摘要
针对石油管道泄漏问题,提出模拟人眼识别和定位管道泄露信号的方法。由动态压力变送器采集的动态压力信号,携带了管道安全信息。正常信号和泄漏信号特征明显,从图像的角度很容易区分。仿照人眼识别信号,将信号存为画面映入大脑,分析时关注画面主要特征,摈弃次要特征。研究以图像处理方法为基础,根据管道信号的图像特征,对基于灰度图的管道泄漏检测和定位方法进行了改进,提出了主流数据区域的概念,提高了泄漏检测和定位的准确度,增加了该方法的适应性。对泄漏定位问题,提出应用基于形态学的图像处理方法,通过腐蚀、膨胀等操作分割信号图像为2个以上的连通区域,第1个连通区域的右边界对应时间即为负压波发生时刻。
For oil pipeline leak problem, a new method to simulate the human eye to identify and locate the pipeline leak signal is pro- posed . Dynamic pressure signals collected by the dynamic pressure transmitter, carry a pipeline safety information. The characteristics of normal signal and leakage signal are significantly different. From the perspective of the image, it is easy to distinguish. Modeled on the human eye recognition signal, the signal is saved as screen catches the brain, and brain concerns the main features of the screen, and abandons other secondary features. The research based on image processing technology to analyze the image characteristics of＇ pipe- line signal, improved the pipeline leak detection and location method based on gray image, and put forward the concept of the main- stream data area, enhanced the accuracy of the leak detection and position, and increased the adaptability. And, to the problem of leakage location, this paper put forward the application of image processing method based on morphology to detect and locate pipeline leakage. Through COITOsion, expansion, such as morphology operation, signal image is divided into two or more connected regions, and the right boundary of the first connected region corresponds to the time that the negative pressure wave occurs.
出处
《控制工程》
CSCD
北大核心
2014年第2期294-297,共4页
Control Engineering of China
基金
陕西学前师范学院青年项目(2012KJ042)
关键词
负压波
管道泄漏
灰度
形态学
negative pressure wave
pipeline leak
gray
morphology
