论文部分内容阅读
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。
It is very important to solve the unit correlation matrix in multi-loop control performance evaluation, which will affect the determination of delay term and the calculation of control performance index. Therefore, it is a meaningful topic to study the calculation of association matrix. Relevant matrix solution generally need to know some of the process model information, this condition is more demanding for practical applications. In order to avoid the constraint of prior knowledge, a method of solving unit correlation matrix based on closed-loop data was proposed. By preprocessing all kinds of transfer models, the diagonal unit correlation matrix can be solved directly from the polynomial matrix or the non-diagonal unit correlation matrix can be iteratively solved. The application results prove the effectiveness of this method.