In the course of driving, due to the influence of environment, signal acquisition device and the degree of unevenness of road surface, the collected signals fluctuate within certain error range and have randomness. To predict the fatigue life of mechanical structures more accurately, the processing method of random load signals for uncertainty is proposed. The collected load signals are converted into interval load signals by Bootstrap error circle method, and based on the generalized Pareto distribution, the peak-over-threshold extrapolation is performed on the interval load signals. Based on the mixed probability distribution model, the one-dimensional and two-dimensional interval load spectrum for shock absorber cylinder are established by analyzing the characteristics of load signals. In addition, based on the uncertain signal problems, the interval life spectrum and interval damage spectrum of one-dimensional and two-dimensional are constructed by the interval algorithm. This paper lays a foundation for load signals processing and fatigue life prediction of parts.

在行驶过程中，受环境，信号采集装置以及路面不平度的影响，采集到的信号在一定的误差范围内波动并具有随机性。为了更准确地预测机械结构的疲劳寿命，提出了不确定性随机载荷信号的处理方法。通过Bootstrap误差圈方法将收集到的负载信号转换为间隔负载信号，并基于广义的Pareto分布，对间隔负载信号执行阈值外推峰。在混合概率分布模型的基础上，通过分析载荷信号的特性，建立了减振器缸体的一维和二维间隔载荷谱。此外，基于不确定的信号问题，利用区间算法构造了一维和二维的区间寿命谱和区间损伤谱。本文为零件的载荷信号处理和疲劳寿命预测奠定了基础。