Contents I. Introduction
Exponentially growing demand for electrical power has rendered existing transmission infra-structure unable to cope up with such a huge burden of load demand [1]. Hence the need arises to either invest in increasing the capacity of transmission system or provide the demand locally to the consumer using distributed generation sources near to their locations. Electrical power generation by small scale generation plants distributed over the grid is termed as Distributed Generation. Development of various small scale power generation technologies has led to a surge of research in the domain of Distributed Generation [2]. Distributed sources unlike conventional large scale generation plants require lesser capital investment, maintenance and operation costs and the negative impact on the environment is less especially if the sources are renewable. Advantages of DG unit installation in the grid include reduced network loss and on-peak operating cost, enhanced system security and reliability, peak load saving, improved voltage stability etc [3] [4]. Additional benefits include diversity of energy sources, modular generating plant availability and shorter construction time, relieving of transmission and distribution congestion and reduced transmission costs due to the generation plants being in the vicinity of the heavy loads. However availing these benefits is subject to the condition that optimal size and size location of DG must be used for this purpose. The effects of placing a DG on the network indices will be differ based on its location, type and load at the connection point [6] [7]. Compared to other optimization techniques in literature such as Genetic Algorithm, differential evolution etc. PSO can be executed and modified for the problem in consideration more easily since its governing equations are simpler. Recent research has focused on newly emerging hybrids of PSO for solving numerous optimization problems. Lalitha et al in [8] proposed a fuzzy and PSO method for DG installation in two stages. In the First stage fuzzy approach is used for determining the location best suited for installation of DG. In the second stage sizing of DG is obtained for maximum reduction of loss using PSO. Amanifar et al in [9] suggested a PSO algorithm to find the optimal size and placement of DG units with objectives such as reduction in total cost, number of DGs to be connected and real power losses in the system. Moradi and Abedini proposed a hybrid Genetic Algorithm (GA) and PSO for optimizing location and sizing of DG in [10]. Gomez et al. in [11] suggested a combination of discrete PSO and optimized power flow (OPF) wherein optimal location of DG is computed using Discrete PSO optimal size is found through OPF. Abu Mouti and El-Hawary proposed an Ant Colony Optimization method using only two tuning parameters in [12]. In [13] Garcia and Mena have used Teacher learner Based optimization (TLBO) to determine location and size optimal for DG.
