Equipment abnormality diagnosis method and system based on artificial intelligence, and program product

The invention belongs to the technical field of fault diagnosis, and particularly discloses an equipment abnormality diagnosis method and system based on artificial intelligence and a program product, and the method comprises the steps: merging current inspection record data with a corresponding historical normal inspection data set according to an inspection data type number, and obtaining a to-be-analyzed inspection data set; then, array feature set extraction is conducted on a to-be-analyzed inspection data set, the extracted array feature set is input into the KNN model for binary classification prediction, data anomaly classification results are obtained, and when abnormal tags exist in all the classification results, all inspection data type numbers and the data anomaly classification results are used for forming a feature matrix; and finally, inputting the feature matrix into a fault classification diagnosis model for fault diagnosis to obtain a fault classification result of the target equipment. Acco