Optimize Location and Capacity of Distributed Sources in Distribution Grids Using Hybrid Neural Networks

This paper, the author studied Improved Particle Swarm Optimization (IPSO) algorithm to determine the optimal location and capacity for installing Distributed Energy Resources (DERs) at buses in the distribution grid system. The IPSO algorithm is combined with a Back-propagation Neural Network (BPNN) to enhance network performance, improve accuracy, and optimize capacity, ensuring the power system maintains stability in terms of voltage, frequency, and power quality. Analysis results on the standard IEEE-14 Bus system identified Bus 12 as the optimal installation point, with an allowable capacity of up to 100 MW and an error rate of 2.16% after 150 iterations. The results were verified and evaluated by using the PowerWorld simulation software, completely suitable for the requirements of optimization problems in planning and developing the power grid considering distributed sources in the distribution grid.