Calculation Analysis on Traction Motor Temperature Rise of EMU Vehicles Based on Fuzzy Neural Network

Abstract

As a major consideration during the traction system design procedure, the traction motor temperature rise is also deemed as an important physical parameter for evaluating performance of the traction motor over the course of long-term service. Due to the operating conditions, ambient temperature and other factors, it is difficult to accurately assess the traction temperature rise during the designing process of traction system. On a basis of extensive data analysis on traction motor temperature rise tests, this paper first adopts the Heuristic method of Sequential Forward Selection to determine main factors that cause the traction motor’s temperature rises in various operating conditions. Then the fuzzy neural network calculation models of traction motor temperature rise is established under different working conditions. Training these fuzzy neural networks with sample data from route test to obtain the traction motor temperature rise calculation model under full operating conditions and the whole climate environment. Taking actual parameters of a certain type EMU (Electric Multiple Units) of the Beijing—Shanghai line as the object, this paper compares the temperature variation of the traction motor obtained from the simulation calculation with the experimental data in a way to justify the correctness and validity of the selected method.