Time series importance measure-based reliability optimization for cellular manufacturing systems

Abstract

Cellular manufacturing systems (CMSs) can improve the quality and efficiency of the manufacturing process by multiple processing cells with different functions and group technology. CMSs require high reliability to complete processing missions successively, so reliability optimization is an important part to guarantee system performance. This paper proposes a binary decision diagram-based three-step evaluation method to analyze CMS reliability. A reliability optimization model of CMS is constructed by considering the limited cost to determine the optimal combination of machine degradation parameters. Considering the advantages of ant colony optimization (ACO) and time series importance measure (TIM), a TIM-based ant colony optimization (TIACO) is developed to solve the optimization model. To verify the performance of TIACO, system reliability and running time are introduced to compare with genetic algorithm (GA), ACO, and time series importance measure-based genetic algorithm (TIGA). (1) System reliability obtained by TIACO is always the best. (2) Running time of TIACO is smaller. A case study of an unmanned aerial vehicle manufacturing company verifies the effectiveness of TIACO, and machines with higher TIMs should be given priority to improving their degradation parameters, which provides a new idea for reliability evaluation and optimization of CMSs.