摘要
提出了一种基于交通流预测的主干道相交路口优化控制方法,将交通流预测与交通信号控制相结合,用于解决交通流量较大的城市主干道相交路口的信号控制问题。建立神经网络预测模型,用变异粒子群优化算法优化网络结构以提高预测精度和速度,将预测的下2个相位队长作为模糊控制的输入以确定下一绿相位时间,在后一绿相位持续时间内放行该相位经预测但尚未全部排队的车辆。仿真实验表明该方法能有效地减小平均车辆延误时间,达到了保持干道交通通畅的目的。
In this paper an optimization control method of urban intersecting arterial roads intersection based on traffic flow forecasting was put forward, which integrated traffic flow forecasting and traffic signal control, with the purpose to resolve traffic signal control problem of urban intersecting arterial roads intersection with heavy traffic volume. A neural network prediction model was established. An algorithm of dissimilation particle swarm optimization was used to optimize network structure and thus improve forecasting precision and speed. Forecasting queue length of next two phases was taken as the input of fuzzy control to fix the next green phase time. Incomplete Forecast queue traffic volume of the next green phase was released during its standing. The simulation experiment indicates that the method reduces the average vehicle delay time efficiently, and achieves the purpose of keeping the traffic smooth in arterial roads.
出处
《交通与计算机》
2008年第4期43-46,共4页
Computer and Communications
基金
国家自然科学基金项目(批准号:60674062)
山东省教育厅科技计划项目(批准号:J07WJ05)
济南大学博士启动基金项目(批准号:B0608)资助
关键词
干道相交路口
神经网络
交通流预测
变异粒子群优化算法
模糊控制
intersecting arterial roads intersection
neural network
traffic flow forecasting
dissimilation particle swarm optimization algorithm
fuzzy control