Research on Fault Early Warning of Mine High-Voltage Explosion-Proof Switch Based on Multivariate State Estimation

In order to solve the problem that most of the fault warnings of high-voltage explosion-proof switches in mines only set thresholds for single variables, a fault early warning model based on Multivariate State Estimation (MSET) was proposed, which used the similarity threshold to determine whether there was a fault and the cumulative error contribution rate to trace the fault point. Firstly, the modeling variables were selected, the data preprocessing was carried out, the historical memory matrix was constructed, and the effectiveness of the MSET model was verified. Then, the sliding window method is used to reduce noise interference, and the similarity function is used as the fault warning criterion, and the cumulative error contribution rate can be traced back to the fault point. Finally, the historical data of 10kV high-voltage explosion-proof switch in a coal mine was used as the test set for verification. The results show that this method can detect the abnormality of the underground explosion-proof switch in time, realize the early warning of the high-voltage explosion-proof switch, and has high accuracy and anti-interference ability.