控制性能评价中单位关联矩阵的求解

Solution of Unitary Interactor Matrix in Controller Performance Assessment

多回路控制性能评价中单位关联矩阵的求解很重要,会影响到时延项的确定与控制性能指标的计算,因此探讨关联矩阵的计算问题是个有意义的话题。关联矩阵求解一般需要知道一些过程模型的信息,这个条件对于实际应用比较苛刻。为避开先验知识的约束,提出了基于闭环数据的单位关联矩阵求解方法。通过对辨识所得各类传函模型的预处理,能从多项式矩阵中直接求解出对角单位关联矩阵或迭代求解出非对角单位关联矩阵。应用结果证明了该方法的有效性。

The calculation of the unitary interactor matrix is important during multivariable control loop performance assessment,which has an impact on determining the time delay term and computing control performance index.Therefore the calculation of the interactor matrix is meaningful.Some information about process model is needed when solving interactor matrix,but this is harsh condition for the practical applications.To avoid the dependence of priori knowledge,a method of obtaining a unitary interactor matrix based on closed-loop data is presented.The polynomial matrix form is derived from the pretreatment of the identified transfer function model,afterwards the diagonal unitary interactor matrix or non-diagonal unitary interactor matrix can be obtained directly or iteratively,respectively.The application examples are illustrated to prove the effectiveness of this method.