Determine the Location for Reactive Power Compensation in the Microgrid Based on the Hybrid Neural Network

Abstract

This paper presents a method to determine the capacity and location of compensating capacitors to reduce power loss and improve voltage quality in the Microgrid. At each bus location, the compensating capacitor capacity is varied to determine the bus location and capacitor capacity. In case of small power loss and good voltage quality, compensation position and capacity will be chosen. The construction of the neural network training dataset is done with load levels from 50% to 100%. For each load level, the reactive power compensation position and the compensation capacity will be determined. The improved PSO algorithm is proposed to improve the traditional neural network structure. The Microgrid 9-Bus power system is used to simulate and test the effectiveness of the proposed method. The results show that power loss and voltage quality achieve positive results. From the simulation results, we can conclude that the proposed neural network model is suitable for controlling the voltage quality of the Microgrid system.