关键词: 结构耐久；大数据；用户使用；载荷谱

Abstract: The vehicle structure durability test, as one of the key pass-through indicators in vehicle development, is crucial for the development of vehicles. In order to improve the relevance between durability specifications and actual user usage, this article proposes a set of methods for developing vehicle structure durability test specifications that are associated with actual user usage. Firstly, based on user big data, machine learning is used to obtain the real proportion of user road type mileage, user driving styles, and typical operating conditions. Secondly, target percentile user load spectra are obtained through user load spectrum collection, enhanced editing, and extrapolation. Then, proving ground condition loads are collected and screened, and through multi-dimensional correlation technology, the proving ground conditions are matched with the user target loads to compile the vehicle structural durability specifications. Finally, the durability specifications are improved and perfected through trial operation and continuous optimization.

Key words: Structural durability; Big data; Customer usage; Load spectrum