摘要 |
<p>Disclosed is a method for identifying a transformer local discharge mode based on a singular value decomposition (SVD) algorithm, comprising a training model and a classification identification process, and comprising: first, building an experiment environment having artificial defects and collecting data samples, and calculating a statistical feature parameter of each sample to form a data sample matrix; performing singular value decomposition on the sample matrix, and determining the order of the optimal retention matrix by determining whether the feature of a retention matrix is clear, so as to obtain a type feature description matrix and a centroid-based description vector group after dimensionality reduction; and performing preprocessing on samples to identify to obtain a sample vector, and performing linear transformation on the sample vector by using a type feature space description matrix, so as to obtain a sample description space vector after dimensionality reduction, and then calculating the degrees of similarity between the vector and each vector in the type vector group, so as to obtain a classification determination result. The algorithm is simple and efficient, so as to implement reliable distinguishing between an interference signal and a discharge signal in local discharge detection, and increase the accuracy of local discharge mode diagnosis.</p> |
申请人 |
STATE GRID CORPORATION OF CHINA;STATE GRID HUBEI ELECTRIC POWER RESEARCH INSTITUTE;HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY |
发明人 |
XIE, QIJIA;LI, CHENGHUA;RUAN, LING;LI, JINBIN;SU, LEI;CHEN, TING;ZHANG, XINFANG |