Research On Knowledge Assistance System And Its Application On Process Planning

With the development of information, networking and intelligence of manufacturing industry, knowledge has become the key resource of development of society and enterprises. And knowledge management has become an important research subject. Product competition requires enterprises to make timely accumulation, rapid discovery and appropriate application of knowledge. Besides, in this information era, efficient knowledge acquirement and application are challenging problems awaiting to be solved. There are many shortcomings in current knowledge- based system, including low knowledge usage ratio and extra-emphasized machine automation concept, which result in system construction and application difficulties. This is the case especially in computer aided process planning.Based on analysis of knowledge management and knowledge engineering, the conception of knowledge assistance system and its corresponding design method have been introduced to solve the problems aforementioned.Firstly, problems are found out based on analyzing development of current research on knowledge-based system. Knowledge in design process is abundant and its contents are complex. To enhance knowledge utility ratio and human-work efficiency, human-machine collaboration strategy is employed as the system general structure, in which human plays a primary role and computer is used to assist people with knowledge.Domain knowledge contains both theoretical knowledge and experiential knowledge, which can be tacit or explicit. Tacit knowledge is difficult to be represented. Through classification of related knowledge, knowledge structure model basing on knowledge units is proposed. Object-oriented representation method combined with case-based and rule-based representation are employed to represent theoretical knowledge, case and rule. Data structure that satisfies system storage and application requirements is established, and thus knowledge can be transferred and shared.The realization mechanism of knowledge assistance is emphasized in this paper. Task-oriented knowledge requirement model is proposed and realized by knowledge pushing mechanism. Related knowledge and cases are provided to human through the combination of rule-based reasoning and case-based reasoning. The methods can provide human with problem solving clue and enhance knowledge utility ratio. In order to make the system organize knowledge assistance reasonably, a method that tests individual knowledge structure is proposed. This system can provide related knowledge closely to user's task by testing and acquiring user's knowledge structure. Then user will get only related knowledge and thus be able to use them to make decisions rapidly.A case study in the filed of machining technology is given, and corresponding prototype system is developed. The application and running effect have been given through analysis of knowledge characteristic and requirements of design tasks.The theory and research have got a good result and validation in the practical case, which provide a new development direction for the research of knowledge-based system.