基于核密度估计的单轨车辆构架载荷谱外推

Extrapolation for A Monorail Vehicle Frame Load Spectrum Based on Kernel Density Estimation

车辆构架疲劳寿命预测准确性受动载荷谱完整性的影响,但线路实测的载荷数据具有短时与局部特征,为了准确预测疲劳寿命需要将短期样本数据外推至全寿命周期范围,本文以单轨车辆转向架构架为分析案例。首先,采用基于自适应带宽的核密度估计方法对实测载荷谱进行概率密度估计,其次,通过蒙特卡洛模拟算法实现了载荷谱外推;最后,基于核密度外推后的载荷谱对构架进行疲劳寿命分析。分析结果表明,与传统线性外推的数据相比,基于核密度估计外推的载荷谱评估结果更加安全,更有利于保障车辆的安全运行。

The accuracy of predicting the fatigue life of vehicle frames is affected by the integrity of the dynamic load spectrum,but the load data measured on the line has short-term and local characteristics.In order to accurately predict the fatigue life,it is necessary to extrapolate the short-term sample data to the full life cycle range.This article takes the bogie frame of monorail vehicles as an analysis case.Firstly,a kernel density estimation method based on adaptive bandwidth was used to estimate the probability density of the measured load spectrum.Secondly,the load spectrum was extrapolated using a Monte Carlo simulation algorithm;Finally,fatigue life analysis of the structure is conducted based on the load spectrum extrapolated from kernel density.The analysis results indicate that compared with traditional linear extrapolation data,the load spectrum evaluation results based on kernel density estimation extrapolation are more secure and more conducive to ensuring the safe operation of vehicles.