基于神经网络的地铁列车运行过程的集成型智能控制  

Automatic subway train operation control based on neural network and other intelligent control methods

作　　者：武妍[1] 施鸿宝[1] 

机构地区：[1]上海铁道大学计算技术研究所,上海200331

出　　处：《铁道学报》2000年第3期10-15,共6页Journal of the China Railway Society

基　　金：上海市教委重点学科项目 !(B971 70 1 )

摘　　要：提出了一种以神经网络为主 ,并综合使用模糊系统、遗传算法等方法的集成型地铁列车智能控制方法。分别采用模糊控制的 BP网络和基于遗传算法的模糊神经网络 ,实行了站间运行控制和定位停车控制 ,仿真取得了令人满意的结果。结果表明 ,将神经网络等计算智能方法综合用于列车运行控制是可行的 ,可以保证较好的舒适性、停车准确性 ,具有精度高 ,适应性、鲁棒性强等特点。A integrated intelligent control method for subway train operation process is proposed. The method which uses fuzzy system and genetic algorithm is mainly based on neural network. A FBP neural network is used to control train operation before the stopping point. A fuzzy neural network based on genetic algorithm is used to control train for stopping. The simulation results are satisfactory. The results indicate that it is feasible for automatic train operation based on neural network and other computing intelligent technology. The method can guarantee quite good comfortbility, stopping accuracy, and has the characteristics of high accuracy, adaptability, and robustness.

关 键 词：ATC ATO 神经网络 遗传算法 模糊逻辑 列车运行 

分 类 号：U231.6[交通运输工程—道路与铁道工程]