爆炸冲击波集合分解排列熵时变峰值降噪算法

Explosion Shock Waves Arrangement Entropy Time-varying Peaks Denoising Algorithm of Integrated Decomposition

针对实测的爆炸冲击波信号中含有大量的噪声信号,严重影响冲击波超压峰值与正压时间的判读以及比冲量的计算等问题,提出了基于完全集合经验模式分解(CEEMDAN)与排列熵(MPE)的时变窗长时频峰值滤波的爆炸冲击波降噪算法,通过构造不同比例距离下的含噪冲击波信号模型和实测数据来进行研究与验证。原始爆炸冲击波数据经CEEMDAN分解为若干个本征模态分量(IMFs);并以IMFs的MPE值作为分类指标,将IMFs分量划分为需滤波和存留两个类别,对含噪模型与实测数据进行降噪处理实验,将降噪处理后的IMFs分量和剩余的IMFs重构。试验结果表明,与贝塞尔低通数字滤波器、CEEMDAN降噪算法相比,该方法能够去除信号中含有的高频噪声,获得较好的降噪指标;同时尽可能地保留了信号中的尖峰与突变信息,是比较理想的爆炸冲击波信号降噪算法。

The measured explosion shock wave signal contains a large amount of noise signals,which seriously affects the interpretation of shock wave overpressure peak and positive pressure time,as well as the calculation of specific impulse.An explosion shock wave denoising algorithm based on fully integrated empirical mode decomposition and arrangement entropy for time-varying window length time-frequency peak filtering(hereinafter referred to as the ensemble decomposition based arrangement entropy time-varying peak explosion shock wave denoising algorithm)e was proposed in this article,which was studied and validated by constructing noisy shock wave signal models and measured data at different proportional distances.The original explosion shock wave data was decomposed into several intrinsic mode components(IMFs)by CEEMDAN;By using the MPE value of IMFs as the classification index,the IMFs components were divided into two categories that require filtering and retention.Noise reduction experiments were conducted on noisy models and measured data,and the denoised IMFs components and remaining IMFs were reconstructed.The experimental results showed that compared with Bessel low-pass digital filter and CEEMDAN denoising algorithm,the method could remove high-frequency noise contained in the signal and achieve better denoising performance;simultaneously preserve the peak and mutation information in the signal as much as possible.