Analysis and Comparison of Economic Load Dispatch Using Genetic Algorithm and Particle Swarm Optimization

Abstract

Economic Load Dispatch (ELD) problem is one of the most important ones in power system operation and planning. The main objective of the ELD problems is to determine the optimal combination of power outputs of all generating units so as to meet the required demand at minimum cost while satisfying the constraints. Conventionally, the cost function for each unit in ELD problems has been approximately represented by a quadratic function and is solved using mathematical programming techniques. Generally, these mathematical methods require some marginal cost information to find the global optimal solution. Unfortunately, the real-world input output characteristics of generating units are highly nonlinear and non-smooth because of prohibited operating zones, valve point loadings, and multi-fuel effects, etc. Thus, the practical ELD problem is represented as a non-smooth optimization problem with equality and inequality constraints, which directly cannot be solved by the mathematical methods. Over the past decade, in order to solve these non-smooth ELD problems, many salient methods have been developed such as hierarchical numerical method, genetic algorithm, evolutionary programming, neural network approaches, differential evolution, particle swarm optimization, and the hybrid method. In this thesis, the two main types evolutionary optimization techniques namely Genetic Algorithm (GA) and Particle Swarm Optimization (PSO), which are generic population based probabilistic search optimization algorithms and can be applied to real world problem are respectively applied to solve an ELD problem. And at the last the comparison between both the methods has been presented. The PSO provides the generation level such that the generation cost is coming out to be lower than the cost resulted with Genetic Algorithm method.