The diesel engine-driven alternator, a critical component in off-grid power generation systems, relies on the automatic voltage regulator (AVR) to maintain stable output voltage under nonlinear loads. This research aims to develop an optimized excitation control strategy to improve the efficiency and reliability of diesel engine-driven alternators under nonlinear loads. The proposed approach incorporates a cognition-based particle swarm optimization (PSO) algorithm to precisely tune the control parameters Kp, Ki, and Kd in real time. By detecting and responding to instantaneous variations in voltage and current, the advanced excitation control system is expected to enhance the alternator's adaptability and maintain desired voltage levels, even during sudden load changes. The research will also investigate the integration of AI-based technologies, such as adaptive load-sensing algorithms, digital signal processors, and predictive maintenance, to further optimize AVR performance. The effectiveness of the proposed excitation control strategy will be evaluated through simulations and experimental validation. The successful implementation of this intelligent PSO-based control system is expected to significantly improve the stability and operational efficiency of diesel engine-driven alternators under nonlinear loading conditions, ultimately contributing to more reliable and cost-effective off-grid power generation.