Multi Objective Optimization of Power Dispatch Problem Using NSGA-II

Abstract

The optimization of power dispatch is a general term that can be used for the optimization of various objective(s) for thermal units, the economic dispatch, generation scheduling or real power dispatch are interchangeable and can be attempted to find the generation level by minimizing fuel cost. The environmental dispatch is the problem of obtaining the generation by minimizing the emission due to thermal units. The reactive power dispatch is attempted by minimizing the losses incurred in the transmission network. Such optimization must be carried out to obtain the solution subjected to equality constraint by power balance equation these problems when attempted separately are bound to give different results and may be conflicting. This situation is similar to most of the practical decision making or optimization. The multi objective optimization of power dispatch problems has been attempted by considering various combinations of objective functions corresponding to real power, emission and reactive power dispatch, the emission dispatch is represented as minimization of SOX, NOX and COX emissions. An algorithm to attempt such multi objective optimization has been formulated using non dominated sorting genetic algorithm (NSGA-II). The results obtained for NSGA-II are indicating optimal pareto. A program is also developed to obtain best compromise solutions. The performance has been studied for various combinations of objectives for systems employing three generators and six generators.