Auto-Tuned Particle Swarm with perturbed velocities - Optimized CLMF Algorithm for DVR based Power Quality Improvement

Problems pertaining to power quality, like voltage sags and swells as well as harmonics, have the potential to impact the reliability of contemporary power systems. An innovative approach that integrates the Composite Least Mean Fourth (CLMF) algorithm with meta-heuristic algorithms like PSO, DE, PSODE and a hybridization Auto-Tuned PSO and DE (APSODE)—to enhance the performance of a Dynamic Voltage Restorer (DVR) is presented in this study. The CLMF algorithm is used mostly to mitigate voltage distortions, while the meta-heuristic algorithms are employed primarily to optimize the DVR's control parameters for improved operation over different power quality conditions. Extensive simulations, conducted in the MATLAB environment, validate the proposed method, showing a substantial improvement in voltage restoration, reduced Total Harmonic Distortion (THD), and faster dynamic response. The findings of the study highlight the effectiveness of combining CLMF with PSO, DE, PSODE and APSODE in addressing power quality issues, offering improved performance and higher adaptability to changing conditions of power system. The proposed APSODE algorithm introduces auto-tuned acceleration coefficients and hybrid velocity mutation mechanisms, offering real-time adaptability and improved convergence over traditional methods, thereby ensuring strong voltage compensation under dynamic power quality conditions.