Sensitivity Guided Genetic Algorithm for Placement of Distributed Energy Resources

This paper introduces an enhanced Genetic Algorithm (GA) that uses the concept of sensitivity analysis to develop an encoding strategy for improving the computational efficiency of the search process. The objective is to determine the optimal placement of distributed energy resources (DERs) with minimum generation cost. Locational Marginal Price (LMP) is employed as an indicator to quantify the need for additional generation at candidate locations. LMP at each node is determined from Lagrange multipliers associated with the power balance equation at that node. By renumbering and encoding the locations based on their LMP ranks, desired candidate locations are gathered and encoded to share more common genes. Then, genetic algorithm is utilized along with the AC optimal power flow model to search for the optimal locations for distributed energy resources with varied sizes. The method is demonstrated on several test systems, including IEEE 14, 30, 57 and 118 bus test systems. The placement of DERs with minimum generation cost is found and the results validate the improvement in convergence speed.