Industrial rotating equipment fault diagnosis method based on correlation detection

The invention discloses an industrial rotating equipment fault diagnosis method based on correlation detection, and the method comprises the following steps: (1) collecting and processing fault sample data of industrial rotating equipment, and dividing the data into source domain data and target domain data; (2) establishing and dividing fault data of a source domain and a target domain; (3) building and pre-training a source domain data model; (4) re-integrating the target domain data set; (5) carrying out source domain data model migration and weight fine tuning; (6) evaluating the performance of the target domain data model; and (7) testing the fault diagnosis effect of the target domain data model in the step (6) by using the data of the target domain data set, and evaluating the operation state of the industrial rotating equipment. According to the method, the target domain data and the weight thereof are adjusted by using correlation detection, so that the deep learning model is wider in application fie