基于静电感应的金属接触损伤诊断研究

Metal Surface Damage Diagnosis Based on Electrostatic Sensing

提出了基于静电感应的金属接触损伤诊断技术和方法。分析了金属接触损伤致电机理和感应模型,利用小波单支重构和短时傅立叶变换方法验证了该技术的可行性,借助传感器阵列和智能算法实现了接触损伤位置的定位。基于因子分析和支持向量机神经网络,对不同损伤程度的试验样本实现了定量识别。该诊断技术和方法可以有效识别金属接触损伤的位置和程度。

Diagnosis technology and method based on electrostatic sensing for metal surface dam- age were proposed. Mechanism of electrostatic in metal contact damage and electrostatic sensing mod- el were produced. Feasibility of this technology was analyzed based on wavelet transform, reconstruc- tion and STFT（short-time Fourier transform） analysis. Damage was located by using sensor array and an algorithm. Quantitative recognition of contact damages were achieved by using support vector machine neural networks and factor analysis. So diagnosis technology and method based on electro- static are capable of locating and quantitating the contact damage.