A Comparative Study on Constraint Handling for Solving Economic Dispatch by Evolutionary Algorithms

Evolutionary algorithm (EA) has been well accepted as a suitable tool for solving non-convex economic dispatch problems. However the major challenge is how to handle both equality and inequality constraints properly. The penalized fitness is commonly used to evaluate quality of candidate solutions. Moreover searching for feasible space is very difficult for a problem with large number of variables and number of constraints. This paper proposes a general framework which can be applied to any EA for handling constraints in ED problems. Four EAs namely differential evolution (DE), genetic algorithm (GA), particle swarm optimization (PSO) and harmony search (HS) were selected to demonstrate effectiveness in solving a non-convex ED problem of a 15 unit power system. Simulation results reveal that with the proposed constraint handling all optimization algorithms converge to the identical result based on 20 independent trials.