulse input or speed is 0.05 m.
Figure 4.39 V-REP Result of Rear Right Suspension with Impulse Input
The result is not oscillating which is the suspension already has its good condition.
After the car drive through the speed bump, the graph shown that it is a little bit
oscillating, that is because the mechanical design is not stable.
Page 96 of 125
Figure 4.40 Position of Car Body before Drive through the Speed Bump
In figure 4.40 shown the position of car body before it goes through the speed bump.
The position in 0.12 m above the V-REP world reference.
Page 97 of 125
In figure 4.41 shown the position of the car body after when it goes through the speed
bump. The position is 0.17 m above the V-REP world reference. It shows that the car
body is lifted up by 0.05 m.
Figure 4.42 V-REP Result of Rear Right Suspension Bottom after First Peak
In figure 4.42 shown the position of the car body after it goes through the speed bump.
Because of the suspension, the car body hold off at 0.113 m which is 0.007 m below its
reference position, and after that it try to goes back to its initial position.
Page 98 of 125
In figure 4.43 shown that the suspension try to place the car body to its initial position,
but overshoot occurred due to the suspension. Overshoot peak position is 0.128 m,
0.008 m above the reference position. After that it goes back to the reference position
which is 0.12 m.
Page 99 of 125
In figure 4.44 below shown the result from MATLAB Simulink. The result almost the
same with V-REP Simulator. The difference is in MATLAB Simulink, the car body
displacement produced oscillating result. The first run, car body has its own reference
position which is 0 m. After that it goes through the speed bump. The car body lifted
up by 0.04 m.
Figure 4.44 MATLAB Simulink Result Rear Right Suspension due to Impulse Input
Figure 4.45 MATLAB Simulink Result after Goes through the speed bump
In figure 4.45 shown that the car body hold off at 0.006 m below its reference position,
and because the suspension, the suspension try to place the car body back to its
reference position.
In figure 4.46 shown that the suspension try to place the car body to its initial position,
but overshoot occurred due to the suspension. Overshoot peak position is 0.009 m,
0.009 m above the reference position. After that it goes back to the reference position
which is 0 m.
All the results detail, provided in Table 4.2.
Table 4.2 Result Comparison V-REP Simulator and MATLAB Simulink for Rear
Right Suspension due to Impulse Input
Program
Parameter
Result
0.05 m above the
V-REP Simulator
reference position
First Peak While at Speed Bump
0.04 m above the
MATLAB Simulink
reference point
0.007 m below the
V-REP Simulator
reference position
After the speed bump
0.006 m below the
MATLAB Simulink
reference position
0.008 m above the
V-REP Simulator
reference position
Overshoot due to Suspension
MATLAB Simulink
The Table above shows that MATLAB Simulink and V-REP Simulator has relatively
same result, but the MATLAB Simulink has more stable data because in V-REP there
is a lot parameter including the mechanical design which can hugely affect the result.
4.8 Comparison V-REP and MATLAB Simulink V-REP Front Right Suspension
of Car Result with Step Input
In figure 4.38 below is the overall result using V-REP Simulator rear right suspension
with impulse input. The step input or speed is 0.05 m.
Figure 4.47 V-REP Result of Rear Right Suspension with Step Input
The result is not oscillating which is the suspension already has its good condition.
After the car drive through the step input, the graph shown that it is a little bit
oscillating, that is because the mechanical design is not stable.
Figure 4.48 Position of Car Body before Drive through the Step Input
In figure 4.48 shown the position of car body before it goes through the step input. The
position in 0.12 m above the V-REP world reference.
In figure 4.49 shown the First peak of the car body displacement of rear right
suspension. The first peak is 0.1665 m which is 0.04 m above the reference point.
Figure 4.50 V-REP Result of Rear Right Suspension Bottom after First Peak
In figure 4.50 shown the position of the car body after it goes through the step input.
Because of the suspension, the car body hold off at 0.1643 m which is 0.002 m below
its first peak.
Figure 4.51 MATLAB Simulink Result Rear Right Suspension due to Impulse Input
Figure 4.52 MATLAB Simulink Result Rear Right Suspension due to Impulse Input
the first peak
In figure 4.52 shown that the first peak after goes through the step input is 0.039 m.
Figure 4.53 MATLAB Simulink Result Rear Right Suspension due to Impulse Input
after the first peak
In figure 4.53 shown the car body displacement after the first peak. It goes to 0.03 m,
which is decreased by 0.009 m from the first peak.
Table 4.3 Result Comparison V-REP Simulator and MATLAB Simulink for Rear
Right Suspension due to Step Input
Program
Parameter
Result
0.039 m above the
V-REP Simulator
reference position
First Peak While at Step Input
MATLAB Simulink
V-REP Simulator
MATLAB Simulink
The modelling of the quarter car passive and active suspension has been
successfully completed in MATLAB Simulink and V-REP Simulator
represented by full car suspension system model
The quarter car simulation response to step and impulse input shows that the car
body displacement can have relatively big oscillating. The oscillating make the
car feel uncomfortable to the passenger while driving the car. However, a
control system has been designed and implemented to the suspension system
and shows acceptable result:
o The oscillating due to the road disturbance is minimized by the PID
controller. The car body displacement is not oscillating and only small
overshoot. The PID controller also reduce the oscillating in wheel
displacement and suspension travel.
o The difference between PID controller and LQR, the LQR controller
reduce the oscillating in car body displacement, but makes no difference
with wheel displacement and suspension travel. The oscillation in wheel
displacement and suspension travel are still occurred.
The LQR is the most known controller for suspension car, but the experiment
shows that PID has better performance than LQR.
The verification of the response in V-REP simulator has shown that in numerical
software which is MATLAB Simulink has accurate result than V-REP.
5.2 Recommendations
The dynamic of force actuator is not developed. For more detail and accurate
result, the force actuator should be develop. The mathematical model and the
physical model of force actuator.
Using another type of controller which can be produce better than LQR and PID
controller.
Physical simulation in real life is not develop. For obtain real life performance,
real physical simulation should be develop.
There are many different type of angle of suspension that connected to the
chassis. Observe with different angel of it for more varieties result and can be
compared to many different car with different type of angel of suspension.
GLOSSARY
REFERENCES
[1] "Repair Pal," Repair Pal, [Online]. Available:
http://repairpal.com/images/managed/content_images/encyclopedia/CM_Steerin
g_Suspension/Suspension_System_07.11.png. [Accessed 17 June 2015].
[2] T. R. M. Rao, G. V. Rao, K. S. Rao and A. Purushottam, "ANALYSIS OF
PASSIVE AND SEMI ACTIVE CONTROLLED SUSPENSION SYSTEMS
FOR RIDE COMFORT IN AN OMNIBUS PASSING OVER A SPEED
BUMP," International Journal of Research and Reviews in Applied Science, vol.
5, no. 1, pp. 7-17, 2010.
[3] "X2 Industries," [Online]. Available: http://www.x2industries.com/images/66T30-4_v1_20091218.jpg. [Accessed 17 June 2015].
[4] "Motorcycle Rider," [Online]. Available: http://www.motorcyclerider.de/wpcontent/uploads/2011/12/tractive_03.jpg. [Accessed 17 June 2015].
[5] "ExtremeTech," PCMag Digital Group, [Online]. Available:
http://www.extremetech.com/wp-content/uploads/2011/09/Bose-suspensionsystem-8_edited.jpg. [Accessed 17 June 2015].
[6] R. Gemilang, "Modeling, Simulating, and Analyzing an Overhead Crane Using
Matlab Simulink and V-Rep Simulator," Swis German University, Tangerang,
2014.
[7] M. Avesh and R. Srivastava, "Modeling Simulation and Control of active
suspension system in Matlab Simulink environment," IEEE, Allahabad, 2012.
[8] A. Agharkakli, G. S. Sabet and A. Barouz, "Simulation and Analysis of Passive
and Active Suspension System Using Quarter Car Model for Different Road
Profile," International Journal of Engineering Trend and Technology, vol. 3, no.
5, pp. 636-644, 2012.
[9] M. F. Ismail, K. Peng, N. Hamzah, Y. M. Sam, M. K. Aripin and M. H. C.
Hasan, "A Linear Model of Quarter Car Active Suspension System Using
Composite Nonlinear Feedback Control," IEEE, Pulau Pinang, 2012.
APPENDICES
Dc1 = [D1];
H2 = ss(Ac1,Bc1,Cc1,Dc1);
%Insert new State Space Components with K for Wheel Displacement
Ac2 = [(A2-B2*K)];
Bc2 = [B2];
Cc2 = [C2];
Dc2 = [D2];
H3 = ss(Ac2,Bc2,Cc2,Dc2);
%Make the amplitude of step input = 0.1
opt = stepDataOptions('StepAmplitude',0.1);
%Plotting the graph
figure
subplot(2,2,1)
step(X1,opt)
title('Car Body Displacement')
subplot(2,2,2)
step(X2,opt)
title('Wheel Displacement')
subplot(2,2,3)
step(X1,opt)
title('Suspension Travel')
figure
subplot(2,2,1)
step(H2,opt)
title('Car Body Displacement')
subplot(2,2,2)
step(H3,opt)
title('Wheel Displacement')
subplot(2,2,3)
step(H1,opt)
title('Suspension Travel')
gamma = 0;
theta = 0;
zs = 1;
zu1 = 0;
zu2 = 0;
zu3 = 0;
zu4 = 0;
gammadot = 0;
thetadot = 0;
zsdot = 0;
zu1dot = 0;
zu2dot = 0;
zu3dot = 0;
zu4dot = 0;
C1 = gamma*[1 0 0 0 0 0 0 0 0 0 0 0 0 0];
C2 = theta*[0 1 0 0 0 0 0 0 0 0 0 0 0 0];
C3 = zs*[0 0 1 0 0 0 0 0 0 0 0 0 0 0];
C4 = zu1*[0 0 0 1 0 0 0 0 0 0 0 0 0 0];
C5 = zu2*[0 0 0 0 1 0 0 0 0 0 0 0 0 0];
C6 = zu3*[0 0 0 0 0 1 0 0 0 0 0 0 0 0];
C7 = zu4*[0 0 0 0 0 0 1 0 0 0 0 0 0 0];
C8 = gammadot*[0 0 0 0 0 0 0 1 0 0 0 0 0 0];
C9 = thetadot*[0 0 0 0 0 0 0 0 1 0 0 0 0 0];
C10 = zsdot*[0 0 0 0 0 0 0 0 0 1 0 0 0 0];
C11 = zu1dot*[0 0 0 0 0 0 0 0 0 0 1 0 0 0];
C12 = zu2dot*[0 0 0 0 0 0 0 0 0 0 0 1 0 0];
C13 = zu3dot*[0 0 0 0 0 0 0 0 0 0 0 0 1 0];
C14 = zu4dot*[0 0 0 0 0 0 0 0 0 0 0 0 0 1];
C = [C1; C2; C3; C4; C5; C6; C7; C8; C9; C10; C11; C12; C13; C14];
D = [0 0 0 0 0 0 0 0; 0 0 0 0 0 0 0 0; 0 0 0 0 0 0 0 0; 0 0 0 0 0 0 0 0; ...
0 0 0 0 0 0 0 0; 0 0 0 0 0 0 0 0; 0 0 0 0 0 0 0 0; 0 0 0 0 0 0 0 0; ...
0 0 0 0 0 0 0 0; 0 0 0 0 0 0 0 0; 0 0 0 0 0 0 0 0; 0 0 0 0 0 0 0 0; ...
0 0 0 0 0 0 0 0; 0 0 0 0 0 0 0 0];
H1 = ss(A,B,C,D);
VC = 100000;
Q = VC*(eye(14));
R = 0.1*eye(8);
[K,S,e] = lqr(A,B,Q,R);
A12 = [(A-B*K)];
B12 = [B];
C12 = [C];
Mochammad Rizky Diprasetya S
D12 = [D];
H2 = ss(A12,B12,C12,D12);
figure
step(H1,H2)
() = [ 11
21
12
] [()] + ()
22
()
= [
]
4
4 ]
11
0
0
0
= 0
0
0
[0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0]
12
1
0
0
= 0
0
0
[0
0
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1]
()
= [
21
2( )
0
0
2( ) 2( + )
1
2( + )
2( )
1
1
]
22
2( )
0
0
=
2( ) 2( + ) 1
2( + )
2( )
1
1
]
0
0
0
0
0
0
0
0
0
() = 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3
]
[4
() = [ 11
21
12
] [ ()] + () + ()
22
0
0
0
0
0
0
0
() =
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
2
[ ]
1
3
1
]
CURRICULUM VITAE
Name
Address
: rsurdia@gmail.com
Mobile Phone
: +628561100059
2011 2015
Mechatronic Engineering
Politeknik Bandung
2009 2011
Science
Work Experiences
Fachhochschule Sudwestfallen
Laboratory Assitant
Mechanic
Skills
Computer Skills: C++ Programming, MATLAB, V-REP Simulator, Microsoft Office
Language Skills : English
Fluent
German
Basic
Indonesian
Mother Tongue