Abnormal/fault positioning detection method based on binomial tree model

Abstract

The invention discloses an abnormal/fault positioning detection method based on a binomial tree model. The method comprises the steps of firstly, converting the physical layout of an industrial pipe network into the binomial tree structural model, and storing the detected data into corresponding nodes of a binomial tree; secondly, according to the character of the binomial tree and the relevance between information of the detection nodes, and utilizing the flow conservation property of the data flow of the industrial pipe network for achieving computation of the data flow of nodes of sensors not to be deployed; thirdly, according to the relation between the actual measuring value of the deployed sensors of the detection nodes of the industrial pipe network and the computation value, analyzing the operation condition of the industry, and judging the abnormal/fault types; finally, utilizing the structure characteristics of the binomial tree model for analyzing the abnormal/fault positions of the industrial pipe network and giving an alarm. According to the method, the design is simple, the cost is low, abnormal/fault detection by a traditional artificial tour gauging method is broken through, and the real-time detection on the working condition of the industrial pipe network in a real time mode under the limiting situations of the detection nodes is achieved.

Description

technical field [0001] The invention relates to an abnormal / fault location detection method based on a binary tree model. According to the data detected by deployed limited sensors, using the properties of a binary tree, the data values ​​of undeployed sensor nodes are inverted and calculated; according to the flow conservation characteristics of industrial pipeline data streams , the relationship between the actual detection data and the calculated data value of the deployment detection point, online judgment of the fault type, and the corresponding location of the fault. This technology breaks through the limitations of the traditional manual inspection method for fault detection and the time lag of data, and realizes the monitoring of pipeline operating conditions under the condition of limited detection points, which belongs to the technical field of fault detection for industrial pipeline network conditions. Background technique [0002] Pipes are widely used in residen...

AI Technical Summary

Problems solved by technology

The negative pressure wave method is more effective in detecting sudden large-scale leaks, but for persistent and small-scale leaks, negative pressure waveforms are not easy to identify and detect

In a multi-branch pipeline network, the negative pressure wave method is difficult to distinguish the difference between the pressure fluctuation caused by the instantaneous diversion of the branch or node and the pressure fluctuation caused by the real leakage; therefore, it is not suitable for the leakage detection of the pipeline

The soft sensing method developed in recent years has been initially applied to detect and locate faults, but a mature application technology system has not yet been formed in China

Images