|
1. A method for safety inspection based on a smart gas geographic information system (GIS), wherein the method is implemented by an Internet of Things system for safety inspection based on the smart gas GIS, the Internet of Things system includes a smart gas user platform, a smart gas service platform, a smart gas pipeline network safety management platform, a smart gas pipeline network device sensing network platform, and a smart gas pipeline network device object platform, wherein the smart gas user platform includes a terminal device, the terminal device is configured to feed back information to a user, receive an inspection instruction, and obtain feedback data input by the user, the smart gas service platform includes a first server, the first server is configured to convey user requirements and control information and send the user requirements and the control information to the terminal device, the smart gas pipeline network safety management platform includes a second server and a smart gas data center, the second server is configured to obtain a geographical feature, a pipeline feature, and pipeline network historical data of a pas pipeline network, and a count of gas pipeline network grades and determine the inspection plan, the smart gas data center is configured to collect and store operation data of the Internet of Things system, the smart gas pipeline network device sensing network platform includes a communication network and a gateway, the communication network and the gateway are configured to implement perceptual information sensing communication and control the information sensing communication, the smart gas pipeline network safety management platform performs data interaction with the smart gas pipeline network device sensing network platform and the smart gas service platform through the smart gas data center, and the smart gas pipeline network device object platform includes various types of gas pipeline network devices and monitoring devices, the monitoring devices include temperature and humidity sensors of gas metering devices, the method comprising: obtaining the geographical feature, the pipeline feature, and the pipeline network historical data of the gas pipeline network through the various types of gas pipeline network devices and the monitoring devices of the smart gas pipeline network device object platform, wherein the gas pipeline network is a pipeline distribution network formed by gas transmission pipelines and devices, the geographic feature refers to a feature of a geographic location where the gas pipeline network is located, the geographic feature includes a terrain environment and a location coordinate, the pipeline feature includes a material of the gas pipeline network, inner and outer diameters of the pipeline gas pipeline network, a count of gas pipeline branches, the pipeline network historical data includes a historical failure possibility and a historical impact range of the gas pipeline network, determining a gridding result through performing gridding processing on the gas pipeline network by a gridding algorithm based on a preset period based on the geographical feature, the pipeline feature, and the pipeline network historical data of the gas pipeline network, wherein the gridding result refers to each of a plurality of divided grid areas after the gridding processing and evaluation result corresponding to each of the plurality of divided grid areas, and the gridding processing refers to a process in which an inspection area of the gas pipeline network is divided by gridding based on a preset gridding standard and a plurality of divided inspection grid areas are evaluated based on a factor including safety of a gas pipeline and a surrounding environment to determine the evaluation result; determining a required inspection area and an optional inspection area based on the gridding result, wherein the required inspection area refers to an area with multiple failures have occurred in the past frequent failures in a short term, and the optional inspection area refers to an area that has already been inspected in the short term and has no failures; generating a target inspection route based on the required inspection area and the optional inspection area, including: determining a grid safety failure possibility by processing an inspection time interval, an accident time interval, the geographical feature, and a gas density in a grid based on a possibility prediction layer, wherein the gas density in the grid is obtained by the monitoring devices of the smart gas pipeline network device object platform, the possibility prediction layer is a deep neural network model, the possibility prediction layer is obtained by training through a gradient descent algorithm, a first training sample for training the possibility prediction layer is determined based on historical data, and a first label for training the possibility prediction layer is a subsequent actual failure possibility corresponding to the first training sample; determining a grid safety score by retrieving in a score preset table through a safety score evaluation layer based on the grid safety failure possibility, wherein the score preset table refers to a table storing grid safety failure possibilities of a plurality different sets of nodes or edges and grid safety scores corresponding to the grid safety failure possibility, and the score preset table is preset according to historical data; and determining whether a grid area is the required inspection area or the optional inspection area by retrieving in an area preset table through a judgment layer based on the grid safety score and a gas pipeline network grade; determining an inspection plan based on the target inspection route, wherein the inspection plan includes an inspection time, an inspection route, the required inspection area, and the optional inspection area; generating the inspection instruction through a preset generation algorithm based on the inspection plan, uploading the inspection instruction to the first server of the smart gas service platform through the communication network and the gateway of the smart gas pipeline network device sensing network platform, and sending the inspection instruction to the terminal device used by an inspection person corresponding to the smart gas user wherein the preset generation algorithm is a machine learning model capable of generating the inspection instruction through training; updating a gas pipeline grid graph based on inspection data after inspecting each inspection area, wherein the inspection data is obtained from the terminal device of the smart gas user platform through the smart gas pipeline network device sensing network platform, wherein the inspection data includes environmental information and gas pipeline failure situation, the gas pipeline grid graph refers to a graph used to represent a connection relationship of the gas pipeline network, the gas pipeline grid graph includes at least one node and at least one edge, wherein the node represents each grid area, and the edge represents a connection between adjacent grid areas, wherein a node feature of the gas pipeline grid graph includes the geographical feature of the grid area, whether a gas pipeline has been inspected, and an edge feature of the gas pipeline grid graph includes an inter-grid correlation coefficient, wherein the inter-grid correlation coefficient refers to a parameter that measures a degree of influence between grid areas at both ends of the edge and the grid area; determining at least one failure possibility of the at least one node or the at least one edge through a failure possibility determination model based on an updated gas pipeline grid graph, wherein the failure possibility determination model is a graph neural network model, the failure possibility determination model is obtained based on the gradient descent algorithm, a second training sample of the failure possibility determination model includes a sample gas pipeline grid graph, and a second label is an actual failure possibility of each node/edge in the sample gas pipeline grid graph, the second training sample and the second label is determined based on the historical data; determining a grid safety score by retrieving in a score preset table based on the at least one failure possibility of the at least one node or the at least one edge through the safety score evaluation layer, and determining an updated inspection plan based on the grid safety score.
|
