Traffic Signal Control Based On Fuzzy Discrete Event Systems And Simulation Study

As traffic increase sharply and traffic flow become more complex, the traffic jams at signalized intersection have become main bottleneck of urban traffic. In traffic control system, many conventional researches have been made to control traffic including fixed-cycle and vehicle-actuated methods. Though it is effective to some extent, these means are not so effective when confronting complex time-variable traffic.This article is mainly aiming at this problem. According to characteristics of complexity, randomicity, difficulty to describe with mathematical model and advantages of the fuzzy discrete event systems(FDES) in controlling without establishing exact mathematic model, this paper presents traffic real-time FDES control method based on vehicles waiting queue number and average waiting time in four-phase isolated intersection. With FDES optimization control method, corresponding control strategy is designed and a simulation experiment performed based on Matlab. The simulation results indicate that the control is effective. The main research contents are as following:1. To perfect basic theories of FDES, the paper researches further on states transit of FDES and perfects and supplements the theorem. In order to solve complex combination of discrete operation, the paper proposes one method which can reduce complexity of synthesized FDES.2. Based on the existed researches, the paper improves computing method of average waiting time. It takes average waiting time of current waiting vehicles as evaluating indicator of pass demand which makes the computation of pass demand of phase more reasonable than before.3. The first model is established based on signalized intersection control taking average waiting time and vehicles waiting queue number of current and successor phase as states, and taking corresponding forecast value of future period of time as events. Then we design the phase sequence and delay strategy of green light. The results of simulate based on Matlab indicate that the model is valid.4. In the second FDES model of signalized intersection, it takes the green light time as the system states, and takes arrival and departure of vehicles as event. Finally simulation is performed and micro-analysis also performed with simulation results to analyze the traffic efficiency. The results indicate that signal control model based on FDES is suitable and effective.The paper proposes methods to control traffic light based on FDES. It is practical and significant to increase pass capability at intersection and decrease traffic jams.