Method and system for monitoring process based on slow feature analysis

Abstract

The invention relates to a method and system for monitoring a process based on slow feature analysis. The method comprises the steps of reading a slow feature model and a monitoring model, reading in an online measurement value of a process variable, forming an input vector, and inputting the input vector into the slow feature model to obtain an online real-time estimation value of a slow feature vector s(t); inputting a vector s1:M (t) formed by front M slow features into the monitoring model to obtain instant evaluation values of T<2> and S<2>; comparing T<2> with Ta<2> and S<2> with Sa<2>, judging that the process is shifted from a steady state operation point if T<2> is greater than or equal to Ta<2>, and judging that the process is subjected to dynamic feature variation if S<2> is greater than or equal to Sa<2>. According to the technical scheme, the shift of the steady state operation point can be monitored, and the dynamic feature variation of the process can be effectively monitored.

Description

Technical field [0001] The present invention relates to the technical field of process processing, in particular, to a process monitoring method based on slow feature analysis and a process monitoring system based on slow feature analysis. Background technique [0002] In the industrial process, faults, disturbances and other abnormalities seriously affect the normal operation of the process. In order to ensure the safe, stable and efficient operation of the production process, real-time monitoring of the overall operation of the process and timely detection of these abnormal conditions are extremely important. In recent years, the multivariate statistical process monitoring (MSPM) method has received full attention from scholars and has become a hot topic in the process control field, and has been widely used in actual industrial production. Because the chemical process mechanism is complex and difficult to perform theoretical analysis, it is more difficult to establish a monit...

AI Technical Summary

Problems solved by technology

However, in this case, traditional process monitoring methods such as Principal Component Analysis (PCA), Independent Component Analysis (ICA), etc., still give false alarm information by monitoring the shift of the steady-state operating point, and cannot Distinguish between normal working condition changes and disturbances and faults that really affect the process characteristics, causing the operators to fall into a large number of invalid false alarm information, which in turn causes the overall failure of the process monitoring module and increases the burden on the operators, affecting the operators to truly solve the problem. Attention to real safety hazards in the production process

[0004] When the steady-state operating point of the process is normally switched, the dynamic characteristics of the process have not changed substantially because the feedback control loop can still work normally; while an important feature of a real process failure is that the control loop cannot compensate for disturbances, The impact of faults, the dynamic characteristics of the process will be significantly different from the dynamic characteristics under normal working conditions, which will affect the smooth operation and stable control of production

Images