论文部分内容阅读
为降低噪声对超声兰姆波检测信号的影响,提高信噪比和增加特征提取的精度,提出了一种赛利斯模型下分数阶微分方法用于超声兰姆波信号去噪.该方法对含噪声的兰姆波信号幅值谱进行各阶分数微分,用赛利斯分布作为待处幅值谱的模型,提出了幅值谱分数阶微分最大值和过零点与微分阶数的拟合三次关系式,建立了幅值谱特征参数的计算式来提取特征参数和重建原始信号的幅值谱,并结合相位谱重构去噪后的兰姆波信号.仿真结果表明,该方法可以有效地提高兰姆波信号甚至微弱兰姆波信号的信噪比,同时降低均方误差和平滑度.实验结果显示,与小波去噪和集合经验模态去噪方法相比,该方法在没有信号先验知识的情况下,可以更有效地去除兰姆波信号的噪声,同时更好地保留主信号的细节特征.因此,本文提出的方法可以有效地去除兰姆波检测信号中混入的噪声.
In order to reduce the influence of noise on the ultrasonic Lamb wave detection signal, improve the signal to noise ratio and increase the accuracy of feature extraction, a fractional differential method based on the Sayles model is proposed for the ultrasonic Lamb wave denoising. The amplitude spectrum of the noisy Lamb wave signal is fractionally differentiated. Using the Saylis distribution as the model of the amplitude spectrum to be used, a method of fitting the fractional maximum and zero crossing of the amplitude spectrum to the differential order Cubic relation, the formula of the characteristic parameters of the amplitude spectrum was established to extract the characteristic parameters and reconstruct the amplitude spectrum of the original signal, and the phase spectrum was used to reconstruct the denoised Lamb wave signal.The simulation results show that this method can be effective Improve the signal-to-noise ratio of the Lamb wave signal or even the weak Lamb wave signal, and reduce the mean square error and smoothness at the same time.The experimental results show that compared with wavelet denoising and ensemble EMD method, Under the circumstance of prior knowledge, the noise of Lamb wave can be removed more effectively and the detail of the main signal can be better preserved.Thus, the proposed method can effectively remove the noises mixed in the Lamb wave detection signal .