An emission constraint capacitor placement and sizing problem in radial distribution systems using modified competitive swarm optimiser approach

Abstract

This paper presents an environment friendly and cost-effective solution procedure to solve the Optimal Capacitor Locations and Sizing (OCLS) problem for power distribution systems. The OCLS problem is formulated as a two-stage Mixed Integer Non-Linear Programming problem. In the first stage, a new Relative Emission Index based formulation is introduced to select the probable candidate buses for the capacitors. In the second stage, the optimum locations and sizes for the capacitor are determined using basic Competitive Swarm Optimiser (CSO) algorithm. Further, a variant of CSO called ‘Opposition based CSO (OCSO)’ is proposed and implemented in this work to get improved results. Due to the introduction of opposition based learning phase, three-fourths of the swarm are updated in each cycle in the proposed method, unlike CSO. The annual cost is considered as an objective function in this problem. The capacitor locations and VAr outputs are taken as binary and discrete control variables respectively. The CSO and its proposed modified version ‘OCSO’ are tested on standard 69 and 118 bus RDS networks with constant power load as well as with practical nonlinear loads. The results obtained with the proposed method are validated by comparing with other reported results in the state-of-art literature and those obtained by other methods.