多设备系统的生产批量、质量控制与预知维护联合优化

针对由多台设备通过串、并混联构成的多阶段生产系统,对其生产批量、质量控制以及预知维护进行了联合建模与优化.系统中的各台设备随时间逐渐发生劣化,可靠性及产品加工质量随之降低.在每个生产批量结束时,对各设备进行状态检测并评估其在下个生产期内的可靠性.预知维护同时以设备的可靠性、结构重要度以及产能比为决策依据,以合理分配维护资源.在每个阶段设置质检台以实时检测各设备加工产品的质量,并根据质量反馈信息,对次品率超过控制阈值的设备进行预防性更换以改善性能提高产品质量.以批量生产周期、质量控制阈值以及预知维护参数为三维联合决策变量,建立了有限时域内的平均费用率模型.基于蒙特卡罗仿真和响应曲面法设计了优化算法,实现了对模型的快速近似求解.最后通过一个实例演示了本模型,并对结果作了相关统计学分析.

Joint optimization of production quantity,quality control and predictive maintenance for production systems with multiple machines

This paper considers an integrated optimization problem of production quantity, quality control and predictive maintenance for multi-stage production systems. Machines in the system deteriorate with usage and age, which result in deteriorations of reliability and product quality. At the end of production runs, the system is inspected to evaluate the condition of each machine and calculate the reliability during the next production run. The predictive maintenance decision-making is simultaneously based on the reliability, structure importance measure and capacity of machines to take good use of maintenance resources. After products are processed in each stage, quality check is conducted. Based on the quality information feedback, a preventive replacement is performed on the machine if the proportion of defectives produced by it reaches a given threshold. A mathematical model of average cost rate is developed to simultaneously determine the optimal values of production up time, quality control threshold and predictive maintenance parameter in a finite time horizon. By coupling Monte Carlo simulation and response surface methodology, a simulation-based optimization approach is devised to obtain a near optimal joint policy. Finally, an illustrative example is provided to demonstrate the proposed model and some statistical analysis are conducted.