| CPC G05B 23/0283 (2013.01) [G16Y 10/35 (2020.01)] | 8 Claims |
|
3. A system for regulating pipeline network maintenance based on a smart gas Internet of Things (IoT), comprising a smart gas user platform, a smart gas service platform, a smart gas safety management platform, a smart gas sensing network platform, and a smart gas object platform interacting in sequence; wherein
the smart gas safety management platform includes a smart gas emergency maintenance management sub-platform and a smart gas data center;
the smart gas sensing network platform is configured to interact with the smart gas data center and the smart gas object platform;
the smart gas user platform is configured as a terminal device, the smart gas object platform is configured as various types of gas-related devices, the gas-related devices include a monitoring device and a regulation device; the monitoring device includes a gas meter, a flow meter, a manometer, a temperature sensor, a humidity sensor, and a pressure sensor; the smart gas object platform is configured to transmit collected information through the smart gas sensing network platform to the smart gas data center;
the system further comprises:
a non-transitory computer-readable storage medium storing a computer instruction; and
at least one processor in communicate with the non-transitory computer-readable storage medium; when executing the computer instruction, the at least one processor is directed to cause the system to:
obtain gas data of at least one faulty pipeline from the gas meter, the flow meter, and the manometer of the smart gas object platform based on the smart gas sensing network platform, wherein the gas data includes at least one of a gas flow rate, a gas flow volume, and a pipeline pressure of the at least one faulty pipeline;
generate at least one set of candidate maintenance schemes based on the at least one faulty pipeline;
predict future pipeline data of the at least one faulty pipeline at a future time point through a prediction model based on a first pipeline network graph and the at least one set of candidate maintenance schemes, the prediction model being a graph neural networks (GNN); wherein the prediction model is obtained by training based on first training samples and first labels, the first training samples are historical first pipeline network graphs constructed based on historical data from a first time period, and the first labels are actual pipeline data of a set of monitoring point locations in the historical data from a second time period; wherein the first time period precedes the second time period, the second time period is a future time period of the first time period, and the training includes:
inputting the first training samples with the first labels into an initial prediction model;
constructing a loss function based on outputs of the initial prediction model and the first labels;
updating parameters of the initial prediction model based on the loss function until meeting preset conditions; and
obtaining the prediction model;
determine a recommended maintenance approach for the at least one faulty pipeline based on the future pipeline data and a preset fault condition; wherein the recommended maintenance approach includes a non-stop transmission with pressure maintenance approach and a stop transmission maintenance approach; the preset fault condition is a maintenance fault probability of each faulty pipeline corresponding to each candidate maintenance scheme that is less than a fault threshold; the fault threshold is inversely related to an importance degree of the at least one faulty pipeline; wherein the importance degree of the at least one faulty pipeline is determined based on a pipeline image, a node of the pipeline image includes a pipeline node and a user node, the importance degree of the at least one faulty pipeline is determined by weighting based on importance degrees of users corresponding to user nodes of different routes, wherein a weighting coefficient is a route coefficient of the pipeline image and the larger the route coefficient, the higher the importance degree of the at least one faulty pipeline, the route coefficient is positively related to a route length of the pipeline image, the route length refers to adjacency between a starting node and an ending node of a route of the pipeline image, wherein the adjacency refers to a count of nodes that need to be traversed in the pipeline image from one node to another node; the future pipeline data includes a maintenance fault type and a maintenance fault probability of the at least one faulty pipeline;
determine the maintenance fault probability based on a preset algorithm in response to a determination that a count of the maintenance fault type is greater than or equal to two; wherein a final maintenance fault probability of the at least one faulty pipeline is determined by weighting maintenance fault probabilities corresponding to different maintenance fault types of the at least one faulty pipeline using the preset algorithm; wherein a weighting coefficient is obtained from a coefficient table; and the coefficient table is a table that reflects combinations of maintenance fault types and corresponding coefficients; when a plurality of candidate maintenance schemes meet the preset fault condition, determining a candidate maintenance scheme with a maximum count of non-stop transmission with pressure maintenance approaches as the recommended maintenance approach;
determine a set of monitoring point locations based on the recommended maintenance approach, wherein the set of monitoring point locations includes a set of preset point locations and a set of expanded point locations;
determine a regulation parameter of a regulation device based on monitoring data of the set of monitoring point locations obtained in real time or intermittently from a plurality of gas meters, flow meters, manometers configured in the smart gas object platform, wherein an output of a regulation parameter determination model includes a preset identifier, and suspend a construction scheme and re-determining the recommended maintenance approach in response to a determination that the output of the regulation parameter determination model is the same as the preset identifier;
determine the construction scheme based on the recommended maintenance approach;
obtain the monitoring data through the monitoring device based on the construction scheme; and
determine, based on a second pipeline network graph, the regulation parameter using the regulation parameter determination model, wherein the regulation parameter determination model is a graph neural networks (GNN), a node of the second pipeline network graph includes a gas pipeline, the monitoring device, and the regulation device, and a node feature includes a gas pipeline node feature, a monitoring device node feature, and a regulation device node feature; and the monitoring device node feature of the node includes the monitoring data; an edge of the second pipeline network graph includes a first-class edge, a second-class edge, and a third-class edge, and the edge is a directed edge; the first-class edge includes a connection relationship between the gas pipelines; a direction of the first-class edge signifies a direction of gas flow within the gas pipelines; an edge feature of the first-class edge represents a distance between gas pipeline nodes; the second-class edge includes a connection relationship between the monitoring device and the gas pipelines; and an edge feature of the second-class edge represents a monitoring relationship between the monitoring device and the gas pipelines; the third-class edge represents a connection relationship between the regulation device and the gas pipelines, and an edge feature of the third-class edge represents a regulation relationship between the regulation device and the gas pipelines; wherein
the regulation parameter determination model is obtained by training using a second training samples and second labels; the second training samples include a sample second pipeline network graph, and the second labels include an actual regulation parameter and the preset identifier corresponding to the second training samples;
at regular intervals, determine the regulation parameter using the regulation parameter determination model;
the smart gas service platform is configured to send the regulation parameter to the smart gas user platform; and
perform real-time regulation of the regulation device by regulating the gas flow rate, a gas flow volume, and a pipeline pressure of the at least one faulty pipeline and a related pipeline based on the determination of the regulation parameter.
|