Traditional design requires designers to envisage a product operating environment in order to identify customer needs. Analyzing product usage context by collecting actual product operating data during the product in use empowers new opportunities for the projection of requirement specifications and understanding of use case scenarios. This paper proposes a data-driven inverse design optimization approach to provide decision support to product personalization design. A closed-loop decision-making framework is formulated by integrating forward design and inverse problem solving within a coherent framework of data-driven analysis. An application to the transmission system personalization design of wheel loaders is presented to demonstrate how personalized product usage contexts are identified through inverse analysis of product operating data under different operating conditions. A particle swarm optimization (PSO) algorithm incorporated with Simulink simulation is developed to solve the multi-objective optimization of power performance and fuel economy for wheel loaders.

传统设计要求设计师设想产品运行环境，以识别客户需求。通过在使用产品期间收集实际的产品运行数据来分析产品使用情况，这为预测需求规范和理解用例场景提供了新的机会。本文提出了一种数据驱动的逆向设计优化方法，为产品个性化设计提供决策支持。通过在数据驱动的分析的一致框架中集成前向设计和逆问题解决方案来制定闭环决策框架。提出了一种轮式装载机传动系统个性化设计的应用程序，以演示如何通过对不同运行条件下的产品运行数据进行反向分析来识别个性化产品使用环境。为了解决轮式装载机动力性能和燃油经济性的多目标优化问题，开发了一种与Simulink仿真相结合的粒子群优化（PSO）算法。