Feature Recognition Method of Pipeline Leakage Sources Based on Factor Analysis and ReliefF
In order to solve the issue of high feature redundancy affecting identification accuracy in pipeline leakage source identification,a feature dimension reduction method combining ReliefF and factor analysis is proposed.Firstly,acoustic emission signals from pipeline leakage sources are collected,and 27 frequency domain features and time domain features are extracted from the signals.Then,the ReliefF method is used to screen the features,so that the features with higher relevance are obtained and used as sensitive features.And factor analysis is used to extract common factors from the sensitive features to form a feature set.Finally,the feature set after the dimension reduction is input into the support vector machine for identification,and the recognition results of the shape and size of the pipeline leak source are obtained.Experimental results show that this method can accurately identify shape characteristics and size of pipeline leakage sources,and effectively reduce feature redundancy and computation time.
vibration and wavepipeline leakagefeature dimension reductionReliefFfactor analysis
万方数据
维普
