Optimal Structural Design of Complicated Mechanical Products based on Data Ming and Muti-level Partitioning Strategy

Abstract

Complicated Mechanical products normally consist of multi-parameter mechanisms. The uncertain factors resulted from the quantity of parameters and complicity of inter-relations shall give rise to problems of large dimensional arrays. Moreover, optimal models often contain functions of distinctive property, therefore feasible optimal methods are hard to find. On the basis of analyzing the optimal design methods and the data mining principles, a multi-level partitioning strategy of overall structural optimal design for mechanical products and an improved ID3 algorithm is hereby put forth to bring out an optimal design method integrated with the data mining technology as well as an optimal structural design system of complicated mechanical products based on the data mining principles, which is demonstrated by the optimal structural design of TBM (Tunnel Boring Machine). The example shows that the data mining technology has fully tackled the issues of the optimal structural design of complicated mechanical products.