Effects of parameter variation on imperialistic competition based economic emission dispatch

Imperialistic competitive algorithm (ICA) is a recently developed meta-heuristic technique that has been successfully implemented for solving various complex optimization problems. Like other meta-heuristic techniques, performance of ICA predominantly depends on the selection of control parameters. However, optimal setting of control parameter for ICA is problem specific. In this paper, a study of control parameter setting for an imperialistic competitive algorithm based economic emission dispatch problem with valve point loading and transmission loss is carried out. Economic emission dispatch problem was formulated by taking equality constraint on power balance and inequality constraint on generator capacity limit and losses into consideration. Proposed approach was examined on a ten generating unit system and effects of parametric variation on results are presented. Analysis of obtained results reveals that an improper selection of algorithm control parameter might lead to premature convergence and poor computation efficiency.