多变量系统的结构辨识及其在电加热炉上的应用

The Structure Identification for Multivariable System with its Application in Electrical Heating Furnace

  • 摘要: 本文在Guidorzi结构辨识方法的基础上,通过引入向量系的标准化、相关性指标和相对相关性指标等概念,对有噪声情况下多变量系统结构辨识提出一种判别准则.采用此准则对电加热炉系统的模型结构进行了辨识,并进行了相应的参数估计.计算结果表明:此判别准则较Guidorzi和文献2所给出的方法,易于确定有噪声情况下模型的结构参数,且基本不受数据数量级及长度N的影响.

     

    Abstract: On the basis of the method of structure identification by Guidorzi,a criterionfor the identification of multivariable system with noise disturbance has beenproposed through the normalization of the vector system and the introduction of theconcept of the correlated and the relative correlated indices.The criterion hasbeen used in the identification of the heating furnace,and the corresponding parameters estimated.The result of the computation shows that it is easier to estimatethe model's parameters with this criterion than with those proposed by Guidorzi and (2).Moreover,it is not affected by the data magnitude and the data length N.

     

/

返回文章
返回