Comparative analysis of PSO and quasi-newton methodsfor economic dispatch of power systems

Abstract

The economic dispatch (ED) is one ofthe important parts of power systems. ED includes the objective function(s) and constraints. Many optimization algorithms including traditional optimization methods and intelligent optimization algorithms have been applied to ED problems. However these two kinds of optimization algorithms are based on different philosophies and they are different from the simplicity ofmethods/algorithms and performance. It is necessary to investigate the efficiency and performance of these two kinds of optimization methods and find the proper situations methods/algorithms for different. In this paper, the Quasi-Newton Method (QNM) is chosen as the representation of the traditional optimization methods and Particle Swarm Optimization (PSO) is taken as the representation of the intelligent optimization algorithms, both QNM and PSO are applied to the ED problems with different numbers of generators, load power demands and constrains. Finally the simulation results show the performance of these two optimization algorithms and the guidance of choosing QNM or PSO is provided for ED problems.