IMPROVED PARTICLE SWARM OPTIMIZATION WITH VARYING PARAMETER SETTINGS FOR ECONOMIC LOAD DISPATCH

Abstract

In this thesis, four evolutionary optimization models (IPSO 1, 2, 3, and 4) based on the particle swarm optimization algorithms for Economic Load Dispatch considering cost of generation. Comparative analysis suggests that IPSO (Improved Particle Swarm Optimization) significantly improves the performance with less no of iteration. In the last version of IPSO, we have moved acceleration coefficient for personal factor Cp and global factor Cg in opposite direction (i.e. Cp maximum to minimum and Cg minimum to maximum), while keeping other parameter with some constant value, which shows that there is tremendous reduction in no of iteration. All different IPSO has been implemented to ECONOMIC LOAD DISPATCH to get optimum value of cost with less no of iteration. A MATLAB program has been developed for Evolutionary Programming and Evolutionary Computation such as Particle Swarm Optimization (PSO) to solve economic load dispatch problem considering cost of generation.