Abstract: To realize the correlation between drive axle user and test field durability, a method was proposed to develop test field durability test specifications based on users and test field road measurement data. Based on the collected user torque loads, combined with the gear and speed signals, a model of the Total load (including axle load and tooth load) distribution of the drive axles was predicted under different gears based on the rotational rainflow counting and nonparametric kernel density estimation methods. The multi-objective optimization model of “user-test field” was established by correlating the damage matrix of the coefficients of each characteristic working condition in the test site. The non-dominated sorting genetic algorithm with elitist strategy(NSGA-Ⅱ) was applied to solve the multi-objective optimization model and select the optimal solution by setting constraints based on the actual use of the test site. The validity of the optimized model solution set was verified from the perspective of relative damage and load distribution for loads. The results show that the total mileage of the developed test range test specification is about 46 133.3 km equivalent to 200 000 km of actual driving by users, and the road reinforcement factor is as 4.34. The study provides a reference and basis for more effective development of the drive axle test field durability test specification and reasonable evaluation of the durability and reliability for the whole vehicles and their components. 

Key words: user-association, durability, test specification, genetic algorithm, drive axle

摘要： 为实现驱动桥用户与试验场耐久性之间的关联，提出了基于用户和试验场道路实测数据制定试验场耐久性试验规范的方法。以采集的用户扭矩载荷为基础，结合挡位和转速等信号，基于旋转雨流计数和非参数核密度估计方法，预测了用户在不同挡位下驱动桥Total（包含轴载荷和齿载荷）载荷分布模型。关联试验场各个特征工况的系数损伤矩阵，建立了“用户试验场”多目标优化模型。运用带精英策略的非支配排序遗传算法（NSGA-Ⅱ），以试验场实际使用情况为依据设置约束条件，求解了多目标优化模型并选取出了最优解。从载荷相对损伤和载荷分布角度，验证了优化模型解的有效性。研究结果表明：所制定的试验场试验规范其总里程约为46 133.3 km等效于用户实际行驶200 000 km，路面强化系数为4.34。本研究为更加有效地制定驱动桥试验场耐久性试验规范、合理评价整车及其零部件耐久性与可靠性提供了参考及依据。

关键词: 用户关联, 耐久性, 试验规范, 遗传算法, 驱动桥