Analysis of Optimal Power Dispatch Using Particle Swarm Optimisation

Abstract

The objective of optimal power dispatch is to schedule the power generation of all generation units in the power system to reach the system constraints and load demand while keeping the total cost of power generation and transmission line loss at a minimum. Optimal power dispatch is multimodal optimization and a non-linear problem with more than one local minima. Convergence to a global minimum from all the local minima is necessary to get the most optimum result. Particle swarm optimization algorithm (PSO) is great in handling non-convex and practical power dispatch problems. This paper proposes to solve optimal power dispatch using the particle swarm optimization (PSO) algorithm. Economic power dispatch, emission power dispatch, and a combination of economic and emission power dispatch are the three types of strategies for optimal power dispatch. In this study, combined dispatch with transmission loss was used as the strategy in optimal power dispatch. This type of power dispatch is a multiple-objective power dispatch as the strategy combines economic and emission power dispatch. Power dispatch can be converted to a single objective with a max-max price penalty factor. The suggested method is tested on a test bus system to validate the effectiveness of optimal power dispatch using the PSO algorithm. The combined power dispatch with and without transmission loss is proven to be effective in minimizing total operating cost when fuel cost and pollutant emission cost are considered in power system. The combined power dispatch results are compared with other conventional methods and metaheuristic methods on the test bus system. The results show that the PSO algorithm is more effective than conventional method in converging the optimal solution, more effective in total fuel cost optimisation and total operating cost but less effective in total pollutant emission optimisation compared to other metaheuristic methods. PSO algorithms have a high convergence rate towards the optimal solution of optimal power dispatch.