Rolling bearing intelligent diagnosis method based on few-label multi-sensor signal fusion

The invention discloses a rolling bearing intelligent diagnosis method based on few-label multi-sensor signal fusion, and relates to the technical field of rolling bearing fault diagnosis, and the method comprises the following steps: S1, based on a distributed sensor, collecting a multi-sensor signal of a rolling bearing in a to-be-monitored device; s2, extracting effective features of the multi-sensor signals, and constructing a multi-source information fusion tensor based on the effective features; s3, constructing a semi-supervised low-rank tensor classifier based on a tensor nuclear norm and a high-order manifold regular term; s4, constructing a training set mixed with a labeled data set and an unlabeled data set, and inputting the training set into a semi-supervised low-rank tensor classifier for model training; and S5, inputting the label-free verification set into the intelligent diagnosis model, and verifying the performance of the model. According to the invention, by constructing the semi-supervise