AI-Driven Optimization of PID-Based AVR Systems for Enhanced Voltage Regulation in Smart Grids

To improve dynamic voltage stability in today's smart grids, this work presents an intelligent design strategy for a two-area AVR system by optimizing PID controller gains with advanced intelligent algorithms. Fast, distributed and adaptive voltage regulation is required in the ever-increasing operational complexity of smart grids, especially in the presence of load disturbances and inter-area oscillations. In this analysis, PID controllers are connected in each of the generating unit. The manual determination, at the early stage serves as an initial parameters for baseline, and then these parameters can be optimized further by particle swarm optimization (PSO), arithmetic optimization algorithm (AOA), Improved Krill Herd algorithm (IKA) and tree-seed algorithm (TSA). TSA among them, has achieved the best performance. For area 1, the TSA-tuned PID decreases the settling time to 1.54 sec and the overshoot to 0.47 %. In Area 2, it settles down at 1.56 sec with an overshoot of 0.70 % having zero steady state error. Effectiveness of the proposed controller is verified by time-domain analysis in a comparative manner with other control schemes. The results justify that TSA-based PID tuning can effectively improve the dynamic voltage regulation performance, and it is a robust and adaptive approach for nextgeneration smart grid control.