YANG Z S, LIU F.A framework of partial discharge detection for gas-insulated switchgear based on LSTM and multi-head attention[J].Journal of Henan Polytechnic University(Natural Science) ,2025,44(2):108-115.

doi:10.16186/j.cnki.1673-9787.2023020042

Received:2023/02/12

Revised:2023/05/06

Published:2025-03-05

A framework of partial discharge detection for gas-insulated switchgear based on LSTM and multi-head attention

YANG Zhengsheng1, LIU Fang2

1.State Grid Jilin Electric Power Co.， Ltd.， Changchun  130000， Jilin， China；2.School of Electric Engineering and Automation， Hefei University of Technology， Hefei  230009， Anhui， China

Abstract: Objectives To achieve long-term stable operation of electrical equipment， a framework for anomaly detection and classification of partial discharge （PD） in gas insulated switchgear （GIS） based on long short-term memory （LSTM） networks and a multi-head attention mechanism was proposed. Methods Firstly， phase resolved partial discharge （PRPD） analysis was used as the input sequence， and LSTM was employed to learn the temporal correlations in PRPD signals. Then， the results from LSTM were fed into a multi-head attention module. By integrating multi-head self-attention with the LSTM network， the framework focused on different representational subspaces corresponding to different phase sets of PRPD. Secondly， the self-attention mechanism identified important information between input and output sequences， while the multi-head self-attention network captured high-order features of faulty PRPD. Finally， a classification layer was utilized for fault detection in GIS.  Results The experimental results were as follows： Linear SVM performed the worst，indicating that traditional machine learning classification algorithms were not effective in capturing subtle differences in the data. CNN+LSTM achieved temporal dependency capture of multivariate time-series data， significantly improving performance over the SVM method. AL+DCNN enhanced the ability to handle dataset imbalances and improved the generality of extracted features through an adversarial learning framework. The proposed method improved the F₁ score by 2.96% compared to the AL+DCNN method， demonstrating that combining LSTM with a multi-head attention mechanism could effectively enhance the performance of PD fault identification in GIS. It achieved the best performance in terms of accuracy， recall， and F₁ score. This was because the proposed method achieved effective performance complementarity by combining LSTM with the attention network， outperforming other advanced methods in switchgear anomaly detection. Conclusions The proposed method could effectively detect PD faults， contributing to ensuring the long-term stable operation of electrical equipment. 

Key words:long short term memory;gas insulated switchgear;phased-resolved partial discharge;anomaly detection;multi-head attention