Attribute Petri Net Modeling In Fault Diagnosis Applications

Qualitative Mapping (QM) is a basic mathematical model of the attribute method, which represents the relationship between the quantity and quality based on philosophy rule of inter-change between quantity and quality. It has been proved to be able to deduce neural network, and is used to solve xor-classfication problem bi-spire problem that are classic difficulty problems in artificial intelligence field. It has been put into practical use in the DSS of nuclear accident and college enrollment system.This paper put forward the Attribute Petri Net in discussing every variety Qualitative Mapping model. APN mathematics structure is defined formally and Qualitative Mapping model can be described vividly by making use of the tool of the APN graphics. In the Attribute Petri Net, the start rule of transition and calculation of place marking are different from traditional Petri Nets. Therefore, we redefined APN transition fired condition, reachability marking set, the incidence matrix and status equation. APN status equation reflected the place marking in different status, which is the description of APN dynamic quality. After analyzing APN the static and the dynamic structure, the quality diagnosis reason algorithm is put forward.Along with the development of science technology and the economic construction, the status monitor and the fault diagnosis of the machine equipment and the manufacturing process have become the key technique and important means that can guarantee stability, reliability and safety of system and raise the product mass and the efficiency. The Attribute Petri Net describes vividly the attribute quantity- quality characteristic conversion relation, so it can be a mold analysis tool of monitor diagnosis system. This paper applied APN to create the fault diagnosis model of the ships host cooling system, and prototype system is implemented by using Visual C++ 6.0.The meaning of the research lies in: putting forward Attribute Petri Net will make the research and the application field of attribute method more abundant.