基于层次分析法的油液诊断特征属性的选择

利用油液监测信息进行故障诊断,由于没有区分监测手段及其特征属性的优势,往往使得通过不同监测手段获得的信息所得到的结论可能不同,甚至矛盾.采用具有这方面独特优势(如整合定性定量信息等)的层次分析法(AHP),通过奇异值分解判断矩阵,成功地为大型冲压自动生产线状态监测解决了油液监测手段及其特征属性的权重优势排序和选择问题,使状态监测和故障诊断更有针对性,从而达到增强诊断力的目的.

Selection of the Characteristic Attributes for Oil Diagnosis Based on Analytic Hierarchy Process

When making fault diagnosis using oil monitoring information,the conclusion drawn from information acquired from different monitoring means can vary even contradict for without distinguishing the dominance between these means and their characteristic attributes.Using analytic hierarchy process(AHP) with special dominance of the conformity of qualitative and quantitative information in that aspect,the weight preference order and choice of oil monitoring means and their characteristic attributes were successfully resolved through application of singular value decomposition(SVD) to judge matrix for condition monitoring of a large press auto-produce line.As a result,the condition monitoring and fault diagnosis of press line becomes more pertinent in order to strengthen the power of diagnosis.