Reliability analysis and optimization of multi-product manufacturing systems based on two types of maintenance mechanisms

Production efficiency, product quality and cost control are the focus issues that manufacturers are extremely concerned about. Multi-product manufacturing systems can mitigate uncertainty and enhance flexibility in the production process. In order to comprehensively explore the characteristics of multi-product manufacturing systems, this paper selects a two-product manufacturing system with setup time as the research object, and introduces a machine maintenance mechanism combining preventive maintenance(PM) with post-failure repair to analyze reliability and optimization problems of the manufacturing system. The Markov process approach is utilized for system modeling and analysis. Based on matrix equations, an iterative solution is employed to calculate the steady-state probabilities. This method yields production management indexes such as system throughput, the average work-in-process of the system and system mission completion rate. Additionally, it helps define and derive reliability indexes of the machine, including mean available time(MAT) and mean time before first unavailable(MTBFU). Based on numerical experiments, the influence degree and underlying reasons of each parameter on the performance indexes of the system and the machine are analyzed. The bi-objective genetic algorithm is used to determine optimal operating cost solutions for the manufacturing system.