An Intelligent Fault Diagnosis Method for Mining Transformers Based on ICOA-LightGBM

The complex and harsh environment of underground coal mines makes coal mining equipment failures frequent, which poses a serious threat to coal mining safety. In order to enhance the safety as well as stability of mining transformer operation, a mining transformer fault diagnosis model based on the improved crayfish optimisation LightGBM algorithm is proposed. The article uses three improvement methods to improve the traditional crayfish algorithm; constructs a transformer fault diagnosis method based on LightGBM, and introduces the ICOA algorithm to optimise the parameters of the LightGBM diagnostic model to form the optimal fault diagnostic model; explores a set of the most suitable hyper-parameter values to further improve the fault identification accuracy of mining transformers, and verifies the diagnosis to prove the effectiveness of the transformer fault diagnostic model constructed by the proposed method. The diagnosis is verified to prove the effectiveness of the fault diagnosis model of mining transformer constructed by the proposed method.