论文部分内容阅读
针对多属性产品实例检索过程中存在一定的主观性问题,影响检索结果,提出一种基于多维关联函数的相似实例检索方法并开发出原型系统.首先根据多维关联函数定义,得到多维关联函数一般计算方法的求解步骤,并基于多维关联函数构建相似度函数.其次,根据客户要求设定产品各属性的重要度,结合重要度和各属性修改难度系数构建实例评价模型.最后,运用多维关联函数计算方法,Access数据库构建技术以及层次分析法构建出包含用户需求输入模块、可拓域输出模块、相似实例检索模块和实例评价模块的原型系统,实现实例输出.以大型螺杆空压机为例来构建基于多维关联函数的相似实例检索系统,验证此理论方法的正确性与可行性.
Aiming at the subjective problems existed in the retrieval process of multi-attribute product instances and the retrieval results are affected, a similar case retrieval method based on multidimensional correlation function is proposed and a prototype system is developed.First, according to the definition of multidimensional correlation function, the general calculation of multidimensional correlation function Method, and construct the similarity function based on the multidimensional correlation function.Secondly, the importance of each attribute of the product is set according to the client’s requirements, and the evaluation model of the instance is constructed by combining the importance degree and the difficulty degree of each attribute modification.Finally, using the multi-dimensional correlation function to calculate Method, Access database construction technology and AHP to build a prototype system which contains user input module, extender output module, similar instance retrieval module and instance evaluation module to realize the output of example.With large screw air compressor as an example, A similar case retrieval system based on multidimensional correlation function verifies the correctness and feasibility of this theoretical method.