劣化系统的生产、质量控制与视情维护联合建模与优化

Joint optimization of production,quality control and condition-based maintenance for imperfect system

针对设备状态信息不能实时获取的问题,以质量反馈信息作为评估系统状态的重要依据制定维修策略。在产品质量与设备劣化状态之间存在明显函数关系时,将批量生产结束后系统状态的检测信息和批量生产过程中的质量反馈信息作为互补,进行预防性维护决策,确保维修活动具有快速响应能力。以生产批量大小、质量控制阈值以及预防维护阈值三者为联合决策变量,利用全概率分解方法建立了平均费用率模型。针对模型的高度复杂性,设计基于蒙特卡洛仿真和响应曲面法的优化算法,实现对模型的快速近似求解。最后通过数值实验演示,验证了仿真算法的有效性,并通过敏感性分析明确了相关参数的变化对最优策略的影响。

Aiming at the problem that the equipment condition information cannot be obtained in real time, the quality information feedback was taken as the important basis of evaluation system to formulate the maintenance policy. Both the equipment information obtained at the end of production lot and the quality information feedback were used to assist maintenance decision-making, which could ensure the fast response of maintenance. By using total probability decomposition method, a mathematical model of average cost rate was developed to simultaneously determine the optimal values of lot size and the thresholds of quality control and maintenance. Due to the high complexity of the model, a simulation-based optimization approach coupling Monte Carlo Simulation and Response Surface Methodology was devised to solve the optimization problem. An illustrative example was provided to demonstrate the effectiveness of proposed model. The influence of the parameters on the optimal policy was investigated by sensitivity analysis.