Lifetime-aware nonlinear model predictive speed control for electric vehicle power converters

This paper presents a nonlinear model predictive control (NMPC) strategy for speed tracking in electric vehicles that accounts for the thermal degradation of the power inverter module. The method integrates a real-time damage predictor based on thermal cycling and cycle counting, enabling adaptive weighting in the NMPC cost function. This allows the controller to balance speed tracking and thermal stress, extending component lifetime based on the actual and predicted health of the power module. Simulations results using a high-fidelity vehicle model under WLTC and ARTEMIS cycles show that the damage-aware controller significantly reduces thermal stress and accumulated damage while maintaining accurate speed tracking.