Power Quantity Enhancement of UPQC-DG Using PSO and ANN

The efficiency of an electrical system largely depends on power quality, making the integration of a Unified Power Quality Conditioner (UPQC) with Distributed Generation (DG) systems essential to mitigate issues such as flicker, harmonics, voltage sags, and swells. In UPQC-DG systems, the Proportional-Integral (PI) controller plays a critical role in stabilizing the DC link voltage, ensuring consistent power quality and seamless grid integration. An effectively tuned PI controller minimizes voltage ripples and offers fast disturbance response, thereby enhancing overall system performance. However, conventional PI tuning techniques, such as the Ziegler-Nichols method, often yield suboptimal results under dynamic and complex operating conditions. This paper proposes a novel real-time PI controller optimization approach for UPQC-DG systems using Particle Swarm Optimization (PSO) in conjunction with Artificial Neural Networks (ANN). The objective is to adaptively fine-tune PI controller parameters in real time, enabling stable and accurate regulation of the DC link voltage under varying system conditions. The proposed hybrid PSO-ANN strategy leverages the global search capability of PSO and the predictive learning strength of ANN to ensure dynamic adaptability and robustness. This study highlights the effectiveness of the PSO-ANN-based adaptive control in maintaining high power quality and improving the reliability and stability of UPQC-DG systems.