Study On The Diagnosis Of Heating Pipe Network Leakage Based On Deep Belief Network

Urban heating is an indispensable basic social service for the lives of our residents.The heating demand is increasing along with the continuous improvement of people’s living standards.At the same time,it is also closely related to the winter haze problem that the whole society has paid close attention to in recent years.In the urban heating system,the leakage failure of the heating pipe network is one of the main failures.The leakage will break the hydraulic balance of the heating network,causing energy waste and affecting the operation of the entire heating system.With the transformation of energy,China’s urban heating system is being upgraded to a clean,low-carbon,smart operation.Therefore,scientific and accurate diagnosis of leaks in heating pipe networks has great economic benefits and practical significance.In the background of industrial big data,the integration of modern artificial intelligence technology and the ability to fully utilize deep learning to mine fault characteristics are the research directions of fault diagnosis in the industrial field in recent years.This paper uses deep learning method Deep Belief Network(Deep Belief Network,DBN)to detect and locate the leakage problem of heating pipe network.The main research contents include:Firstly,combine the graph theory algorithm to abstract the heating pipe network into points and edges.The form is expressed in the form of a matrix.The operating data of the pipe network can be obtained by solving the corresponding mathematical equations,which provides a mathematical basis for the hydraulic calculation of the heating pipe network leakage conditions.Secondly,taking the central heating secondary pipe network as the research object,under the quality adjustment condition of constant and small supply flow,a two-level diagnosis model of the leakage fault of the heating pipe network based on the deep confidence network is established.The pressure change value of the pressure monitoring point in the pipeline network is used as the input variable of the diagnostic model.The first-level model realizes the identification of the leaking pipe section,and then establishes its own unique second-level diagnostic model for each leaking pipe section to determine the location of the leak point.The model was verified using branched pipe network and looped pipe network respectively.In the actual engineering experiment,a heat exchange station in a certain area is selected as the experimental object,and the validity of the established model is verified from the three aspects of leakage pipe section,leakage location,and leakage rate.Finally,based on the proposed leak diagnosis method of the heating pipe network,the unique visual expression and spatial analysis functions of the geographic information system are integrated into the management of the heating system,and when the leakage accident occurs,an intuitive image display and Analyze the leakage accident.