Industrial equipment abnormal state detection method based on sensor monitoring technology

The invention relates to the technical field of industrial state data processing, in particular to an industrial equipment abnormal state detection method based on a sensor monitoring technology. According to the method, the data sensitivity degree is obtained through the data change of each dimension, and then the reference dimension is screened out; acquiring local direction consistency according to the mean shift vector of each data in a preset time range; obtaining a dimension angle according to data differences between the reference dimension and other dimensions; obtaining mean value mapping vectors of other dimensions by combining the mean value drift vectors of other dimensions; obtaining a correction vector by using the vector difference between other dimensions and the reference dimension, the data sensitivity degree and the local direction consistency, and obtaining a correction mean shift vector of each datum in the reference dimension; clustering is carried out, and anomaly recognition is carried