Genetic - Fuzzy system of power units maintenance schedules generation

Article type: Research Article

Authors: Pająk, Michał

Affiliations: Department of Thermal Technology, Radom University of Technology, ul. Krasickiego, Radom, Poland

Note: [] Corresponding author. Michał Pająk, Department of Thermal Technology, Radom University of Technology, ul. Krasickiego 54, 26-600 Radom, Poland. Tel.: +48 48 3617149; E-mail: m.pajak@pr.radom.pl

Abstract: Quality of power units' maintenance schedule is one of the most important factors, which influences reliability and efficiency of power plant exploitation process. The schedule is created during a planning process taking into consideration a lot of different criteria. These criteria result from bidirectional connections between power plant exploitation system and surrounding systems. Therefore they are formulated in completely different domains. The criteria often are defined approximately. Frequently they are contradictory so it is not possible to fulfil all the criteria simultaneously. Because of all these difficulties, the creation process of coherent assessment system, comprising all considered criteria, is very complicated. Additionally, the number of maintenance schedule possible variants is very big. Therefore, to reach high quality of maintenance management, it is necessary to apply automatic system of power units maintenance schedules generation. This paper describes the implementation of artificial intelligence techniques in the area of maintenance management. The automatic system of power units maintenance schedules generation was created using genetic algorithms as optimization technique and developed fuzzy assessment system. In the paper, the industrial implementation of the system is also presented.

Keywords: Fuzzy system models, genetic algorithms, decision analysis, multiple criteria evaluation, large-scale systems

DOI: 10.3233/IFS-141442

Journal: Journal of Intelligent & Fuzzy Systems, vol. 28, no. 4, pp. 1577-1589, 2015