ISSN : 2230-7109(Online) | ISSN : 2230-9543(Print) IJECT Vol. 2, Issue 3, Sept.

2011

Introduction of Traffic Light Controller with

Fuzzy Control System
Sandeep Mehan
Dept. of ECE, RIEIT, Railmajra, Punjab, India

Abstract control system that sozlves these problems by continuously

Vehicular travel is increasing throughout the world, particularly sensing and adjusting the timing of traffic lights according to
in large urban areas. With the increasing use of automobiles the actual traffic load is called an Intelligent Traffic control
in cities traffic congestion occurred. So there is requirement System (ITS) [3,12]. ITS moves a broad range of applications
for optimizing traffic control methods to better accommodate from basic traffic signal control system to advanced systems
this increasing demand. So as the transportation system will that provide operational benefits to the transportation system:
continue to grow, intelligent traffic controls have to be employed reduce congestion, reduce operational costs, provide alternate
to face road traffic congestion problems. Fuzzy controllers have routes to travelers, enhances productivity and increase the
been widely used in many consumer products and industrial capacity of infrastructure [2]. Under these circumstances, the
applications with success over the past two decades. For traffic traffic control on urban expressways is generally dependent
control, however, fuzzy controllers have not been widely applied. upon the operators’ judgment, For example, fully experienced
This paper proposes a fuzzy traffic lights controller to be used operators are always working to optimize the traffic conditions,
at a complex traffic junction.The real time parameters such at the console desk in the traffic control center. The problems
as traffic density and queue length are obtained by using in the case of human traffic control are as follow:
image processing techniques. So the on and off timings for • Only skilled operators can make suitable judgments and
the green, red and orange lights are adjusted as per the actual decision, because the situation is very complicated and
road conditions. Fuzzy logic has been widely used to develop a many factors should be considered at control;
traffic signal controller because it allows qualitative modeling • The work load of skilled operators is very high, because
of complex systems. This paper describes a fuzzy logic signal they always make decisions according to traffic condition
controller for a four–way intersection suitable for mixed traffic, at very short time intervals;
including a high proportion of motorcycles. This paper discusses • It is very difficult to improve the process of traffic control,
the traffic control strategy, which dictates the design criteria because the actual process of the operators’ judgment
for the fuzzy logic controller. The components of fuzzy logic is not described clearly.
controller-the fuzzifier, the fuzzy rule base formulated by human To solve these problems, it becomes necessary to formulate
experts, the fuzzy inference engine and the defuzzifier. the operators’ judgment and to develop an automatic decision-
making system in place of operators. If the system is establish,
Keywords a more effective traffic control process can be constructed
Traffic Control, Fuzzy Logic, membership function, Intelligent and the actual situation of traffic control is investigated.
Transportation Systems (ITS), Defuzzification Secondly, the decision process of the operator is described
by the method of fuzzy logic control that is useful to express
I. Introduction human fuzziness.
Traffic congestion is a severe problem in many modern cities
around the world. Traffic congestion has been causing many II. The Present Day Technique
critical problems and challenges in the major and most Today we live in the world of automation. Microcontrollers
populated cities. To travel to different places within the city is control most things around us. The control of traffic lights is well
becoming more difficult for the travelers in traffic. Due to these known area where this type of control system is incorporated,
congestion problems, people lose time, miss opportunities, which controls the four sets of traffic lights at the traffic crossing.
and get frustrated. Traffic congestion directly impacts the . But, the control is not flexible, based on the condition of traffic
companies. Due to traffic congestions there is a loss in at the crossing. Rather, the on and off time periods are fixed for
productivity from workers opportunities are lost, delivery gets the red, green and orange lights. These timing durations are
delayed, and thereby the costs goes on increasing. To solve varied as per the day, the day of the week etc. But there are
these congestion problems, we have to build new facilities and no real time adjustments of the on/off times as per the traffic
infrastructure. The only disadvantage of making new roads on conditions on the crossing. This is ordered to work such that
facilities is that it makes the surroundings more congested. the traffic in only those directions , which do not cross each
So for that reason we need to change the system rather than others, is allowed to move at any given time while in all the
making new infrastructure twice. The Main goals of this paper other directions, the traffic is forbidden to move. For example
are improving safety, minimizing travel time and increasing the in INDIA (left hand drive system) the traffic may be allowed to
capacity of Infrastructures. Such improvements are beneficial move both from north to south and from south to north, while
to health, economy and the environment. Delay reduction at the east to west direction, the east to north direction, the west
city intersections and travel time savings are major goal of to east direction, the west to south direction and the south to
Intelligent Transportation Systems (ITS) Traffic load is highly east direction are blocked at that time as shown in Fig. 1 [9]
dependent on parameters such as time, day, season, weather
and unpredictable situations such as accidents, special events
or constructional activities. If these parameters are not taken
into account, the traffic control system will create delays. A traffic

w w w. i j e c t. o r g   International Journal of Electronics & Communication Technology  119

IJECT Vol. 2, Issue 3, Sept. 2011 ISSN : 2230-7109(Online) | ISSN : 2230-9543(Print)

and with a particularly high proportion of motorcycles. Due to

lack of lane discipline, queues at intersections are built up
based on the optimum road space utilization which means
vehicles can occupy any position across the road based on
the available space. It is obvious that traffic behavior in
mixed traffic conditions is different to that in non-mixed traffic
conditions. Therefore, the main objective of this research is to
design an adaptive fuzzy logic signal controller for an isolated
four-way intersection with reference to mixed traffic (including
high proportion of motorcycles) [14]. Simulations are used to
examine and analyze the effectiveness of the proposed FLSC.
Fig. 1: Present day technique for control or traffic light (Left Then, the performance of the proposed controller is contrasted
hand side traffic) with an optimized fixed time controller.

So the traffic lights which exist in the junction work according to IV. Degisn Creteria And Constraints
an algorithm which does not take into account the number of In the development of the fuzzy traffic light control system the
vehicles which arrive at the crossroad. Inefficient configuration following assumptions are made:
of traffic lights based in a fixed cycle protocol can lead to • The junction is an isolated four-way junction with traffic
unnecessarily long waiting times for vehicles and even to traffic coming from the north, west, south and east directions;
congestion. Simplifying, in this case, this paper presents a • When traffic from the north and south moves, traffic from
possible solution using the fuzzy logic. It is an approach that the west and east stops, and vice versa;
allows the implementation of real-life rules by using computer • No right and left turns are considered;
algorithm. • The fuzzy logic controller will observe the density of the
north and south traffic as one side and the west and east
III. The Proposed Technique traffic as another side;
Traffic signal control is one measure that is commonly used • The east-west lane is assumed as the main approach.
at road intersections to minimize vehicular travel times and
delays [5]. Traffic signal control at road intersections allows V. Design For Fuzzy Logic Traffic Light Controller
vehicle movements to be controlled by allocating time intervals, A fuzzy logic controller is designed for an isolated 4-Lane traffic
during which separate traffic demands for each approach of intersection: North, South, East and West [4,5]. In this method
the intersection can make use of the available road space. we consider two fuzzy input variables and one output fuzzy
Traffic signal control in most signalized intersections is done variable. These input variables are:
with fixed time signal control. Under fixed time control, all signal • Quantity of the traffic on the arrival side (Arrival);
timing parameters are pre-computed and kept constant. These • Quantity of traffic on the queuing side (Queue).
parameters are calculated based on historical traffic data. This If the north and the south side is green then this would be the
method usually shows good results in normal traffic conditions, arrival side while the west and east side would be considered
but sometimes they fail to cope with complex, time varying as the queuing side, and vice- versa.
traffic conditions. Vehicle actuated (VA) control presents an On the other side the output variable would be the extension
improvement over fixed time control [6]. The VA control principle time needed for the green light on the arrival side. So based on
aims to adjust the length of green time in response to the real the current traffic conditions the fuzzy rules can be formulated
traffic flow variations. VA control requires vehicle detectors to so that the output of the fuzzy controller will approaches to
provide accurate information of traffic in real-time. This method increase or not increases the current green light time. During
has limited ability to respond to real-time traffic demand, these methods some points to be remember, which are:
where its performance generally deteriorates with heavy traffic • The four traffic lights work in four sequences, every
conditions. To overcome such problems adaptive traffic signal light having a variable sequence from 10 to 130 seconds
controllers are designed to address those deficiencies. Fuzzy depending on the congestion (the number of vehicles from
logic has been used widely to develop an adaptive traffic signal the queue bur also the number of one’s which arrive every
controller, because it allows qualitative modeling of complex minute);
systems, where it is not easy to solve using mathematical models • All the four traffic lights will be controlled by this same
and is good for systems that have inherent uncertainties. Many mechanism;
researchers have proposed the traffic signal control systems If there is no extension of the current green light time, the
using fuzzy logic. They proposed a FLSC for an isolated state of the traffic lights will immediately change to another
intersection of four-way east-west/north-south without turning side which allows the traffic from the alternate traffic flow.
movement. This gives generally a better performance of the In Fuzzy controller structure input 1 (I/P 1) is the arrival of
FLSC when compared to fixed time and actuated controllers. the vehicles and input 2 (I/P 2) is the queuing of vehicles
However, all existing research has developed FLSC based on are two parameters which are used to set the extension time
non-mixed traffic conditions (developed countries), [11] where for green light are fuzzified and then these parameters are
they considered the passenger car only and neglect motorcycles given to fuzzy inference system which actually sets the time
in their traffic. It is quite different to that in mixed traffic but fuzzy in nature which are actually different parameters
conditions (developing countries), where the traffic streams are (decrease, constant, increase) so to convert these performance
heterogeneous, consisting of different types of vehicle with wide parameters in crisp we use defuzzification method as shown
variation in their static, dynamic and operating characteristics, in Fig. 2 which gives the actual time for what time the green

120  International Journal of Electronics & Communication Technology w w w. i j e c t. o r g

ISSN : 2230-7109(Online) | ISSN : 2230-9543(Print) IJECT Vol. 2, Issue 3, Sept. 2011

light is ON or OFF [7] [8]. are arrive at the crossroad. A possible membership function
for the allocated time for the extension time for the green light
is below allocated to the Green light in Fig. 5.

`
Fig. 2: Fuzzy controller structure Fig. 5: The membership function for allocated time for the
extention time or green light.
When the transportation needs is large, the signal cycle should
be increased. The upper bound should not be more than 130 VII. Fuzzy Rule Base And Establishment
seconds generally, because the driver on the opposite direction The basic function of the fuzzy rule base is to represent the
may not tolerate that the signal cycle is more than 130 second expert knowledge in a form of IF-THEN rule structure combined
India. So minimum and maximum signal cycle is set for each with AND/OR operators. For e.g. IF traffic from the north of
phase. When transportation needs is small, the minimum signal the city is more AND traffic from the west is less THEN allow
cycle is run. On the other side when the transportation needs movement of traffic from the north side [5,13].
is large, the maximum signal cycle is the present cycle and The fuzzy rule base is set of fuzzy rules. It maps the combination
traffic congestion is avoided [1]. of fuzzy inputs (arrival, queuing linguistic variables) to the
corresponding fuzzy output. In this paper we consider A stands
VI. Membership Functions for arrival, Q stands for queuing and T stands for time required
The types of membership function used in this paper are for green light. They are divided into different fuzzy subsets,
triangular and trapezoidal due to the computational efficiency. A= {VS, S, M, L, VL}
Membership function of each input and output fuzzy variable Q= {VS, S, M, L, VL}
of FLSC is as follows: T= {D, C, I}
• Membership function for the number of vehicles on the Where VS is Very Small, S is Small, M is Medium, L is Large,
queuing side; VL is Very Large, D is Decrease, C is Constant, I is Increase.
• Membership function for the number of vehicles which There are normally many rules for each decision. There are
arrive at the crossroad; lot of rules for each decision. Some of the fuzzy rules defined
The quantity of the traffic on the arrival and queuing side have presented below:
{very small, small, medium, large, very large} and extension time • If the number of vehicles waiting in line or queuing (Q) is
has {decrease, constant, increase} linguistic variables [3]. medium and the number of vehicles which arrive or arrival
A possible membership function for number of vehicles (A) is small then the allocated time for the green light (T)
which are standing or queuing in the line at the traffic lights is decreases.
represented as shown in Fig. 3. • If the number of vehicles waiting in line or queuing (Q) is
large and the number of vehicles which arrive or arrival
(A) is small then the allocated time for the green light (T)
constant.
• If the number of vehicles waiting in line or queuing (Q)
is very large and the number of vehicles which arrive or
arrival (A) is medium then the allocated time for the green
light (T) increases.
Fig. 3: The membership function for the number of vehicles • If the number of vehicles waiting in line or queuing (Q)
which are standing in queue at the traffic is very small and the number of vehicles which arrive or
arrival (A) is medium then the allocated time for the green
The membership functions for the number of vehicles which light (T) decreases.
are standing in queue at the traffic. A possible membership
function for number of vehicles which arrive at the crossroad From Table 6 we can easily understand the different fuzzy rules
is represented as shown in Fig. 4. with different parameters i.e arrival and queuing.

Table 1: Fuzzy control rules

Queue
VS S M L VL
Arrival
VS D D D D C
S D D D C C
Fig. 4: The membership function for the number of vehicles M D D C C I
which arrived at the crossroad
L D C C I I
The membership functions for the number of vehicles which VL C C I I I

w w w. i j e c t. o r g   International Journal of Electronics & Communication Technology  121

IJECT Vol. 2, Issue 3, Sept. 2011 ISSN : 2230-7109(Online) | ISSN : 2230-9543(Print)

VIII. Inferrence Engine and Defuzzification of the 37th Midwest Symposium pp.1507-1510 Vol.2,
Membership functions are used to retranslate the fuzzy output August,1994.
into a crisp value. This method is known as Defuzzification [4]. [5] Hoyer, R., Jumar, U., “Fuzzy Control of Traffic Lights”, Proc.
The fuzzy inference evaluates the control rules stored in the IEEE International Conference on Fuzzy Systems, 1994,
fuzzy rule base. Defuzzification is a process to convert the fuzzy pp 1526-1531.
output values of a fuzzy inference to real crisp values. Fist a [6] Jee-Hyong Lee, Hyung Lee Kwang, “Distributed and
typical value is computed for each term in the linguistic variable Cooperative Fuzzy Controllers for Traffic Intersections
and finally a best compromise is determined by balancing out Group”, IEEE Transactions on Systems, Man, and
the results using different methods like center of sum, center Cybernetics- Part C : Applications and Reviews, Vol. 29,
of area, center of area mean of maximum etc. But for this No. 2, May, 1999
application we use Center of Sum to process defuzzification [7] Kosonen, T., “Fuzzy logic based traffic signal control”.
of the output variable extension time. This method is mostly Helsinki University of Technology, Transportation
used because this method has better performance in terms Engineering. Master’s thesis, 2002. pp.146
of continuity, computer complexity and counting. [8] Lin, Q., Kwan, B.W., Tung, L.J., “Traffic Signal Control Using
Fuzzy Logic”, Proc. IEEE International Conference on Fuzzy
IX. Conclusion and Future Works Systems 1997, pp. 1644-1649.
The system proposed here is very flexible. The feedback of the [9] Lisakki Kosonen , “Multi-Agent fuzzy signal control based
queue length and traffic densities can be taken from images on real-time simulation”, Helsinki University of Technology,
taken from cameras. Because of the flexibility of the fuzzy logic Transportation engineering, Finland. Transportation
in dealing with uncertainty, it can be used advantageously for research Part C 11 (2003) pp. 389-403
traffic light controlling systems. The proposed FLSC and fixed [10] Niittymaki, J., “Fuzzy traffic signal control--Principles
time controller produces little difference in results in terms of and applications”. Doctoral thesis. Helsinki University of
constant traffic flow while in the case of time varying traffics, Technology, Transportation Engineering pp.103, 2002
the proposed FLSC is superior to the fixed time controller [2] [8] [11] Niittymaki, J., Nevala, R., “Multi-level and multi-objective
[11]. This controller gives a suitable green time to improve the traffic signal control using fuzzy methods”. Conference
traffic capacity effectively and reduced the intersection delay, Proceedings of the 6th International Conference on
which can insure vehicles don’t allow waiting too long on the Applications of Advanced Technologies in Transportation
road. While in the case of fixed time controller when green time Engineering. Technical session 13, Real-time traffic
is finished, this will give the green time to the next phase even control. Singapore, 28-30 June 2000. pp.11
if the current vehicle flow is large. So arriving vehicles must [12] Pappis, C., Mamdani, E.,“A fuzzy logic controller for a traffic
wait for the next cycle to leave. The performance of the FLSC junction”. IEEE Trans. Systems, Man Cybernetics SMC-7
is affected by the configuration of the membership functions (10), pp. 707-717, 1977.
of the input and output variables and the rule base . It can [13] Qinghui Lin, B.W. Kwan, L.J. Tung, “Traffic Signal Control
be observed that fuzzy logic control system provides better Using Fuzzy Logic”, Systems, Man and Cybernetics, 1997.
performance in terms of improving the safety and efficiency by Computational Cybernetics and Simulation, 1997 IEEE
reducing the waiting delay of vehicles on signals. Less traffic International Conference pp. 1644-1649, Vol. 2. October.
congestion and less waiting time at red traffic lights will reduce 1997.
the fuel consumption, air pollution, sound pollution, and time [14] Xu Li-qun, Jiang Fu, “Fuzzy Reasoning Algorithm for Control
and energy waste. Furthermore, with the comparative ease of Traffic in Congestion Route[J]”, Information and Control,
terms like weather conditions, environments aspects etc. can Apr.2003, Vol.32, No.2, pp. 132-141.
be added to the fuzzy system. A lot of research work has to
be done to verify the expected features by simulation. The
definitions of the fuzzy sets of the antecedents are also very
easily changeable. This is a very promising application of fuzzy Sandeep Mehan was born in Chandigarh,
logic in practical areas, and will be highly useful in traffic control Punjab, India on feb.20,1988. He received his
in the today congestion traffic. B-Tech degree in Electronics and Comm. From
RIEIT, Railmajra (Ropar),Punjab in 2009 and
References pursuing the M-tech in Electronics and
[1] Akçelik, R., Besley, M., Roper, R., “Fundamental Comm. From RIEIT (Ropar), Punjab. He has
Relationships for Traffic Flows at Signalized Intersections, already published several papers in national
ARRB Transport Research Ltd.,” Research Report No. and international conferences. His area of interest includes
340,pp. 268 , ISBN 0869107968, 1999 instrumentation and PLC.
[2] A. Hegyi, B. De Schutter, R. Babuska, S. Hoogendoorn,
H. VanZuylen, H. Shuurman, “A Fuzzy Decision Support
System for Traffic Control Centers[J]”, Faculty of Information
Technology and Systems, Aug. 2001, pp.358-363.
[3] Beauchamp-Baez,G., Rodriguez-Morales, E., Muniz
Marrero, E.,L., “A Fuzzy Logic Based Phase Controller for
Traffic Control”, Proc. IEEE International Conference on
Fuzzy Systems 1997, pp.1533-1539.
[4] Dr. Devinder Kaur, Elisa Konga, Esa Konga, “Fuzzy Traffic
Light Controller, Circuits and Systems” 1994, Proceedings

122  International Journal of Electronics & Communication Technology w w w. i j e c t. o r g