����������Ľ���ڿɱ�����ģ�����Ʒ���

针对城市交叉口交通流的特点，提出了一种自适应可变相序的多相位控制算法。该算法依据绿灯相位车队长度和红灯相位车队长度的比较决定绿灯相位是否转移，在不需要绿灯相位转移时，利用模糊神经网络控制器控制绿灯延时长度。不但结合了模糊控制和神经网络控制的优点，而且所给出的算法相序可变，实现了道路交叉口多相位相序可变控制。仿真结果表明，本文设计的模糊神经网络控制器能够有效降低车辆平均延误，满足实时控制的要求。

An Alterable-phase Fuzzy Control Based on Neutral Network

Based on the traffic flow characteristics in urban intersections, this paper presents a control algorithm which is adaptive and can transform the multi-phase sequence accordingly. This algorithm can decide whether the phase needs to be transferred based on the comparison of the length of the green-lights phase motorcade and the red-lights phase motorcade. When the green light phase does not need transfer, the fuzzy neural network can be used to control the length of green-light delay. This paper not only combines the fuzzy control and neural network control advantages, but also provides an algorithm whose phase sequence is variable was. And then the multi-phase and variable phase sequence control can be achieved in the intersection. Simulation results show that the design of the fuzzy neural network controller can reduce the average delay of vehicles effectively, and meet the demand for real-time control.