【摘 要】
：
Pressure is the main source signal in real-time leak diagnosis of long pipeline.And it is the most widely used signal on real-time diagnosis of long pipeline leakage.In common,there are three necessar
【机 构】
：
School of Information Science and Engineering,Northeastern University,Shenyang 110004,China
论文部分内容阅读
Pressure is the main source signal in real-time leak diagnosis of long pipeline.And it is the most widely used signal on real-time diagnosis of long pipeline leakage.In common,there are three necessary steps,such as detecting anomaly,identifying anomaly and the leakage location,which should be completed by the signal of pressure.In this paper we propose a rapid and accurate method to deal with the problem of leak diagnosis.In order to deal with the problem of leak diagnosis rapidly and accurately,we utilize a real-time diagnosis based on pressure.Firstly,a combination filter algorithm is designed to realize noise reduction.Then,an anomaly detection algorithm is designed to detect abnormal pressure on the basis of the head and the tail pressure of the pipeline,in which an adaptive threshold updating method is used.Finally,the relevancy of the detected novelties is computed by Pearson correlation coefficient to identify the leakages.The experimental result shows that the proposed method can detect the leakage with few false alarms rapidly and locate the position of the leakage accurately.
其他文献
Focused on the problem that the take-off of a small unmanned helicopter(SUH)mainly depends on the control of the hand-operated,an automatic take-off method for SUH is proposed and the circuit is imple
To improve the training speed of traditional support vector regression(SVR)algorithm on large datasets,an adaptive resample strategy is proposed to reduce the training dataset without loss of performa
Images restored by using dark channel prior theory have haze edges.A novel improved image clearness method based on dark channel prior theory is proposed.First of all,characteristics of pixel distribu
In this work,an improved stacked partial least square regression(VIP-SPLS)algorithm based on variable importance in the projection(VIP)is introduced for near-infrared spectroscopy quantitative analysi
This paper presents a Lagrangrian relaxation(LR)approach with machine capacity relaxation and operation precedence relaxation for solving a flexible job shop(FJS)scheduling problem from the steelmakin
This paper proposes a novel batch process monitoring method that is based on Kernel Entropy Independent component analysis(KEICA).The main idea of KEICA is to carry out an independent component analys
To identify the irregular fluctuation of the spinning quality,and to control abnormal behavior of influencing factors in the spinning process,first,through massive spinning data,fluctuation reasons of
A novel two-degree-of-freedom(2-DOF)PID control scheme for a magnetic levitation system was proposed in this paper.To make the set-point tracking performance and the disturbance rejection characterist
In order to predict the wind power generation capacity accurately,an optimal combination forecasting technique combining Artificial Neural Network(ANN)model with gray model(GM)is proposed in this pape
Our research concerns about basic oxygen furnace(BOF)steelmaking process,according to the analysis of actual industry data,the soft measure modeling method is proposed for BOF steelmaking process.The