Process Fault Diagnosis Based On Graph Theory

In modern process industries, with the rapid development of mass production and complexity, reliability and security become emergent needs to avoid large economical loss brought by accidents and even breakdowns of industrial productions. Therefore, it is stringent for industrial intergrated automation to research and develop process supervisory systems which intergrate the functions of control, supervision and diagnosis. Fault diagnosis (FD) technology supervise the status of production process, detect the fluctuation and faults, and locate the fault sources, then isolate and eliminate them. In this way, calamities are prevented and fluctuations of quality of products are reduced.In this thesis, important aspects of process fault diagnosis (FD) based on graph theory which is an important branch of FD technology are presented and studied systematically. The main work could be stated as following:1. The main contents of FD, the classification of various diagnostic methods and prospects are summarized, especially the history and status in quo of diagnostic method based on graph theory.2. The concept and diagnostic procedure of FD based on graph theory are presented systematically and detailedly. Furthermore, a method based on fault propagation digraph, including its fault modeling, level structuring and fault location, is studiedintegrality.3. Aiming at developing practical diagnostic systems, how to build the simulative models characterized by propagation is studied. The designs and implements of fault propagation digraph classes using OOP is presented, followed by a simple improved diagnostic algorithm.4. By pointing out the shortcoming of propagation digraph based method, a diagnostic method based on Signed Directed Graph (SDG) is studied. The idea of combining SDG with expert systems and how to produce the diagnostic rules from SDG are presented by examples.5. A practical software tool for building fault propagation simulative models is developed.Finally, various diagnostic methods are compared, pointing out some aspects deserving to focus and further studies.