Impact of Thermal Aging on Winding Insulation Loss-of-Life Fraction Using H∞ Algorithm for Integrated permanent Magnet In-Wheel Motor

To improve the reliability of in-wheel structures based on permanent magnet synchronous motors (PMSM), thermal management is a crucial requirement. However, online thermal analysis for in-wheel motors which drives electric vehicles (EVs) specially autonomous electric vehicles (A-EVs) is a highly challenging task due to the complex thermal distribution affected by geometrical and material parameters and intricate boundary conditions. In the pursuit of avoiding winding insulation to get overheated and the motor lifetime to be shortened, real-time temperature estimation is needed. In this contribution we focus on a model-based robust temperature estimation for a PMSM on the stator and rotor parts, specifically to evaluate gradual winding insulation deterioration. The proposed temperature estimation method is based on H ∞ algorithm applied on a low-order lumped parameter thermal network (LPTN). Then, the insulation loss-of-life fraction is investigated when an over-current occurs during a short-time thermal overload. Finally, a comparison between numerical simulation results and experimental results confirms the robustness, feasibility and effectiveness of the presented model.