Abstract: In view of the technical problem that it is difficult to obtain the dynamic characteristic equation of the main engine unit of most pumping stations in China,a method is proposed to obtain the dynamic characteristic equation of the main unit of the pumping station through the comprehensive characteristic curve of the model operation provided by the manufacturer of the main unit of the pumping station.Firstly,the discrete characteristic data is obtained through the comprehensive characteristic curve of the model operation,and then the dynamic characteristic equation of the main unit model of the pumping station is formed by the double multiplication fitting. Secondly,the static operation data samples are obtained through the real computer monitoring historical database and the cloud database of the pump station operation management,and the static point correction is carried out by using the cubic index matrix data processing method. Finally,the extension theory and neural network calculation method are combined to create an extension neural network training method to realize the dynamic correction of the dynamic characteristic equation of the original main unit of the pumping station,and the method has the characteristics of self adaptability,self-learning habit and extension. The application practice shows that the innovative combination of the cubic index matrix data processing method and the extension neural network training method can accurately,effectively and reliably obtain the dynamic characteristic equation of the main unit of the pumping station,which provides a scientific decision-making basis for the safe and reliable operation of the whole pumping station,the combined dispatching of the main units and the optimal load distribution.