A Birnbaum-importance based genetic local search algorithm for component assignment problems

• Published in: Annals of operations research Vol. 212; no. 1; pp. 185 - 200
• Main Authors: Yao, Qingzhu, Zhu, Xiaoyan, Kuo, Way
• Format: Journal Article
• Language: English
• Published: Boston Springer US 01.01.2014; Springer; Springer Nature B.V
• Subjects: Algorithms; Assignment problem; Business and Management; Combinatorial optimization; Combinatorics; Engineering research; Genetic algorithms; Heuristic; Literature reviews; Mathematical research; Operations research; Operations Research/Decision Theory; Reliability; Reliability (Engineering); Studies; Theory of Computation; Local search; Birnbaum importance; Component assignment problem; Reliability; Genetic algorithm
• ISSN: 0254-5330; 1572-9338
• DOI: 10.1007/s10479-012-1223-1

This paper considers the component assignment problem (CAP) of finding the optimal assignment of n available components to n positions in a system such that the system reliability is maximized. To solve the CAP, an important type of problems in reliability, we propose a Birnbaum-importance based genetic local search (BIGLS) algorithm in which a local search using the Birnbaum importance is embedded into the genetic algorithm. This paper presents comprehensive numerical tests to compare the performance of the BIGLS with a general genetic algorithm and a Birnbaum-importance based two-stage heuristic. The testing results show that the BIGLS is robust (with respect to its random operations) and effective, and outperforms two benchmark methods in terms of solution quality. It demonstrates the effectiveness of embedding the Birnbaum importance in the local search under the genetic evolutionary mechanism.