DESIGN AND IMPLEMENTATION OF A CONTROL STRATEGY FOR INCREASING THE LIFETIME OF POWERTRAIN COMPONENTS OF ELECTRIC VEHICLES.

The increasing diffusion of battery electric vehicles (BEVs) raises critical challenges concerning the reliability and lifetime of power electronic converters, which are exposed to severe thermal and electrical stress. This thesis investigates lifetime-aware control strategies for speed regulation in electric vehicles, aiming to mitigate component degradation while preserving drivability. As an initial step, a rule-based controller has been implemented to demonstrate the fundamental principles of lifetime-aware design. This approach is mainly intended as a proof of concept and preliminary experiment to evaluate the effects of controlling certain parameters on lifetime. Building on these insights, a Nonlinear Model Predictive Control (NMPC) framework is developed. The NMPC framework explicitly integrates cumulative damage models and online predictors based on cycle counting. This enables adaptive penalisation of thermal stress and real-time trade-offs between tracking performance and semiconductor lifetime. The strategies are tested through high-fidelity simulations of a Fiat E-Doblò electric vehicle under standard driving cycles (WLTC and ARTEMIS MW130). Results show that the NMPC achieves a significant reduction in peak temperature and thermal cycling, leading to slower damage accumulation compared to baseline controllers, while ensuring accurate speed tracking. The main contribution of this work is to demonstrate that embedding reliability objectives into predictive control frameworks can effectively extend component lifetime and reduce maintenance costs, supporting the long-term adoption of BEVs.