基于模糊神经网络的驼峰推送速度控制研究

对影响驼峰机车推送速度的各种因素进行了探讨,将模糊神经网络理论引入到驼峰变速推峰系统中,结合神经网络和模糊控制的各自优点,将专家或熟练操作人员的控制策略转换为控制功能函数,利用神经网络自学习、自校正的能力调整、修改模糊控制规则,从而实现对推送速度的优化控制。

Study of Hump Pushing Speed Control Based on Fuzzy Neural Network

This paper discusses each kind of factors which influence pushing speed of hump engine,and introduces fuzzy and neural network theories into the pushing speed changing system of hump engine.The system is unified respective merit of the neural network and fuzzy control.The experts and skilled operator＇s strategies are transformed into functions of control,and the fuzzy control rules may be adjusted or revised by using the self studying and self correcting ability of neural network so that accurate control is realized.