现场实测的碰摩声发射信号常淹没在各种噪声中,阻碍了碰摩声发射信号的识别,因此声发射信号降噪是声发射应用的核心技术。基于数学形态学的基本形态变换及其组合形式,该文提出了一种采用梯度法实现广义形态开、闭滤波器自适应加权组合的广义形态滤波器,将其应用于碰摩声发射信号的降噪中。实验结果表明,通过选取恰当的结构元素与步长参数,基于梯度法的自适应广义形态滤波器对碰摩声发射信号能取得较好的的降噪效果,相比于传统形态滤波器能够获得更高的信噪比,为碰摩声发射信号降噪提供了一种有效的途径。
A number of noises usually cover up the field-measured rubbing acoustic emission (AE) signals, which induce the difficulty in AE recognition. So the de-nosing of rubbing AE signals is the core technique by the application of AE. Based on basic morphological transformations and their combination modes, a generalized morphological filter using the gradient method to achieve the combined generalized morphological opening and closing filter adaptively is proposed. This filter is applied in the de-nosing of rubbing acoustic emission signal. Results of the experiment show that by selecting appropriate structure elements and step parameters, the adaptive generalized morphological filter based on the gradient method has better performance than traditional morphological filters, and it is effective in the de-nosing of rubbing AE signal.