Fuzzy Rule Bad Traffic Signal Control System for Oversaturated
送朋友的祝福语
Interctions
赞美夕阳的诗句Syed Ali Abbas Syed Muhammad Sheraz Dr. Humera Noor
ali. sheraz@neduet.edu.pk humera@neduet.edu.pk MEngg Scholar MEngg Scholar Associate Professor Department of Computer and Information Systems Engineering
NED University of Engineering and Technology, Karachi Pakistan
化妆品市场Abstract – A Fuzzy Rule Bad Control System is prented in this paper. This control system controls Traffic Signals for regulating traffic on oversaturated interctions with the integration of left and right turns. Bad on the Fuzzy Rules, the system decides, whether to extend the current green signal or terminate it. The control system also controls the continuous and safe flow of emergency vehicles.
Keywords – Fuzzy Controller; Traffic Control; Cycle Time; Queuing Length; Emergency Vehicle.
I.I NTRODUCTION
As the number of vehicles and the needs for greater transportation has grown in recent years, city streets and highways frequently face rious road traffic congestion problems. Due to this factor, traffic signals now become a common feature of cities controlling heavy traffic. Careful planning of the signals is important to increa the efficiency of traffic flow on road. Controlling traffic on oversaturated interctions is a big issue.
Conventional methods for traffic signal control bad preci models fail to deal efficiently with the complex and varying traffic situations. They are modeled bad on the pret cycle time to change the signal without any analysis of traffic situation. Due to fixed cycle time, such systems do not consider that which interction has more load of traffic, so should kept green more or should terminate earlier then complete cycle time. In ca of interctions, conventional control systems only consider waiting time of signals on different directions but not the vehicle directions. Such situations can be en in various areas of Karachi like Shahrah-e-Faisal where traffic flow varies in different hours and heavy traffic flows in morning and evening timings becau of large number of offices on that route. Also, in different interctions, traffic flow abruptly changes in schools timings then other daily hours. Pret Cycle Time Controllers fail in such scenarios becau they could not get complete information of vehicles earlier. Also, sometimes situation aris, when some VIP move
ment is there, the traffic flow has to divert and control different interctions. In such situations, efficiency of human decision-making is unprecedented efficiency of human becau decision-making objectives are unclear [11].
Fuzzy bad controllers are proved to be well manager of traffic system in such scenarios [1, 10]. Fuzzy controllers have the ability to take decision even with incomplete information. More and more sophisticated controllers are being developed for traffic control [2, 3, 4, 5, 6, 7, 8, and 9]. The algorithms are continually improving the safety and efficiency by reducing the waiting delay of vehicles on signals [1]. This increas the tempo of travel and thus makes signals more effective and traffic flow smooth.
The key motivation towards Fuzzy Logic in traffic signal control is the existence of uncertainties in signal control. Decisions are taken bad on impreci information and the effect of evaluation is not well known [8].
The objective of this rearch study is to design a Fuzzy Rule Bad System for oversaturated interctions with left and right turns. Also, the system controls the smooth flow of emergency vehicles.
In earlier studies, the fuzzy controllers do not consider left and right turns simultaneously. Lin Zhang and Honglong Li developed Fuzzy Traffic Controller for Oversaturated Interctions [8]. They designed an algorithm to control over-saturated interctions of two-way streets with left turning movements.
Jee-Hyong Lee and Hyung Lee-Kwang also designed a Fuzzy Control Model. The goal of controller is to decrea the average time delay in the whole traffic network. They assumed that special establishments named right-turning lane in the interction allow right-turning traffic flow to pass the interction without disturbing the other traffic flows at the same interction. Under this assumption, right-turning traffic flow is out of the consideration of fuzzy control [9].
决心In this paper, all the limitations and assumptions have been removed. This work includes not only straight turns and single turns but all single, left and right turns. Moreover, this paper has merged the controlling of Emergency Vehicles with the interction control algorithm to provide smooth flow to the vehicles towards their destination. This was not considered earlier. As Police Vehicle or Ambulance need flush flow, emergency vehicle control has been incorporated in this study. The signal should always be Green after sudden entrance of emergency vehicle.
This paper describes a fuzzy rule bad approached designed to regulate traffic flow for oversaturated interctions. The fuzzy controller decides whether to terminate the currently Green signal or extend it for some period. The asssments are made using t of fuzzy rules. The rules consider the Queuing Lengths and Arriving Rates of current Green signal and the compared waiting signal. In this paper, rules are bad on not only “Current Green Signal Arriving Rates” but also take into
2009 International Conference on Computational Intelligence and Natural Computing
consideration, the “Arriving Rates of Compared Signals”. In Lin Zhang and Honglong Li rearch, Arriving Rates of Current Green Signal was considered only [8]. Here, rules are made more accurate by adding new parameters and making membership functions more preci and sophisticated. This is done to achieve more demanding results and effective incorporation of Emergency Vehicle Control.
Fuzzy logic technology allows the implementation of real-life rules similar to the way humans would think. In Traffic Control System, humans would think in the following way to control traffic situation at a certain junction: “if the traffic is heavier on the north or south lanes and the traffic on the west or ea
st lanes is less, then the traffic lights should stay green longer for the north and south lanes”. Such rules can now be easily accommodated in the fuzzy logic controller. In this consideration, we can say that it is replaceable to Traffic Police Officers. Fuzzy Logic works glowing when traffic flow in different directions is highly uneven as compared to Pretimed Controller [8, 11].
II.L ITERATURE R EVIEW
Traffic flow is usually characterized by randomness and uncertainty. Fuzzy logic is known to be well suited for modeling and control such problems. Applications of fuzzy logic in traffic signal control has been made since the 1970s [10].
The first attempt made to design Fuzzy Traffic Controller was in 70s by Pappis and Mamdani [2]. After that Niittymaki, Kikuchi, Chui and other rearchers [4, 5] developed different algorithms and logic controllers to normalize traffic flow.
Kely and Bist [3] also designed a simulator for signal controlling of an isolated interction with one lane. Same work was also done by Niittymaki and Pursula [6]. They obrved that Fuzzy Controller reduces the vehicle delay when traffic volume was heavy.
Niittymaki and Kikuchi [4] developed Fuzzy bad algorithm for pedestrians, crossing the road.
Nakatsuyama, Nagahashi, and Nishizuka [7] applied fuzzy logic to control two adjacent interctions on an arterial with one-way movements. Fuzzy control rules were developed to determine whether to extend or terminate the green signal for the downstream interction bad on the upstream traffic.
经济订货量Chui was the first who us Fuzzy Logic to control traffic in multiple interctions [5]. In this attempt, only two way streets are evaluated without considering any turnings.
In recent years, Lin Zhang and Honglong Li [8] also worked on designing Fuzzy Traffic Controller for Oversaturated interctions.
Jee-Hyong Lee and Hyung Lee-Kwang [9] prented direction-varying traffic signal control but assume that right turn traffic flow do not disturb any other traffic flows in an interction.
III.F UZZY L OGIC
The concept of Fuzzy Logic (FL) was conceived by Lotfi Zadeh, a professor at the University of California at Berkley, and prented not as a control methodology, but as a way of processing data by allowing partial t membership rather than crisp t membership or non-membership. FL is a problem-solving control system methodology that lends itlf to implementation in systems ranging fr
om simple, small, embedded micro-controllers to large, networked, multi-channel PC or workstation-bad data acquisition and control systems. It can be implemented in hardware, software, or a combination of both.
FL provides a simple way to arrive at a definite conclusion bad upon vague, ambiguous, impreci, noisy, or missing input information. FL's approach to control problems mimics how a person would make decisions, only much faster. FL can be built into anything from small, hand-held products to large computerized process control systems. It us an impreci but very descriptive language to deal with input data more like a human operator. It is very robust and forgiving of operator and data input and often works when first implemented with little or no tuning.
IV.F UZZY T RAFFIC S IGNAL C ONTROL
The aim of using fuzzy methods is attempt to model expert’s thinking ion the situations where development of an exact mathematical model of the phenomenon is very difficult or even impossible. The main goals of fuzzy logic in the traffic signal control, and a matter of fact, also in traffic signal control in general, are
1.Improving of traffic safety in the interction.
2.Maximizing the capacity of the interction.
3.Minimizing the delays.
4.Clarifying the traffic environment.
5.Influencing the route choices.
V.M ETHODOLOGY
The fuzzy logic controller determines whether to extend or terminate the current green pha bad on a t of fuzzy rules. The fuzzy rules compare traffic conditions with the current green pha and traffic conditions with the next candidate green pha. The flow diagram of a controller is shown in figure 1.
Figure 1: Fuzzy Traffic Signal Control
Cycle Time indicates the period to which current signal is green. If the cycle time is small, it denotes that it is usual time of green in normal situations; the cycle time of each signal is small. If traffic situation is such that it needs extension, then it can extend to long extension. If the cycle time is medium, it denotes that some extension has already been done, and now long extension cannot be done. If cycle time is already large then no extension is allowed, no matter what the situation of traffic. This is done to avoid starvation of any signal. In ca of large cycle time, short extension is allowed only when emergency vehicle is in the current green signal.
Controlling of emergency vehicle is done by investigating their prence. If the emergency vehicle is in the current green signal then extend the signal for short extension only. When there is a situation, emergency vehicles are both in current signal and any other signal then extend the current signal. If emergency vehicle is in any signal which is not green then terminate the current signal.
VI.F UZZY P ARAMETERS AND THEIR M EMBERSHIP
FUNCTIONS
The t of control parameters is:
EV= Emergency Vehicle
CT = Cycle Time of Current Green Signal
Q C = Queue Length of Current Green Signal
Q N = Queue Length of Next Signal to be Green
AR C = Arrival Rate of Current Green Signal
AR N = Arrival Rate of Next Signal to be Green
CS=Current Signal
NS=Next Signal
TABLE 1: I NPUT &O UTPUT V ARIABLES AND THEIR M EMBERSHIP
F UNCTIONS
Input Variables
EV CT Q C/Q N AR C/AR N
Prent Not Prent
Small
Medium
Large
V Low
Low
Medium
High
V High
Low
Medium
High Output Variable
Extension
Zero
Short
Medium
Large
VII.F UZZY R ULE S ET
The fuzzy controller decides an extension of current
green signal on the basis of rules t. The rules will act
differently on the basis of Cycle Time. As mentioned
earlier, by default, the cycle time is ‘Small’ for every
signal. Extension from small CT will be on certain rules.
‘Small’ CT will allow extension to ‘Long’. If the cycle
time is ‘Medium’, it shows, it already had some extension,
so that signal can extend only to ‘Medium’. In ca of
already ‘Large’ cycle time, the extension will be given
only in the prence of Emergency Vehicle. When
Emergency vehicle is there on any signal then it will
ignore all other traffic factors and extension will be given
to that signal.
Some of the rules are shown below:
1.If EV in CS is prent then Short extend.
2.If EV in NS is prent then Zero extend.
If CT is ‘Medium’ then extension can be ‘Zero’,
‘Short’ or ‘Medium’. No Long extension.
3.Q C is V Low and AR C is low then Zero extend.
4.Q C is Medium and AR C is Low and Q N is
Medium and AR N is High then Zero extend.
5.Q C is V High and AR N is Low and Q N is Low
then Medium extend.
If CT is ‘Small’ then all values from extension t
are extension can be done to ‘Long’.
6.Q C is V Low and AR C is Medium and Q N is V
Low then Short extend.
7.Q C is L Low and AR C is High and AR N is Low
and Q N is V low then Long extend.
8.Q C is High and AR C is Low and Q N is V High
then Zero extend.
9.QC is V High and ARC is High and QN is
Medium then Long extend.
10.Q C is V High and Q N is V Low then Long
extend.
VIII.C ONVENTIONAL T ECHNIQUES USED FOR
I NTELLIGENT T RAFFIC L IGHT C ONTROL
A.E XPERT S YSTEMS
An expert system us a t of given rules to decide
upon the next action [12, 15]. The expert systems can
communicate to allow for synchronization. Performance
on the network depends on the rules that are ud. For
each traffic light controller, the t of rules can be
optimized by analyzing how often each rule fires, and the
success it has. The system could even learn new rules.
B.E VOLUTIONARY A LGORITHMS
Taale [14] compare using evolutionary algorithms
evolution strategy to evolve a traffic light controller for a
宝宝大便发白single simulated interction to using the common traffic
light controller in the Netherlands. They did not try their
system on multiple coupled interctions, since dynamics
of such networks of traffic nodes are much more complex
and learning or creating controllers for them could show
additional interesting behaviors and rearch questions.
C.R EINFORCEMENT L EARNING
Reinforcement learning for traffic light control has first been studied by Thorpe [13], but Thorpe’s approach
is different from approach in this paper. He ud a traffic light bad value function, and we ud a car bad one. Thorpe ud a neural network for the traffic light bad value function which predicts
the waiting time for all cars standing at the junction. A neural network is ud to
predict the Q values for each decision, bad on the number of waiting cars and the time since the lights last changed. The goal state is the state in which there are no
cars waiting.
Thorpe trained only a single traffic light controller, and tested it by instantiating it on a grid of 4 X 4 traffic lights. The system outperformed both fixed and rule bad controllers in a realistic simulation with varying speed.
IX.C ONCLUSION
A basic fuzzy logic control algorithm for full interctions and left and right turns lanes was developed. The fuzzy logic controller makes the decision to what extent the current green pha has to be extended bad on a t of fuzzy rules and real-time traffic information.
A large number of improvements are planned for the future; they include the following:
•Additional simulation tests on interctions with different levels of geometric complexity, phasing
自己编童话and demand.
•The cycle time may be further fuzzified to get better results.
•Expansion of the fuzzy logic controller strategy to arterial and network applications.
R EFERENCES
[1]Niittymäki J., Mäenpää M. (2001). The role of
fuzzy logic public transport priority in traffic signal
control. Traffic Engineering and Control,
International Journal of Traffic Management and
京本かえでTransportation Planning, January 2001, Hemming-
Group Ltd. By permission.
[2]Pappis, C. P., and E. H. Mamdani. A Fuzzy Logic
Controller for a Traffic Junction. IEEE
Transactions on Systems, Man, and Cybernetics,
Vol. SMC-7, No. 10, October 1977, pp. 707-717. [3]Kely, R. L., and K. R. Bist. Simulation of
Traffic Flow and Control Using Fuzzy and
Conventional Methods. Fuzzy Logic and Control:
Software and Hardware Applications, Prentice
Hall, Englewood Cliffs, New Jery, 1993, pp.
262-278.
[4]Niittymaki, J., and S. Kikuchi. Application of
Fuzzy Logic to the Control of a Pedestrian
Crossing Signal. In Transportation Rearch
Record: Journal of the Transportation Rearch
Board, No. 1651, TRB, National Rearch Council,
Washington, D.C., 1998, pp. 30-38. [5]Chiu, S. Adaptive Traffic Signal Control Using
Fuzzy Logic. Proceedings of the IEEE Intelligent
Vehicles Symposium, 1992, pp. 98-107.
[6]Niittymaki, J., and M. Pursula. Signal Control
Using Fuzzy Logic. Fuzzy Sets and Systems, Vol.
116, 2000, pp. 11-22.
[7]Nakatsuyama, M., H. Nagahashi, and N. Nishizuka.
Fuzzy Logic Pha Controller for Traffic Junctions
in the One-Way Arterial Road. Proceedings of the
IFAC Ninth Triennial World Congress, 1984, pp.
2865-2870.
[8]Li, H., P. D. Prevedouros, and L. Zhang. Signal
Control for Oversaturated Interctions Using Fuzzy LogicSubmitted for consideration for prentation at the 2005 Annual Meeting of the TRB and publication in the Transportation Rearch Record.
[9]Hong Wei, Wang Yong, Mu Xuanqin and Wu Yan.
A cooperative fuzzy control method for traffic
Lights. 2001 IEEE Intelligent Transportation Systems Conference Proceedings - Oakland (CA),
USA - August 25-29, 2001.
[10]Yi Hu, CQU, Peter Thomas, Member, IEEE, and
Rusl J. Stonier, Member, IEEE. Traffic Signal
Control using Fuzzy Logic and Evolutionary Algorithms. 2007 – IEEE.
[11]J Niittymaki, R Nevala, E Turunen. Fuzzy Traffic
Signal Control and a New Inference Method!
Maximal Fuzzy Similarity - Fuzzy Sets and Systems, 2003 - it
[12]Marco Wiering, Jelle van Veenen, Jilles Vreeken,
Arne Koopman. Intelligent Traffic Light Control.
Institute of information and computing sciences,
Utrecht university technical report UU-CS-2004-
029.
[13]Thorpe, T. (1997). Vehicle traffic light control
using sarsa. Master’s thesis, Department of
Computer Science, Colorado State University. [14]Taale, H., B¨ack, T., Preuß, M., Eiben, A. E., de
Graaf, J. M., and Schippers, C. A. (1998).
Optimizing traffic light controllers by means of
evolutionary algorithms. In EUFIT’98.
[15]Findler, N. and Stapp, J. (1992). A distributed
approach to optimized control of street traffic
signals. Journal of Transportation Engineering,
118-1:99–110.
