An Intelligent Approach Based on Metaheuristic for Generator Maintenance Scheduling

TLDR: This chapter comprehensively applies GA, which is a popular evolutionary-based optimization method, to solve a power system planning problem, that is, generator maintenance scheduling of generators, with the basics of evolutionary process such as encoding, function evaluation, parent selection, genetic operation, and replacement.

Abstract: To guarantee an efficient and reliable operation of power system components, optimization techniques are used at every stage of planning and operations. For instance, they are used for planning power system expansion, generator scheduling, regulating control devices, evaluating security margins, and for several other critical tasks. However, most of the traditional optimization algorithms applied have limitations. If there is limited knowledge about the nature of the objective function, it is worthwhile to use the Metaheuristic technique (MHT), for getting better solutions. Some of the popular MHTs are tabu search, simulated annealing, harmony search, genetic algorithms (GA), evolutionary programming, ant colony optimization, particle swarm optimization, differential evolution, etc. They imitate natural evolutionary principles or group behavior of animals to carry out the search and optimization efficiently. These methods, in fact, choose their path through the parameter space randomly. They can get along with the function value of the objective, so that one does not have to bother about the continuity of the objective function or its gradient during the iteration process. This fact enables to find the region of the global optimum with the high probability. The aim of this volume is to offer a sample work on optimization of power system problem using evolutionary algorithm (EA). This chapter comprehensively applies GA, which is a popular evolutionary-based optimization method, to solve a power system planning problem, that is, generator maintenance scheduling (GMS). It discusses the step-by-step procedure for GA-based maintenance scheduling of generators, with the basics of evolutionary process such as encoding, function evaluation, parent selection, genetic operation, and replacement. It also provides the performance of GA in terms of savings in computation time and improvement in solution quality with respect to the classical method. This work will be useful for research scholars facing any optimization problem related to planning and operation of electric power systems.