基于涡流技术的高压电缆铅封缺陷识别仿真研究

Simulation Study on Identification of Lead Sealing Defects of High Voltage Cable Based on Eddy Current Technology

铅封是保护电缆接头部位的重要部件。为确保铅封的有效性,基于涡流技术设计了一种高压电缆铅封缺陷识别方法,并展开仿真分析。首先,设置一个含噪声的一维涡流检测信号,基于离散傅里叶变换得到频域信号,对所得数据进行多项式拟合,利用去噪的方式对涡流检测数据实施预处理。然后基于涡流技术设计电缆铅封缺陷判别方法,训练缺陷识别分类器,通过求解凸二次规划得到最优分类决策函数,结合分类结果设计缺陷识别算法。仿真实验中,高压电缆铅封中分别模拟感应电流与感应磁场的分布情况。结果显示,该方法的识别准确率始终保持在87.0%以上,其在无缺陷以及缺陷深度大于1 mm时的识别精度在95%左右,可见该方法识别精度较高。

Lead seal is an important part to protect cable joints. In order to ensure the effectiveness of lead sealing, a lead sealing defect identification method for high-voltage cables is designed based on eddy current technology, and simulation analysis is carried out. First, set a one-dimensional eddy current testing signal with noise, obtain the frequency domain signal based on discrete Fourier transform, perform polynomial fitting on the obtained data, and preprocess the eddy current testing data by denoising. Then, based on eddy current technology, a method for identifying cable lead seal defects is designed, and a defect recognition classifier is trained. The optimal classification decision function is obtained by solving convex quadratic programming, and a defect recognition algorithm is designed by combining the classification results. In the simulation experiment, the distribution of induced current and induced magnetic field in the lead seal of high-voltage cable are simulated respectively. The results show that the recognition accuracy of this method is always above 87.0 %, and the recognition accuracy is about 95 % when there is no defect and the defect depth is greater than 1mm, which shows that the recognition accuracy of this method is high.