由于火工冲击试验条件的限制,测试所得冲击加速度信号中通常含有低频趋势项,导致试验数据不准确。因此需要对信号进行有效性分析,校正信号,消除积分零漂。现有的校正方法大都是基于经验的定性方法,具有一定的局限性。基于离散小波分解,提出了一种冲击数据有效性分析方法,该方法通过分析信号的频谱与相关性,能够定量确定小波分解层数的范围,还讨论了Db小波基函数的消失矩对数据校正结果的影响。通过对某分离冲击试验数据进行有效性分析与校正,证明该方法能够有效抑制积分零漂,消除趋势项对冲击谱的影响,提高试验数据的准确性和有效性。
Acceleration data measured in pyroshock experiment usually contain a low-frequency trend term,which makes the signal inaccuracy. Therefore the pyroshock data require validation and correction to avoid the integral zero shift.Traditional correction methods are mostly qualitative methods based on experience,making some limitation in application.A method based on the discrete wavelet decomposition was proposed. In the method wavelet correction parameters were chosen quantitatively via analying the frequency spectrum and correlation coefficients. The influence of vanishing moments on the result of correction was discussed. The proposed method was applied to the validation and correction of experimental data of pyroshock. It is revealed that the method can remove integral zero shift,correct the shock spectrum and improve the accuracy and effectiveness of pyroshock test systems.