论文部分内容阅读
研究了汽轮发电机组轴承转子系统摩擦学设计知识获取的神经网络方法,通过分析BP网络权重,得到网络输入与输出的函数关系,并定义了关系系数C.由关系系数C的大小和正负可以确定输入与输出的关系。
The neural network method for obtaining the tribological design knowledge of the bearing rotor system of a steam turbine generator set is studied. By analyzing the weight of the BP network, the functional relationship between the input and the output of the network is obtained and the relation coefficient C is defined. The relationship between the input and output can be determined by the size and positive and negative of the coefficient of relationship C.