基于超声特征值的微泄漏孔径识别算法研究

Research on micro leakage aperture recognition algorithm based on ultrasonic eigenvalue

针对真空容器微泄漏孔径无法准确识别的问题,提出采用泄漏超声信号特征值数据模型结合欧氏距离算法的解决方法。根据泄漏超声信号强度与微泄漏孔径的对应关系,选用特征值建立数据模型,并利用欧氏距离算法取最小值对微泄漏孔径进行识别,有效避免了因泄漏超声信号随机性引起的误判,提高了微泄漏孔径识别的准确率。通过多次实验验证,以泄漏超声信号有效值和互相关有效值建立的数据模型,对微泄漏孔径的辨识效果最好,其对微泄漏孔径的辨识正确率高达92%以上,达到了真空容器微泄漏孔径的检测识别要求。

Aiming at the problem that the micro leakage hole diameter of vacuum vessel can not be identified accurately, a solution based on the eigenvalue data model of leakage ultrasonic signal and Euclidean distance algorithm is proposed.According to the corresponding relationship between leakage ultrasonic signal strength and micro leakage aperture, the eigenvalue is selected to establish the data model, and the minimum value is taken by Euclidean distance algorithm to identify the micro leakage aperture, which effectively avoids the misjudgment caused by the randomness of leakage ultrasonic signal and improves the accuracy of micro leakage aperture identification.Through many experiments, the data model based on the effective value of leakage ultrasonic signal and cross-correlation effective value has the best identification effect on the micro leakage aperture, and the identification accuracy of the micro leakage aperture is more than 92 %,which meets the detection and identification requirements of the micro leakage aperture of vacuum container.