| 赵燕锋,冯早,朱雪峰,伍林峰.复杂工况下基于时频图像和CNN SVM的管道堵塞识别研究[J].电子测量与仪器学报,2021,35(2):161-170 |
| 复杂工况下基于时频图像和CNN SVM的管道堵塞识别研究 |
| Research on blocking recognition of drainage pipeline under complicated conditions based on time frequency image and CNN SVM |
| |
| DOI： |
| 中文关键词: 复杂工况 声信号时频图像 自适应特征提取 管道堵塞故障诊断 |
| 英文关键词:complicated conditions acoustic signal time frequency image adaptive feature extraction fault diagnosis of pipeline |
| 基金项目:国家自然科学基金(61563024，51765022)项目资助 |
|
|
| Author | Institution |
| Zhao Yanfeng | 1.Faculty of Information Engineering ＆ Automation, Kunming University of Science and Technology, Kunming 650500，China;2.Yunnan Key Laboratory of Artificial Intelligence，Kunming 650500,China |
| Feng Zao | 1.Faculty of Information Engineering ＆ Automation, Kunming University of Science and Technology, Kunming 650500，China;2.Yunnan Key Laboratory of Artificial Intelligence，Kunming 650501,China |
| Zhu Xuefeng | 1.Faculty of Information Engineering ＆ Automation, Kunming University of Science and Technology, Kunming 650500，China;2.Yunnan Key Laboratory of Artificial Intelligence，Kunming 650502,China |
| Wu Linfeng | 1.Faculty of Information Engineering ＆ Automation, Kunming University of Science and Technology, Kunming 650500，China;2.Yunnan Key Laboratory of Artificial Intelligence，Kunming 650503,China |
|
| 摘要点击次数: 1208 |
| 全文下载次数: 4 |
| 中文摘要: |
| 针对复杂工况下管道系统堵塞状态识别模型精度出现偏差的问题，提出一种基于时频图像和卷积神经网络（CNN）对管道内的堵塞物和三通件个体识别方法。首先，利用声波检测管道得到不同工况的低频声压信号，滤波处理后进行平滑伪Wigner Ville时频分析得到声信号时频分布图；然后，采用大津阈值分割法对单一和复杂工况时频分布图像进行自适应分割，得到堵塞物和三通件时频图像；最后，将单一工况下轻度堵塞、重度堵塞、三通件和管道尾端4种物体的时频图像输入至卷积神经网络 支持向量机（CNN SVM）模型进行训练，将训练好参数的模型应用于复杂工况下不同程度堵塞物和三通件的自动识别。实验结果表明，所提出的方法对4类目标在复杂工况下的识别率均达到96%以上，识别精度高于传统人工特征提取的模型。验证了堵塞物在不同工况下对声波的影响具有共性，与三通件具有差异性；对复杂工况下管道中不同程度堵塞物和三通件个体进行分析，可以有效克服工况分布差异对模型识别精度造成的偏差。 |
| 英文摘要: |
| Aiming at the deviation of the recognition accuracy for pipeline system blocking recognition model under complicated working conditions, a method is proposed for recognizing blockage and lateral connection in pipeline individually based on time frequency image and convolution neural network algorithm. Firstly, acoustic wave detection method is used to obtain low frequency sound pressure signals under different working conditions, smooth pseudo Wigner Ville time frequency analysis method is performed to the filtered signal to obtain the time frequency distribution map. Then, the Otsu threshold segmentation method is applied to adaptively segment the time frequency distribution images to obtain time frequency images of blockages and lateral connection under single and complicated working conditions. At last, the time frequency images of light blockage, heavy blockage, lateral connection and pipe end under a single working condition are entered into the CNN SVM model for training, the trained parameter model is applied to the automatic recognition of blockages and pipe components under complicated working conditions. The experimental results show that the recognition rate of the proposed method for four kinds of targets under complicated conditions is over 96%, and the recognition accuracy is higher than that of the traditional artificial feature extraction model, which verified that the influence of the blockage on the acoustic wave under different working conditions is common and different from that of the lateral connection. individual analysis of different degrees of blockage and lateral connection under complicated working conditions individually, can effectively avoid the deviation of model recognition accuracy owing to the difference of working condition distribution. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
