Oil and gas pipeline data classification transmission method and system based on wireless communication

Abstract

The invention discloses an oil and gas pipeline data classification transmission method and system based on wireless communication, relates to the technical field of digital information classification transmission, and aims at a long-distance oil and gas pipeline wireless monitoring scene, multiple features are extracted at edge nodes, divided into A / B / C / D classes and mapped to different bearing slices; by combining multiple paths and opportunity links such as a line patrol unmanned aerial vehicle, emergency alarm and rapid diagnosis data are transmitted in a priority, reliable and low-delay manner, stable redundant data are compressed and transmitted in a low-priority manner, and narrowband and satellite link bandwidth occupation and node energy consumption are reduced while the leakage and damage event discovery speed and diagnosis precision are improved.

Description

Technical Field

[0001] This invention relates to the field of digital information classification and transmission technology, and in particular to a method and system for classification and transmission of oil and gas pipeline data based on wireless communication. Background Technology

[0002] Long-distance oil and gas pipelines often span hundreds of kilometers, traversing uninhabited deserts, hilly and mountainous areas, and sparsely populated regions. For large sections, wired communication links cannot be laid, and monitoring data can only be transmitted back to the control center via solar-powered edge communication nodes and relay nodes through narrowband wireless, public cellular networks, or satellite links. To ensure pipeline safety, pressure sensors, flow meters, acoustic emission sensors, accelerometers, strain gauges, cathodic protection potential measuring points, temperature and humidity sensors, and camera terminals are typically deployed along the pipeline to generate multi-source raw sensing data. While existing technologies can utilize wireless networks to upload this raw sensing data to the central terminal, they generally treat all monitoring data as the same service type, queuing them in the same wireless link and transmission queue. They lack data classification mechanisms based on event urgency and diagnostic value, cannot spatially reconfigure resources according to the risk level and link quality of different pipeline sections, and for high-bandwidth raw sensing data such as acoustic emission, vibration, and high-frame-rate video, they often only upload low-frequency statistics and rarely incorporate temporary high-bandwidth opportunistic links formed by patrol drones and patrol vehicles into the overall transmission design. Furthermore, they fail to automatically rearrange the transmission queue and prioritize the transmission of critical diagnostic data when opportunistic links occur.

[0003] Currently, Chinese invention patent application number 202210991659.2 discloses a dynamic transmission and loading method and system for multi-business data of oil and gas pipelines. The dynamic transmission and loading method includes: automatically configuring the number of data transmission channels, concurrency, and threads based on the size and data generation speed of the multi-business data of oil and gas pipelines; collecting or synchronizing the multi-business data of oil and gas pipelines to obtain dynamically synchronized data; storing the dynamically synchronized data in a dynamic data storage module; reading the dynamically synchronized data and providing a dynamic data sharing interface; and loading the dynamically synchronized data through the dynamic data sharing interface in the upper-layer application to realize the data processing of multi-business data of oil and gas pipelines. This invention, by setting up a dynamic data loading intermediate service, can not only solve the problem of unified data collection for multi-source heterogeneous business systems with high timeliness, large data volume, and diverse structures, but also streamline the design of the upper-layer application, alleviate the dynamic loading pressure on the upper-layer application, and facilitate the later maintenance of the upper-layer application.

[0004] The aforementioned technologies are insufficient to perform fine-grained classification and cross-layer resource allocation based on the urgency and diagnostic value of data in the wireless communication environment of long-distance oil and gas pipelines with limited bandwidth, thereby ensuring low-latency and reliable transmission of emergency alarms and critical diagnostic data while taking into account overall bandwidth and energy consumption. Summary of the Invention

[0005] The technical problem solved by this invention is that existing technologies are difficult to perform fine classification and cross-layer resource allocation based on the urgency and diagnostic value of data in the wireless communication environment of long-distance oil and gas pipelines with limited bandwidth, so as to simultaneously ensure low-latency and reliable transmission of emergency alarms and key diagnostic data while taking into account overall bandwidth and energy consumption.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: A method for classifying and transmitting oil and gas pipeline data based on wireless communication includes the following steps: Step S1: Construct the pipeline topology along the pipeline based on the oil and gas pipeline location and wireless coverage information, divide the oil and gas pipeline into several pipe segments, assign a pipe segment identifier to each pipe segment, and send a synchronization clock to the terminal in the pipe segment. In step S2, the sensing terminal in the pipe section collects raw sensing data according to the synchronous clock and sends it to the edge communication node. The edge communication node performs time alignment and preprocessing on the raw sensing data to form a raw sensing dataset, calculates features and outputs them as feature data. Step S3: The edge communication node analyzes the feature data and corresponding associated data according to the preset data classification rule set, determines the data category, priority, latency constraint and reliability level, generates classification labels and organizes them into a classification label table. Step S4: Divide the bearer slices on the wireless interface according to the classification label table, map different categories of data to the corresponding bearer slices, and set queue scheduling parameters; Step S5: Combining pipeline topology and queue scheduling parameters, select transmission paths for each bearer slice and perform multi-hop relay or multi-path aggregation transmission to obtain transmission results; In step S6, the central end adaptively updates the data classification rule set and the resource allocation of each bearer slice based on the transmission results and historical alarm records.

[0007] Preferably, step S1 includes the following sub-steps: Step S101: Obtain the location information of oil and gas pipelines and the coverage information of base stations, satellites or public cellular networks, and divide the oil and gas pipelines into several segments according to the wireless reachability and terrain conditions to form the segment division result; Step S102: Deploy edge communication nodes and relay nodes along the pipeline, and associate the edge communication nodes with the pipeline segments according to the pipeline segment division results to construct the pipeline topology along the pipeline. Step S103: Assign a pipe segment identifier to each pipe segment, and broadcast the pipe segment identifier and synchronization clock to the sensing terminal of the corresponding pipe segment by the corresponding edge communication node.

[0008] Preferably, step S2 includes the following sub-steps: Step S201: Each pipe segment sensing terminal collects raw sensing data according to the synchronization clock. The raw sensing data includes pressure data, flow data, acoustic emission data, vibration data, strain data, cathodic protection potential data, ambient temperature and humidity data, and video image data, and sends them to the corresponding edge communication node along with the pipe segment identifier to form the raw sensing dataset. In step S202, the edge communication node aligns the original sensing dataset according to the timestamp and pipe segment identifier, calculates pressure gradient features, flow change features, acoustic emission energy features, vibration frequency band features, strain growth features, cathodic protection potential shift features, and image motion features based on the original sensing data. Based on the pre-calibrated pipe corridor area mask of each camera terminal, the image motion features are calculated only within the pipe corridor area mask. An index relationship is established between the original sensing dataset and the pressure gradient features, flow change features, acoustic emission energy features, vibration frequency band features, strain growth features, cathodic protection potential shift features, and image motion features. Step S203: Output the pressure gradient features, flow change features, acoustic emission energy features, vibration frequency band features, strain growth features, cathodic protection potential shift features, image motion features, pipe segment identification and timestamp as feature data, and store the feature data together with the original sensing dataset and index relationship in the edge communication node.

[0009] Preferably, step S3 includes the following sub-steps: Step S301: A data classification rule set is pre-set in the edge communication node. The data classification rule set provides the determination conditions for data category, priority, latency constraint and reliability level based on the correspondence between features and historical events. Step S302: The edge communication node analyzes each feature data and its associated original sensing data in the original sensing dataset according to the data classification rule set. When any of the following combinations of conditions are met simultaneously, the corresponding data will be marked as Class A emergency alarm data: The pressure gradient characteristics exceed the severe pressure anomaly threshold within several consecutive sampling periods, and the flow rate mutation characteristics are lower than the flow rate drop threshold. The acoustic emission energy characteristics show a sudden increase near the inherent frequency band of the pipe structure. Or the vibration frequency band characteristics continuously exceed the vibration threshold corresponding to illegal mechanical operations in the low frequency band, and the image motion characteristics show that there is continuous large target movement within the pipeline corridor; Or the strain growth characteristics show a step-like increase within a preset length of pipe section and exceed the strain abnormality threshold, and the cathodic protection potential deviation characteristics continue to deviate from the normal range. When the event urgency score is lower than the Class A threshold, but the diagnostic value score is higher than the preset diagnostic threshold, and the corresponding original perception data is high-sampling-rate acoustic emission data, high-sampling-rate vibration data, or video image data related to the area around the pipe, the image motion feature is the motion vector density calculated based on the pre-calibrated pipe corridor area mask of the corresponding pipe segment in the original video image captured by the camera terminal. The pipe corridor area mask represents the set of pixels in the image corresponding to the area around the pipe. The corresponding data is marked as Class B rapid diagnostic data, and the original perception data within the preset time window before and after the event occurs is retained. When the pressure gradient feature, flow change feature, strain growth feature and cathodic protection potential shift feature corresponding to the feature data are all within their respective normal ranges, and only contain the mean or extreme value obtained by statistics at fixed time intervals, the corresponding data will be marked as Class C periodic state data. When all features corresponding to the feature data are in the normal range, and the difference between the data and the original sensing data of the same pipe section in multiple consecutive time windows is lower than the preset steady-state difference threshold, the corresponding data will be marked as D-type redundant data. Step S303: Generate classification labels containing category, priority, delay constraint and reliability level for Class A emergency alarm data, Class B rapid diagnosis data, Class C periodic status data and Class D redundant data, and organize them into a classification label table.

[0010] Preferably, step S4 includes the following sub-steps: Step S401: The edge communication node divides multiple bearer slices on the wireless interface according to the classification label table. The bearer slices include an emergency bearer slice for carrying Class A emergency alarm data, a diagnostic bearer slice for carrying Class B rapid diagnostic data, and a periodic bearer slice for carrying Class C periodic status data and Class D redundant data. Step S402: Map Class A emergency alarm data to emergency bearer slices, Class B rapid diagnostic data to diagnostic bearer slices, Class C periodic status data and Class D redundant data to periodic bearer slices, and set queue modulation and coding schemes, maximum retransmission counts and reserved transmission time slots for emergency bearer slices, diagnostic bearer slices and periodic bearer slices according to the priority, delay constraints and reliability levels in the classification labels to form queue scheduling parameters; In step S403, edge communication nodes queue and transmit according to queue scheduling parameters. Emergency bearer slices obtain fixed high-priority transmission resources, diagnostic bearer slices obtain elastic bandwidth as needed, and periodic bearer slices use opportunistic transmission when the link is idle.

[0011] Preferably, step S5 includes the following sub-steps: Step S501: The relay node periodically collects link quality information, which includes the signal-to-noise ratio, retransmission rate and duty cycle of each hop transmission path, and feeds back the link quality information to the corresponding edge communication node. In step S502, the edge communication node selects both ground multi-hop transmission path and satellite or public cellular transmission path for multi-path copying and transmission of Class A emergency alarm data and selected Class B rapid diagnostic data based on the pipeline topology, classification label table, queue scheduling parameters and link quality information. For Class C periodic status data and Class D redundant data, the edge communication node prioritizes ground multi-hop transmission path for batch transmission and records the transmission results of each type of data at the central end. Step S503: When the inspection drone or inspection vehicle arrives at a certain pipe section and establishes a temporary high-bandwidth opportunity link with the corresponding edge communication node, the edge communication node rearranges the diagnostic bearer slice queue according to the queuing time of the Class B rapid diagnostic data in the queue and the latency constraints in the classification label table, and prioritizes sending the Class B rapid diagnostic data that meets the conditions through the opportunity link.

[0012] Preferably, step S6 includes the following sub-steps: Step S601: The central end calculates the end-to-end delay distribution, packet loss rate and retransmission count of different types of data in each segment based on the transmission results, and forms the link performance evaluation result. Step S602: The central end performs correlation analysis between the link performance evaluation results and historical alarm records, and counts the number of false alarms, the number of missed alarms, and the fault location accuracy of Class A emergency alarm data and Class B rapid diagnosis data to form alarm effect evaluation results; In step S603, the central terminal generates data classification rule set adjustment instructions and resource allocation adjustment instructions for each pipe segment based on the alarm effect evaluation results. The data classification rule set adjustment instructions are used to adjust the mapping conditions from features to data categories and priorities. The resource allocation adjustment instructions are used to adjust the bandwidth ratio, modulation and coding scheme and maximum retransmission count of each bearer slice, and are sent to the corresponding edge communication nodes through the control channel.

[0013] Preferably, the logic for establishing the data classification rule set in step S301 is as follows: The urgency score of an event is calculated based on the correspondence between acoustic emission energy characteristics, vibration frequency band characteristics, strain growth characteristics and historical leakage events. The diagnostic value score is calculated based on the contributions of pressure gradient characteristics, flow rate mutation characteristics, and cathodic protection potential offset characteristics to the fault location accuracy. Calculate the bandwidth usage score based on the amount of raw sensing data and the sampling frequency in the raw sensing dataset. Calculate the reconfigurability score based on the time window in which the original sensing data can be repeatedly constructed in the edge communication node; The urgency score, diagnostic value score, bandwidth usage score, and reconfigurability score are weighted and superimposed to form a comprehensive score. Based on the comprehensive score range, the feature data and its associated original sensing data are respectively classified into Category A emergency alarm data, Category B rapid diagnostic data, Category C periodic status data, and Category D redundant data.

[0014] Preferably, the opportunity link scheduling logic in step S503 is as follows: When a line inspection drone or line inspection vehicle establishes a temporary high-bandwidth opportunity link, the edge communication node first reads the classification label table and queue scheduling parameters, calculates the queuing time and corresponding event occurrence time of each B-type rapid diagnostic data in the queue, and compares it with the time delay constraints specified in the classification label. If a certain type B rapid diagnostic data cannot be sent within the preset diagnostic delay in the original diagnostic bearer slice, then the type B rapid diagnostic data will be migrated from the diagnostic bearer slice queue to the opportunity link queue, and the sending priority of type C periodic status data and type D redundant data in the same pipe segment will be reduced, so that the opportunity link bandwidth is used first to upload the original sensing data and its feature data of the most recent event.

[0015] The oil and gas pipeline data classification and transmission system based on wireless communication includes a pipeline topology construction module, a feature data calculation module, a correlation data analysis module, a scheduling parameter setting module, a transmission operation implementation module, and a central alarm update module. The pipeline topology construction module is used to construct the pipeline topology along the pipeline based on the oil and gas pipeline location and wireless coverage information, divide the oil and gas pipeline into several pipe segments, assign a pipe segment identifier to each pipe segment and send a synchronization clock to the terminal in the pipe segment. The feature data calculation module is used by the sensing terminal in the pipe section to collect raw sensing data according to the synchronous clock and send it to the edge communication node. The edge communication node performs time alignment and preprocessing on the raw sensing data to form a raw sensing dataset, calculates features and outputs them as feature data. The associated data analysis module is used by edge communication nodes to analyze feature data and corresponding associated data according to a preset data classification rule set, determine data category, priority, latency constraint and reliability level, generate classification labels and organize them into a classification label table; The scheduling parameter setting module is used to divide the bearer slices on the wireless interface according to the classification label table, map different categories of data to the corresponding bearer slices, and set queue scheduling parameters. The transmission operation implementation module is used to combine pipeline topology and queue scheduling parameters to select a transmission path for each bearer slice and perform multi-hop relay or multi-path aggregation transmission to obtain the transmission result. The central alarm update module is used by the central end to adaptively update the data classification rule set and the resource allocation of each bearer slice based on the transmission results and historical alarm records.

[0016] The beneficial effects of this invention are as follows: This invention is designed for wireless monitoring scenarios of long-distance oil and gas pipelines. It extracts multiple features at edge nodes and classifies them into A / B / C / D categories and maps them to different bearer slices. Combined with multi-path and opportunistic links such as pipeline inspection drones, it enables priority, reliable, and low-latency transmission of emergency alarm and rapid diagnostic data. Stable redundant data is compressed and sent with low priority. While improving the speed of detection and diagnostic accuracy of leakage and damage events, it reduces the bandwidth occupation of narrowband and satellite links and the energy consumption of nodes. Attached Figure Description

[0017] Figure 1 A flowchart illustrating the steps of a method for classifying and transmitting oil and gas pipeline data based on wireless communication, provided in one embodiment of the present invention. Figure 2 This is a basic flowchart of a wireless communication-based oil and gas pipeline data classification and transmission system provided in one embodiment of the present invention. Detailed Implementation

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0019] Example 1, referring to Figure 1 This paper provides a method for classifying and transmitting oil and gas pipeline data based on wireless communication, including the following steps: Step S1: Construct the pipeline topology along the pipeline based on the pipeline location and wireless coverage information, divide the oil and gas pipeline into several pipe segments, assign a pipe segment identifier to each pipe segment, and send a synchronization clock to the terminal in the pipe segment.

[0020] In step S2, the sensing terminal in the pipe section collects raw sensing data according to the synchronous clock and sends it to the edge communication node. The edge communication node performs time alignment and preprocessing on the raw sensing data to form a raw sensing dataset, calculates features and outputs them as feature data.

[0021] Step S3: The edge communication node analyzes the feature data and corresponding associated data according to the preset data classification rule set, determines the data category, priority, latency constraint and reliability level, generates classification labels and organizes them into a classification label table.

[0022] Step S4: Divide the bearer slices on the wireless interface according to the classification label table, map different categories of data to the corresponding bearer slices, and set queue scheduling parameters.

[0023] Step S5: Combining pipeline topology and queue scheduling parameters, select transmission paths for each bearer slice and perform multi-hop relay or multi-path aggregation transmission to obtain transmission results.

[0024] In step S6, the central end adaptively updates the data classification rule set and the resource allocation of each bearer slice based on the transmission results and historical alarm records.

[0025] Step S1 includes the following sub-steps: Step S101: Obtain the location information of oil and gas pipelines and the coverage information of base stations, satellites or public cellular networks. Divide the oil and gas pipelines into several segments according to the wireless reachability and terrain conditions to form the segment division result.

[0026] Step S102: Deploy edge communication nodes and relay nodes along the pipeline, and associate the edge communication nodes with the pipeline segments according to the pipeline segment division results to construct the pipeline topology along the pipeline.

[0027] Step S103: Assign a pipe segment identifier to each pipe segment, and broadcast the pipe segment identifier and synchronization clock to the sensing terminal of the corresponding pipe segment by the corresponding edge communication node.

[0028] In this embodiment, the pipeline's alignment information, along-route topography information, operator base station coverage information, and satellite coverage information are first acquired. Based on the wireless coverage radius of the edge communication nodes and terrain obstruction, the pipeline is divided into n segments, resulting in a segmentation outcome.

[0029] For each pipe segment, one edge communication node and several relay nodes are deployed in the corresponding valve chamber or encryption point to construct the pipeline topology along the line through a short-range wireless ad hoc network. Each pipe segment is assigned a unique pipe segment identifier "Seg01~Segn". The edge communication node has a built-in clock synchronization module. After being synchronized with the timekeeping satellite or the upper-level clock, it periodically sends the synchronization clock and the corresponding pipe segment identifier to the sensor terminals within the pipe segment in the form of broadcast frames, so that the raw sensing data collected by each sensor terminal carries a unified pipe segment identifier and timestamp.

[0030] Step S2 includes the following sub-steps: In step S201, each pipe segment sensing terminal collects raw sensing data according to the synchronization clock. The raw sensing data includes pressure data, flow data, acoustic emission data, vibration data, strain data, cathodic protection potential data, ambient temperature and humidity data, and video image data, and sends them along with the pipe segment identifier to the corresponding edge communication node to form the raw sensing dataset.

[0031] In step S202, the edge communication node aligns the original sensing dataset according to the timestamp and pipe segment identifier, calculates pressure gradient features, flow change features, acoustic emission energy features, vibration frequency band features, strain growth features, cathodic protection potential shift features, and image motion features based on the original sensing data. Based on the pre-calibrated pipe corridor area mask of each camera terminal, the image motion features are calculated only within the pipe corridor area mask. An index relationship is established between the original sensing dataset and the pressure gradient features, flow change features, acoustic emission energy features, vibration frequency band features, strain growth features, cathodic protection potential shift features, and image motion features.

[0032] Step S203: Output the pressure gradient features, flow change features, acoustic emission energy features, vibration frequency band features, strain growth features, cathodic protection potential shift features, image motion features, pipe segment identification and timestamp as feature data, and store the feature data together with the original sensing dataset and index relationship in the edge communication node.

[0033] In this embodiment, pressure sensors, flow meters, acoustic emission sensors, accelerometers, strain gauges, cathodic protection potential measuring points, temperature and humidity sensors, and fixed camera terminals are installed along each pipe section. Each sensing terminal is connected to the corresponding edge communication node through a short-range wireless link.

[0034] Under the control of the synchronous clock, the sampling frequency of each sensor is set, and the sensing terminal sends the raw sensing data with pipe segment identification and sampling timestamp to the corresponding edge communication node to form the raw sensing dataset.

[0035] The edge communication nodes align the original sensing dataset by timestamp and pipe segment identifier, and calculate the following features by data type within each fixed time window: Pressure gradient characteristics: Within this time window, the absolute value of the pressure difference between adjacent sampling points is taken and averaged to obtain the average pressure gradient, which is used to reflect the rate of pressure change. Traffic flow mutation characteristics: The maximum drop and maximum rise in traffic flow within the time window are statistically analyzed, and the maximum value is used to characterize the traffic flow mutation characteristics; Acoustic emission energy characteristics: The waveform of the acoustic emission sensor within the window is enveloped and the energy is integrated within the preset tube's inherent frequency band to obtain the acoustic emission energy characteristics; Vibration frequency band characteristics: Perform spectrum analysis on the acceleration signal, calculate the frequency band energy in the low frequency band representing mechanical operation and the mid-to-high frequency band representing pipe vibration, and construct the vibration frequency band characteristics in the form of energy ratio; Strain growth characteristics: Compare strain values ​​at the beginning and end of the window to obtain strain increment, and divide the increment by the window duration to obtain the strain growth rate, which is used as the strain growth characteristic. Cathodic protection potential deviation characteristics: The difference between the mean cathodic protection potential and the historical normal potential value is calculated within the window and used as the cathodic protection potential deviation characteristics; Image motion features: For each camera terminal, a mask for the pipe corridor region is pre-defined in the video frame during the calibration phase. This mask corresponds to the set of pixels in the frame related to the area around the pipe. During operation, inter-frame difference or optical flow calculations are performed on adjacent frames within the mask region to count the number and average amplitude of motion vectors. This statistical result is used as the image motion feature, and only motion information within the mask region is retained.

[0036] The edge communication node establishes an index relationship between the original sensing dataset and the aforementioned features, ensuring that the corresponding original sensing data can be quickly located based on the feature data. Then, each feature, along with the pipe segment identifier and timestamp, is encapsulated into feature data, and the feature data and index relationship are stored in the edge communication node.

[0037] Step S3 includes the following sub-steps: Step S301: A data classification rule set is pre-set in the edge communication node. The data classification rule set provides the determination conditions for data category, priority, latency constraint and reliability level based on the correspondence between features and historical events.

[0038] The logic for establishing the data classification rule set in step S301 is as follows: The urgency score of an event is calculated based on the correspondence between acoustic emission energy characteristics, vibration frequency band characteristics, strain growth characteristics, and historical leakage events.

[0039] The diagnostic value score is calculated based on the contributions of pressure gradient characteristics, flow change characteristics, and cathodic protection potential offset characteristics to the fault location accuracy.

[0040] The bandwidth usage score is calculated based on the amount of raw sensing data and the sampling frequency in the raw sensing dataset.

[0041] The reconfigurability score is calculated based on the time window that can be repeatedly constructed in the edge communication node using the original sensing data.

[0042] The urgency score, diagnostic value score, bandwidth usage score, and reconfigurability score are weighted and superimposed to form a comprehensive score. Based on the comprehensive score range, the feature data and its associated original sensing data are respectively classified into Category A emergency alarm data, Category B rapid diagnostic data, Category C periodic status data, and Category D redundant data.

[0043] Step S302: The edge communication node analyzes each feature data and its associated original sensing data in the original sensing dataset according to the data classification rule set. When any of the following combinations of conditions are met simultaneously, the corresponding data will be marked as Class A emergency alarm data: The pressure gradient characteristics exceed the severe pressure anomaly threshold within several consecutive sampling periods, and the flow rate mutation characteristics are lower than the flow rate drop threshold. The acoustic emission energy characteristics show a sudden increase near the inherent frequency band of the pipeline structure.

[0044] Or the vibration frequency band characteristics continuously exceed the vibration threshold corresponding to illegal mechanical operations in the low frequency band, and the image motion characteristics show that there is continuous large target movement within the pipeline corridor.

[0045] Or the strain growth characteristics show a step-like increase within a preset length of pipe section and exceed the strain abnormality threshold, and the cathodic protection potential deviation characteristics continue to deviate from the normal range.

[0046] When the event urgency score is lower than the Class A threshold, but the diagnostic value score is higher than the preset diagnostic threshold, and the corresponding original sensing data is high-sampling-rate acoustic emission data, high-sampling-rate vibration data, or video image data related to the area around the pipe, the image motion feature is the motion vector density calculated based on the pre-calibrated pipe corridor area mask of the corresponding pipe segment in the original video image captured by the camera terminal. The pipe corridor area mask represents the set of pixels in the image corresponding to the area around the pipe. The corresponding data is marked as Class B rapid diagnostic data, and the original sensing data within the preset time window before and after the event occurs is retained.

[0047] When the pressure gradient feature, flow change feature, strain growth feature, and cathodic protection potential shift feature corresponding to the feature data are all within their respective normal ranges, and only contain the mean or extreme value obtained by statistics at fixed time intervals, the corresponding data will be marked as Class C periodic state data.

[0048] When all features corresponding to the feature data are within the normal range, and the difference between the data and the original sensing data of the same pipe section in multiple consecutive time windows is lower than the preset steady-state difference threshold, the corresponding data will be marked as D-type redundant data.

[0049] Step S303: Generate classification labels containing category, priority, delay constraint and reliability level for Class A emergency alarm data, Class B rapid diagnosis data, Class C periodic status data and Class D redundant data, and organize them into a classification label table.

[0050] In this embodiment, a data classification rule set is pre-set in the edge communication node. First, based on historical operational data and records of abnormal events caused by past leaks, illegal excavations, and geological disasters, the typical ranges of the above characteristics when various events occur are statistically analyzed to determine the normal range, alarm threshold range, and severe abnormality threshold range. Weights are then assigned to pressure gradient characteristics, flow mutation characteristics, acoustic emission energy characteristics, vibration frequency band characteristics, strain growth characteristics, cathodic protection potential shift characteristics, and image motion characteristics for subsequent calculation of event urgency scores and diagnostic value scores.

[0051] In this embodiment, the data classification rule set is defined with the following decision logic: Category A Emergency Alarm Data: When the feature data within a certain time window meets any of the following conditions, the edge communication node marks the feature data and its associated original sensing data as Category A emergency alarm data: The pressure gradient characteristics exceeded the severe pressure anomaly threshold within three consecutive windows, and the flow rate mutation characteristics showed a significant decrease within the same time period. The acoustic emission energy characteristics showed a continuous surge near the pipe's inherent frequency band. The vibration frequency band features show that the energy in the low frequency band continuously exceeds the vibration threshold corresponding to illegal mechanical operations, and the motion vector density of the image motion features in the mask of the pipeline corridor area exceeds the threshold for large targets, indicating that there is continuous large mechanical or vehicle movement near the pipeline corridor. The strain growth characteristics increase in a stepwise manner along the pipeline direction at several adjacent measuring points and exceed the strain anomaly threshold. At the same time, the cathodic protection potential shift characteristics continuously deviate from the normal range within the same pipe section.

[0052] Category B rapid diagnostic data: When the event urgency score calculated from the feature data is lower than the Category A threshold, but the diagnostic value score (weighted by acoustic emission energy features, vibration frequency band features, strain growth features, and image motion features) is higher than the preset diagnostic threshold, and the associated original sensing data is high-sampling-rate acoustic emission data, high-sampling-rate vibration data, or video image data from within the mask in the pipeline corridor area, the edge communication node marks the feature data and its associated original sensing data as Category B rapid diagnostic data, and additionally retains the original sensing data within a preset time window before and after the event for detailed analysis at the central end.

[0053] Type C periodic state data: When the pressure gradient feature, flow change feature, strain growth feature and cathodic protection potential shift feature corresponding to the feature data are all within their respective normal ranges, and only contain statistical quantities such as the mean, maximum and minimum values ​​obtained from the original sensing data according to a fixed sampling period, without containing the high-frequency original waveform, the edge communication node marks the data as Type C periodic state data.

[0054] Type D redundant data: When all features corresponding to the feature data are in the normal range, and the difference between the original sensing data of the same pipe segment and the original sensing data of the same pipe segment is lower than the preset steady-state difference threshold in multiple consecutive time windows, the edge communication node will mark the corresponding original sensing data as Type D redundant data, and only retain the compressed statistical summary and no longer retain the original waveform.

[0055] Edge communication nodes generate classification labels for each piece of feature data, including data category, priority, latency constraints, and reliability level. For example, category A emergency alarm data is assigned the highest priority, shortest latency constraint, and highest reliability level, while category D redundant data is assigned the lowest priority and discardable attribute. Subsequently, the classification labels are summarized according to pipeline segments to form a classification label table for use in subsequent bearer slice division and queue scheduling.

[0056] Step S4 includes the following sub-steps: In step S401, the edge communication node divides multiple bearer slices on the wireless interface according to the classification label table. The bearer slices include emergency bearer slices for carrying Class A emergency alarm data, diagnostic bearer slices for carrying Class B rapid diagnostic data, and periodic bearer slices for carrying Class C periodic status data and Class D redundant data.

[0057] Step S402: Map Class A emergency alarm data to emergency bearer slices, Class B rapid diagnostic data to diagnostic bearer slices, and Class C periodic status data and Class D redundant data to periodic bearer slices. Based on the priority, delay constraints, and reliability levels in the classification labels, set the queue modulation and coding scheme, maximum retransmission count, and reserved transmission time slots for emergency bearer slices, diagnostic bearer slices, and periodic bearer slices to form queue scheduling parameters.

[0058] In step S403, edge communication nodes queue and transmit according to queue scheduling parameters. Emergency bearer slices obtain fixed high-priority transmission resources, diagnostic bearer slices obtain elastic bandwidth as needed, and periodic bearer slices use opportunistic transmission when the link is idle.

[0059] In this embodiment, the edge communication node supports multiple wireless bearer methods, including narrowband low-speed bearers (such as LoRa or similar low-power wide-area networks), medium-speed bearers (such as industrial Wi-Fi), and high-speed bearers for accessing public cellular or satellite terminals. Based on a classification label table, the edge communication node divides its local wireless interface into at least three types of bearer slices: Emergency bearer slice: used for transmitting Class A emergency alarm data; Diagnostic carrier slice: used for transmitting Class B rapid diagnostic data; Periodic bearer slice: used to transmit Class C periodic status data and Class D redundant data.

[0060] Edge communication nodes configure queue scheduling parameters for different bearer slices: emergency bearer slices use lower modulation order and strong error correction coding, reserve a fixed proportion of transmission time slots, and support multiple retransmissions; diagnostic bearer slices use medium-to-high modulation order and medium error correction coding, dynamically allocating available bandwidth as needed; periodic bearer slices use higher modulation order and fewer retransmissions, primarily transmitting when the link is idle. In the queue scheduling algorithm, the emergency bearer slice queue has the highest priority, followed by the diagnostic bearer slice, and the periodic bearer slice has the lowest priority, allowing the discarding of some Class D redundant data during link congestion.

[0061] Step S5 includes the following sub-steps: Step S501: The relay node periodically collects link quality information, which includes the signal-to-noise ratio, retransmission rate and duty cycle of each hop transmission path. The link quality information is then fed back to the corresponding edge communication node. Each hop transmission path refers to the link corresponding to each hop in multi-hop transmission.

[0062] In step S502, the edge communication node selects both ground multi-hop transmission path and satellite or public cellular transmission path for multi-path copying and transmission of Class A emergency alarm data and selected Class B rapid diagnostic data based on the pipeline topology, classification label table, queue scheduling parameters and link quality information. For Class C periodic status data and Class D redundant data, the edge communication node prioritizes the ground multi-hop transmission path for batch transmission and records the transmission results of each type of data at the central end. The ground multi-hop transmission path refers to the transmission route or link path formed by data being forwarded hop by hop through two or more relay nodes in the ground wireless network.

[0063] Step S503: When the inspection drone or inspection vehicle arrives at a certain pipe section and establishes a temporary high-bandwidth opportunity link with the corresponding edge communication node, the edge communication node rearranges the diagnostic bearer slice queue according to the queuing time of the Class B rapid diagnostic data in the queue and the latency constraints in the classification label table, and prioritizes sending the Class B rapid diagnostic data that meets the conditions through the opportunity link.

[0064] The opportunity link scheduling logic in step S503 is as follows: When a line inspection drone or line inspection vehicle establishes a temporary high-bandwidth opportunity link, the edge communication node first reads the classification label table and queue scheduling parameters, calculates the queuing time and corresponding event occurrence time of each B-class rapid diagnostic data in the queue, and compares it with the latency constraints specified in the classification label.

[0065] If a certain type B rapid diagnostic data cannot be sent within the preset diagnostic delay in the original diagnostic bearer slice, then the type B rapid diagnostic data will be migrated from the diagnostic bearer slice queue to the opportunity link queue, and the sending priority of type C periodic status data and type D redundant data in the same pipe segment will be reduced, so that the opportunity link bandwidth is used first to upload the original sensing data and its feature data of the most recent event.

[0066] In this embodiment, relay nodes periodically collect link quality information such as signal-to-noise ratio, retransmission rate, and duty cycle for each hop transmission path and feed it back to the corresponding edge communication nodes. The edge communication nodes, considering the pipeline topology, classification label table, queue scheduling parameters, and link quality information, simultaneously select one terrestrial multi-hop transmission path and one satellite or public cellular transmission path for both Category A emergency alarm data and some Category B rapid diagnostic data, employing a multi-path replication method for transmission. The central end processes the data based on its first arrival, discarding subsequent duplicate data to shorten end-to-end latency and improve the reliability of critical data transmission.

[0067] For Class C periodic status data and Class D redundant data, this embodiment prioritizes sending them in batches via ground multi-hop transmission paths when the link is idle. Only when a certain segment cannot establish a stable ground link for a long time will the critical Class C periodic status data be retransmitted in batches via satellite or public cellular channels. Class D redundant data can be discarded according to the strategy.

[0068] When a patrol drone or patrol vehicle equipped with a wireless communication terminal enters a high-risk pipe section and establishes a temporary high-bandwidth opportunity link with the edge communication node of that pipe section, the edge communication node rearranges the transmission queue of the diagnostic bearer slice according to the latency constraints in the classification label table and the queuing time of Class B rapid diagnostic data in the queue. It migrates Class B rapid diagnostic data that is about to exceed the diagnostic latency threshold to the opportunity link transmission queue, and prioritizes uploading the original acoustic emission waveform, vibration waveform and video images around the pipe body within the corresponding time window through the opportunity link. At the same time, it reduces the transmission priority of Class C periodic status data and Class D redundant data in this pipe section.

[0069] Step S6 includes the following sub-steps: In step S601, the central end calculates the end-to-end delay distribution, packet loss rate, and retransmission count of different types of data in each segment based on the transmission results, and forms the link performance evaluation results.

[0070] In step S602, the central end performs correlation analysis between the link performance evaluation results and historical alarm records, and counts the number of false alarms, the number of missed alarms, and the fault location accuracy of Class A emergency alarm data and Class B rapid diagnostic data to form alarm effect evaluation results.

[0071] In step S603, the central terminal generates data classification rule set adjustment instructions and resource allocation adjustment instructions for each pipe segment based on the alarm effect evaluation results. The data classification rule set adjustment instructions are used to adjust the mapping conditions from features to data categories and priorities. The resource allocation adjustment instructions are used to adjust the bandwidth ratio, modulation and coding scheme and maximum retransmission number of each bearer slice, and are sent to the corresponding edge communication nodes through the control channel.

[0072] In this embodiment, the central system summarizes the transmission results of each category of data for each pipeline segment according to a fixed statistical period (e.g., daily or hourly), and statistically analyzes the end-to-end latency distribution, packet loss rate, and retransmission count of Category A emergency alarm data and Category B rapid diagnostic data to form a link performance evaluation result. The central system also correlates the above evaluation results with historical alarm records, statistically analyzes the number of false alarms and missed alarms of Category A emergency alarm data in different pipeline segments, and the spatial error of fault location based on Category B rapid diagnostic data to form an alarm effectiveness evaluation result.

[0073] Based on the link performance evaluation results and alarm effect evaluation results, the central terminal generates data classification rule set adjustment instructions and resource allocation adjustment instructions for each pipe segment. For example, when a pipe segment experiences excessive latency or high packet loss rate of Class A emergency alarm data for an extended period, the central terminal increases the priority of Class A emergency alarm data for that segment, increases the bandwidth proportion occupied by emergency bearer slices, and appropriately lowers the alarm thresholds for acoustic emission energy characteristics and vibration frequency band characteristics in the data classification rule set to improve anomaly detection sensitivity. When a pipe segment remains in a stable state for a long time and has very little Class A and Class B data, the central terminal reduces the resource proportion of emergency bearer slices and diagnostic bearer slices for that segment, increases the resource proportion of periodic bearer slices, and raises the steady-state difference threshold to reduce the storage and transmission volume of Class D redundant data.

[0074] The aforementioned adjustment instructions are sent to the corresponding edge communication nodes via the downlink control channel. The edge communication nodes update their local data classification rule sets and queue scheduling parameters online, completing the policy adjustment without downtime and restart, thus achieving adaptive optimization of data classification and transmission strategies based on the actual operating conditions of the pipeline.

[0075] Example 2, refer to Figure 2 It provides a data classification and transmission system for oil and gas pipelines based on wireless communication, including a pipeline topology construction module, a feature data calculation module, a correlation data analysis module, a scheduling parameter setting module, a transmission operation implementation module, and a central alarm update module.

[0076] The pipeline topology construction module is used to construct the pipeline topology along the pipeline based on the oil and gas pipeline location and wireless coverage information, divide the oil and gas pipeline into several pipe segments, assign a pipe segment identifier to each pipe segment, and send a synchronization clock to the terminal in the pipe segment.

[0077] The feature data calculation module is used by the sensing terminal in the pipe section to collect raw sensing data according to the synchronous clock and send it to the edge communication node. The edge communication node performs time alignment and preprocessing on the raw sensing data to form a raw sensing dataset, calculates features and outputs them as feature data.

[0078] The associated data analysis module is used by edge communication nodes to analyze feature data and corresponding associated data according to a preset data classification rule set, determine data category, priority, latency constraint and reliability level, generate classification labels and organize them into a classification label table.

[0079] The scheduling parameter setting module is used to divide the bearer slices on the wireless interface according to the classification label table, map different types of data to the corresponding bearer slices, and set queue scheduling parameters.

[0080] The transmission operation implementation module is used to combine pipeline topology and queue scheduling parameters to select transmission paths for each bearer slice and perform multi-hop relay or multi-path aggregation transmission to obtain transmission results.

[0081] The central alarm update module is used by the central end to adaptively update the data classification rule set and the resource allocation of each bearer slice based on the transmission results and historical alarm records.

[0082] By constructing a raw sensing dataset from raw sensing data at edge communication nodes, pressure gradient characteristics, flow mutation characteristics, acoustic emission energy characteristics, vibration frequency band characteristics, strain growth characteristics, cathodic protection potential shift characteristics, and image motion characteristics are calculated. A data classification rule set is established to clearly classify the data into Class A emergency alarm data, Class B rapid diagnostic data, Class C periodic status data, and Class D redundant data. At the same time, classification labels and classification label tables are generated for each type of data. This distinguishes the importance and timeliness of different services throughout the entire process from the acquisition end to the transmission end, avoiding the coarse-grained processing of all data in the existing technology.

[0083] This invention divides the wireless interface into emergency bearer slices, diagnostic bearer slices, and periodic bearer slices based on a classification label table. Corresponding queue scheduling parameters are configured for each bearer slice, ensuring that Category A emergency alarm data receives fixed bandwidth, strong error correction coding, and the highest transmission priority on the emergency bearer slice; Category B rapid diagnostic data receives on-demand elastic bandwidth on the diagnostic bearer slice; and Category C periodic status data and Category D redundant data are transmitted opportunistically on the periodic bearer slice. By binding data categories to bearer slices and queue scheduling parameters, cross-layer collaborative allocation from the feature layer, service layer, to the underlying wireless resource layer is achieved, significantly reducing queuing latency and packet loss risk of critical data on the wireless link.

[0084] Based on pipeline topology and link quality information, this invention simultaneously selects ground multi-hop transmission paths and satellite or public cellular transmission paths for both Class A emergency alarm data and some Class B rapid diagnostic data. It employs a multi-path replication method for transmission, with the central end receiving only the first arriving data. This reduces end-to-end latency of emergency data and improves transmission reliability. When patrol drones or patrol vehicles establish temporary high-bandwidth opportunity links in high-risk pipeline sections, edge communication nodes automatically rearrange the transmission queue based on latency constraints in the classification labels and the queuing time of Class B rapid diagnostic data. Data about to expire is migrated to the opportunity link for priority transmission. This ensures timely transmission of high-sampling-rate acoustic emissions, vibration waveforms, and video images around the pipeline even in scenarios with extremely limited bandwidth resources. This provides sufficient data support for accident tracing and detailed diagnosis at the central end, something difficult to achieve with traditional static wireless transmission strategies.

[0085] The central system periodically analyzes the end-to-end latency distribution, packet loss rate, and retransmission count of different categories of data for each pipe segment based on transmission results. This data is then correlated with historical alarm records to obtain link performance evaluation results and alarm effectiveness evaluation results. Based on these evaluation results, data classification rule set adjustment instructions and resource allocation adjustment instructions are generated and sent to edge communication nodes to update local data classification rule sets and queue scheduling parameters online. This enables the system to continuously adjust the classification boundaries and carrying resource allocation of Class A, B, C, and D data according to changes in pipeline operating conditions, risk distribution, and the wireless environment. This forms a closed-loop adaptive optimization mechanism aimed at improving safety monitoring effectiveness, avoiding the long-term performance degradation caused by static configuration based on manual experience in existing technologies.

[0086] By classifying highly repetitive, stable-range raw sensing data into Class D redundant data, retaining only necessary statistical summaries, and transmitting them opportunistically with low priority in periodic bearer slices or selectively discarding them during link congestion, this invention significantly reduces the amount of invalid data that needs to be transmitted over narrowband wireless links and satellite links. Simultaneously, by employing a batch-based, low-retransmission strategy for Class C periodic status data, the consumption of wireless bandwidth and node power is further reduced. While ensuring the quality of Class A emergency alarm data and Class B rapid diagnostic data transmission, this invention effectively extends the battery life of solar-powered edge communication nodes and relay nodes, improving the overall operational efficiency and engineering feasibility of the long-distance oil and gas pipeline wireless monitoring system.

[0087] In summary, this invention, in the application scenario of wireless monitoring of long-distance oil and gas pipelines, solves the problems of difficulty in guaranteeing alarm data latency, difficulty in timely uploading of key diagnostic data under limited bandwidth, and mismatch between wireless resources and pipeline risk distribution in existing technologies through multi-source feature-driven data classification, cross-layer mapping of carrier slices, joint transmission of multi-path and opportunistic links, and closed-loop adaptive optimization at the central end. It has good engineering application value and promotion significance.

[0088] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0089] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the protection scope of the present invention.

Claims

1. A method for classifying and transmitting oil and gas pipeline data based on wireless communication, characterized in that, Includes the following steps: Step S1: Construct the pipeline topology along the pipeline based on the oil and gas pipeline location and wireless coverage information, divide the oil and gas pipeline into several pipe segments, assign a pipe segment identifier to each pipe segment, and send a synchronization clock to the terminal in the pipe segment. In step S2, the sensing terminal in the pipe section collects raw sensing data according to the synchronous clock and sends it to the edge communication node. The edge communication node performs time alignment and preprocessing on the raw sensing data to form a raw sensing dataset, calculates features and outputs them as feature data. Step S3: The edge communication node analyzes the feature data and corresponding associated data according to the preset data classification rule set, determines the data category, priority, latency constraint and reliability level, generates classification labels and organizes them into a classification label table. Step S4: Divide the bearer slices on the wireless interface according to the classification label table, map different categories of data to the corresponding bearer slices, and set queue scheduling parameters; Step S5: Combining pipeline topology and queue scheduling parameters, select transmission paths for each bearer slice and perform multi-hop relay or multi-path aggregation transmission to obtain transmission results; In step S6, the central end adaptively updates the data classification rule set and the resource allocation of each bearer slice based on the transmission results and historical alarm records.

2. The method for classifying and transmitting oil and gas pipeline data based on wireless communication as described in claim 1, characterized in that, Step S1 includes the following sub-steps: Step S101: Obtain the location information of oil and gas pipelines and the coverage information of base stations, satellites or public cellular networks, and divide the oil and gas pipelines into several segments according to the wireless reachability and terrain conditions to form the segment division result; Step S102: Deploy edge communication nodes and relay nodes along the pipeline, and associate the edge communication nodes with the pipeline segments according to the pipeline segment division results to construct the pipeline topology along the pipeline. Step S103: Assign a pipe segment identifier to each pipe segment, and broadcast the pipe segment identifier and synchronization clock to the sensing terminal of the corresponding pipe segment by the corresponding edge communication node.

3. The method for classifying and transmitting oil and gas pipeline data based on wireless communication as described in claim 2, characterized in that, Step S2 includes the following sub-steps: Step S201: Each pipe segment sensing terminal collects raw sensing data according to the synchronization clock. The raw sensing data includes pressure data, flow data, acoustic emission data, vibration data, strain data, cathodic protection potential data, ambient temperature and humidity data, and video image data, and sends them to the corresponding edge communication node along with the pipe segment identifier to form the raw sensing dataset. In step S202, the edge communication node aligns the original sensing dataset according to the timestamp and pipe segment identifier, calculates pressure gradient features, flow change features, acoustic emission energy features, vibration frequency band features, strain growth features, cathodic protection potential shift features, and image motion features based on the original sensing data. Based on the pre-calibrated pipe corridor area mask of each camera terminal, the image motion features are calculated only within the pipe corridor area mask. An index relationship is established between the original sensing dataset and the pressure gradient features, flow change features, acoustic emission energy features, vibration frequency band features, strain growth features, cathodic protection potential shift features, and image motion features. Step S203: Output the pressure gradient features, flow change features, acoustic emission energy features, vibration frequency band features, strain growth features, cathodic protection potential shift features, image motion features, pipe segment identification and timestamp as feature data, and store the feature data together with the original sensing dataset and index relationship in the edge communication node.

4. The method for classifying and transmitting oil and gas pipeline data based on wireless communication as described in claim 3, characterized in that, Step S3 includes the following sub-steps: Step S301: A data classification rule set is pre-set in the edge communication node. The data classification rule set provides the determination conditions for data category, priority, latency constraint and reliability level based on the correspondence between features and historical events. Step S302: The edge communication node analyzes each feature data and its associated original sensing data in the original sensing dataset according to the data classification rule set. When any of the following combinations of conditions are met simultaneously, the corresponding data will be marked as Class A emergency alarm data: The pressure gradient characteristics exceed the severe pressure anomaly threshold within several consecutive sampling periods, and the flow rate mutation characteristics are lower than the flow rate drop threshold. The acoustic emission energy characteristics show a sudden increase near the inherent frequency band of the pipe structure. Or the vibration frequency band characteristics continuously exceed the vibration threshold corresponding to illegal mechanical operations in the low frequency band, and the image motion characteristics show that there is continuous large target movement within the pipeline corridor; Or the strain growth characteristics show a step-like increase within a preset length of pipe section and exceed the strain abnormality threshold, and the cathodic protection potential deviation characteristics continue to deviate from the normal range. When the event urgency score is lower than the Class A threshold, but the diagnostic value score is higher than the preset diagnostic threshold, and the corresponding original perception data is high-sampling-rate acoustic emission data, high-sampling-rate vibration data, or video image data related to the area around the pipe, the image motion feature is the motion vector density calculated based on the pre-calibrated pipe corridor area mask of the corresponding pipe segment in the original video image captured by the camera terminal. The pipe corridor area mask represents the set of pixels in the image corresponding to the area around the pipe. The corresponding data is marked as Class B rapid diagnostic data, and the original perception data within the preset time window before and after the event occurs is retained. When the pressure gradient feature, flow change feature, strain growth feature and cathodic protection potential shift feature corresponding to the feature data are all within their respective normal ranges, and only contain the mean or extreme value obtained by statistics at fixed time intervals, the corresponding data will be marked as Class C periodic state data. When all features corresponding to the feature data are in the normal range, and the difference between the data and the original sensing data of the same pipe section in multiple consecutive time windows is lower than the preset steady-state difference threshold, the corresponding data will be marked as D-type redundant data. Step S303: Generate classification labels containing category, priority, delay constraint and reliability level for Class A emergency alarm data, Class B rapid diagnosis data, Class C periodic status data and Class D redundant data, and organize them into a classification label table.

5. The method for classifying and transmitting oil and gas pipeline data based on wireless communication as described in claim 4, characterized in that, Step S4 includes the following sub-steps: Step S401: The edge communication node divides multiple bearer slices on the wireless interface according to the classification label table. The bearer slices include an emergency bearer slice for carrying Class A emergency alarm data, a diagnostic bearer slice for carrying Class B rapid diagnostic data, and a periodic bearer slice for carrying Class C periodic status data and Class D redundant data. Step S402: Map Class A emergency alarm data to emergency bearer slices, Class B rapid diagnostic data to diagnostic bearer slices, Class C periodic status data and Class D redundant data to periodic bearer slices, and set queue modulation and coding schemes, maximum retransmission counts and reserved transmission time slots for emergency bearer slices, diagnostic bearer slices and periodic bearer slices according to the priority, delay constraints and reliability levels in the classification labels to form queue scheduling parameters; In step S403, edge communication nodes queue and transmit according to queue scheduling parameters. Emergency bearer slices obtain fixed high-priority transmission resources, diagnostic bearer slices obtain elastic bandwidth as needed, and periodic bearer slices use opportunistic transmission when the link is idle.

6. The method for classifying and transmitting oil and gas pipeline data based on wireless communication as described in claim 5, characterized in that, Step S5 includes the following sub-steps: Step S501: The relay node periodically collects link quality information, which includes the signal-to-noise ratio, retransmission rate and duty cycle of each hop transmission path, and feeds back the link quality information to the corresponding edge communication node. In step S502, the edge communication node selects both ground multi-hop transmission path and satellite or public cellular transmission path for multi-path copying and transmission of Class A emergency alarm data and selected Class B rapid diagnostic data based on the pipeline topology, classification label table, queue scheduling parameters and link quality information. For Class C periodic status data and Class D redundant data, the edge communication node prioritizes ground multi-hop transmission path for batch transmission and records the transmission results of each type of data at the central end. Step S503: When the inspection drone or inspection vehicle arrives at a certain pipe section and establishes a temporary high-bandwidth opportunity link with the corresponding edge communication node, the edge communication node rearranges the diagnostic bearer slice queue according to the queuing time of the Class B rapid diagnostic data in the queue and the latency constraints in the classification label table, and prioritizes sending the Class B rapid diagnostic data that meets the conditions through the opportunity link.

7. The method for classifying and transmitting oil and gas pipeline data based on wireless communication as described in claim 6, characterized in that, Step S6 includes the following sub-steps: Step S601: The central end calculates the end-to-end delay distribution, packet loss rate and retransmission count of different types of data in each segment based on the transmission results, and forms the link performance evaluation result. Step S602: The central end performs correlation analysis between the link performance evaluation results and historical alarm records, and counts the number of false alarms, the number of missed alarms, and the fault location accuracy of Class A emergency alarm data and Class B rapid diagnosis data to form alarm effect evaluation results; In step S603, the central terminal generates data classification rule set adjustment instructions and resource allocation adjustment instructions for each pipe segment based on the alarm effect evaluation results. The data classification rule set adjustment instructions are used to adjust the mapping conditions from features to data categories and priorities. The resource allocation adjustment instructions are used to adjust the bandwidth ratio, modulation and coding scheme and maximum retransmission count of each bearer slice, and are sent to the corresponding edge communication nodes through the control channel.

8. The method for classifying and transmitting oil and gas pipeline data based on wireless communication as described in claim 7, characterized in that, The logic for establishing the data classification rule set in step S301 is as follows: The urgency score of an event is calculated based on the correspondence between acoustic emission energy characteristics, vibration frequency band characteristics, strain growth characteristics and historical leakage events. The diagnostic value score is calculated based on the contributions of pressure gradient characteristics, flow rate mutation characteristics, and cathodic protection potential offset characteristics to the fault location accuracy. Calculate the bandwidth usage score based on the amount of raw sensing data and the sampling frequency in the raw sensing dataset. Calculate the reconfigurability score based on the time window in which the original sensing data can be repeatedly constructed in the edge communication node; The urgency score, diagnostic value score, bandwidth usage score, and reconfigurability score are weighted and superimposed to form a comprehensive score. Based on the comprehensive score range, the feature data and its associated original sensing data are respectively classified into Category A emergency alarm data, Category B rapid diagnostic data, Category C periodic status data, and Category D redundant data.

9. The method for classifying and transmitting oil and gas pipeline data based on wireless communication as described in claim 8, characterized in that, The opportunity link scheduling logic in step S503 is as follows: When a line inspection drone or line inspection vehicle establishes a temporary high-bandwidth opportunity link, the edge communication node first reads the classification label table and queue scheduling parameters, calculates the queuing time and corresponding event occurrence time of each B-type rapid diagnostic data in the queue, and compares it with the time delay constraints specified in the classification label. If a certain type B rapid diagnostic data cannot be sent within the preset diagnostic delay in the original diagnostic bearer slice, then the type B rapid diagnostic data will be migrated from the diagnostic bearer slice queue to the opportunity link queue, and the sending priority of type C periodic status data and type D redundant data in the same pipe segment will be reduced, so that the opportunity link bandwidth is used first to upload the original sensing data and its feature data of the most recent event.

10. A wireless communication-based oil and gas pipeline data classification and transmission system, applied in the wireless communication-based oil and gas pipeline data classification and transmission method as described in any one of claims 1-9, characterized in that, It includes a pipeline topology construction module, a feature data calculation module, a correlation data analysis module, a scheduling parameter setting module, a transmission operation implementation module, and a central alarm update module; The pipeline topology construction module is used to construct the pipeline topology along the pipeline based on the oil and gas pipeline location and wireless coverage information, divide the oil and gas pipeline into several pipe segments, assign a pipe segment identifier to each pipe segment and send a synchronization clock to the terminal in the pipe segment. The feature data calculation module is used by the sensing terminal in the pipe section to collect raw sensing data according to the synchronous clock and send it to the edge communication node. The edge communication node performs time alignment and preprocessing on the raw sensing data to form a raw sensing dataset, calculates features and outputs them as feature data. The associated data analysis module is used by edge communication nodes to analyze feature data and corresponding associated data according to a preset data classification rule set, determine data category, priority, latency constraint and reliability level, generate classification labels and organize them into a classification label table; The scheduling parameter setting module is used to divide the bearer slices on the wireless interface according to the classification label table, map different categories of data to the corresponding bearer slices, and set queue scheduling parameters. The transmission operation implementation module is used to combine pipeline topology and queue scheduling parameters to select a transmission path for each bearer slice and perform multi-hop relay or multi-path aggregation transmission to obtain the transmission result. The central alarm update module is used by the central end to adaptively update the data classification rule set and the resource allocation of each bearer slice based on the transmission results and historical alarm records.

Images