基于BP神经网络的瓷绝缘子振动声学检测结果分类

Classification of the Insulator Inspection Data by Acoustic Vibration Based on BP Neural Network

摘要: 为便于对采用振动声学法检测瓷绝缘子的数据进行判读评估，提出了采用BP神经网络对检测结果进行分类的方法。使用Matlab编程语言编制BP神经网络分类代码，用已分类的三组检测数据对神经网络进行训练。通过对128组检测结果进行分类比较，分类的平均准确率达到93.5%。证实BP神经网络对振动声学法检测绝缘子的检测结果分类的有效性和实用性。

Abstract: For the convenience of reading the data of the insulator inspection by the acoustic vibration , the method in grouping testing data is proposed by using the BP neural network. By classifying the 128 groups of detection results ,the average accuracy of classification has achieved 93.5%. The BP neural network of classifying the results of insulator inspection has been proved to be an effective and practical method.