摘要
针对城市交叉口交通流的特点,提出了一类交叉口多相位自适应控制算法,介绍了基于改进的多层神经网络实现交叉口模糊控制的方法,给出了学习算法。仿真结果表明,设计模糊神经控制器具有学习和泛化能力,实时性好,能在车流大小随机变化的单个路口中得到应用。
In light of the characteristics of the traffic flow in urban intersections, this paper presents a kind of intersection multi-phase self-adaptive calculation. It introduces the approaches of realizing intersection fuzzy control based on improved multi-layer neural net and then raises the study calculation method. The simulation result shows that the designed fuzzy neural controller has the capability of study and generalization with real time features and can be applied to single intersections with random traffic flows.
出处
《广西交通科技》
2003年第5期20-23,27,共5页
Guangxi Communication Science & Technology
基金
国家自然科学基金 (60 0 640 0 1 )
江西省自然科学基金 (0 0 1 1 0 0 6)
广东省自然科学基金 (2 0 0 1 1 70 7)资助课题
