化工过程数据校正中的数据分类问题

Observability and Redundancy Classification in Process Data Rectification

讨论了化工过程数据校正中数据分类的作用和现有的算法 ,证明了用投影矩阵法对含有未测回路的流程网络进行数据分类的局限性。提出了一种新的利用关联矩阵中各物料流列向量间的线性相关关系的数据分类算法 :线性相关法。

The existing algorithms of observability and redundancy classification in process data rectification are discussed. It shows the limit of the method of matrix projection for data classification of the flow network with unmeasured loop.A new correlative algorithm is proposed for classification by identifying correlation of column vectors of incidence matrix.It gives some theoretical insights into the design for a sensor network.