高重合度人字齿轮副啮合刚度影响规则的知识发现

为进行高重合度人字齿轮系统设计，采用知识发现方法获得多因素对啮合刚度的影响规律。在多次仿真数据的基础上，采用模糊聚类方法对人字齿轮副啮合刚度进行分类。基于粗糙集理论完成给定仿真数据集的离散化、属性约简以及属性值约简，从而获得综合考虑多因素的决策影响规则集，最终建立各结构设计因素与高重合度人字齿轮副啮合刚度之间的知识规则。新样本的综合判断表明，所获得的设计知识对高重合度齿轮系统的设计具有一定理论指导作用和实用性。

Influences of Design Parameters on Meshing Stiffness of Herringbone Gear with High Contact Ratio

In order to help to design herringbone gear system with high contact ratio, a knowledg discovery method was proposed, which combined fuzzy mathmatic with rough sets theory (RST). At first, a fuzzy clustering method was applied on the simulation results and helped to acquire the classification of the tooth meshing stiffness. Then, the design decision rules were deduced by discretization of simulation parameters, the reduction of attributes and values which were based on the RST. And these rules show the relationship among the meshing stiffness and design factors. At last, the illustration of three new samples demonstrates the validity of this new method in application of knowledge discovery of simulation data.