在过程监控中,使用现代工业系统中的变量进行准确有效的监控诊断仍然是一个具有挑战性的任务.本文以多元指数加权移动平均(MEWMA)策略结合一种有监督分类器(“one plus epsilon”,简称OPE分类器),提出OPE-MEWMA控制图.在考虑不同模型、...在过程监控中,使用现代工业系统中的变量进行准确有效的监控诊断仍然是一个具有挑战性的任务.本文以多元指数加权移动平均(MEWMA)策略结合一种有监督分类器(“one plus epsilon”,简称OPE分类器),提出OPE-MEWMA控制图.在考虑不同模型、偏移模式和偏移大小的情况下,探究了控制图对均值偏移的检测能力,通过比较平均运行长度等多个指标衡量控制图的性能表现.仿真结果表明,所开发的OPE-MEWMA控制图能够快速检测到均值偏移,灵敏度较高.展开更多
This paper studies the robust stochastic stabilization and robust H∞ control for linear time-delay systems with both Markovian jump parameters and unknown norm-bounded parameter uncertainties. This problem can be sol...This paper studies the robust stochastic stabilization and robust H∞ control for linear time-delay systems with both Markovian jump parameters and unknown norm-bounded parameter uncertainties. This problem can be solved on the basis of stochastic Lyapunov approach and linear matrix inequality (LMI) technique. Sufficient conditions for the existence of stochastic stabilization and robust H∞ state feedback controller are presented in terms of a set of solutions of coupled LMIs. Finally, a numerical example is included to demonstrate the practicability of the proposed methods.展开更多
文摘在过程监控中,使用现代工业系统中的变量进行准确有效的监控诊断仍然是一个具有挑战性的任务.本文以多元指数加权移动平均(MEWMA)策略结合一种有监督分类器(“one plus epsilon”,简称OPE分类器),提出OPE-MEWMA控制图.在考虑不同模型、偏移模式和偏移大小的情况下,探究了控制图对均值偏移的检测能力,通过比较平均运行长度等多个指标衡量控制图的性能表现.仿真结果表明,所开发的OPE-MEWMA控制图能够快速检测到均值偏移,灵敏度较高.
文摘This paper studies the robust stochastic stabilization and robust H∞ control for linear time-delay systems with both Markovian jump parameters and unknown norm-bounded parameter uncertainties. This problem can be solved on the basis of stochastic Lyapunov approach and linear matrix inequality (LMI) technique. Sufficient conditions for the existence of stochastic stabilization and robust H∞ state feedback controller are presented in terms of a set of solutions of coupled LMIs. Finally, a numerical example is included to demonstrate the practicability of the proposed methods.
