基于混合推理的高速切削数据库系统的建立

高速切削包含大量相互作用的参量，详细地讨论了这些参量之间的相互关系和作用，并对工程中常用的智能方法进行了分析比较。在此基础上，提出了基于规则和实例混和推理建立高速切削数据库系统。在该系统中，实例推理起主要作用，当实例库中没有积累相关数据时，采用规则推理求解。根据各属性的特点和高速切削加工知识，确定了实例属性局部相似度的计算方法和各属性整体相似度的权值。为了快速查找实例，采用全部属性值索引和单个属性值索引两种索引方式，并采用规则推理来补充和验证实例推理的结果。规则和实例混合推理的采用，使系统的数据采集和知识更新简单易行。

Development of high-speed cutting database system based on hybrid reasoning

High-speed cutting was a complex manufacturing process with a large number of interacting variables. Relationships and interactions among these variables were discussed in detail and the existing artificial intelligence methods mainly applied in engineering were analyzed. Based on these discussions and analysis, a hybrid approach combining rule-based and case-based reasoning was presented for development of high-speed cutting database system. In the system, case-based reasoning was employed as the main problem-solving agent. Rule-based reasoning was employed whenevel relevant data have not yet accumulated enough in the case base. With analysis of the characteristics of attributes and domain knowledge of high-speed cutting, the calculation methods of the local similarities of case attributes and the weight of global similarity of case attributes were provided. In order to retrieve case quickly, two index modes were employed, namely index by values of all case attributes and index by value of single case attribute, And the result from case-based reasoning was complemented and verified by rule-based reasoning. Using hybrid reasoning in high-speed cutting database system has made data collection and knowledge update simpler and easier.