Performance of Mixed Integer Non-linear Programming and Improved Harmony Search for optimal placement of DG units

Distribution system loss reduction is one of the prime objectives for planning of distributed generation. To minimize the losses, optimal sizing and siting of distributed generators (DGs) is critically important. In this paper, the nonlinear and non-convex optimization problem of placement DG units is solved using two methods, namely Mixed Integer Non-linear Programming (MINLP) and heuristic approach based Improved Harmony Search (IHS). Both MINLP and IHS based algorithms are implemented for optimal placement of single and multiple DG units capable of delivering either real or both real and reactive power. In MINLP formulation, the potential locations are obtained by sensitivity analysis to reduce the search space and then the problem is solved for optimal sizes and locations. The results are also obtained with IHS due to its non-monotonic solution surface. The proposed optimization approaches are tested on IEEE 33-bus distribution system. The results of proposed formulations are compared with Improved Analytical (IA) approach. Improved solutions and moderate computational time confirms the potential of MINLP based formulation.