Method for establishing abnormal event early warning system based on expert experience and historical data

Abstract

The invention discloses a method for establishing an abnormal event early warning system based on expert experience and historical data. The method includes three steps; a model is established offline in the first two steps; and the model is applied in real time and offline in the third step. According to the method, alarm data are continuously acquired, and the expert experience is used in combination, and therefore, an abnormal event early warning system is updated regularly; the system is implemented online, and the state change of variables are continuously monitored; the probability of abnormal events which may occur is predicted in real time, and root causes causing deviation are analyzed continuously; and an operator can be timely informed of the change of the system, and therefore, the operator can be assisted to take appropriate measures in time. With the Bayesian network-based early warning system management method adopted, the randomness and uncertainty of a real process can be embodied, and complexity of consideration on all variables in process knowledge modeling can be omitted, and correlativity between monitoring variables can be reflected, and discovery and judgment of fault sources can be benefitted for the operator.

Description

Technical field [0001] The present invention relates to the field of chemical engineering, in particular to a method for establishing an abnormal event early warning system based on expert experience and historical data. Background technique [0002] In the early industrial enterprises, the control system was relatively simple, and each alarm signal was directly connected to the central control room through a hard wire, but this method was costly and complicated to maintain, and could not achieve good results. With the widespread use of programmable logic controllers (PLC), field bus control systems (FCS) and distributed control systems (DCS), industrial production has become increasingly complex and large-scale. At the same time, the losses caused by production accidents have also attracted more and more attention, because they will not only affect the reputation and benefits of enterprises, but also greatly affect the overall development of related industries. However, countri...

AI Technical Summary

Problems solved by technology

[0005] However, with the expansion of system scale and the simplicity of signal acquisition, operators need to face a large amount of system information

Images