中文摘要：

针对质量屋存在内在模糊性和不确定性的问题,引入了粗糙集理论中的基本原理.提出了粗糙数的概念,用于得到客观的用户需求重要度排序.将用户需求与技术特性的关联性测度和技术特性自相关性确定问题转化为决策系统中条件属性重要度的确定问题,并结合条件熵与互信息的概念,提出了一种可提高计算精度的属性重要度算法.为了实现技术特性的最优配置,建立并求解以可用资源决策值与目标值的偏差最小为优化目标的整数规划模型,从而实现由用户需求到技术特性的映射.最后,通过推力轴承设计实例验证了所提方法的可行性和有效性.

英文摘要：

As HoQ has the problems of inherent fuzziness and uncertainty,basic principles of rough set theory have been introduced.And the concept of rough number has been put for ward for the impersonal importance rating of customer requirements.On basis of the definitions of condition entropy and mutual information,and with the transformation of the calculation of the relationship between customer requirements and technical characteristics,and of the self-correlation of technical characteristics into the importance calculation of condition attributes in decision system,an improved algorithm of attributes importance,which can raise the calculation accuracy,has been proposed.In order to optimize the configuration of technical characteristics,an integer-programming model has been established with the objective to minimize deviation variable between decision value and target value of usable resources so that the mapping from customer requirements to technical characteristics is realized.The case of the design of a thrust bearing verifies the feasibility and effectiveness of the proposed method.