Method and system for determining maintenance time of pipe networks of natural gas

Abstract

The present disclosure provides a method and a system for determining a maintenance time of a pipe network of natural gas. The method may comprise: obtaining pipe network information of natural gas in at least one area, the pipe network information including a running time of the system and gas leakage information of the pipe network; extracting feature information based on the running time and the gas leakage information; generating a pipe network maintenance value through a maintenance value prediction model based on pipe network maintenance information and pipe network environment information, the pipe network maintenance value reflecting a priority of pipe network maintenance processing; and predicting the maintenance time of the pipe network based on the feature information and the pipe network maintenance value using a maintenance time prediction model, the maintenance time prediction model being a machine learning model.

What is claimed is: 
     
         1 . A method for determining a maintenance time of a pipe network of natural gas, implemented by a processor, the method comprising:
 obtaining pipe network information of natural gas in at least one area, the pipe network information including a running time of a system for determining a maintenance time of a pipe network of natural gas and gas leakage information of the pipe network;   extracting feature information based on the running time and the gas leakage information;   generating a pipe network maintenance value through a maintenance value prediction model based on pipe network maintenance information and pipe network environment information, the pipe network maintenance value reflecting a priority of pipe network maintenance processing; wherein
 the maintenance value prediction model is a Graph Neural Network model, a plurality of nodes of the Graph Neural Network model include a plurality of historical maintenance locations of the pipe network and historical pipe network environmental information, a plurality of edges of the Graph Neural Network model include one or more pipes between the plurality of historical maintenance locations of the pipe network, features of the nodes include at least one of replacement pipe material, a maintenance time, a maintenance location, gas leakage after maintenance, a vibration detection result, a vibration frequency of the pipe network, and natural gas usage environment information, and features of the edges include at least one of pipe material, a diameter, a connection manner, and a relationship between the historical pipe network environment information and the historical maintenance locations of the pipe network; and 
   predicting the maintenance time of the pipe network based on the feature information and the pipe network maintenance value using a maintenance time prediction model, the maintenance time prediction model being a machine learning model.   
     
     
         2 . The method of  claim 1 , wherein the pipe network environment information comprises at least one of the vibration frequency of the pipe network or the natural gas usage environment information; and
 the pipe network maintenance information comprises at least one of the replacement pipe material, the maintenance time, the maintenance location, the gas leakage after maintenance, or the vibration detection result.   
     
     
         3 . The method of  claim 2 , wherein the vibration frequency of the pipe network includes a natural frequency of a pipe and an external vibration frequency, the natural frequency of the pipe is a frequency of vibration generated due to changes in an elbow or a diameter of the pipe, or due to a flow of the natural gas, and the external vibration frequency is a frequency of vibration caused by a surrounding construction site, traffic, or an unstable pipe support. 
     
     
         4 . The method of  claim 1 , wherein an input of the maintenance time prediction model further comprises a vibration fatigue factor of the pipe network; and the vibration fatigue factor of the pipe network is calculated based on the vibration frequency of the pipe network and a vibration time of the pipe network. 
     
     
         5 . The method of  claim 4 , wherein the vibration fatigue factor of the pipe network is determined by processing the vibration frequency of the pipe network, the vibration time of the pipe network, and pipe material strength using a second model, the vibration fatigue factor of the pipe network reflects strength of pipe fatigue due to vibration, and the second model is a machine learning model. 
     
     
         6 . The method of  claim 5 , wherein the second model is obtained by training based on a historical vibration frequency of the pipe network, a historical vibration time of the pipe network, and historical pipe material strength. 
     
     
         7 . The method of  claim 1 , wherein the maintenance value prediction model is obtained by training based on training data, the training data includes feature information corresponding to a historical running time and historical gas leakage information, a historical pipe network maintenance value, and a historical vibration fatigue factor of the pipe network; the historical vibration fatigue factor of the pipe network is determined based on a historical vibration frequency of the pipe network and a historical vibration time of the pipe network, and a label of the training data is a maintenance time of the pipe network corresponding to the training data. 
     
     
         8 . A system for determining a maintenance time of a pipe network of natural gas, comprising:
 an information obtaining module configured to obtain pipe network information of natural gas in at least one area, the pipe network information including a running time of the system and gas leakage information of the pipe network;   a feature extracting module configured to extract feature information based on the running time and the gas leakage information;   a time prediction module configured to:   generate a pipe network maintenance value through a maintenance value prediction model based on pipe network maintenance information and pipe network environment information, the pipe network maintenance value reflecting a priority of pipe network maintenance processing; wherein
 the maintenance value prediction model is a Graph Neural Network model, a plurality of nodes of the Graph Neural Network model include a plurality of historical maintenance locations of the pipe network and historical pipe network environmental information, a plurality of edges of the Graph Neural Network model include one or more pipes between the plurality of historical maintenance locations of the pipe network, features of the nodes include at least one of replacement pipe material, a maintenance time, a maintenance location, gas leakage after maintenance, a vibration detection result, a vibration frequency of the pipe network, and natural gas usage environment information, and features of the edges include at least one of pipe material, a diameter, a connection manner, and a relationship between the historical pipe network environment information and the historical maintenance locations of the pipe network; and 
   predict the maintenance time of the pipe network based on the feature information and the pipe network maintenance value using a maintenance time prediction model, the maintenance time prediction model being a machine learning model.   
     
     
         9 . The system of  claim 8 , wherein the pipe network environment information comprises at least one of the vibration frequency of the pipe network or the natural gas usage environment information; and
 the pipe network maintenance information comprises at least one of the replacement pipe material, the maintenance time, the maintenance location, the gas leakage after maintenance, or the vibration detection result.   
     
     
         10 . The system of  claim 9 , wherein the vibration frequency of the pipe network includes a natural frequency of a pipe and an external vibration frequency, the natural frequency of the pipe is a frequency of vibration generated due to changes in an elbow or a diameter of the pipe, or due to a flow of the natural gas, and the external vibration frequency is a frequency of vibration caused by a surrounding construction site, traffic, or an unstable pipe support. 
     
     
         11 . The system of  claim 8 , wherein an input of the maintenance time prediction model further comprises a vibration fatigue factor of the pipe network; and the vibration fatigue factor of the pipe network is calculated based on the vibration frequency of the pipe network and a vibration time of the pipe network. 
     
     
         12 . The system of  claim 11 , wherein the vibration fatigue factor of the pipe network is determined by processing the vibration frequency of the pipe network, the vibration time of the pipe network, and pipe material strength using a second model, the vibration fatigue factor of the pipe network reflects strength of pipe fatigue due to vibration, and the second model is a machine learning model. 
     
     
         13 . The system of  claim 12 , wherein the second model is obtained by training based on a historical vibration frequency of the pipe network, a historical vibration time of the pipe network, and historical pipe material strength. 
     
     
         14 . The system of  claim 8 , wherein the maintenance value prediction model is obtained by training based on training data, the training data includes feature information corresponding to a historical running time and historical gas leakage information, a historical pipe network maintenance value, and a historical vibration fatigue factor of the pipe network; the historical vibration fatigue factor of the pipe network is determined based on a historical vibration frequency of the pipe network and a historical vibration time of the pipe network, and a label of the training data is a maintenance time of the pipe network corresponding to the training data. 
     
     
         15 . A non-transitory computer readable medium storing instructions, when executed by at least one processor, causing the at least one processor to implement a method comprising:
 obtaining pipe network information of natural gas in at least one area, the pipe network information including a running time of a system for determining a maintenance time of a pipe network of natural gas and gas leakage information of the pipe network;   extracting feature information based on the running time and the gas leakage information;   generating a pipe network maintenance value through a maintenance value prediction model based on pipe network maintenance information and pipe network environment information, the pipe network maintenance value reflecting a priority of pipe network maintenance processing; wherein
 the maintenance value prediction model is a Graph Neural Network model, a plurality of nodes of the Graph Neural Network model include a plurality of historical maintenance locations of the pipe network and historical pipe network environmental information, a plurality of edges of the Graph Neural Network model include one or more pipes between the plurality of historical maintenance locations of the pipe network, features of the nodes include at least one of replacement pipe material, a maintenance time, a maintenance location, gas leakage after maintenance, a vibration detection result, a vibration frequency of the pipe network, and natural gas usage environment information, and features of the edges include at least one of pipe material, a diameter, a connection manner, and a relationship between the historical pipe network environment information and the historical maintenance locations of the pipe network; and 
   predicting the maintenance time of the pipe network based on the feature information and the pipe network maintenance value using a maintenance time prediction model, the maintenance time prediction model being a machine learning model.