Study On Abnormal Event Intelligent Diagnosis Of Chemical Process

Abnormal event is referred to as a departure from an acceptable range of an observed variable or a calculated parameter associated with a process. The systematic failures may cause abnormal events, while mal-operation and larger process disturbances may also lead to them. Abnormal events of varying magnitudes always result in incipient faults, so it is important to establish a monitor system for dignosing and prognosticating the chemical process to have an effective and stable operation.This paper is organized as follows: firstly, According to the characteristic of the abnormal event in chemical process, a new type of timed Fuzzy Petri net model which is suitable for describe the abnormal event is proposed, and the definition, firing rules, knowledge represention, analytical method, simulatical method are also given. Secondly, abnormal event propagation algorithm and abnormal event diagnosis algorithm based on timed Fuzzy Petri net are detailed explored, which are examined correctly and compared with each other. Thirdly, based on the object of production process on polypropylene, an intelligent diagnosis method about abnormal event is presented which includes two parts: prognostication and diagnosis, it can find the case and forecast the result of abnormal event based on the trend of the initial abnormal event. The method is also simulated on artificial equipment. Finally, the means is given to construct an intelligent diagnosis expert system of the abnormal event in chemical process based on what have been studied.