基于遗传模糊算法的高速公路入口匝道自适应控制

摘要

The problem of on-ramp metering of freeway network is dealt with. In order to reduce the peak hour congestion, an adaptive control way is proposed to solve this problem. In a macro-dynamic model, a genetic algorithm is used to tune the fuzzy sets parameters so that the total time spent in the network remains minimum.In an adaptive scheme, the destination independent MetaNet model is used to tune the controller' s parameters and present the simulation results. To evaluate the efficiency of the method, the test results were examined and compared with traditional Alinea controller and genetic-fuzzy ramp metering without adaptive controlling. The results shows that the proposed adaptive genetic-fuzzy control is expected to enhance the performance of the freeway traffic network control while keeping the computational simplicity of the problem.%研究了在高速公路路网中入口匝道的控制问题.为达到在车流高峰时减少交通拥塞的目的使用了一种自适应控制方法:在基于高速公路的宏观动态模型上,用遗传算法调整模糊集合的参数来校准模糊控制器,从而使得网络中总的时间消耗(TTS)保持最小.在自适应控制的总体设计框架下,用基于目的地独立的MetaNet模型来调整控制器,最后给出了仿真结果.为了证明该方法的有效性,把该方法和传统的Alinea控制器方法和只有模糊控制器而没有自适应控制的方法分别作了比较,并且通过实例对控制器进行了评价.结果表明:自适应遗传模糊控制器在保持计算简单性特点的同时,很好地增强了高速公路交通网络的控制性能.