Rotor Temperature Estimation of Permanent Magnet Synchronous Motors Considering Electro-Thermal Bidirectional Coupling

Accurate and real-time rotor temperature estimation is critical for the safe and reliable operation of permanent magnet synchronous motors (PMSMs) in high-speed trains. This paper presents a real-time estimation method based on electro-thermal bidirectional coupling. Firstly, a temperature-dependent electromagnetic loss model is developed to account for the thermal characteristics of motor materials. Secondly, a bidirectional coupled finite element model is constructed to simulate the interaction between electromagnetic and thermal fields. The resulting temperature profiles serve as reference data for model parameter identification. Subsequently, a lumped-parameter thermal network model is built, with detailed node segmentation along heat transfer paths. A particle swarm optimization (PSO) algorithm is applied to identify thermal resistances and capacitances by minimizing the error between the network output and finite element results. Finally, the proposed method is validated through simulation using real-world current, speed, and torque data from high-speed train operation. The estimated rotor temperatures closely match the finite element benchmark, with errors within ±3%, demonstrating both accuracy and real-time capability for practical traction motor monitoring.