Performance-driven industrial process anomaly monitoring method

• Main Authors: GAO MING, JI HONGQUAN, GUO ENTAO, ZHU JIFENG, WU DEHAO, CHEN MAOYIN, YAN FEI, ZHENG SHUIMING, ZHOU DONGHUA
• Format: Patent
• Language: Chinese; English
• Published: 31.03.2020
• Subjects: CALCULATING; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; HANDLING RECORD CARRIERS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention discloses a performance-driven industrial process anomaly monitoring method, and particularly relates to the technical field of industrial process monitoring. The method includes: on thebasis of fault detectability analysis, defining a detection performance index, and carrying out component selection according to the index; and then describing a component selection problem as a random optimization problem, and giving an optimal solution of the problem in statistical significance. The method comprises two parts of off-line training and on-line monitoring. In an offline training stage, the detection statistics of each sample are calculated by utilizing historical data under a normal working condition, and a monitoring control limit is determined under a given significance level; in the online monitoring stage, detection statistics are calculated according to a new sample, and if the detection statistics exceed a control limit, it is considered that the process is abnormal.Compared with a principal