Abstract: Combining particle swarm optimization (PSO) with deep learning (DL), the present work provides a new method for the hybrid optimization of the best placement and real-time control of unified power flow controllers (UPFCs) in power systems. The suggested technique combines the global search capacities of PSO to identify the ideal position for UPFCs with the adaptive control power of DL to dynamically update UPFC settings in response to system changes, hence overcoming the limits of current approaches. This method is special in that it considers transient and voltage stability margins, hitherto neglected in earlier research. This element of the advised approach reduces UPFC costs, fuel, and power loss. The efficacy of the hybrid DL-PSO approach is confirmed by means of the IEEE-26 bus test system; the obtained results are compared with those of conventional algorithms like Genetic Algorithm (GA), PSO, and GSA. The advantages of the suggested technique in terms of speed and accuracy for UPFC deployment and control are shown by the simulation results, thereby stressing its possibility of raising power system stability and efficiency. This creative technique opens the path for further advancements in AI-driven power system management by providing a consistent answer for challenging power system optimization.

Keywords: UPFC, Particle swarm optimization, Control, IEEE-26 bus, AI-Driven Hybrid Optimization, Power Systems

Pages: 665-674