Reactive Power Optimization of DG-Integrated Active Distribution Network Based on an Improved Multi-Objective Particle Swarm Optimization

Abstract

Active distribution networks (ADNs) achieve the secure and cost-effective operation of distribution grids by coordinating controllable assets integrated in the grid. However, against the backdrop of the persistent rise of the grid-connected capacity of distributed generation (DG), ADNs are confronted with considerable challenges in addressing DG output randomness and coordinating multi-type reactive power resources. To address the operational challenges of distribution networks under different DG penetration rates, and by comprehensively considering the operational characteristics and constraint conditions of various adjustable reactive power devices in the network, a reactive power and voltage optimal control model is established to achieve the synergistic optimization of network loss, voltage stability, and the switching cost of discrete devices. To improve the operational stability of distribution networks, an Improved Multi-objective Particle Swarm Optimization (IMOPSO) algorithm is proposed in this study. By combining random opposition-based learning (ROBL) with Orthogonal Latin Hypercube Sampling (OLHS), adaptive inertia weight, differential evolution algorithm-based co-evolution and two-stage leader selection, this algorithm effectively reduces the likelihood of trapping in local optima. Simulation outcomes derived from the modified IEEE 33-node test feeder indicate that the presented approach achieves trade-off and collaborative optimization across multiple objectives, exhibits enhanced stability and improved solution precision, and thereby provides reliable technical support to facilitate the secure and cost-effective operation of ADNs.