Transmission Line Fault-Cause Identification Based on Hierarchical Multiview Feature Selection

Fault-cause identification plays a significant role in transmission line maintenance and fault disposal. With the increasing types of monitoring data, i.e., micrometeorology geographic information, multiview learning can be used to realize information fusion for better fault-cause identification. To reduce redundant different this paper, hierarchical feature selection (HMVFS) method is proposed address challenge combining waveform contextual features. enhance discriminant ability model, an ε-dragging technique introduced enlarge boundary between classes. effectively select useful subset, two regularization terms, namely l2,1-norm Frobenius norm penalty, are adopted conduct data. Subsequently, iterative optimization algorithm developed solve our method, its convergence theoretically proven. Waveform features extracted from yield data evaluate HMVFS. The experimental results demonstrate effectiveness combined reveal superior performance application potential