CONTROL DESIGN AND ANALYSIS OF CRUISE

CONTROL SYSTEM

1. Introduction
A system is said to be a cruise control system if it has technical accuracy with an average error of 0.065m achieved.
skill to maintain a certain set speed for a particular length of Advanced cruise control that is already in use includes
time in as much there won’t be any human intervention. This adaptive cruise control, antilock braking, automatic parking,
phenomenon would be studied in a car as a system. Also, this pedestrian protection system and blind spot detection. The
tremendous technology is very important at highway where a contributions of designing a control system for cruise control
constant speed is to be maintained, and as such will reduce was achieved by which involved modeling the adaptive
driving fatigue and as the case may be, the driver won’t need cruise control in Petri Nets, analyzing the functionality of
to tamper with any car components as at when the cruise Petri nets, verifying the adaptive cruise control process, and
system is activated. In real life, cruise control can be activated implementing controller and fault tolerance techniques.
when a desired speed is reached, and immediately the driver Cruise control can also be employed to control the speed of
presses the ON button on the steering mounted controls, the robotic systems .
system is automatically deactivated due to interruption from This work is aimed at the modeling and control design of
the driver. Cruise control system has a lot of advantages and a cruise control system to achieve a fixed set of design
some of these are: Comfort diving as the driver does not need specifications such as steady state error, settling time,
to press accelerator during long journey; the system makes percentage overshoot.
fuel consumption to be more economical and the driving
fatigue equally reduced a great deal. Although, the system is 2. Methodology
not good when there is heavy traffic on the road. This is so
because; the driver cannot keep the vehicle at constant speed Automatic cruise control is achieved by measuring the speed
for long when there is heavy traffic . Cruise control is applied of a car and also monitors the speed by comparing it to the
to several automatic cars such as Honda to control the speed. reference speed no matter what the external disturbance is all
proposed in their studies, feed forward controller that take about. At the same time the throttle is to be adjusted according
into serious consideration, elimination of gravitational and to the established control law. From Figure 1, the mass of the
wind disturbances effect. This system was implemented with car is represented by m, which is acted upon by the force, pull-
AMR microcontroller based robotic application and C++ u that is generated at the interface between the tire and the
programing language. The use of GPS was later proposed not surface of the road. It is also assumed of the model that ‘u’ can
to control the speed alone but to operate the vehicle in safe be controlled while the dynamic of the power train is
speed mode. The GPS monitored the state of the vehicle neglected. The oppositional pull, R is assumed to be varied
speed at normal. ARM Processor used to reduce the vehicle linearly with velocity of the car and it acts opposite to the
speed to normal state whenever it receives signal of motion of the car.
abnormality from GPS. Presently, the state of the art according Horizontal resolution of vectored forces and application of
to, involves a cruise control system to predict future hazard Newton’s second law, the final model is given as: 𝒎𝒗̇ + 𝒃𝒗̇ =
and automatically adjust the system parameters, so that such 𝒖, and also, the output equation is assumed to be 𝒚 = 𝒗̇ since,
accident will not occur. Evaluation was done for the model the action of interest is controlling the speed of the car.
and the model detected changes in lane with 99.2%
 Fig. 1 Free Body Diagram and Side view diagram of a Car for Fig. 2 Simulation of Open loop control system
Cruise Control.
The system reached its set value of 10m/s at no oscillation and
Where m is the mass and b is the damping factor. Let no overshoot. Although the rise time is approximately 50s, this
these parameters of interest be, 𝑚 = 1500𝐾𝑔 𝑎𝑛𝑑 𝑏 = value is low, and it is required for a better closed-loop control
75𝑁. 𝑠/𝑚. The state model for the system is therefore: system.
𝑏 1
𝑣̇ = [− ] [𝑣̇] + [ ] [𝑢] (1)
𝑚 𝑚
and
𝑦 = [1][𝑢] (2)
The model arrived at is according to the FBD broken
down into the form below:
𝑑𝑣̇
𝑏𝑣̇ + 𝐹 = 𝑢, 𝑏𝑣̇ + 𝑚𝑎 = 𝑢, 𝑏𝑣̇ + 𝑚 = 𝑢, 𝑏𝑣̇ + 𝑚𝑣̇
𝑑𝑡
= 𝑢 (3)
Also, the transfer function for this system is derived below:
𝑉(𝑠) −𝑏
𝐺(𝑠) = = 𝐶(𝑠. 𝐼 − 𝐴)−1𝐵 + 𝐷, 𝑤ℎ𝑒𝑟𝑒 𝐴 = ,𝐵
𝑈(𝑠) 𝑚
1
= , 𝐶 = 1, (4)
𝑚
𝑏 −1 1 𝑏 −1 1
𝐺(𝑠) = 1 × (𝑠. 𝐼 + ) ( ) + 0 = (𝑠. 𝐼 + ) ( ) Fig. 3 The pole-zero plot of the system
𝑚 𝑚 𝑚 𝑚
1
1 (5)
= 𝑚 =
𝑏 𝑚𝑠 + 𝑏
𝑠+𝑚
G(S) = Transfer Function, U(S) = Input and V(S) = Output.

3. Analysis of the model

The beauty of any model is to be able to be mathematically
analyzed. The deductions arrived at will then help to fully
explain the nature of such system in real sense. Also, it will
help to see the system responsiveness with respect to time and
frequency. Besides, we can choose design parameters to
amend the system for stability, speed of response, steady state
error, and oscillatory- free system. The analysis of this system
is based on the following parameters and specifications:
steady state error is less than 1%, overshoot is < 5% and rise
time < 1.5 𝑠𝑒𝑐𝑜𝑛𝑑𝑠. The response of the system to a step
input force of 750N is first simulated and the graphical
illustration is shown in Figure 2.

Fig. 4 The Bode plot of the system.

The pole-zero plot of the equivalent open loop system in
Figure 3 shows that there is no oscillation due to the fact that
only a pole exists as real and negative. Now, a controller to
meet the larger value of b/m should be designed because, the
greater the value of 𝑏 , the faster the system becomes to reach
𝑚
the steady state.
As shown in Figure 4, the Bode plot exhibited a system
property with -61.6dB of magnitude and -87.1° phase at the
corner frequency of 0.803 rad/s and it roll-off at high
frequencies of 0.992 rad/s.

3.1 PID Controller for the System

Figure 5 is a typical PID closed loop representation of the
system, and the consideration is given to the unity system.
The PID of this system has a standardized mathematical Fig. 6 Root locus plot of the system
equation:
𝑑𝑒 If overshoot of 5% is required, a 𝜉 > 0.7 would be needed
𝑢(𝑡) = 𝐾𝑝𝑒( 𝑡) + 𝐾𝑖 ∫ 𝑒( 𝑡)𝑑𝑡 + 𝐾𝑝 (6) while the rise time of 1s is possible; a natural frequency above
𝑑𝑡 1.8rad/s would be fine as shown in Figure 7. The desired poles
on the locus is chosen, a P-only controller is needed for the
poles to the be in the desired region. The value chosen for K is
2822.

Fig. 5 The PID controller for the system

From Figure 5, the reference input and measured output are

compared to generate the error (e) which is fed to the
controller, at which the error is administrated through
proportional parameter, also integrated and finally
differentiated to generate control output (u) that is later
presented to be new input to the plant (or system or model).
The final actual output generated is released to be further
compared with the desire input. The process continues on and
on like that.
Fig. 7 Root locus plot of the system for a required overshoot of 5%
The Laplace function of the PID equation above is
manipulated through to generate the result shown below,
when all initial conditions are assumed zero
𝐾𝑖 𝐾𝑑𝑠2 + 𝐾𝑝𝑠 + 𝐾𝑖
𝐾𝑝 + + 𝐾𝑑𝑠 = (7)
𝑠 𝑠
The PID controller will be able to do the task: the
proportional part will reduce 𝑡𝑟 and 𝑒𝑠𝑠, the integral part will
eliminate the error but the transient response will slow down,
and the derivative part will drive the system to stability,
reduce the overshoot and improve the transient response.

3.2 Root Locus Control Design of the System

Figure 6 shows the root locus of the system.

Fig. 8 The closed loop system response has 𝑀𝑝 < 5% and a 𝑡𝑟 of

0.178s.

From Figure 8, the response has an overshoot (𝑀𝑝) < 5% and

a 𝑡𝑟 <1 second (0.178 Seconds). The design specifications
relating to the overshoot and rise time have been therefore
achieved using root locus control.

3.3 Frequency Domain for the System Design

In this section, the system and parameter specifications are
unchanged to carry out control design using frequency
domain methods. A unity feedback system is equally
considered. The specifications and design parameters are
given as: step input force = 750N; m=1500Kg; b=75N.s/m;
desired speed of 10𝑚𝑠−1. The performance specifications
are given as: 𝑡𝑟 < 5 sec; 𝑀𝑝 < 10%; and 𝑒𝑠𝑠< 2%. The
frequency response is examined by the use of open loop
system. The transfer function for the system is:
𝑌(𝑆) 𝐾𝑝
= , 𝑤ℎ𝑒𝑟𝑒 𝐾𝑝 = 1 (8)
𝐸(𝑆) 𝑚𝑠 + 𝑏
Also, the system must be stable for Bode plot to be examined
with the system. For the system, the low frequency gain is
37.5dB = 0.00101 which would be:
1 1
𝑠𝑠 𝑒𝑟𝑟𝑜𝑟 = × 100% × 100% =
1+𝑀ѡ→0 1+0.00101 Fig. 10 Step-response of the system having met all the design
0.9989 × 100% = 99.89% (8), specifications.

4. Conclusion
This work involved the modeling and control design of a
cruise control system with set specifications of steady state
error less than 1%, overshoot less than 5% and rise time less
than 1.5 seconds achieved using the developed model with
the aid of MATLAB. A future work on cruise control system
will be to predict future incidence, suggest possible solution
to solve the problem, and also, interact with neighborhood
vehicles on the highway and automatically adjust itself to the
form that would be safe for the passengers, driver of the
vehicle and other road users. The use of more sophisticated
sensors with higher switching speed semiconductor materials
with the knowledge of AI and machine learning techniques
will be required. In fact, such a system having inherent
attributes will be suitable to be used as a self-driving robotic
vehicle.

Fig. 9 Bode plot of the system

Low frequency gain needs to be increased, so that the steady-

state error could still be improved. Specifically, the error is
less than 1.2% approximately. But it is equally good if it is
lesser than that. Let 1/(1+M) = 0.11 and it implies M = 8.0909
= 17.98dB. Therefore, to reach the desired ss error using P
controller only requires a 𝐾𝑝 > 82.02 dB = 12616.76 = 12617
approximately. Figure 9 shows the Bode diagram of the
compensated open-loop system.
From Figure 9, the low frequency magnitude is
44.5dB. The behavior of the step-response diagrammatically
is shown in Figure 10. As shown, the system could be further
improved by a lag compensator