中文摘要：

由于受限空间瓦斯爆炸冲击波压力信号具有突变性和瞬态非平稳性特点，采用小波包分析进行去噪处理。选取典型的Daubechies小波包系和Symlets小波包系各小波包基对信号的重构均方根误差，并计算了原始信号与重构信号的相关系数，确定Sym8为处理瓦斯爆炸冲击波压力信号的最佳小波包基。去噪结果显示：小波包分析在降低噪声的同时，仍然保留信号突变部分的信息，能尽可能不失真得重现有效信号，受限空间瓦斯爆炸冲击波压力信号小波包去噪能够得到更可靠地结果。

英文摘要：

The signal of shock wave pressure by gas explosion in confined space has the characteristics of abrupt changing and transient non-stationary, so the wavelet packet analysis was applied to conduct signal de-noising.The typical wavelet packets basis of Daubechies and Symlets were selected to calculate the reconstructed RMSE （ Root Mean Square Error） to the signal and the correlation coefficient of original and reconstructed signal.The wavelet packets basis sym8 was determined as the optimal one to dispose the signal of shock wave pressure by gas explo-sion.The results showed that the abrupt changing part of pressure signal was still reserved when the noise was re-duced, and reproduced the effective signal without distortion.So the wavelet packet analysis technique has the fa-vorable de-noising effect.