摘要
利用以真值流描述知识流为基础的动态模糊推理神经网络，构造了一个模糊推理控制系统用于控制列车在区间恒速区的匀速运行。通过一个权值矩阵生成模块，利用司机的驾驶经验来获取相应的控制策略逻辑蕴含强度，生成一族对应的权值矩阵，并利用该权值矩阵和当前列车走行速度为输入变量，经过动态模糊神经网络的推理，最后可以得到某个收敛的控制输出变量，该输出变量能够较好地体现司机的控制思维过程和控制策略。
Using dynamic fuzzy reasoning neural network based Onknowledge flow described by truth- value flow,the authors construct a fuzzy reasoning control system which makestrain running at constant speed.With the weight matrix forming module,the authors get thestrength of the logical control strategy and a group of weight matrixes from driver's experiences,Using this weight matrix and current speed as input parameters,after reasoning bydynamic fuzzy neuraI network,a convergence control output parameter which perfectly expresses the dirver's control strategy may be obtained.
出处
《铁道学报》
EI
CAS
CSCD
北大核心
1995年第S2期74-78,共5页
Journal of the China Railway Society
关键词
模糊推理
神经网络
列车自动控制
模糊控制
fuzzy reasoning
neural network
automatic train control
fuzzy control
