基于人工神经网络预测控制的交通信号调度

在传统的交通信号控制中，信号的变化周期一般是固定的，由于车流量随时间的不确定性，引起了道路负荷的不均衡，容易造成道路拥塞或闲置现象。对基于人工神经网络的预测控制算法进行介绍。根据预测结果对整个路况进行决策判断，实现交通灯信号周期的自适应调节，从而实现交通流量的负荷均衡。根据城市交通系统的特点，设计一个基于神经网络的单个交叉路口的交通灯预测控制系统，得出相关不同时间段内的交通灯控制周期。分析表明，该方法能有效提高车辆通行效率，增强道路的吞吐能力。

Traffic Signal Scheduling Based on Artificial Neural Network Prediction Control

The cycle of traffic signal in crossroads is determined in traditional control. According to uncertainty in vehicles flow, road load is unbalanced. It can easily lead to load jamming or idling. A predictive control algorithm based on arti- ficial neural network is introduced in the paper. The technology can be used to analysis the situation and decide the variation of traffic signal cycle. The traffic load balance can be achieved. Base on urban road intersection, this paper designs a predictive control system for intersection transportation lights control based on neural network. The different transportation lights control cycles in different periods are given. The method can improve road transportation efficiency and increase the load throughout.