Relative-transform-based industrial process monitoring method of information increment matrix

The invention belongs to the field of process monitoring, and mainly relates to a relative-transform-based industrial process monitoring method of an information increment matrix. The conventional process monitoring method of the information increment matrix can overcome the defect that a traditional principal component analysis method cannot recognize faults caused by a severe model compound effect, but a dimensional effect is neglected by the method, consequently, variables which play important roles in a system cannot detect the faults with much smaller absolute values due to the fact that the absolute values of the variables are smaller, and the small faults of the important variables frequently play extremely crucial roles for safety and stability of the system. According to the importance degree of the variables in an actual production process, the invention discloses the improved process monitoring method of the information increment matrix on the basis that relative transform is performed on the system; and the method can detect original faults effectively, and can detect faults of a certain variables which are smaller but play important roles in an original system simultaneously.