Water plant substation electrical equipment fault diagnosis method based on bi-kmean-LDA-KELM

The invention provides a bi-kmean-LDA-KELM-based water plant transformer substation electrical equipment fault diagnosis method, which comprises the following steps of S1, collecting data when water plant transformer substation electrical equipment is abnormal, and then performing data cleaning, duplicate removal, standardization and normalization; s2, carrying out clustering processing on the normalized abnormal data of the power equipment through a bi-kmeans algorithm; s3, performing dimension reduction operation on the clustered features by adopting linear discriminant analysis; s4, taking the data after the linear discriminant analysis processing as the input of the KELM, and performing model training on the KELM; and S5, inputting abnormal data of the water plant transformer substation to be detected into the trained KELM, and diagnosing the fault of the electrical equipment of the transformer substation to obtain a fault diagnosis result. According to the method, large-scale and complex electrical equip