Design and Simulation of a fuzzy Controller for a busy Intersection

Design and Simulation of A Fuzzy Controller for A
Busy Intersection
Abdollah Amirkhani Shahraki*¹, Meisam Niazi Shahraki², Mohammad Reza Mosavi^3
1Dept. of Electrical Engineering, Iran University of Science and Technology, Narmak,Tehran, Iran
²Dept. of Electrical Engineering, University of Sistan and Baluchestan, Zahedan, Iran
*[e-mail]

Abstract—Traffic in large cities is one of the biggest problems that can lead to excess utilization of fuel by motor vehicles, accidents, and the waste of time of citizens. To have an effective and efficient city management system, it is necessary to intelligently control all the traffic light signals. For this reason, many researchers have tried to present optimal algorithms for traffic signal control. Some common methods exist for the control of traffic light signal, including the preset cycle time controller and vehicle-actuated controller. Results obtained from previous works indicate that these traffic light signal controllers do not exhibit an effective performance at moments of traffic peak. So to resolve this dilemma at such moments, traffic cops are employed. The application of fuzzy logic in traffic signal controllers has been seriously considered for several decades and many research works have been carried out in this regard. The fuzzy signal controllers perform the optimization task by minimizing the waiting time of the vehicles and maximizing the traffic capacity. A new fuzzy logic based algorithm is proposed in this article, which not only can reduce the waiting time and the number of vehicles behind a traffic light and at an intersection, but can consider the traffic situations at adjacent intersections as well. Finally, a comparison is made between the designed fuzzy controller and the preset cycle time controller.
Keywords—Fuzzy logic; Traffic intersection; Traffic control; Simulation

1. 
   Introduction
   

Traffic congestion of streets and roads constitutes a critical problem which is aggravated by the rise in the number of vehicles and by greater urbanization. The slow pace in the development of new highways and roads and public opposition to the widening of existing streets in some locations has forced the city managers to optimally use the existing infrastructures in order to effectively manage the flow of traffic. Moreover, the loss of valuable time during traffic congestion can directly affect the production, productivity, performance and the utilization of fuel [1, 2]. In some countries, a university curriculum called the ‘urban traffic management’ has been established for studying and calculating the number of required highways, roads, bridges and overpasses. Principally, the development of an area has a close relationship with the ease of transportation in that region. The control of traffic light signal is one of the subjects of intelligent systems being investigated by researchers; because this kind of control has a direct impact on the effectiveness of urban transportation systems. Numerous investigators have focused their studies on algorithms for the optimal control of traffic signals. Webster [3] presented some equations for optimizing the cycle duration and determining the time duration of the green phase based on a fixed time control, which were used extensively. Akcelik modified Webster’s theory for the supersaturated scenario in a new time signal algorithm called ARRB [4, 5]. Although these methods work well and with low computation costs, they cannot respond instantaneously to time variations. With the progress in the development of sensors, computers and communications technologies, many advanced methods have been devised for regulating the signal time instantaneously according to the received traffic data. The SCOOT, SCATS, OPAC and RHODES are some of the adapted traffic light signal control systems in the world, which were successfully applied to large road networks, and formed the initial traffic management system in many countries [6]. Urban management systems are complex dynamic systems which are truly very hard to model. The modeling of these systems is needed for the development of an effective control system and for the simulation of a physical process.
In some cities, programmed software is used for the control of traffic. Although these software programs don’t perform adequately during the rush hour traffic, this problem can be solved to some extent by using traffic guide officers during these congested hours.There are several approaches for the control of traffic at intersections, and two of the most common of these, which have been installed and used in many big cities, include the preset cycle time controller and the vehicle- actuated controller [7].
The existing software programs’ inability in controlling the traffic, especially at the peak hours, has made it necessary to seek new controllers with more effective performance. The fuzzy method has been used extensively in recent years in the designing of products and industrial equipments. Numerous research works based on the technique of fuzzy logic have been conducted with the aim of controlling the traffic lights [8]. In most of these investigations, only one intersection has been taken into consideration. However, to effectively solve the traffic congestion problem, these systems should be able to consider the adjacent controllers and the peak traffic circumstances.
Fuzzy logic enables the implementation of rules very similarly to what goes on in the human thinking process. For example, based on the thinking of a human being under traffic conditions in a specific intersection, if there is heavy traffic in the northbound or southbound lanes and less traffic in the eastbound and westbound lanes, then the traffic lights for the north and southbound lanes should remain green for longer periods. Such rules can be easily accommodated in a fuzzy logic controller. A major advantage of the fuzzy logic is that in it, the conditions and terms such as ‘heavy’, ‘less’, ‘shorter’ and ‘longer’ can be represented by numbers to make them understandable to a computer. This type of controller derives the control logic from human thinking and reasoning.
In this article, a fuzzy traffic lights control system is designed which can be effectively used for a traffic junction with several intersections. The fuzzy rules are mostly determined by the conditions of traffic. The traffic controller uses a selection phase and an extension phase, while the waiting time of vehicles is optimized by means of a timer installed at the intersection. This system is able to manage the traffic instantaneously and to effectively reduce the waiting time of vehicles and the traffic volume.
The first efforts to use fuzzy logic in traffic control were made by Mamdani and Pappis [8]. In fact, Mamdani was the first individual to benefit from fuzzy logic in a control application. He and his colleagues simulated an isolated traffic intersection as a simple two-way street. Based on the previous works, Chen carried out more involved investigations; however these studies mostly focused on the control of traffic with a constant phase [9]. Bisset and Kelsey simulated the traffic control for a single intersection with one lane as a two-phase system [10, 11]. Also, Pursula and Niittymaki simulated a 4- lane intersection [12]. They realized that a fuzzy logic controller shortens the delay and stoppage time of vehicles. Trabia et al. designed a fuzzy logic controller for an intersection with right and left turns where the traffic volume and queue length of vehicles were measured by installed detectors [11]. Niittymaki and Kikuchi developed a fuzzy logic algorithm for the control of pedestrian movement on sidewalks [13]. Their simulation showed that the fuzzy logic algorithm performs better than the actuated controller. Chen studied the fuzzy control of a freeway and demonstrated that it reduces traffic congestion as well as the number of accidents [14]. Fuzzy logic has been used in [15] for the control of vehicular movement in two adjacent intersections. Chiu employed the fuzzy logic to control several distinct intersections where the vehicles don’t exhibit turning moves [16]. All these works indicate that the fuzzy logic controller has a better performance compared to the preset cycle time controller or the vehicle- actuated controller.

2. 
   Different Types of Traffic Light Controllers
   

Basically, there are two types of control systems for the traffic light signal. The first type uses a preset cycle time in the traffic light, and the second type employs a combination of the first type controller and a proximity sensor, which can actively alter the green time duration of the lights. The first type creates a traffic light control mechanism with a fixed operating cycle [17], whereas the second type results in a controller with a variable operating period. Fuzzy logic can be incorporated into the second type controller for the adjustment of the operating period so that the experiences of an expert in traffic control can also be taken into consideration.

1. 
   Preset cycle time controller
   

This type of traffic controller includes an initially adjusted time. In fact, the duration of each phase in a operating period is adjusted according to the existing traffic pattern [18]. The major drawback of this controller is its lack of conformity with the changing traffic flow conditions. In other words, if a traffic jam occurs at an intersection, the green light duration doesn’t increase and the next phase continues regardless of the volume of vehicles at each intersection. Because this controller is greatly dependent on traffic data, its performance is hindered by any sudden change in the traffic pattern. These limitations of performance can be remedied by real-time vehicle-actuated methods such as SOS, LHVORA and MOVA [18].

2. 
   Vehicle-actuated controller
   

This controller includes detectors that can cause a change in phase duration. In this approach, in every street leading to an intersection, there is a proximity detector which shows the existing vehicles at that intersection. In this method, three parameters of initial time interval, extension unit and extension limit are used, which will be explained below.
If the traffic light is in the green phase, first, the initial time interval should elapse and then the pause duration of the green light will be prolonged by an amount equal to the extension unit. If during this time, the installed detectors detect the presence of a vehicle, the pause time of the green light will again be extended by the amount of extension unit. This procedure will continue until the extension limit is reached. Fig. 1 shows an example of a vehicle in a particular phase (particular route) in a vehicle-actuated traffic light control system. The major flaw of this method is that, with regards to traffic conditions, sometimes it is necessary for the green light to stay on for more than the specified extension limit or that there may be a few vehicles and yet, for the green phase to end, the extension unit time should elapse.

Figure 1. Vehicle-actuated controller with three parameters of initial time interval, extension unit and extension limit

3. 
   Description of The Fuzzy Controller
   

Before describing the fuzzy controller, it is necessary to explain certain expressions that are used in this type of control system. A ‘link’ represents a route that connects two intersections. The ‘Length of Link Num’ indicates the number of vehicles on a link between two intersections; while the ‘Length of Queue Num’ denotes the number of vehicles on a lane which stay behind the traffic light during a red phase. An example of a link has been shown in Fig. 2(a). A ‘time cycle’ denotes the time it takes for an intersection’s traffic light to turn all four routes ON or OFF. A ‘phase’ indicates all the possible directions for a vehicle during a complete time cycle. An example of a time cycle along with the phases has been illustrated in Fig. 2(b) for a typical intersection. It should be noted that number of phases in each cycle and also the routes of each phase can be chosen arbitrarily. For example, in Fig. 2, only three routes for each phase and four phases for each time cycle have been selected for the sake of simplicity. The maximum number of vehicles that can be accommodated between two intersections is defined as the capacity of the corresponding link.
The fuzzy control system comprises three stages, which have been shown in Fig. 3. These three stages include the next green phase, green phase extender, and the decision stage. In view of Fig. 3, the inputs are applied through the green phase selector. The next green phase stage selects the most urgent phase from the phases waiting to become green. If necessary, the green phase extender increases the duration of the green light. In the decision making stage, by deciding either to increase the green light duration or to change to another phase, the most urgent stage is selected from the two stages of next green phase and green phase extender. For example, if the green phase extender stage is more urgent than the next green phase stage, then the fuzzy control system will extend the green light. Conversely, if the next green phase stage is more urgent



(b)
Figure 2. (a) Some expressions and parameters used in the design of fuzzy
controller; (b) An example of phases for a typical intersection



Figure 3. The designed fuzzy control system.

than the green phase extender stage, the decision stage goes from one green phase to another green phase.