Leakage detection method for large complex natural gas pipeline network system

Abstract

The invention discloses a leakage detection method for a large complex natural gas pipeline network system. The method comprises the steps of 1, performing static simulation to obtain a pressure along a line, flow and initial distribution of temperature in a pipeline; 2, performing online simulation to obtain a pressure along the line, flow and real-time distribution of temperature in the pipeline; 3, performing parameter learning, and proposing a pipeline transient flow state space model-based pipeline parameter real-time online estimation and learning model based on an implicit method and characteristic line method-based gas pipeline dynamic simulation state space model according to a sensitivity analysis theory; and 4, performing leakage detection, calculating a deviation between a simulated pressure and an actually measured pressure of each pressure measurement point, and realizing leakage state identification and diagnosis of the large complex oil and gas pipeline system through correlation analysis of simulated and actually measured pressure deviation distribution. The limitation of sectional detection is broken through, and a leakage detection system for the CPM (Computational Pipeline Monitoring)-based large complex natural gas pipeline system is established.

Description

technical field [0001] The invention relates to the technical field of pipeline system leakage detection, in particular to a large-scale and complex natural gas pipeline network system leakage detection method. Background technique [0002] At present, the online pipeline leakage detection technology and system are developing along two lines of thought: one is the pipeline analysis calculation detection method CPM (Computational Pipeline Monitoring) based on the internal flow and leakage characteristics of the pipeline, and the other is the external detection with the help of external systems and equipment Methods (optical fiber, cable and infrasonic method, etc.). [0003] The leak detection method based on the pipeline dynamic model makes full use of all the characteristics of the flow and leakage of the pipeline system, can identify the normal transient flow of the pipeline and the disturbance caused by small leaks, and is sensitive to the small leaks that occur but is no...

AI Technical Summary

Problems solved by technology

However, in China, its theoretical research and technical development work is relatively limited. Due to the high difficulty of theoretical research and application development, domestic pipeline leak detection technology developers have not set foot in this field

Images