论文部分内容阅读
针对动态稀疏保局投影考虑数据的局部相似度信息,忽视局部差异度信息的问题,在综合考虑数据相似度信息和差异度信息的基础上,融合数据类标信息,提出了一种新的基于动态鉴别结构保持投影的故障诊断方法。该方法首先构造原始数据的扩展矩阵,利用稀疏表示获取数据的全局稀疏重构关系,并融入到相似度信息中;同时,考虑数据差异度信息和鉴别信息,构建的结构保持投影目标函数进行数据降维,最后利用鉴别函数值进行故障诊断。田纳西-伊斯曼过程的仿真结果表明,与保局鉴别分析方法相比,所提方法具有更好的诊断精度和更强的稳定性。
Considering the local similarity information of dynamic sparse projection and projection and neglecting the information of local difference degree, this paper proposed a new method based on the information of similarity and difference of data, Fault Diagnosis Method of Dynamic Projection Structure for Projection. In this method, the sparse representation of the original data is constructed firstly, and then the sparsity representation is used to obtain the global sparse refactoring relation of the data, which is integrated into the similarity information. At the same time, considering the difference of data and the identification information, the constructed structure holds the projection objective function for data Dimension reduction, and finally use the value of the discriminant function for fault diagnosis. The simulation results of the Tennessee-Eastman process show that the proposed method has better diagnostic accuracy and stronger stability than the discriminant analysis method.