Application research of PID neutral network control for traction drive inverter

This paper presents a study on PID neural network control used in traction inverter drive system. The analysis is based on structure of three level inverter, motor dynamic d-q equivalent circuit, and direct rotor field-oriented vector control mathematical equation. A direct vector control system of synchronous current control with feedforward CEMF compensation is built. Speed loop and flux loop controllers are realized with PID neural network. Training methods of feedforward neural network and back propagation are also given. MATLAB simulation results show good performance of the system in transient and stable stage.