一类线性时变多变量系统的广义最小方差控制和性能评价

针对一类具有可确定"时滞"的线性时变多变量控制系统的性能评价方法进行研究.利用特别定义的关联矩阵和正常相乘的方法,计算出LTV MIMO过程广义最小方差控制律的解析解,并由此获得了广义最小方差基准.为避免估计关联矩阵所带来的困难,利用过程数据和关联矩阵阶数建立了LTV MIMO GMV基准上界.理论和仿真研究表明,上述方法能够正确和有效地评价LTV MIMO系统.

Generalized Minimum Variance Control and Performance Assessment of a Class of Linear Time-variant Multivariable System

Study of performance assessment methods of a class of linear time-varying multiinput-multioutput(LTV MIMO) control systems was focused on,which time delays can be decided. Using a special defined interactor matrix and normal multiplication,an explicit solution of a generalized minimum variance(GMV) control law for LTV MIMO processes was worked out and GMV benchmarking was acquired. In order to avoid the trouble of estimating interactor matrix,an LTV MOMO GMV benchmarking upper bound was built using process data and interactor matrix rank. Research on theory and simulation shows the methods above can assess LTV MIMO processes accurately and effectively.