Optimization of system reliability using chaos-embedded self-organizing hierarchical particle swarm optimization
This paper addresses a reliability optimization problem, where the motive is to select the best components for series and series-parallel systems such that system reliability becomes maximized while simultaneously minimizing the cost, weight, and volume. Previous formulation of the problem has implicit restrictions, i.e. it either maximizes system reliability or minimizes the cost. Thus, in order to give a realistic view to the model, a comprehensive objective function has been formulated by combining the normalized values of reliability, cost, weight, and volume. In this paper, a chaos-embedded hierarchical particle swarm optimization (CE-HPSO) algorithm has been proposed to solve the problems arising in the optimization of system reliability using redundancy. The salient features of the proposed algorithm are the use of chaotic sequences and time-varying acceleration coefficients which are responsible for diversifying the search space. Moreover, to restrict the premature convergence, a hierarchical particle swarm optimizer has been used in the proposed algorithm. The performance of the CE-HPSO algorithm has been tested on three benchmark problems and the comparisons are made with genetic algorithm results. In order to check the scalability of the proposed solution methodology, small and large problems are also considered. The results demonstrate the benefits of the proposed algorithm for solving this type of problem.