-
ISSN: 2395-0587 International Research Journal of Electronics
& communication Engineering
www.irjece.com Volume 1, Issue 4 of May 2015
_____________________________________________________________________________________
2015, IRJECE -All Rights Reserved Page -1
Regulation, Pole Placement & Tracking (RST) Robust
Controller for Automatic Highway System
Zain Anwar Ali1, Li Ning2, Iftikhar-uddin H.Farooqui3, Faheem H.
Rizvi4, Muhammad Faraz5.
1. College of Automation Engineering, Nanjing University of
Aeronautics & Astronautics, Nanjing, China. 2. College of
Electronic Information Engineering, Nanjing University of
Aeronautics & Astronautics, Nanjing, China.
4. Department of Telecommunication Engineering, Sir Syed
University of Engineering& Technology, Karachi, Pakistan 3
&5. Department of Electronic Engineering, Sir Syed University
of Engineering& Technology, Karachi, Pakistan.
Abstract The goal of this research is to develop a system which
can detect the vehicle, detect the lane and tracking by using
Automatic Highway System (AHS) model. For achieving this goal by
using a video processing technique as well as image processing to
detect the vehicle and lane tracking. The video processing is used
to detect the vehicle in a road and maintain a safe distance
between vehicles as well as image processing tool Hough transform
is used to detect the lanes of the road all the detection mechanism
is done by using color threshold checking and detect the vehicle
and a lane. Moreover, Regulation, Pole-Placement and Tracking
controller is designed for the non-linear dynamics in order to
control the vehicle dynamics and decision making. The overall model
of the system is designed in Matlab which shows that system
successfully works in the required scenario.
Keywords RST Controller, AHS System, Video & Image
processing.
1. Introduction China is the 4th biggest country in the Universe
with respect to its huge land area about 9327489.90 square km
according to World Bank [1]. According to the geographic area and
population, Chinas road network in 2002 extended roughly 1.76
million km, with some 25,130 km of expressways and about 27,468 km
of other high-grade highways in operation [2]. The daily basis
accident ratio rate is about 20.8 due to the human error moderate
rate of accidents as compared to other countries. The Smart Road
concept and Automatic Highway System (AHS) pays a great benefit to
the Economy. The Automatic Highway System is now become an active
research area in many years. AHS is an important area to avoid
accident and ensure passenger safety. By automating all the
vehicles on the road and by making those coordinating with each
other seamlessly will improve the passenger comfort in traveling.
Another advantage of a system which makes sure that the vehicle is
driving at the lawful speed and following all the other law
enforced by the state. This will avoid any fines and point on the
driver's license other than just improving the road safety [3].
Initially, AHS will probably be deployed and operated on
high-priority routes in high-demand major urban and intercity
freeway corridors [4]. And an AHS car will look like a normal car.
But both facility and road will be outfitted with sophisticated
control and communication devices that will essentially put the
vehicle in communication with the roadside. The car will "know"
what roadway conditions are like. The road will "offer" each
vehicle options, navigation, and advisories based on its
conditions. While on the AHS facility, the vehicle will be operated
under automated control--similar to the autopilot control in
aircraft [5]. AHS will be a collection of different systems working
together to achieve a collective task of automating a car's drive
on any highway [5]. In the long run, these different systems like
cruise control, GPS navigation, lane detection system, ABS system,
RF based interaction between vehicles and collision avoidance
system, etc, to interact with each other and a master system to get
guidelines and updates on road traffic in order to route the
journey effectively and drive on its own to the destination
[6].
-
ISSN: 2395-0587 International Research Journal of Electronics
& communication Engineering
www.irjece.com Volume 1, Issue 4 of May 2015
_____________________________________________________________________________________
2015, IRJECE -All Rights Reserved Page -2
The theme of this research is to develop an automated system for
highway that can detect the vehicle as well as lane track and
regulate the speed of the vehicle. The developed algorithm of this
paper is divided into two parts the detection of vehicle which is
done by using video processing and other is detection of lane
tracking by using image processing. The division of this article is
as follows. In section 2 visions of the complete system are defined
along with detection of vehicle and detection of lane tracking.
Section 3 defines the Automatic, Cruise Controller and in last
section covers the discussion and results are as follows.
2. Visioning of Complete System The color camera is equipped
with a car that record, the video of every 5 seconds regularly and
after recording that video the video reader will read that video
and taking all parameters of the video and convert it into video
frames and gives images after that convert RGD to grayscale and
checking the threshold. If the threshold value is greater than 55
it detects the dark object and neglect it and again goes to the
first step. If the threshold of the image is in between 45-55 it
means detect light color and detect a vehicle and dynamic model of
AHS start working and gives the signal to the vehicle controller
and maintain a safe distance between another vehicle. If the
threshold value is between 35-44 it will detect lane tracking and
after that detection of the edges in that image and by applying
Hough transform boundaries of the road is found and it gives the
signal to the AHS model and after that car controller will adjust
the car in a lane or follow a lane tracking. As shown in figure 1
before Conclusion.
2.1 Detection of Vehicle The vehicle detection is one of the
major tasks in this research by using video processing. Object
detectors such as face and pedestrian detection are among the
well-researched domains. But the vehicle detection algorithm
typically uses extracted features and learning algorithm to
recognize instances of a vehicle. Vehicle detection has many
applications in different areas like traffic controlling and
surveillance, etc. The reference object in a scene using feature
characteristic and similarity. The "ransac" command can be used to
calculate approximately the locality of the object in the test
image. Many other approaches like an image segmentation gradient
based, derived based and template based approaches may also be used
for that.
Taking "avi" format video file as an input from the camera
storage. After reading that file and getting information from
video. Matlab provides a platform for reading a video and create an
object that can take every information about that video clip. Then
this video clip can play using implay command. Two gray scale
pictures from the 74th frame is taken one is for the image
processed and other is for input. The threshold value is used to
process known as keyword. The intensity of some pixel lower than
the threshold will be discarded and processed into dark or black.
55 is the threshold value, and it is a modest bit above the
middling intensity of the dark color object. For processing the
image through most of the dark objects removed, but there is some
leftover of the dark object.
The lane marks are not touching at all because their pixel
values are above the threshold level. Any object size is smaller
than the defined size, length and width will be discarded. Not only
is the small objects also the dark objects and lane marks is also
obsolete due to the width smaller than the "disk". The disk is
created by the function "steel" and its function is sizing the
structure of elements such as the shape, square and lines of the
ball, etc. "imopen" command open the morphological binary image or
grayscale with the "SE" element structure. The structure of the
element "SE" should be single and opposite to an array of an
object. The whole procedure is used to detect multiple cars in a
road. After that maintain a safe distance between them.
-
ISSN: 2395-0587 International Research Journal of Electronics
& communication Engineering
www.irjece.com Volume 1, Issue 4 of May 2015
_____________________________________________________________________________________
2015, IRJECE -All Rights Reserved Page -3
Figure No: 02 Detection of Vehicle
Figure No: 03Converting into Binary image
2.2 Detection of Lane Tracking Firstly, we find out the edges of
the image using edge detection algorithm. Many techniques are there
for edge detection like Sobel, Canny, Prewitt or Roberts. But we
use canny to convert an image output into binary, it gives the
matrix of Boolean values equivalent to the edges. After that step
using Hough transform to detect the lines.
= x * cos () + y * sin() (1)
The equation (1) help to map points in the Cartesian picture
co-ordinate. Where "" and " " represent the rows and column
respectively. The Hough transform output is used for line segment
adjustment to contact with the image boundary outlines and after
that calculate the Hough lines. The Cartesian coordinate finds by
the Hough line in image or video processing by locating the
collision between the lines that is characterized by the parameters
"" and " " and reference image boundaries. The image is
reconstructed by computing endpoints to draw the polygon. The
detection of lines from the sides of the polygon. The original
video or image is overlaid and simulate the AHS model to verify the
detection and tracking of the vehicle.
Figure No: 04 Image of China Highway
-
ISSN: 2395-0587 International Research Journal of Electronics
& communication Engineering
www.irjece.com Volume 1, Issue 4 of May 2015
_____________________________________________________________________________________
2015, IRJECE -All Rights Reserved Page -4
Figure No: 05 Gray Scale Conversion
Figure No: 06 Binary Image
Figure No: 07 Hough Transform
3. Automatic Cruise Controller In this part of our paper
convinced that the vehicle maintains a secure space from the master
vehicle. This could
be done by pursuing the velocity of the leading vehicle and the
maximum speed of the road which we were taken from the highway is
120km/hr. The dynamics of master and slave vehicle is taken from
[7].
x = [, v,f] (2) The "" sign shows the space between master and
slave vehicle the velocity of the vehicle is defined by "v" the
driving and force is defined by "f". . Equation (3) defines the
spacing between vehicle 1 & 2 and equation (4) defines the
spacing between vehicle 2 & 3. Same procedure apply as
follows.
-
ISSN: 2395-0587 International Research Journal of Electronics
& communication Engineering
www.irjece.com Volume 1, Issue 4 of May 2015
_____________________________________________________________________________________
2015, IRJECE -All Rights Reserved Page -5
= x x (3) = x x (4)
The positive sign shows the force between vehicles and the
negative sign shows the braking between vehicles.
The dynamics of the system in longitudinal displacement shown as
below.
= v v (5) 1 = v v (6) v1 =
(Av d + f) (7)
v2 =
(Av d + f) (8) f1 = (f + u) (9) f2 =
(f + u) (10)
The above equations shows the orientation of master vehicle
called as vehicle "one" along with secondary vehicle
called as vehicle "two". After that secondary vehicle act as
master vehicle, then secondary vehicle called as a vehicle "three"
follow the same placement and distance. Taking mass as constant
about 1500kg for all vehicles. The engine time constant is 0.15
seconds.
Step signal is applied to the input of the system, it defines
the system is velocity dependent. A reasonable spacing
of the vehicle is defined by using master slave configuration
and it spacing between 18kph to 20kph. The major system
implementations are done by using simulink Matlab. The robust RST
controller is used as a main controller, which can help us to run
the system smoothly. The RST controller is used for linear system
dynamics, but also used for non-linear dynamics. The responses of
the system show stability.
Figure No: 08 Simulink Model of System
-
ISSN: 2395-0587 International Research Journal of Electronics
& communication Engineering
www.irjece.com Volume 1, Issue 4 of May 2015
_____________________________________________________________________________________
2015, IRJECE -All Rights Reserved Page -6
Figure No: 09 Master & Secondary Vehicle Tracking w.r.t
Velocity & Time.
Figure No: 01 Flow Diagram of Complete System
Video Recording
Video Parameter Reading
Video Reading
Frame in to Image
RGB Convert to Grayscale
Threshold Color > 55
Detection of dark object
Threshold > = 45-55
Threshold > = 35-44
Threshold > = 0-34
Vehicle Detect
Dynamic Model AHS
Vehicle Controller
Lane Tracking
Edge Detection
Hough Transform
Boundaries Found
-
ISSN: 2395-0587 International Research Journal of Electronics
& communication Engineering
www.irjece.com Volume 1, Issue 4 of May 2015
_____________________________________________________________________________________
2015, IRJECE -All Rights Reserved Page -7
CONCLUSION In this article, a video processing based vehicle
spacing or vehicle distance maintain technique is presented
along
with image processing is also used for detecting the lanes and
tracking of the system by using Hough transform canny technique is
used to detect lanes of the road. The proposed system works
identically in a practical scenario and vehicle follow master slave
mechanism successfully. Where robust RST controller also helps to
remove steady state errors in the system.
REFERENCES
[1] Trading Economics, Table 1: "
china/land-area-sq-km-wb-data.html , 2013." [2] World Bank, Data
base,
worldbank.org/transport/transportresults/regions/eap/eap-china-output.pdf.
[3] Fahad A. Siddiqui, Samreen Amir, Muhammad Asif, and Zain Anwar
Ali, LANE TRACKING AND
AUTONOMOUS CRUISE CONTROL FOR AUTOMATIC HIGHWAY SYSTEM, IEEE
19th Conference on Signal Processing and Communications
Applications (SIU), 2011.
[4] "Request for Applications Number DTFH61-94-X-0001 to
Establish a National Automated Highway System Consortium," Federal
Highway Administration, Washington, D.C., December 1993.
[5] Horowitz, R. and Varaya, P., "Control Design of Automated
Highway System", Proc of IEEE , vol 88, issue 7, July 2000
[6] Ashley, S. "Smart Cars and Smart Highways", Magzine:
Mechanical Engineering, The American Society of Mechanical
Engineers, May 1998