Statistic monitoring and fault diagnosis method of super-speed packaging equipment

The invention discloses a statistic monitoring and fault diagnosis method of super-speed packaging equipment. In the prior art, the source code and the library file that are opened to an bottom PLC program of H1000 super-speed packaging equipment are limited; the bottom PLC original data can not be collected and stored in real time; the existing data report focuses on simple recording of measurement point data of the equipment; and potential fault information that contains an incidence relation change between a process variable dynamic nature and a variable and is hidden into data can not be dug and discovered based on deep analyses of data. However, with the method, the problems can be solved. An off-line monitoring model is established based on a principal component analysis to represent potential correlated characteristics between process variables; on-line calculation of T and SPE monitoring statistical magnitudes and real-time monitoring are carried out; when any statistical magnitude exceeds a limit, a contribution chart is used for determining a main process variable causing a fault. According to the invention, on the basis of modeling and analyzing of report data of the H1000 super-speed packaging equipment, timeliness and reliability of the equipment monitoring and diagnosis result can be improved effectively.