Performance Improvement of Active Distribution Systems Using Adaptive and Exponential PSO Algorithms

Abstract

This paper presents heuristic-based approaches to enhance the voltage profile and reduce the power losses of Radial Distribution Systems (RDS) equipped with Distributed Generation (DG) sources. These approaches determine the size and location of the DG using modified versions of the Particle Swarm Optimization (PSO) algorithm. Standard, adaptive, and exponential PSO algorithms are derived and compared to each other. It is found that exponential PSO converges faster than adaptive PSO, although they give nearly the same final results. These algorithms are applied and used for single-objective PSO and Multi-Objective PSO (MOPSO) problems to minimize the losses, enhance the voltage stability, and reduce the total voltage deviation of the studied distribution systems. Different methods, including aggregated sum and Pareto-optimal set, are used for MOPSO with a fuzzy-based function to determine the best compromise solution. The proposed algorithms are applied to the balanced IEEE 33- and unbalanced 37-bus radial distribution systems. The obtained results are compared with other algorithms, and it is found the effectiveness of the proposed methodologies.