|
With the development of aviation, aerospace, robotics and other high-tech people to explore the natural areas of expansion, the level of automation of the system is increasing, the increasing size and complexity of the rapid increase. For many large complex systems, using the traditional fault diagnosis technology or the use of a single fault diagnosis mode can not meet the practical requirements. In recent years, with the continuous progress of artificial intelligence techniques, diagnostic techniques have begun to enter a new phase, that intelligent diagnosis stage. The expert system is a branch of artificial intelligence research and applications, a very wide range of applications. In this paper, based on expert systems and case-based reasoning model two methods of graph theory, combined leakage fault diagnosis of a launch vehicle models. For graph theory model, the representation of knowledge representation based on the knowledge of the structure and behavior of the system model, and based on information entropy \Object-oriented methods for case-based reasoning, knowledge representation, knowledge acquisition is a diagnostic procedure automatically obtain mainly by knowledge engineers get supplemented based relational database access methods. This paper, a process-based, class election, rougher, Featured, preferably the \query method used in the search of knowledge, the reasoning process more simple and fast. In addition, on the basis of the diagnostic expert system and case-based reasoning model of two diagnostic methods of graph theory, put forward the idea of ??fusion diagnosis. In the diagnostic process, for the failure phenomenon, the first man-machine interaction module with several steps after diagnosis method based on graph theory model, a certain knowledge of new cases, and then turn the case-based reasoning for diagnosis, if from the case knowledge library search for the same or similar to the case of the failure phenomenon, Click here to the case of diagnostic results can locate the point of failure, the diagnosis is successful, do not have to enter the next step, otherwise continue with the diagnosis method based on graph theory model of human-computer interaction by accurate diagnosis and rapid diagnostic unified. This method is applied to a certain model rocket launch system leakage detection system to improve the efficiency of the electric leakage detection, and to achieve the purpose of quickly locating a fault point. Article also combined with software engineering, use of the VC 6.0 programming tools and the Access database, developed a model of carrier rockets leakage fault diagnosis system, given a diagnosis instance, the software to run properly, accurate fault location, efficient, good test results.
|