See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/307959592

Dynamic Modelling and Control Design for a Vehicle in its Longitudinal Motion

Article in Indian Journal of Science and Technology · August 2016

DOI: 10.17485/ijst/2016/v9i30/99019

CITATIONS READS
2 3,532

2 authors:

Kirti swarup Jena Vijay Joseph

SRM Institute of Science and Technology SRM Institute of Science and Technology
2 PUBLICATIONS 9 CITATIONS 9 PUBLICATIONS 18 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Vijay Joseph on 05 February 2017.

The user has requested enhancement of the downloaded file.

 ISSN (Print) : 0974-6846
Indian Journal of Science and Technology, Vol 9(30), DOI: 10.17485/ijst/2016/v9i30/99019, August 2016 ISSN (Online) : 0974-5645

Dynamic Modelling and Control Design for a Vehicle

in its Longitudinal Motion
Kirti Swarup Jena1* and Arockia Vijay Joseph2
Department of Electronics and Control Engineering, SRM University, Chennai - 603203, Tamil Nadu, India;
1

[email protected]
2
Department of Electronics and Instrumentation Engineering, SRM University, Chennai - 603203, Tamil Nadu, India;
[email protected]

Abstract
Background/Objectives: The purpose of developing a model based control study is to understand the longitudinal
dynamics in detail. This paper develops mathematical model of longitudinal dynamics of the vehicle and develops a suitable
control methodology to have a reliable safe system. Methods/Statistical Analysis: Here the longitudinal dynamics of the
system is modeled using white box modeling and suitable PID control algorithm is developed to have a stable dynamics
of the systems. To develop the required model Physical laws of equations are used. Findings: For a tuned PID controller,
the system stayed within the set point of acceleration and the change in throttle has not affected the system’s velocity
and acceleration. Applications/Improvements: This model can be imparted in the driverless car models for complete
analysis. Modern control methodologies can be developed for the developed model to have a better performance.

Keywords: Longitudinal Dynamics, Longitudinal Motion, Mathematical Model, Physical Laws of Equations, PID Control,
White Box Modeling

1. Introduction velocity using PI controller. In general, studies on the

design of autonomous vehicle involves two types of con-
Research in the field of Advanced Vehicle Control System trol,
(AVCS) has increased in large numbers due to the devel- • The steering control (lane keeping)
opment of Intelligent Transportation Systems (ITS) in last • The longitudinal control
decade. Integration of modern model-based automatic This paper exclusively deals with the longitudinal con-
control systems is an important research area. Model trol of the vehicle. The proposed technique will reduce the
based automatic control system describes the dynamics space between vehicles and also avoids accidents due to
behavior of an automobile and accesses the unknown frontal collision. In this paper we use the conventional
parameters. Vehicles with close spacing require a high controller (PID) to control the velocity of the vehicle in
performance based longitudinal controller. Safety of the order to avoid collision and accidents.
passengers is the prime motive in developing any lon- This paper is categorized as follows. In section 2
gitudinal controller. As safety is the important concern we describe the approach towards the modeling of the
for designing the controller, uniform spacing should be required vehicle model. After analyzing the vehicle
ensured during the whole duration of driving. In this model, in section 3 we design the control algorithm and
paper we are focusing on a single vehicle using the con- test in the developed vehicle model. We discuss the
ventional control algorithm. Jullierme Emiliano1 presents experimental results in section 4 and finally section 5
the longitudinal model identification which benefits to concludes the paper with suggestions to be considered for
design the longitudinal vehicle model and control the future research.

*Author for correspondence

 Dynamic Modelling and Control Design for a Vehicle in its Longitudinal Motion

2. Modeling to the vehicle is the summation of different individual

forces
r such as Viscous force
r ( F viscuss ), Aerodynamic force
The proposed system is designed in order to regulate the ( Faero ), Braking force ( Fbrake ) and Climbing force or the
r
vehicle speed so as to avoid collision and accident. The Downgrade force ( F g ).This force is also called as gravi-
system is used to follow the command based on the feed- tational force. These forces plays a role of deciding factor
back signal. The free body diagram of the Car is shown for calculating the force exerted in the vehicle to move the
in Figure1. The equation from this free body diagram is vehicle or for the motion of the vehicle. The mathematical
represented in equation 1. The vehicle has different forces equation to explain the law of motion is given in equation
such as Aerodynamics force, Braking force, Viscous Force 1.
r r r r r r
or frictional force and engine force. Except the engine t ( mv ) = FVISC + FAERO + FBRAKE + FGRAV + FENGINE
d
d
dynamic force all the forces opposes the movement of the (1)
vehicle that means all other forces are opposite in direc- From equation 1 ‘M’ represents the mass of the vehi-
tion to that of the engine force. So in order to move the cle and ‘V’ represents the longitudinal velocity. The force
vehicle the engine force should be much higher in com- equation is posturized in the free body diagram in Figure1.
parison to all other forces shown in free body diagram. Aerodynamics force, viscous force and braking force as
As we are concern about the design of the controller so always act on the opposite direction of the movement of
the modeling of the vehicle will not take principal role. the vehicle. To design the required system assumption has
Our aim is to design the controller in order to fulfill the to be made for a flat and leveled surface and the brake
required task, so by considering the system from2 and condition has been eliminated. For the above condition
adapting certain derivation, we can realize the simple the force related to gravitational, brake, andr aerodynam-
linear transfer function model for designing the control- ics are neglected .Now the engine force ( FENGINE ) is the
ler. Frank3 describes Longitudinal Control for Automated only force used to run the car. Which affects the motion
Automobiles and Trucks Operating on a Cooperative of the vehicle. The transmission system consists of gear
Highway. presents adaptive control for interconnected box, clutch, speed and torque conversions etc. The trans-
nonlinear dynamics system4. Cemhatipoglu5 presents mission system of a vehicle is a very complicated system.
a smooth Intelligent Cruise Control (ICC) structure. For driving the wheels the transmission system plays an
Reader’s wants a depth study on the model can refer to6 important role by relating the output. So it is a matter of
which deals with the different types and methods of mod- challenge to design the mathematical model of the trans-
eling technique7. The longitudinal vehicle model by using mission system.
different physical equations. The longitudinal dynamics In this paper simplification proposed by Jullierme
of the vehicle model using Simulink platform8. solves the Emiliano and Khairuddin Osman is adapted by taking in
problem in automated highway by designing the vehicle to account to the difficulties faced in transmission system.
model and the controller9. Jullierme tells the use of the first order linear system .The
system consists of a time constant ‘T’, static gain ‘K’, and a
time delay ‘𝞽’.The input of the system also has been limited
to a certain range .The throttle signal which correlate with
the model’s input has been restricted to the sluggish speed
of the vehicle and to the maximum value which is most
appropriate for the safety purpose. Analyzing equation
1 in frequency domain the longitudinal dynamics of the
car has been presented in the block diagram in Figure1.
Now simplifying the block diagram shown in Figure2, the
transfer function of the system is presented in equation 2.
Figure 1. Various forces acting on a vehicle.
k −t s
e
The longitudinal dynamics of a vehicle is designed by v (s ) mT (2)
G( s ) = =
referring the Newton’s law of motion. As per the consid- u(s) 1 B
(s + )(s + )
eration of Newton’s law of motion the force being exerted T m

2 Vol 9 (30) | August 2016 | www.indjst.org Indian Journal of Science and Technology
 Kirti Swarup Jena and Arockia Vijay Joseph

Now the transfer functions we get by solving these

equations are

C1 −t s
e
v (s ) mT
= (9)
u(s) 2c v 1
(s + a )(s + )
M T
The time delay also has been simplified by using the
Figure 2. Process block diagram.
power series expansion.
where v(s) and u(s) are the output and the input of the 1
e-𝞽s = (10)
system respectively. V(s) is the velocity of the vehicle and (1 + ts )
u(s) is the throttle input to the system. Again considering
Substituting above equations in the transfer function
the symbols represent in the study that is ‘m’ shows the
and by substituting the adapted constant values we got the
mass of the vehicle and the passengers. The above trans-
transfer function as
fer function is the final transfer function of the system.
Different parameter values are adopted from the research v( s ) 2.4767
G (s) = (11)
done by frank. These parameters constant values are dis-
=
u ( s ) ( s + 0.0476)( s + 1)( s + 5)
played in Table.1 .After getting the simplified model the
From the above transfer function it shows a third
controller design has been started. The vehicle longitudi-
order type‘0’ system. Due to the time delay approxima-
nal dynamics obtained by Khairuddin Osman is shown in
tion, the important fact in realizing the system is that the
Figure 2. Now assume all the initial condition to be zero.
system has been linearized, as the system has been linear-
Applying the initial conditions that are zero wind gusts
ized so the further calculation will become much simpler
and no grading the model is converted to only the for-
as compared to the nonlinear system.
ward path gain and the feedback as shown in Figure 2.
Now the output state equations are represented in equa-
tion 3 and equation 4.
1
v/ = ( Fd − c a v 2 ) (3)
m

1
Fd = (c1 u (t − T ) − Fd ) (4)
T
Y=v (5)
Due to the square term present in equation 3 nonlin- Figure 3. System Block diagram.
earity arises, to make this nonlinear equation in to linear,
we have to differentiate each terms. After differentiating
the equations it becomes as represented in equation 6 and
equation 7.
d 1
v ′ = (−2ca vd v + d Fd ) (6)
dt m
Figure 4. Switching logic to control the process.
d ′ 1
Fd = (c1d u(t − T ) − Fd ) (7)
dt T
III. Control Schemes
And the output equation becomes
In this research the velocity controller has been proposed
Y = d v (8) using conventional control techniques. The focus of this

Vol 9 (30) | August 2016 | www.indjst.org Indian Journal of Science and Technology 3
 Dynamic Modelling and Control Design for a Vehicle in its Longitudinal Motion

paper is to control the velocity of the vehicle as per the 1 t de (t )

compared values between the set point and the feed- u (t ) = k (e(t ) + ∫ e(t )dt + Td )
Ti 0 dt
back velocity. The set point limit has been obtained from ------ (12)
the literature survey10 which gives the different comfort From the Figure5 ‘y’ is the measured process vari-
limit. The model has been analyzed in the previous sec- able, ‘r’ is the reference variable, ‘u’ is the control signal
tion which is used as the process for the controller. As and ‘e’ is the error. The control signal generated from the
per the controller we need to control the input variables controller has the combination of 3 terms that is P-term,
to the process and so from the previous section. The con- I-term and the D-term.
trol algorithm uses vehicle throttle position and brake as
two control parameter for the control of the velocity of
the vehicle. The controller used in this technique is the
PID controller which behaves linearly. This chosen PID
controller is adaptable in structure, flexible and tuning
compatible. The block diagram of the proposed model
has been shown in Figure 3. The controller schemes has
been classified in conventional, intelligent category. PID
controller is one of the conventional control schemes
used most commonly. The Invoke of throttle setting and Figure 6. Overview analysis of the system.
the brake setting has been controlled by a switching logic
shown in Figure 4. It decides the required controller to Different control parameters are proportional gain
switch in to the said parameters for the motion of the (K), Integral gain (Ti), and derivative time (Td). These
car. The development of dynamic and robust traffic man- three parameters belong to proportional, integral and
agement system based on fuzzy logic approach11. Basic derivative part respectively which deals with past, pres-
methods for computer formulation and solution of the ent and future. The controller block diagram is shown in
equations of motion12. Figure 5 with Ti =1, Td=1 and K=1. The visualization of
adaptive cruise control has been presented in Figure 6.
The speed of the vehicle has been set to a particular speed
after detecting a vehicle in front of the host Vehicle.
By designing the controller with the mathematical
equations will not fulfill the required task. Some parame-
ters have been considered in order to achieve the required
control action. These parameters are
• Noise filtering and high frequency roll-off
• Set point weighting
Fig.5. Block diagram of PID controller • Wind up
• Tuning
In 1940 when process control introduced it has been • Computer implementation
become the most common element for different con- The PID controller transfer function can be described
trol algorithm. Application of PID controllers is used by equation 13.
more commonly. Different function blocks such as logic 1
sequential functions, selectors and simple function blocks G(s) = k(1 + + sTd ) (13)
sTi
are combined in order to generate the PID controller.
Different applications of PID controller are energy pro- The equation 13 describes the non-interacting control-
duction, Transportation and manufacturing. In the lowest ler. Sometimes it is claimed that the interacting controller
level of the system the controller has been used in order is easier to tune manually. If a different controller changed
to compensate the output. PID controller is an important the PID controller then the controller also has to change.
component of every design box. The PID controller is Another representation of the PID controller is shown
illustrated by the equation. in equation 14

4 Vol 9 (30) | August 2016 | www.indjst.org Indian Journal of Science and Technology
 Kirti Swarup Jena and Arockia Vijay Joseph

ki
G ′′(s) = K + + sK d
s (14)

Figure 8. Control unit with different modes.

Figure 7. Speed of vehicle with respect to Throttle setting. of the controller which is the key goal of the paper.
In order to design the controller, the first experiment was
The equation 14 is similar that of the equation 13. to validate the model. After getting the vehicle model
Only the parameters used are different. This equation is we implement the PID controller in order to control the
basically used for analytical calculation because of the vehicle speed. The diagram shows in Figure 3 defines the
linear appearance. Proportional, Integral or derivative block diagram of the controller and the vehicle mathe-
action can be prevail from the above given equation by matical model. The controller has been designed in order
varying the parameter values. Using integral action in a to assure safety and comfort to the driver and the pas-
controller will lead to increasing error, which results the sengers. The output has been obtained by running the
integral term will become very large. Wind up technique simulation. The switching logic is used to switch between
is used to reduce the integral action error, results due to the throttle and brake controller. Different values of Kp, Ti
the above conditions based on the Figure 5. Throttle and and Td are used in order to get the settling time quickly. In
breaking action has been selected by the switching logic this research the input is the throttle percentage that is the
technique. The switching criterion used in this work con- accelerator to increase the speed and the brake applied
siders the required acceleration to achieve the velocity. is used to control the speed of the vehicle. The analysis
The throttle loop is choosen when the required velocity is has two different modes. One is manual mode another
high and the brake is selected when the required velocity is cruise control mode. In the manual mode the vehicle
is very low. Between certain ranges the hysteresis loops is controlled by the driver. In automatic mode if certain
comes in to play. The error between the set point velocity condition arises like the driver is tired and wants to put
and the actual velocity is calculated and then processed as the vehicle speed in automatic condition, then cruise con-
the variable to the PID controller with KP =1, Ti=1, Td=1 trol mode is activated. In cruise control mode the driver
respectively. These are used because of the limit comfort will not able to control the speed of the vehicle until he
to the passenger. Filter is required because the vehicle applies the brake. When the driver applies the brake the
doesn’t permit the high frequency noise. cruise control mode will automatically switched to man-
ual mode. This situation is well known as Ideal condition.
3. Result Analysis Figure 7 shows the throttle opening and closing and
corresponding vehicle speed. The road profile has been
The whole system is modeled and controlled using the described in Figure 9.This has been designed in order to
graphical programming language LabVIEW. The vehicle develop the longitudinal controller. Figure 8 shows the
model is the first part of the paper. The vehicle modeled different modes of the system. The simulation have been
has been designed by two conventional methods. The first verified for certain range of speed.
method is by implementing the traditional Newton’s law
of motion and the Second method is to find the trans- Table 1. Parameter Estimation
fer function from1,4. This model depends on friction, slip, Constant parameters Assumed values
torque, aerodynamics. As our target of this paper is to
M 1500 Kg/
design the controller so these Parameters are neglected
.we basically focus on the designing Ca 1.19N/(m/s)2

Vol 9 (30) | August 2016 | www.indjst.org Indian Journal of Science and Technology 5
 Dynamic Modelling and Control Design for a Vehicle in its Longitudinal Motion

Constant parameters Assumed values Anindya gourav panigrahi. The authors would also like
T 1s
to thank Mrs. Sailabala Samal for her support during the
development of this work.
G 9.8m/s2.
C1 743,
6. References
τ 0.2s,
1. Emiliano J, Dias A, Pereira GAS. Longitudinal model
identification and velocity control of an autonomous car.
IEEE Trans Intelligent Transportation System. 2015 Apr;
16(2):776-86.
2. Osman K, Rahmat MF, Ahmad MA. Modeling and control-
ler design for a cruise control system. IEEE International
Colloquium on Signal Processing and Its Application. 2009
Mar; p. 254-8.
3. Frank AA, Liu SJ, Liang SC. Longitudinal control concepts
Figure 9. Road profile. for automated automobiles and trucks operating on a coop-
erative highway. Society of Automotive Engineers Technical
Paper Series.1989 Aug.
4. Conclusion 4. Sheikholeslam S, Charles A. Desoer in-direct adaptive con-
trol of a class of interconnected nonlinear dynamics system.
The vehicle model has been developed by using the physi- International Journal of Control. 1993; 57: p. 743-65.
cal law of equations like Newton’s law of motion. The 5. Hatipoglu C, Ozguner U, Sommer ville M. Longitudinal
developed model is controlled using the designed conven- headway control of autonomous vehicles. IEEE International
tional PID controller, which is used to control the speed Conference on Control Application; 1996 Sep. p. 721-6.
of the vehicle. From the result analysis, we could find the 6. Rajamani R. Vehicle Dynamics and Control. 2nd ed.
developed controller has better control over the longitu- Minneapolis, MN, USA: Springer-Verlag; 2012.
dinal dynamics in terms of speed than that of the PI based 7. Short M, Pont MJ, Huang Q. Safety and Reliability of
controller. The vehicle model used in this paper doesn’t Distributed Embedded System: Simulation of Vehicle
involve external forces like aerodynamics forces, viscous Longitudinal Dynamics. 2004 Oct. p. 1-21.
forces etc. So the future work may be focused on develop- 8. Shakouri P, Ordys A, Askari M, Lajla DS. Longitudinal
vehicle dynamics using Simulink/matlab. UKACC-
ing an appropriate model considering the all above said
International Conference on Control; 2010 Sep. p. 1-6.
parameters for better representation of the actual con-
9. McMahon DH, Hedrick JK, Shladover SE. Vehicle mod-
sidered vehicle. The simulation has been verified using eling and control for automated highway system. IEEE
the rapid prototyping software called LabVIEW. The American Control Conference; 2002. p. 297-303.
results are analyzed in section 5. Research can be done to 10. Yi K, Chung J. Nonlinear brake control for vehicle CW/
improve the speed profile of the vehicle by using intelli- CA systems. IEEE/ASME Trans Mechatronics. 2001 Mar;
gent controller such as Advanced Control Systems which 6(1):17-25.
may result in assured safety, stability and improved per- 11. Mittal P, Sing Y. Development of intelligent transportation
formance. system for improving average moving and waiting time
with artificial intelligence. Indian Journal of Science and
Technology. 2016 Jan; 9(3):1-7.
5. Acknowledgment 12. Prem JKM, Gopalakrishnan SJ, Satheesh B, Anbazhagan
R. Computer modeling of a vehicle system. Indian Journal
The authors gratefully acknowledge the support by P.
of Science and Technology. 2013 May; 6(5S):4620-8.
Ramani Ranjan Senapati, Prasant Kumar Maharana and

6 Vol 9 (30) | August 2016 | www.indjst.org Indian Journal of Science and Technology

View publication stats