城市高速公路交通的神经网络建模与控制

Urban Expressway Traffic Flow Modeling and Control Using Neural Networks

从城市高速公路交通流的宏观、动态特性出发 ,分析了交通流控制中常用的宏观、动态、确定性模型 在此基础上 ,利用人工神经网络技术建立了城市高速公路的神经网络模型 ,并提出了入口匝道放行和路段速度相结合的多变量神经网络控制策略 利用该控制策略建立的自适应神经网络控制器 ,可以使高速公路上的交通密度维持在理想的密度值附近 .进一步分析可以得到 ,该控制器是一个状态和控制作用均可跟踪的伺服系统 .以杭州某高架高速公路为背景的仿真结果表明 :该控制器具有较强的鲁帮性 ,控制效果令人满意 .

From the viewpoint of macro and dynamic character is tics of urban freeway traffic flow, a commonly used macroscopic, dynamic and det erministic traffic flow model for traffic control is developed. Furthermore, the neural network model for urban freeway traffic flows, and the urban freeway mul ti-variable neural control strategy with both the on-ramps control and the roa d speeds control are also presented simultaneously. The developed adaptive neura l controller is used to control the traffic density and force it to follow a des ired one. This control strategy is a servo system of which the states and the co ntrol effect can be followed. Finally computer simulation on Hangzhou urban free way shows that the controller is robust and the result is satisfactory.