Evolutionary generator maintenance scheduling in power systems

Abstract

This chapter considers the development of metaheuristic and evolutionary-based solution methodologies to solve the generator maintenance scheduling (GMS) problem of a centralized electrical power system. The effective maintenance scheduling of power system generators is very important to power utilities for the economical and reliable operation of the power system. To demonstrate the application and capabilities of the proposed algorithms a GMS test problem is formulated as an integer programming problem using a reliability based objective function and typical problem constraints. The implementation of a genetic algorithm (GA) and a simulated annealing (SA) heuristic and the effect of varying the GA and SA parameters on the performance of these approaches are presented. The application of a GA/SA hybrid approach is also investigated. This approach uses the probabilistic acceptance criterion of SA within the GA framework. The implementation and performance of the proposed solution techniques are discussed. The application of an inoculated initial population with some heuristically developed solutions are also demonstrated. Results contained in this chapter demonstrate that the GA/SA hybrid technique is more effective than approaches based solely on GA or solely on SA, offering an effective alternative for solving the GMS problems within a realistic timeframe. © Springer-Verlag Berlin Heidelberg 2007.