A data-driven optimization model to collaborative manufacturing system considering geometric and physical performances for hypoid gear product

Abstract

Abstract A data-driven optimization model to collaborative manufacturing system considering geometric and physical performances is proposed to improve competitiveness of hypoid gear product development facing with economic globalization. Firstly, to deal with the vagueness or impreciseness of the voice of customer (VOC), the fuzzy quality function deployment (fuzzy-QFD) using fuzzy weighted average method in the fuzzy expected value operator is introduced into hypoid gear manufacturing. It can convert them into the critical to qualities (CTQs), and the technical geometric and physical performance requirements. And then, the multi-objective optimization (MOO) modification of machine settings is proposed to establish a basic data-driven model for collaborative system. Different with the conventional modification only considering geometric performance, it provides an improved modification also considering the physical performances. Finally, a double-curved shell model of hypoid gear finite element is used to perform the numerical loaded tooth contact analysis (NLTCA), as well as to establish the data-driven relationships between machine settings and physical performance evaluations. Immediately, whole development is divided into three sub-problems: i) optimal operations of the noise factors by measurement and numerical control (NC) compensation, ii) identification of the prescribed ease-off topography by multi-objective optimization using iterative reference point approach and iii) modification considering geometric performance by a trust region algorithm with step strategy. The numerical instance in practical applications is given to verify the proposed methodology.