Reliability-based preventive maintenance scheduling in power generation systems: A lévy flight and chaotic local search-based discrete mayfly algorithm

Abstract

An adequate and reliable precautionary upkeep plan in power generation systems is required to reduce failures, to improve the generator's lifespan, to diminish repair costs and to ensure consistent power supply to consumers with managing the energy flows in power systems. The Generators’ Precautionary Upkeep Planning (GPUP) problem is a complex optimization problem. It is a critical challenge in the power generation industry, involving the optimization of maintenance schedules for power generators to minimize the generation reserve and maximize the reliability. This problem consists of several important restrictions which include the load power demand and the labour force restrictions. In this research paper, a Discrete Chaotic Mayfly Optimization (DCMFO) algorithm which uses Lévy flight random walk for female mayflies and chaotic local search move rule for male ones, is adapted for designing an appropriate precautionary upkeep scheme of a list of generators in power generation systems. The DCMFO algorithm is evaluated using 21-unit test thermal power system. The results indicate that unlike the classical DMFO algorithm, the DCMFO algorithm has proven to have superior optimization capabilities and to surpass all earlier adopted algorithms in performance. This reinforces DCMFO's standing as the current leading optimization algorithm for solving this particular problem, ever since its initial inception. The DCMFO's efficiency and reliability have been demonstrated with different cases through several statistical tests.