摘要
大量的计算以及对一般关联矩阵先验知识的需求导致用户自定义指标在实际工业应用中存在诸多不便,鉴于此,为了减少对过程模型信息的需求,避免估计一般关联矩阵,研究了如何简化用户自定义指标的计算,并提出了两种改进的用户自定义指标的计算方法.该方法仅需过程常规数据和对过程输入/输出时延或关联矩阵阶次的估计便可完成指标的计算.最后通过仿真验证了两种方法的优越性和有效性.
The complicacy of computation procedures and the requirement of a general interactor pose a challenging issue for industrial applications of user-specified benchmark based performance monitoring techniques.Therefore,to reduce the model requirement of a general interactor,two improved computational methods are presented in this paper,which only require routine operating data and the estimate of input/output delays or the order of interactor matrix.Simulation examples show the superiority and effectiveness of the two proposed methods.
出处
《控制与决策》
EI
CSCD
北大核心
2011年第7期1117-1120,共4页
Control and Decision
关键词
控制性能评估
多变量控制系统
对角关联矩阵
用户自定义基准
control performance assessment
multivariable control systems
diagonal interactors
user-specified benchmark
