随机振动响应峰值因子的预计理论研究

简要介绍了超越概率理论、超越频次理论、损伤等效理论和功率谱密度(PSD)的时域拟合理论等4种常见的峰值因子预计理论,并基于三角级数提出了一种新的预计理论。结合试飞加速度数据样本,对比分析超越频次理论、PSD时域拟合理论和三角级数理论的预估精度。研究表明,上述4种常见的预计理论本质上属于统计学理论;PSD时域拟合理论预计的峰值因子波动较大,峰值因子与归一化次数满足高斯分布;三角级数理论的预估精度较高,但缺乏离散峰个数的合理判据。

Theoretical Research for Predicting Crest Factor of Random Vibration Response

The work aims to briefly introduce four common theories for predicting crest factor including exceeding probability theory, exceeding frequency theory, damage equivalent theory and time-domain fitting theory of power spectral density (PSD) and propose a new prediction theory based on trigonometric series. According to the acceleration data sample during flight test, exceeding frequency theory, time-domain fitting theory of PSD and trigonometric series theory were analyzed contrastively in order to evaluate their accuracy. The analysis results showed that the above four common theories belonged to statistics theory. The crest factors predicted by time-domain fitting theory of PSD fluctuated greatly and satisfied Gauss distribution as the number of normalization. The predicting accuracy of trigonometric series theory was better, but lacked reasonable criterion to determine the number of discrete peak values.