Research on Traction and Levitation Characteristics of High-Speed Maglev Train Based on BP Neural Network

The finite element model of long stator linear synchronous motor (LLSM) applied in high speed maglev train is built, and then FEM results of different working conditions are calculated as sample data. With these sample data, the fitted model between the electromagnetic forces and two currents is erected by BP neural network, and the fitted model between Back-EMF, current and operating speed is also obtained. Based on fitted model, considering the feeder cable and the limit of converter, the output of converters and the residual acceleration are calculated under two kinds of power supply modes. The calculation results can be helpful for the design of traction system of LLSM due to consider the coupling effect of traction and levitation.