Hierarchical method of fault diagnosis based on extended SDG

Compared with the traditional signed directed graph (SDG) model, the extended SDG model can enlarge the information of edges and nodes, reduce false solution and improve resolution. Therefore, it possesses good information completeness. However, the multi-node information and complex relationships of edges in the model result in a long search time and low efficiency. To solve these problems, a hierarchical method of fault diagnosis based on extended SDG is proposed. According to the qualitative relationship of fault propagation path between nodes, the fault source nodes can be found by backtracking alarm nodes and searching compatible branches. In this paper, the chlorine gas pump of the chemical process is used as an example to carry out modeling and reasoning in fault diagnosis. The results show that this diagnosis algorithm is not only feasible in improving the search efficiency of fault sources, but is also accurate and valuable.