基于对角递归网络的气体管道泄漏检测与定位的研究OA北大核心CSCDCSTPCD
Research on Gas Pipeline Leakage Detection and Localization Based on Diagonal Recurrent Neural Network
利用对角递归网络能够描述非线性动态系统的特点,设计了用于气体管道泄漏检测定位的动态网络模型.网络的训练样本由管道各种工况下的压力、流量数据组成.试验结果表明,与传统的静态前馈神经网络模型相比,本研究建立的网络能更好的反映气体在管道中的流动特性,实现气体管道的泄漏检测与定位.
A new method is put forward to detect and locate the gas pipeline leakage, which designed a dynamic network model DRNN. The samples of the network are composed of pressure and flowrate data of all kinds of working conditions. The experiment results verify that the model can describe the gas' flow-characteristic in the pipeline, as well as detect the leakage and locate the leakage position.
姚志英;彭光正;张雅丽
北京理工大学信息学院,北京100081北京理工大学信息学院,北京100081中国人民公安大学安全防范系,北京100038
石油、天然气工程
管道泄漏检测与定位DRNN
pipeline leakagedetection & localizationDRNN
《中山大学学报(自然科学版)》 2007 (z1)
93-94,2
评论