Intelligent Integrated Approach for Voltage Balancing Using Particle Swarm Optimization and Predictive Models

In this paper, an intelligent integrated approach is proposed to control the reactive power and restore voltage balance in a three-phase system using particle swarm optimization (PSO), Gaussian process regression (GPR), support vector machine (SVM). The PSO algorithm used offline mode determine optimal set of firing angles for thyristor-controlled-reactor (TCR) compensator according smallest fitness value required balancing. optimum are then train GPR SVM models. models finally as real-time controller retrieve online mode. A simulation model experimental setup electrical built. modeled consists 500 km long transmission line. line divided into three-pi sections guarantee real response. Several practical case studies have been conducted test validate capability solving unbalance problem. results revealed supreme ability quickly (within 20 ms) wide range factors (VUFs) (3.90–8.42%).