Design Optimization for Minimizing Performance Deviations of Complex Vehicle Door Systems Using Virtual Manufacturing Big Data and Axiomatic Design

This study introduces an innovative framework aimed at minimizing performance deviations in complex vehicle door systems by leveraging the principles of axiomatic design and virtual manufacturing big data. Utilizing the independence axiom of axiomatic design theory, an optimal design sequence is established for a vehicle door system. Analytical models for door opening and closing are developed, and surrogate models are constructed for weatherstrips in conjunction with machine learning techniques. Monte Carlo simulations are performed, enabling the generation of virtual manufacturing data and thereby facilitating a comprehensive analysis. The application of genetic algorithms with information content as the objective function can minimize vehicle performance variability, offering a promising approach for design optimization. This methodology not only demonstrates the potential for significantly reducing performance deviations but also highlights the effectiveness of integrating computational techniques with axiomatic design principles to enhance system predictability and quality control.