A modified hybrid multi-objective GA and LSF algorithm for optimal siting and sizing of PV-based distributed generation in distribution networks considering different types of loads

Abstract

Many methodologies have been developed for the problem of optimal siting and sizing of photovoltaic (PV) distributed generation in distribution networks. However, clear solar potential assessment and variability of load constitutions at buses are aspects that are still missing in current optimization formulations. In this paper, a new methodology that uses the Multi-Objective Genetic Algorithm (MOGA) is proposed for the optimal allocation of PV-based distributed generation units (PV-DGs) in distribution networks. The method aims at minimizing active power losses, voltage deviations, and energy costs using a practical model that considers both the solar potential of each bus and variable load classifications. For this purpose, a modified MOGA algorithm that incorporates a Loss Sensitivity Factor (LSF) of solar potential in the optimal allocation problem at peak load is developed and compared to the traditional MOGA algorithm. The hybrid MOGA-LSF algorithm involves a two-step optimization problem. In the first step, the candidate buses for optimal allocation of PV DGs are specified using the LSF algorithm and in the second step, the radial power flow nested within the MOGA formulation is used to determine the sizes of the DGs to be installed. The time variations of loads are modeled by dividing a typical summer day into six intervals, each having different solar irradiance levels. Loads profiles of different classifications are also modeled for each candidate bus. The proposed technique is applied to the IEEE-15 bus radial distribution network to illustrate its effectiveness.