| With our understanding in knowledge engineering and the use of object-oriented design method, we construct a fault diagnose expert system for hydraulic system based on the Expert System Platform (ESP).Orderly we discuss the feasibility of developing a fault diagnose ES for a fork-lift truck, objective to achieve, knowledge acquisition, knowledge base construction and implementation.During the process of knowledge base design, we represent our acquired knowledge in a hierarchical manner and classify them into three subdivisions: description knowledge, processing knowledge and control knowledge. This kind of classification helps improve our comprehension with the whole knowledge base and key component. It also enhances the reusibility of the knowledge base.Based on the knowledge classification mentioned above, we use the method of modularization to construct our knowledge base. Following our design principle, we view the typical component and circuit as an independent module and implement it separately. When we construct a larger system, the well-designed module can be added into it. With the use of this method, we can improve our work efficiency greatly. During the process of production rule extraction, we use a special method to implement it. We use trouble diagram to represent the relation between all kinds of events, which include top events, intermediate events and bottom events. In this way, we can recognize the cause and effect between all the relevant events. The introduction of the trouble diagram makes the production rule extraction much more easy and precise.We adopt the theory of software engineering to design our classes, in the end we figure out our classes and its relations with E-R diagram. Finally we complete our evaluation of our system in five aspects, which includes the quality of the advice proposed by the system, the correctness of reasoning, the quality of the interactivity, efficiency and cost.
|
PDF Full Text Request