参数未知自相关过程的自适应边界调整策略

Adaptive Deadband Adjustment Policy for Autocorrelation Process with Unknown Parameters

针对参数未知的自相关过程的质量控制问题,研究了每次调整成本固定下的自适应调整策略。建立了过程的状态空间方程模型以及未知参数的贝叶斯模糊先验分布,利用序贯蒙特卡洛法估计未知参数值进而得到了过程每个阶段的调整边界线,从而确定使过程总体损失最小的边界调整策略;给出了基于序贯蒙特卡洛法的边界调整策略的算法步骤,并通过算例验证了调整策略的有效性。结果表明,该调整策略具有较好的调整效果。

Aiming at the quality control problem of autocorrelation process with unknown parameters,the adaptive adjustment policy for the situation of fixed adjustment cost is studied.The state space process control model and Bayesian vague prior distribution of unknown parameters are built.To use sequential Monte Carlo（SMC） method estimates the values of unknown parameters,then the deadband of each stage is achieved,furthermore the deadband adjustment policy for minimum the total process loss is obtained.The algorithm to derive the deadband adjustment policy based on SMC is introduced.An example is given to verify the effectiveness of the approach.The results show that the adjustment policy can achieve good adjustment performance.