AI‐Driven and Cyber‐Secure Embedded Inverter Control for Energy Optimization in Electric Vehicles

The rapid evolution of intelligent electric vehicle (EV) powertrains requires inverter‐control systems capable of simultaneously optimizing energy efficiency and ensuring cyber resilience under real‐time embedded constraints. This work proposes a unified AI‐driven and cyber‐secure inverter‐control framework for permanent magnet synchronous motor (PMSM)‐based EV systems. The proposed methodology integrates a Deep Q‐Network (DQN)‐based deep reinforcement learning (DRL) adaptive switching controller with encrypted command validation and an embedded intrusion detection system (IDS). The DRL controller minimizes switching losses, torque ripple, and total harmonic distortion (THD) while satisfying thermal and operational constraints through multi‐objective optimization. The framework is implemented on an ARM Cortex‐M7 (STM32H7, 400 MHz, 2 MB Flash, 1 MB SRAM) platform connected to a three‐phase inverter prototype and validated using publicly available WLTP and UDDS driving‐cycle datasets in MATLAB/Simulink and hardware‐in‐the‐loop environments. Experimental results obtained across repeated simulation and hardware‐in‐the‐loop evaluations under WLTP and UDDS conditions demonstrate an average peak efficiency of 95.4%, representing a 4.2% improvement over conventional field‐oriented control (FOC), along with a 34.5% reduction in switching losses, 67% reduction in torque ripple, and THD reduction to 3.2%. The embedded IDS achieves 98.6% attack‐detection accuracy with an average detection latency of 3.2 ms and only 1.6% computational overhead. The results confirm that the proposed integrated framework provides an effective balance between adaptive energy optimization, embedded feasibility, and cybersecurity for next‐generation EV power‐electronic systems.