Optimization of Technical System by Using FTA and Genetic Algorithm

Abstract

The main aim of the paper lies in providing a proposal of technical system optimization in terms of maximizing its reliability when a certain sum of maintenance costs is given, or minimizing maintenance costs to achieve the defined reliability. The technical system is analyzed using the method called fault tree analysis (FTA) and the final diagram is represented by a tree structure. At the top of the diagram, there is an undesired event which represents the system failure. This top event is analyzed and branched into several other events in order to obtain all possible faults of system elements and their combinations which lead to the system failure. Each primary event contains basic information as minimal and maximal reliability and the corresponding maintenance costs. As a result, it is possible to calculate the total probability of system failure, respectively to its reliability, and the total costs needed to ensure the required system reliability. Genetic algorithms are used and applied in the diagram created by FTA. The paper contains the whole proposal of using a genetic algorithm in order to optimize costs and system reliability on the basis of defined conditions. The described proposal is implemented within a functional application and then tested using real data. The results form an integral part of this paper.