钢板腐蚀声发射信号的分形方法

摘要

In view of the seriousness of the steel corrosion,the paper proposes a method of using Fractal Dimension combined with Independent Component Analysis to j udge the degree of steel corrosion.Applying the method of Independent Component Analysis can separate the mixed signals into independent components to extract and denoise the received signals.And then the fractal dimension of the denoised signals was calculated.By this combined method,the effect of noise on fractal dimension can be reduced,and the result will be more accurate.The comparison of the simulation results showed that the fractal dimension of the extracted denoised signal by Independent Component Analysis was more accurate corresponding to the acoustic emission event number.The experimental data of steel corrosion indicated that the fractal dimension of AE signals was less affected by the corrosion parameters.Furthermore,the numeric value reflected the different degrees of steel corrosion well.%针对钢板的腐蚀检测，提出了一种运用分形维结合独立分量分析法来判断钢板腐蚀程度的方法。利用独立分量分析法将含噪混合信号分离成各自独立的成分进行去噪提取，再对提取到的去噪声发射信号进行分形维计算以降低噪声对分形维数的影响，使结果更准确。通过仿真计算结果的比较可看出，运用独立分量分析法对提取后的信号进行去噪，其分形维对声发射事件数的对应更准确。而钢板腐蚀的试验数据表明，声发射信号的分形维数受腐蚀参数的影响较小，其大小能较好地反映钢板的不同腐蚀状态。