Risk detection method and system for natural gas station based on visual detection

Abstract

The invention discloses a risk detection method and system for a natural gas station based on visual detection, and relates to the technical field of visual detection, and the method comprises the steps: determining corresponding risk features based on the detection of images shot by each visual detector; according to the form of each risk feature, the current position and the environment type of the natural gas station, the plurality of risk events of the natural gas station are estimated, and the detection accuracy of the plurality of risk events of the natural gas station is improved. Therefore, the risk management and control area of the natural gas station is determined according to the plurality of risk events and the overall form of the natural gas station, and the risk coefficient of the risk management and control area is marked; and determining a corresponding risk management and control path according to the area position of the risk management and control area, the risk coefficient and the natural gas station risk management and control logic, thereby realizing accurate management and control of the risk management and control path, and performing corresponding risk management and control measures on the risk management and control area. And the risk detection effect and the risk management and control effect of the natural gas station are further improved.

Description

Technical Field

[0001] This invention relates to the technical field of visual inspection, and more particularly to a risk detection method and system for natural gas stations based on visual inspection. Background Technology

[0002] With the development of technology, natural gas stations have become a key component of natural gas transmission systems, used for receiving, storing, pressurizing, and transmitting natural gas. These stations store large quantities of natural gas, which is then transported along pipelines laid at various locations within the station. Current technology establishes multiple risk monitoring points at these stations, determining the station's risk status based on data from these monitoring points. However, this approach does not consider the risk characteristics of on-site images of the natural gas station, affecting the accuracy of risk event detection and neglecting precise control of risk management pathways. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art. This invention provides a risk detection method and system for natural gas stations based on visual inspection.

[0004] This invention provides a risk detection method for natural gas stations based on visual inspection, comprising: Collect a distribution map of natural gas stations; Based on the distribution map of natural gas stations, multiple risk monitoring points were identified, and corresponding visual detectors were configured at these points. Based on the detection of images captured by each visual detector, the corresponding risk characteristics are determined, and multiple risk events of the natural gas station are predicted according to the shape of each risk characteristic, its current location, and the environmental type of the natural gas station. The risk control area of ​​the natural gas station is determined based on multiple risk events and the overall shape of the natural gas station, and the risk coefficient of the risk control area is marked. Based on the regional location, risk coefficient, and risk management logic of the risk management area, the corresponding risk management path is determined, and corresponding risk management measures are implemented in the risk management area along the risk management path.

[0005] This invention provides a vision-based risk detection system for natural gas stations. The vision-based risk detection system is applied to the aforementioned vision-based risk detection method for natural gas stations. The vision-based risk detection system for natural gas stations includes: The data acquisition module is used to collect distribution maps of natural gas stations; The risk monitoring point module is used to determine multiple risk monitoring points based on the distribution map of natural gas stations, and to configure corresponding visual detectors at multiple risk monitoring points; The risk event module is used to determine the corresponding risk features based on the detection of images captured by each visual detector, and to predict multiple risk events of the natural gas station based on the shape of each risk feature, its current location, and the environmental type of the natural gas station. The risk management area module is used to determine the risk management area of ​​a natural gas station based on multiple risk events and the overall shape of the natural gas station, and to mark the risk coefficient of the risk management area. The risk management path module is used to determine the corresponding risk management path based on the regional location, risk coefficient, and risk management logic of the risk management area, and to implement corresponding risk management measures along the risk management path in the risk management area.

[0006] Compared with the prior art, the beneficial effects of the present invention are: In this embodiment of the invention, the method determines the corresponding risk features based on the detection of images captured by each visual detector, and predicts multiple risk events of the natural gas station based on the morphology of each risk feature, the current location, and the environmental type of the natural gas station. This method takes into account the overall consideration of the morphology of each risk feature, the current location, and the environmental type of the natural gas station, thereby improving the detection accuracy of multiple risk events of the natural gas station and ensuring the risk detection effect of the natural gas station.

[0007] Therefore, risk control areas of natural gas stations are determined based on multiple risk events and the overall shape of the stations, and risk coefficients of these areas are marked. Corresponding risk control paths are then determined based on the location, risk coefficients, and risk control logic of the natural gas stations. This introduces risk control areas and enables precise control of risk control paths, facilitating corresponding risk control measures and further improving the risk detection and control effectiveness of natural gas stations. Attached Figure Description

[0008] Figure 1 This is a flowchart illustrating the risk detection method for natural gas stations based on visual detection in an embodiment of the present invention. Figure 2 This is a flowchart illustrating step S11 of the risk detection method for natural gas stations based on visual detection in an embodiment of the present invention. Figure 3 This is a flowchart illustrating step S12 in the visual detection-based risk detection method for natural gas stations according to an embodiment of the present invention. Figure 4This is a flowchart illustrating step S13 in the visual detection-based risk detection method for natural gas stations according to an embodiment of the present invention. Figure 5 This is a flowchart illustrating step S14 of the risk detection method for natural gas stations based on visual detection in an embodiment of the present invention. Figure 6 This is a flowchart illustrating step S15 of the risk detection method for natural gas stations based on visual detection in this embodiment of the invention. Figure 7 This is a schematic diagram of the structural composition of a risk detection system for a natural gas station based on visual detection, according to an embodiment of the present invention. Detailed Implementation

[0009] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0010] Please see Figures 1 to 7 A vision-based risk detection method for natural gas stations is proposed and applied to risk detection scenarios at natural gas stations. This vision-based risk detection method for natural gas stations includes: Step S11: Collect a distribution map of natural gas stations; Step S12: Determine multiple risk monitoring points based on the distribution map of the natural gas station, and configure corresponding visual detectors at the multiple risk monitoring points; Step S13: Determine the corresponding risk features based on the detection of images captured by each visual detector, and predict multiple risk events of the natural gas station based on the shape of each risk feature, its current location, and the environmental type of the natural gas station. Step S14: Determine the risk control area of ​​the natural gas station based on multiple risk events and the overall shape of the natural gas station, and mark the risk coefficient of the risk control area; Step S15: Determine the corresponding risk control path based on the regional location, risk coefficient, and risk control logic of the risk control area, and implement corresponding risk control measures along the risk control path in the risk control area; refer to Figure 2 In step S11, a distribution map of natural gas stations is collected; In the specific implementation of this invention, the specific steps are as follows: S111: Collect the name of the natural gas station, determine the data space of the natural gas station based on the name of the natural gas station and the town database, determine the location of the natural gas station based on the traversal of the data space of the natural gas station, at this time, the UAV will detect the location of the natural gas station to collect the current state of the natural gas station. S112: Determine the preset form of the natural gas station based on the traversal of the data space of the natural gas station, and determine the consistency of the natural gas station based on the matching of the preset form and the current form of the natural gas station. At this time, when the preset form and the current form of the natural gas station are consistent, determine the distribution map of the natural gas station based on the detection of the data space of the natural gas station.

[0011] In the embodiments of this application, the names of natural gas stations are collected, and the data space of natural gas stations is determined based on the names of natural gas stations and town databases. The location of natural gas stations is determined by traversing the data space of natural gas stations. At this time, the UAV detects the location of the natural gas station to collect the current form of the natural gas station. This takes into account the overall consideration of traversing the data space of natural gas stations and ensures the accuracy of the location of natural gas stations.

[0012] At this point, the names of natural gas stations are collected. The town database is a collection of data containing various geographical information, building information, facility information, etc. By using the name of the natural gas station, all information related to that station can be retrieved from this database. This information constitutes a "data space".

[0013] The data space usually contains the location information of the station, such as latitude and longitude coordinates and address description. By traversing the data space, this information is extracted to determine the specific location of the station. At this time, in the data space of "Lantian Gas Station", the station's latitude and longitude coordinates (e.g., 30.67°N, 104.07°E) and detailed address (e.g., No. 123 Lantian Road, Wuhou District, Chengdu, Sichuan Province) were found. This information helped to determine the specific location of the station.

[0014] After determining the location of the gas station, a drone was used to conduct aerial reconnaissance of the location. The drone will capture high-definition images or videos to capture the current state of the station, including equipment layout, environmental changes, etc. At this time, a drone was dispatched to the location of the "Blue Sky Gas Station". The drone took high-definition images of the station from the air, which clearly showed the equipment layout, pipeline routing and changes in the surrounding environment.

[0015] Furthermore, the preset form of the natural gas station is determined by traversing the data space of the natural gas station, and the consistency of the natural gas station is determined by matching the preset form with the current form. At this point, when the preset form and the current form of the natural gas station are consistent, the distribution map of the natural gas station is determined by detecting the data space of the natural gas station. This takes into account the overall consideration of matching the preset form and the current form of the natural gas station, ensuring the accuracy of the consistency of the natural gas station.

[0016] At this point, in addition to the location information of the station, the data space usually also contains the station's preset form information. The preset form is based on historical data, design drawings, or planning documents, and represents the layout and form of the station in an ideal or planned state. By traversing the data space, this information is extracted to determine the preset form of the station. At this point, in the data space of "Blue Sky Gas Station", the station's design drawings and historical layout diagrams were found. These drawings and images show the preset form information of the station, such as the layout, equipment location, and pipeline routing during construction and planning.

[0017] After obtaining the preset and current forms of the gas station, it is necessary to match the two to determine whether they are consistent. This step involves image recognition, feature extraction, and comparison algorithms to assess the degree of difference between the current form and the preset form. At the same time, the current form image of the "Blue Sky Gas Station" taken by the drone was compared with the design drawings and historical layout diagrams in the data space. Through image recognition algorithms, it was found that the current form and the preset form are basically consistent in terms of equipment layout, pipeline routing, etc., with no obvious differences.

[0018] If the current state is consistent with the preset state or the difference is within an acceptable range, the station is considered to be in a normal state. At this time, based on the information in the data space and the verification of the current state, a distribution map of the natural gas station is generated or updated. The distribution map should mark in detail the key information such as the location of equipment, pipeline routes, and safety zones within the station. Since the current state of the "Blue Sky Gas Station" is consistent with the preset state, a distribution map of the station is generated based on the design drawings and historical layout maps in the data space, combined with the current state images taken by drones. The distribution map clearly shows the equipment layout, pipeline routes, and the division of safety zones within the station, providing an important basis for subsequent risk monitoring and risk management.

[0019] In some embodiments of this application, a morphology matching table is collected, as shown in Table 1: Table 1 Morphological Matching Table Key elements Pre-set morphology description Current morphology description Matching result Device A Located in the northeast corner of the site, near the main entrance Located in the northeast corner of the site, near the main entrance Consistent Device B Located south of Device A, near the storage tank area Located south of Device A, near the storage tank area Consistent Pipeline 1 Connects Device A and Device B, arranged along the east wall Connects Device A and Device B, arranged along the east wall Consistent Safety area Located in the center of the site, with clear markings Located in the center of the site, with clear markings Consistent refer to Figure 3 In step S12, multiple risk monitoring points are determined based on the distribution map of the natural gas station, and corresponding visual detectors are configured at the multiple risk monitoring points. In the specific implementation of this invention, the specific steps are as follows: S121: Based on the division of the distribution map of natural gas stations, multiple risk monitoring areas are determined. In each risk monitoring area, the corresponding risk monitoring points are determined based on the location and current state of the risk monitoring area, so as to collect multiple risk monitoring points. S122: Among multiple risk monitoring points, collect the location of each risk monitoring point, and determine the monitoring range of each risk monitoring point based on the location of each risk monitoring point and the distance between two adjacent risk monitoring points; S123: Determine the visual detector corresponding to each risk monitoring point based on the monitoring range and the surrounding environment of each risk monitoring point, so that multiple risk monitoring points are configured with corresponding visual detectors.

[0020] In the embodiments of this application, multiple risk monitoring areas are determined based on the division of the distribution map of natural gas stations. In each risk monitoring area, corresponding risk monitoring points are determined based on the location and current form of the risk monitoring area to collect multiple risk monitoring points. This approach takes into account the overall consideration of the location and current form of the risk monitoring areas, ensuring the accuracy of the corresponding risk monitoring points.

[0021] At this point, the distribution map of the natural gas station is analyzed. This is usually a plan view containing information on all key equipment, pipelines, and safety areas within the station. Based on the station's design specifications, equipment operating principles, potential risk points, and other factors, the entire station is divided into multiple logically independent but interconnected risk monitoring areas. These areas are divided based on equipment type (such as compressor area, tank area), functional area (such as gas intake area, processing area, and outlet area), or safety risk level. At this point, geographic information system (GIS) software or manual drawing tools are used to divide the areas, ensuring that each area has clear boundaries and markings.

[0022] A detailed analysis of the location of each risk monitoring area should be conducted, considering its positional relationship relative to site entrances / exits, main roads, other critical equipment, or potential hazards. Based on the current site layout, such as equipment arrangement, pipeline routing, and the location of safety facilities, identify the risk points requiring special attention within each area. According to the location analysis and current layout, select an appropriate number of monitoring points within each risk monitoring area. These monitoring points should comprehensively cover the critical equipment and potential risk points within the area, while also considering factors such as the field of view, installation conditions, and cost of the monitoring equipment. Simultaneously, clearly mark each risk monitoring area and its corresponding monitoring points on the distribution map to provide clear guidance for subsequent risk monitoring work.

[0023] Furthermore, the location of each risk monitoring point is collected among multiple risk monitoring points, and the monitoring range of each risk monitoring point is determined based on the location of each risk monitoring point and the distance between two adjacent risk monitoring points. This takes into account the overall consideration of the location of each risk monitoring point and the distance between two adjacent risk monitoring points, ensuring the accuracy of the monitoring range of each risk monitoring point.

[0024] At this point, the location of each risk monitoring point is collected, and the monitoring range of each risk monitoring point is determined based on its location and the distance between two adjacent risk monitoring points. Then, the distance between each risk monitoring point and its adjacent monitoring points is calculated, which is done using measurement tools or the distance calculation function in GIS software. Based on the distance calculation results and monitoring requirements, a reasonable monitoring range is set for each risk monitoring point. The size of the monitoring range should ensure that blind spots between adjacent monitoring points are covered, while also considering the performance of the monitoring equipment (such as field of view, resolution, etc.) and the characteristics of the monitored target (such as size, movement speed, etc.). In practical applications, the monitoring range needs to be adjusted according to the site conditions or monitoring effect; for example, if the monitoring effect of a certain monitoring point is not ideal, its monitoring range needs to be expanded or adjacent monitoring points added.

[0025] Specifically, suppose there is a natural gas station that has identified multiple risk monitoring points and collected their location data; each risk monitoring point is precisely located using a GPS locator, and the location data is recorded in a spreadsheet; for example, the location data of monitoring point A is (X1, Y1), the location data of monitoring point B is (X2, Y2), and so on.

[0026] GIS software was used to calculate the distance between each risk monitoring point and its adjacent monitoring points. Taking monitoring point A as an example, the distances between it and its adjacent monitoring points B and C were calculated, which are D1 and D2, respectively. Based on the distance calculation results and monitoring requirements, a circular monitoring range with A as the center and a radius of R was set for monitoring point A. The selection of radius R took into account factors such as the field of view of the monitoring equipment, the safety distance requirements of the natural gas station, and the distribution of potential risk points. The overlap between monitoring ranges was also considered to ensure that blind spots between adjacent monitoring points are covered. For example, there is a certain overlap between the monitoring range of monitoring point A and the monitoring range of monitoring point B to ensure that any abnormalities in this area can be detected in a timely manner.

[0027] In practical application, it was found that the monitoring range of monitoring point A had blind spots in some areas. Therefore, it was decided to expand its monitoring range or add adjacent monitoring points to eliminate the blind spots. The adjusted monitoring range is more reasonable and complete, and can better meet the risk monitoring needs of natural gas stations. Through the implementation of the above steps, a reasonable monitoring range was determined for each risk monitoring point, providing an important basis and guarantee for subsequent risk monitoring work.

[0028] Therefore, by determining the visual detectors corresponding to each risk monitoring point based on its monitoring range and surrounding environment, multiple risk monitoring points can be configured with corresponding visual detectors. This approach takes into account both the monitoring range and the surrounding environment of each risk monitoring point, ensuring the accuracy of the visual detectors corresponding to each risk monitoring point.

[0029] At this point, review and understand the monitoring range of each risk monitoring point, which is usually determined in step S122; identify the characteristics of the monitoring range, such as size, shape, whether it contains complex terrain or obstacles, etc.; based on the characteristics of the monitoring range, clarify the monitoring requirements, such as the type of target to be monitored (personnel, vehicles, equipment), monitoring time (daytime, nighttime), monitoring accuracy (resolution, frame rate), etc.

[0030] Consider environmental factors around the monitoring point, such as lighting conditions (natural light, artificial lighting), weather conditions (rain, snow, fog), obstructions (trees, buildings), electromagnetic interference, etc.; assess the security risks of the surrounding environment to the monitoring equipment, such as whether it is easily damaged or susceptible to malicious interference; analyze the adaptability and performance of the monitoring equipment in specific environments, such as the performance of infrared cameras at night or in low light conditions, and the performance of waterproof and dustproof cameras in severe weather.

[0031] Based on the assessment results of monitoring needs and the surrounding environment, select appropriate visual detector types (such as cameras, infrared sensors, thermal imagers, etc.) and specifications (such as resolution, focal length, frame rate, protection level, etc.); at the same time, develop a detailed equipment configuration plan, including the installation location, installation angle, connection method (wired, wireless), power supply, etc.; ensure that the selected equipment is compatible with the existing monitoring system, network architecture, etc., to avoid technical obstacles during subsequent integration.

[0032] According to the configuration plan, the visual detectors are installed at the designated risk monitoring points; the installed equipment is debugged, including image quality adjustment, focus calibration, network connection testing, etc., to ensure that the equipment can work normally and meet the monitoring requirements; the debugged equipment is integrated into the existing monitoring system to realize real-time data transmission, storage and analysis.

[0033] Specifically, suppose there is a natural gas station that has identified multiple risk monitoring points and their monitoring ranges, and assessed the surrounding environment. Taking monitoring point A as an example, its monitoring range is a circular area with a radius of 50 meters, covering the key equipment in the tank area and the surrounding area. The monitoring requirements include: personnel activities, vehicle entry and exit, equipment operating status, and monitoring capabilities under nighttime and low light conditions within the monitoring area.

[0034] The lighting conditions around monitoring point A are good, but nighttime illumination is insufficient; the weather conditions are changeable, with occasional rain and snow; there are a few trees and buildings nearby, but they do not affect the monitoring field of view; the security risk assessment shows that monitoring point A is relatively concealed and not easily damaged, but waterproof, dustproof, and electromagnetic interference resistance capabilities need to be considered; at the same time, based on the monitoring requirements and the assessment results of the surrounding environment, a camera with high-definition night vision, high waterproof and dustproof rating, and strong electromagnetic interference resistance was selected as the visual detector; the camera specifications are: 1080P resolution, adjustable focus, 30fps frame rate, and IP66 protection rating.

[0035] Install the camera at the designated location at monitoring point A, adjust the focus and image quality to ensure clear coverage of the entire monitoring range; connect the camera to the existing monitoring system, transmit data via wireless network to achieve real-time monitoring and video recording storage; perform regular maintenance and inspection of the camera to ensure its long-term stable operation; through the implementation of the above steps, a suitable visual detector is configured for each risk monitoring point, and effective monitoring of key areas of the natural gas station is achieved.

[0036] In some embodiments of this application, a visual detector matching table is acquired, as shown in Table 2: Table 2. Visual Detector Matching Table Risk monitoring point Monitoring range characteristics Peripheral environment characteristics Visual detector A Circular, radius 50 meters Good lighting, no obstruction High-definition camera, day and night B Rectangular, length 100 meters, width 50 meters Insufficient night lighting, with a small amount of tree obstruction Infrared camera, waterproof and dustproof C Complex terrain, with height difference Variable climate, more rain and snow Thermal imaging camera, designed to resist rain and snow D Narrow space, requires close-range monitoring Electromagnetic interference source Explosion-proof camera, anti-interference design In this visual detector matching table, the most suitable visual detector is recommended based on the monitoring range (such as shape and size) and surrounding environment (such as lighting conditions, obstructions, weather conditions, electromagnetic interference, etc.) of each risk monitoring point.

[0037] refer to Figure 4 In step S13, the corresponding risk features are determined based on the detection of images captured by each visual detector, and multiple risk events of the natural gas station are predicted based on the shape of each risk feature, its current location, and the environmental type of the natural gas station. In the specific implementation of this invention, the specific steps are as follows: S131: Real-time monitoring of multiple visual detectors, acquisition of images captured by each visual detector, determination of each risk area based on the recognition of multiple images, determination of corresponding risk features based on the synchronous detection of each risk area, and acquisition of multiple risk features; S132: Mark the location of the natural gas station, collect multiple environmental parameters based on environmental monitoring of the location of the natural gas station, and determine the environmental type of the natural gas station based on the multiple environmental parameters, the location of the multiple environmental parameters, and the shape of the natural gas station; S133: Determine a first risk set based on the form of each risk characteristic and the environmental type of the natural gas station; determine a second risk set based on the current location of each risk characteristic and the environmental type of the natural gas station; and predict multiple risk events of the natural gas station based on the synthesis of the first and second risk sets. The multiple risk events include weather risk events, collision risk events, vibration risk events, and indoor risk events.

[0038] In the embodiments of this application, multiple visual detectors are monitored in real time, and images captured by each visual detector are acquired. Each risk region is determined based on the recognition of multiple images, and the corresponding risk features are determined based on the synchronous detection of each risk region. This process of acquiring multiple risk features takes into account the overall consideration of synchronous detection of each risk region and ensures the accuracy of the corresponding risk features.

[0039] At this point, multiple visual detectors are monitored in real time. The system periodically or on demand captures image frames from each camera. These image frames are stored on a local server or in cloud storage for subsequent analysis and processing. Image acquisition is usually achieved through video stream processing software, which can extract individual image frames from the video stream and preprocess them as needed (such as noise reduction, contrast enhancement, etc.).

[0040] Image recognition algorithms (such as deep learning models) are used to analyze acquired images to identify potential risk areas. These algorithms can detect abnormal or dangerous elements in images, such as people, vehicles, flames, leaks, etc. At the same time, image recognition usually relies on pre-trained deep learning models, which are trained on a large amount of image data and can identify specific objects and scenes. In the application of natural gas stations, these models need to be fine-tuned for specific risk characteristics.

[0041] After identifying risk areas, the system further analyzes the images within these areas to determine specific risk characteristics. These characteristics include personnel behavior (such as not wearing safety equipment, illegal operation, etc.), equipment status (such as overheating, leakage, malfunction, etc.), and environmental factors (such as smoke, flames, water accumulation, etc.). At the same time, the determination of risk characteristics usually relies on more sophisticated image analysis techniques, such as target detection, semantic segmentation, and behavior recognition. These techniques can extract key information from the images and match it with predefined risk characteristics.

[0042] The system will summarize and record all identified risk characteristics, which will be used for subsequent risk assessment, early warning and response. At this time, the collection of risk characteristics is usually achieved through databases or data warehouses. These storage systems can efficiently store, retrieve and analyze large amounts of images and data.

[0043] Specifically, suppose a visual inspection system is deployed in a critical area of ​​a natural gas station. This system includes multiple high-definition cameras for real-time monitoring of the area. The system begins receiving real-time video streams from each camera and displays them on a central control console. The system periodically captures image frames from each camera and stores them in cloud storage. Using a deep learning model to analyze the acquired images, the system identifies a potential risk area where a worker is approaching a high-voltage device.

[0044] Further analysis of the images within the risk area revealed a specific risk characteristic: the worker was not wearing necessary protective equipment (such as insulated gloves and a safety helmet). This risk characteristic was recorded and stored in the database along with other identified risk characteristics. Based on these characteristics, the system further assessed the overall risk level of the natural gas station and triggered corresponding early warning or response mechanisms. This example demonstrates the crucial role of step S131 in practical applications; it helps natural gas stations promptly identify and recognize potential risks, enabling them to take effective measures to ensure safety and operational stability.

[0045] Furthermore, the location of the natural gas station is marked, and multiple environmental parameters are collected based on environmental monitoring at the location of the natural gas station. The environmental type of the natural gas station is determined based on the multiple environmental parameters, their locations, and the shape of the natural gas station. This comprehensive consideration of multiple environmental parameters, their locations, and the shape of the natural gas station ensures the accuracy of the environmental type of the natural gas station.

[0046] At this point, accurately marking the geographical location of the natural gas station in a Geographic Information System (GIS) or related monitoring system typically involves using GPS coordinates, map annotations, or other positioning technologies to determine the station's precise location. This is accomplished using specialized GIS software or platforms, which usually provide a map interface that allows users to input or import coordinate data and display the station's location on a map.

[0047] Once the site location is determined, a series of environmental sensors need to be deployed around it to monitor and collect various environmental parameters in real time, including temperature, humidity, air pressure, wind speed, wind direction, rainfall, air quality (such as oxygen content and concentration of harmful gases), soil moisture, etc. At this time, the environmental sensors are independent devices, but also integrated into a more complex monitoring system. These sensors are connected to the data acquisition system via wired or wireless means to transmit monitoring data in real time.

[0048] After collecting environmental parameters, it is necessary to analyze these data to determine the type of environment in which the site is located. This usually involves comparing the actual monitored parameter values ​​with preset thresholds or standards, and considering the interactions between parameters and the physical form of the site (such as open, closed, or semi-open structures). At the same time, data analysis software or algorithms are used to process the environmental parameter data to identify specific environmental patterns or trends.

[0049] Specifically, GIS software was used to accurately mark the GPS coordinates of the natural gas station on the map and display its location. Multiple environmental sensors were deployed around the station, including temperature sensors, humidity sensors, wind speed and direction sensors, and air quality sensors. These sensors monitored and collected environmental parameters such as temperature, humidity, wind speed, wind direction, oxygen content, and concentration of harmful gases in real time.

[0050] Analysis of the collected environmental parameter data revealed high temperatures, moderate humidity, significant variations in wind speed and direction, and normal oxygen levels, although the concentration of harmful gases occasionally exceeded standards. Considering the station's location in a coastal industrial area and its semi-open structure, it is susceptible to the influence of marine climate and industrial emissions. Based on the above information, the environmental type of the station was determined to be "semi-open, high-temperature, and variable environment in a coastal industrial area." This environment places higher demands on the corrosion resistance and stability of equipment, and also requires attention to the impact of harmful gas emissions and meteorological conditions on station safety. This example demonstrates the importance of step S132 in practical applications; it helps to accurately understand the environmental conditions of the natural gas station, thus providing crucial information for subsequent risk assessment and safety management.

[0051] Therefore, a first risk set is determined based on the form of each risk characteristic and the environmental type of the natural gas station. A second risk set is determined based on the current location of each risk characteristic and the environmental type of the natural gas station. Based on the synthesis of the first and second risk sets, multiple risk events of the natural gas station are predicted. These multiple risk events include weather risk events, collision risk events, vibration risk events, and indoor risk events. This approach takes into account the overall consideration of the current location of each risk characteristic and the environmental type of the natural gas station, ensuring the accuracy of the second risk set. At the same time, it also takes into account the overall consideration of the form of each risk characteristic, its current location, and the environmental type of the natural gas station, improving the detection accuracy of multiple risk events of the natural gas station and ensuring the effectiveness of risk detection at the natural gas station.

[0052] At this point, a detailed analysis is conducted on the previously identified risk characteristics (such as personnel behavior, equipment status, environmental factors, etc.), and the risk event set, i.e. the first risk set, is initially determined by combining the environmental type of the natural gas station (such as coastal industrial areas, high-temperature desert areas, mountainous areas, etc.). The key here is to understand the interaction and potential impact between risk characteristics and specific environments. At this time, a preset risk assessment model is used. The preset risk assessment model is usually based on historical data and expert knowledge and can automatically analyze the form of risk characteristics (such as size, shape, color, movement trajectory, etc.) and environmental type, and output a list of potential risk events.

[0053] After determining the first risk set, it is necessary to further consider the current location of the risk characteristics (such as whether they are near critical equipment, flammable and explosive areas, or densely populated areas) and the impact of environmental types on the event-related nature of the risk, in order to determine a more specific and precise risk event set, namely the second risk set. This step emphasizes the spatial relationship between risk characteristics and site layout, equipment distribution, and personnel activities. At this point, spatial analysis techniques, such as buffer analysis and overlay analysis in Geographic Information Systems (GIS), are used to assess the impact of the location of risk characteristics on site safety. Simultaneously, the assessment results of risk events are dynamically adjusted by combining real-time monitoring data of environmental parameters.

[0054] By synthesizing the first and second risk sets, and comprehensively considering the impact of the risk characteristics, location, and environmental type on risk events, multiple risk events faced by natural gas stations are predicted. These risk events include weather risk events (such as rainstorms, lightning, and high temperatures), collision risk events (such as vehicle collisions and falling objects), vibration risk events (such as vibrations caused by equipment failure and earthquakes), and indoor risk events (such as fires, explosions, and leaks). At the same time, a comprehensive risk assessment system or platform is used. This system can integrate data and information from different sources (such as risk characteristics, environmental parameters, and station layout) and apply advanced algorithms and models to predict risk events. These systems typically provide an intuitive user interface and reporting functions so that users can easily understand and respond to potential risks.

[0055] Specifically, suppose a natural gas station is located in a mountainous area and has recently experienced continuous rainfall. The first risk set is determined as follows: the identified risk characteristics include: complex mountainous terrain, heavy rainfall, high soil moisture, and the risk of landslides in some areas; some old equipment within the station, posing a risk of leakage or malfunction; frequent personnel activity, posing a risk of unauthorized operation or insufficient safety awareness. Based on the environmental type (mountainous area, rainfall), the first risk set is preliminarily determined as: landslide risk, equipment leakage or malfunction risk, and personnel unauthorized operation risk.

[0056] Identifying the second risk set: Analyzing the current location of risk characteristics: It was found that areas with landslide risk are close to the station entrance and gas storage facilities; old equipment is located in the core area of ​​the station, and the surrounding area has dense personnel activity; the risk of non-compliance mainly exists in the station's maintenance and operation areas; Combining environmental type and location information, the second risk set was identified as: the risk of station entrance blockage and gas storage facility damage caused by landslides, the risk of fire and explosion caused by equipment leakage or failure, and the risk of safety accidents caused by personnel non-compliance.

[0057] Based on information from both the first and second risk sets, several risk events facing the natural gas station were predicted: landslides caused by continuous rainfall blocking the station entrance, affecting emergency evacuation and rescue; leaks or malfunctions in aging equipment in humid environments, leading to fires or explosions; and personnel violations resulting in equipment damage or safety accidents. Corresponding preventative measures and emergency plans were developed to address these risk events and ensure the safe operation of the station. This example demonstrates the importance of step S133 in practical applications; it helps to comprehensively consider the impact of the risk characteristics, location, and environmental type on risk events, thereby predicting risk events facing the natural gas station and developing corresponding preventative and response measures.

[0058] In some embodiments of this application, risk feature weights are introduced, where the risk feature form is 40%; the risk feature location is 30%; and the environmental type is 30%. Assuming a risk feature is identified: large-area water accumulation, located near the gas storage tanks of a natural gas station, and the station is located in a humid / rainy region, its score is calculated according to the following steps: Risk characteristic form score: The risk of large-scale water accumulation and flooding is matched, scoring 40 points (out of 40). Risk characteristic location score: The water accumulation was located near the gas storage tank, which matched the explosion risk, earning 20 out of 30 points (because location is a key factor, but not the only determining factor, so it did not receive full marks). The site is located in a humid / rainy region, matching the risk of flooding, and scores 30 points (out of 30). The total score = risk characteristic form score + risk characteristic location score + environmental type score. Therefore, the total score = 40 + 20 + 30 = 90 points. Based on the total score, the most dangerous event caused by this risk characteristic is estimated to be: flooding risk in weather risk events, accompanied by explosion risk in indoor risk events (due to water accumulation near the gas storage tank).

[0059] refer to Figure 5In step S14, the risk control area of ​​the natural gas station is determined based on multiple risk events and the overall shape of the natural gas station, and the risk coefficient of the risk control area is marked. In the specific implementation of this invention, the specific steps are as follows: S141: Collect the overall shape of the natural gas station and mark the locations of multiple risk events. Based on the overall shape of the natural gas station and the locations of multiple risk events, determine the three-dimensional risk model of the natural gas station. S142: The risk control areas of the natural gas station are determined based on the risk three-dimensional model of the natural gas station; within each risk control area, multiple sub-risk areas are determined based on the risk detection of each risk control area; S143: Based on the risk identification of multiple sub-risk areas, the corresponding risk category is determined; the sub-risk coefficient of the sub-risk area is determined according to the regional environment and risk category of the multiple sub-risk areas; and the risk coefficient of the risk control area is determined based on the sub-risk coefficient and risk coefficient mapping relationship of the multiple sub-risk areas.

[0060] In the embodiments of this application, the overall shape of the natural gas station is collected, and the locations of multiple risk events are marked. Based on the overall shape of the natural gas station and the locations of multiple risk events, a three-dimensional risk model of the natural gas station is determined. This approach takes into account both the overall shape of the natural gas station and the locations of multiple risk events, ensuring the accuracy of the three-dimensional risk model of the natural gas station.

[0061] At this stage, a comprehensive survey and mapping of the natural gas station is conducted to obtain its overall morphological information. The survey includes the station's layout, building heights, equipment distribution, road directions, and the location of green belts. Mapping methods include drone aerial photography, satellite image analysis, and on-site measurements. Drones equipped with high-definition cameras are used to take aerial photos to obtain an aerial view of the station. GIS (Geographic Information System) software is used to process and analyze satellite images to extract the station's layout information. On-site measurements are conducted using laser rangefinders or total stations to obtain the precise dimensions and locations of buildings and equipment.

[0062] Based on the overall shape of the site, it is necessary to mark the locations of known or potential risk events within the site according to historical data, real-time monitoring data, or expert assessments. These risk events include equipment failures, leaks, fire hazards, and explosion risks. At this time, GIS software is used to mark the locations of risk events on the site plan, using different forms of marking such as points, lines, and polygons to represent different types of risk events. At the same time, attribute information is added to each risk event, such as risk type, risk level, and occurrence time.

[0063] Based on the overall shape of the gas station and the location of multiple risk events, a three-dimensional risk model of the gas station is constructed using 3D modeling technology. This model not only includes the station's plan layout information but also the 3D shape of buildings and equipment, the 3D location of risk events, and other information. Simultaneously, the station's plan layout and risk event location information are imported using 3D modeling software (such as AutoCAD, SketchUp, Revit, etc.), and a 3D model is constructed based on parameters such as the height and shape of buildings and equipment. In the model, different colors or textures are used to represent different levels of risk areas to intuitively display the risk distribution of the station.

[0064] Specifically, drones were used to take aerial photos of the natural gas station, obtaining clear overhead images; GIS software was used to process satellite images and extract the station's plan layout information, including building locations, road directions, and equipment distribution; and on-site measurements were conducted to obtain precise dimensions and height information of the buildings and equipment.

[0065] Based on historical data, known leak points and equipment malfunction locations within the site were marked. Based on real-time monitoring data, existing fire hazards and explosion risk areas were marked. GIS software was used to mark the locations of these risk events on the site's floor plan, and attribute information, such as risk type and risk level, was added to each risk event. Then, 3D modeling software was used to import the site's floor plan and risk event location information. A 3D model of the site was constructed based on parameters such as the height and shape of buildings and equipment. Different colors were used in the model to represent different levels of risk areas, such as red for high-risk areas, yellow for medium-risk areas, and green for low-risk areas. The model provides a clear visual representation of the risk distribution within the site, offering a scientific basis for subsequent risk management.

[0066] Furthermore, the risk control areas of the natural gas station are determined based on the risk three-dimensional model of the natural gas station. Within each risk control area, multiple sub-risk areas are determined based on the risk detection of each risk control area, which takes into account the overall consideration of risk detection in each risk control area and ensures the accuracy of multiple sub-risk areas.

[0067] At this point, an in-depth analysis of the three-dimensional risk model of the natural gas station is conducted. Based on multiple dimensions such as the distribution of risk events, equipment type, operating procedures, and environmental factors, the station is divided into different risk control zones. Each risk control zone should have clear boundaries and characteristics to facilitate subsequent risk management and control. Then, using 3D modeling software and GIS analysis tools, the risk three-dimensional model is sliced, layered, or divided into regions. Division criteria are set based on factors such as the density of risk events, the importance of equipment, and the complexity of operating procedures. Through algorithms or manual judgment, the station is divided into several risk control zones, and each zone is assigned a unique identifier and descriptive information.

[0068] After identifying the risk control areas, further detailed risk detection is required for each area. This includes real-time monitoring of equipment status, analysis of operational safety, and assessment of the impact of environmental factors on risks. Based on these detection results, each risk control area is further subdivided into multiple sub-risk areas. Each sub-risk area should target specific risk events or risk factors to facilitate targeted risk control and mitigation measures. At this stage, tools such as sensor networks, video surveillance, and data analysis software are used to monitor and analyze data in real time for each risk control area. Based on the detection results, high-risk points, potential hazard areas, or abnormal operational behaviors are identified. Through algorithms or expert judgment, these high-risk points or hazard areas are divided into sub-risk areas, and information such as risk level, risk type, and scope of impact is assigned to each sub-risk area.

[0069] Specifically, assuming there is a natural gas station that has already constructed a 3D risk model through step S141, the risk control area and sub-risk areas will be further determined according to step S142. At this point, 3D modeling software and GIS analysis tools are used to slice and divide the risk model into regions. Based on the equipment type (such as gas storage tanks, compressors, pipelines, etc.) and the distribution of risk events (such as leaks, fires, explosions, etc.), the station is divided into multiple risk control areas such as the gas storage tank area, compressor area, pipeline transportation area, and office area. Each risk control area has clear boundaries and characteristics; for example, the gas storage tank area contains multiple large gas storage tanks, and the compressor area contains compressor equipment and related pipelines.

[0070] In the gas storage tank area, a sensor network is used to monitor parameters such as pressure, temperature, and liquid level in the gas storage tanks in real time. By analyzing this data, it was found that the pressure of a certain gas storage tank was abnormally high, indicating a risk of leakage. Therefore, the gas storage tank and its surrounding area were classified as a sub-risk area and marked as high-risk level and leakage risk type. In the compressor area, through video monitoring and analysis of the safety of the operation process, it was found that the operator of a certain compressor equipment had violated operating procedures, leading to equipment failure or accident. Therefore, the compressor equipment and its operating area were classified as a sub-risk area and marked as medium-risk level and operation risk type.

[0071] Similarly, in other risk management areas such as pipeline transportation areas and office areas, multiple sub-risk areas were identified based on real-time monitoring and data analysis results. Each sub-risk area was assigned information such as risk level, risk type, and scope of impact. This example demonstrates the importance of step S142 in practical applications. It helps to further determine risk management areas and sub-risk areas based on the three-dimensional risk model of the natural gas station, providing a scientific basis for subsequent risk control and mitigation measures. Simultaneously, real-time monitoring and data analysis enable the timely detection of potential risks, improving the safety and reliability of the station.

[0072] Therefore, the risk category is determined based on the risk identification of multiple sub-risk areas, the sub-risk coefficient of the sub-risk area is determined according to the regional environment and risk category of the multiple sub-risk areas, and the risk coefficient of the risk control area is determined based on the sub-risk coefficient and risk coefficient mapping relationship of multiple sub-risk areas. This approach takes into account the overall consideration of the sub-risk coefficient and risk coefficient mapping relationship of multiple sub-risk areas, ensuring the accuracy of the risk coefficient of the risk control area.

[0073] At this point, risk identification is performed for each sub-risk area, that is, the main types of risks faced by the area are determined. Risk identification involves the analysis of various aspects such as equipment status, operation process, environmental factors, and historical data. Based on the analysis results, each sub-risk area is classified into the corresponding risk category, such as leakage risk, fire risk, explosion risk, mechanical failure risk, etc. At this point, risk identification is performed for each sub-risk area. By comparing historical data and real-time monitoring data, the changing trends of equipment status, the compliance of operation process, and the fluctuation of environmental factors are analyzed to determine the risk category of each sub-risk area.

[0074] After determining the risk category of a sub-risk area, it is necessary to further assess the risk level of that area. This involves an in-depth analysis of the regional environment, including equipment status, operator qualifications, implementation of safety management systems, and emergency response capabilities. Based on the characteristics of the regional environment and risk category, the sub-risk coefficient of the sub-risk area is calculated using a risk assessment model or algorithm. The sub-risk coefficient reflects the likelihood and severity of a specific risk event occurring in that area. At this point, considering the actual situation of the regional environment and the characteristics of the risk category, the sub-risk coefficient of each sub-risk area is calculated. This involves a quantitative analysis of multiple influencing factors, such as the aging of equipment, operator training, and the completeness of safety management systems. The sub-risk coefficient of each sub-risk area is obtained through methods such as weighted summation and fuzzy comprehensive evaluation.

[0075] After obtaining the sub-risk coefficients for each sub-risk area, the risk coefficient mapping relationship needs to be used to determine the risk coefficient of the entire risk management area. The risk coefficient mapping relationship is established based on historical data, expert experience, or industry standards, and it describes the correlation between the sub-risk coefficients and the risk coefficients of the risk management area. By applying this mapping relationship and comprehensively considering the risk levels of all sub-risk areas, the risk coefficient of the entire risk management area is obtained. At the same time, the sub-risk coefficients of each sub-risk area are converted into the risk coefficients of the risk management area. This involves the application of weighted averages, maximum value selection, or other statistical methods for the sub-risk coefficients. Through this method, an indicator that comprehensively reflects the risk level of the entire risk management area is obtained.

[0076] Specifically, suppose there is a risk management area for a natural gas station, and multiple sub-risk areas have been identified through step S142. Now, step S143 will be used to further determine the risk category, sub-risk coefficient, and overall risk coefficient of these sub-risk areas. In sub-risk area A, through real-time monitoring and historical data analysis, problems such as equipment aging and seal wear have been identified, leading to gas leaks. Therefore, sub-risk area A is classified as a leak risk category. In sub-risk area B, through video surveillance and operator behavior analysis, problems such as improper operation and equipment overload have been identified, leading to fire or explosion accidents. Therefore, sub-risk area B is classified as a fire / explosion risk category.

[0077] For sub-risk area A (leakage risk category), considering factors such as equipment aging, seal wear, and gas pressure, the sub-risk coefficient calculated using the risk assessment model is 0.6 (assuming 0 is no risk and 1 is extremely high risk); for sub-risk area B (fire / explosion risk category), considering factors such as operator skills, equipment overload operation, and emergency rescue capabilities, the sub-risk coefficient calculated using the same risk assessment model is 0.8.

[0078] Assuming the risk coefficient mapping relationship is a weighted average of sub-risk coefficients with equal weights (i.e., each sub-risk area contributes equally to the risk coefficient of the risk control area), the risk coefficient of the entire risk control area = (sub-risk coefficient of sub-risk area A + sub-risk coefficient of sub-risk area B) / total number of sub-risk areas = (0.6 + 0.8) / 2 = 0.7. This example demonstrates the importance of step S143 in practical applications. It helps determine the sub-risk coefficient of each sub-risk area based on the risk identification results and regional environmental characteristics, and further derives the risk coefficient of the entire risk control area. This provides a scientific basis for subsequent risk control and mitigation measures, contributing to improved safety and reliability of natural gas stations.

[0079] In some embodiments of this application, sub-risk area A is identified as the "leakage risk" category; sub-risk area B is identified as the "fire / explosion risk" category; and sub-risk area C is identified as the "mechanical failure risk" category.

[0080] For sub-risk area A (leakage risk): Environmental factor 1 (severity): Medium (score 2); Environmental factor 2 (detection frequency): High (score 3); Environmental factor 3 (emergency response): Medium (score 2); Sub-risk coefficient = 0.42 + 0.33 + 0.3*2 = 1.4 + 0.9 + 0.6 = 2.9 / 3 (because the total weight is 3) = 0.97 (rounded to two decimal places); For sub-risk area B (fire / explosion risk): Environmental factor 1 (flammable material inventory): High (score 3); Environmental factor 2 (fire prevention measures): Medium (score 2); Environmental factor 3 (personnel training): Good (score 3); Sub-risk coefficient = 0.43 + 0.32 + 0.3*3 = 1.2 + 0.6 + 0.9 = 2.7 / 3 = 0.90; For sub-risk area C (mechanical failure risk): Environmental factor 1 (equipment aging): medium (score 2); Environmental factor 2 (maintenance frequency): low (score 1); Environmental factor 3 (spare parts reserve): poor (score 1); Sub-risk coefficient = 0.42 + 0.31 + 0.3*1 = 0.8 + 0.3 + 0.3 = 1.4 / 3 = 0.47.

[0081] Assuming that A, B, and C have equal weights, each 1 / 3: Risk coefficient of the risk control area = (0.97 * 1 / 3 + 0.90 * 1 / 3 + 0.47 * 1 / 3) / (1 / 3 + 1 / 3 + 1 / 3) = (0.97 + 0.90 + 0.47) / 3 = 2.34 / 3 = 0.78; Through the above steps, based on the risk identification results of multiple sub-risk areas, the regional environment, and risk categories, the sub-risk coefficient of each sub-risk area was successfully determined, and the risk coefficient of the entire risk control area was further calculated. This provides an important scientific basis for subsequent risk control and mitigation measures.

[0082] refer to Figure 6 In step S15, the corresponding risk control path is determined based on the regional location, risk coefficient, and risk control logic of the risk control area and the natural gas station, and corresponding risk control measures are carried out along the risk control path in the risk control area. In the specific implementation of this invention, the specific steps are as follows: S151: Collect the risk management logic of natural gas stations, determine the first risk path based on the regional location of the risk management area and the risk management logic of the natural gas stations, and determine the second risk path based on the risk coefficient of the risk management area and the risk management logic of the natural gas stations; S152: In the risk management logic of natural gas stations, the corresponding risk management path is determined based on the first risk path, the second risk path and the risk management mapping relationship. At this time, the risk management path passes through multiple sub-risk areas of the risk management area in sequence, and the corresponding path segment is adjusted according to the different sub-risk coefficients of the sub-risk areas. S153: Collect risk control paths, mark multiple risk control nodes based on the risk control paths, and determine corresponding sub-risk control measures according to the location of each risk control node, the risk event corresponding to the risk control node, and the actual location of the natural gas station. At this time, each risk control node is matched with a sub-risk control measure, and the risk control measures of the risk control area are triggered according to the location of each risk control node and the corresponding sub-risk control measure.

[0083] In the embodiments of this application, the risk management logic of natural gas stations is collected, a first risk path is determined based on the regional location of the risk management area and the risk management logic of the natural gas stations, and a second risk path is determined based on the risk coefficient of the risk management area and the risk management logic of the natural gas stations. This approach takes into account both the risk coefficient of the risk management area and the overall consideration of the risk management logic of the natural gas stations, ensuring the accuracy of the second risk path.

[0084] At this point, the risk management logic of the natural gas station is collected. After understanding the risk management logic of the natural gas station, it is necessary to determine the first risk path based on the regional location of the risk management area and the risk management logic of the station. This is usually the path with the greatest risk event or the greatest risk impact.

[0085] When determining the primary risk path, consider the following factors: High-risk areas: Which areas are considered high-risk due to dense equipment, complex operations, or frequent historical accidents? Critical points in the process flow: Which links play a crucial role in the process flow, and whose problems could paralyze the entire system? Potential risk events: Based on historical data and expert judgment, which risk events are most likely to occur and cause serious consequences? Optionally, in a natural gas pressurization station, the primary risk path may be the pipeline system from the gas storage tank to the compressor. This is because: the gas storage tank and compressor are critical equipment in the station, and the connecting pipeline between them is subjected to the impact of high-pressure natural gas; once a pipeline leaks or ruptures, it will lead to a large-scale natural gas leak, causing serious consequences such as fires or explosions; historically, there have been accidents where compressor failures have led to pipeline leaks, indicating that this path carries a high risk.

[0086] After determining the first risk path, it is necessary to determine the second risk path based on the risk coefficient of the risk control area and the risk control logic of the site. This is usually the suboptimal path, with a slightly lower risk level than the first risk path, but it still needs to be closely monitored. When determining the second risk path, the following factors should be considered: low-to-medium risk areas: although the risk coefficient of these areas is low, there are still potential risk points; auxiliary links in the process flow: although these links are not the core of the process flow, their failure will affect the stability of the entire system; chain reaction of risk events: certain risk events trigger chain reactions, leading to an increase in the risk of other paths.

[0087] Optionally, in a natural gas pressurization station, the second risk path is identified as the output pipeline from the compressor to the high-pressure pipeline network. This is because: although the risk factor of this path is slightly lower than that of the first risk path, a leak in the pipeline from the high-pressure natural gas output by the compressor can still have serious consequences; the output pipeline is a critical link connecting the station and the external high-pressure pipeline network, and its failure will affect the stability of the entire natural gas supply system; if the first risk path fails, causing the compressor to shut down or natural gas to leak into the second risk path, the risk will increase; through step S151, the critical risk path in the natural gas station can be clearly identified, providing strong support for subsequent risk management measures.

[0088] Furthermore, in the risk management logic of natural gas stations, the corresponding risk management path is determined based on the first risk path, the second risk path, and the risk management mapping relationship. At this time, the risk management path passes through multiple sub-risk areas of the risk management area in sequence, and the corresponding path segments are adjusted according to the different sub-risk coefficients of the sub-risk areas. This takes into account the overall consideration of the first risk path, the second risk path, and the risk management mapping relationship, and ensures the accuracy of the corresponding risk management path.

[0089] At this point, based on the previously determined first risk path, second risk path, and risk control mapping relationship, the risk control path of the natural gas station is clarified. These paths will traverse multiple risk control areas, and will be adjusted and controlled accordingly based on the sub-risk coefficient of each sub-risk area.

[0090] Using the first and second risk paths as benchmarks, a general framework for risk management is outlined. These paths typically represent the main flow direction of natural gas within the station and potential high-risk areas. Next, more details are added to these paths, such as specific equipment, valves, and monitoring points, to form a complete risk management path. This path details the risk points faced by natural gas throughout its entire process from input to station, through processing and storage, and finally to output.

[0091] Optionally, suppose there is a natural gas processing station where the first risk path is the natural gas pipelines entering the station from the receiving station, which connect to multiple gas storage tanks; the second risk path is the pipelines from the gas storage tanks to the processing equipment; on the risk management path, the inlet and outlet pipelines of each gas storage tank, as well as the connecting pipelines between them and the processing equipment, will be marked; in addition, monitoring points will be set up at key locations, such as pressure monitoring, temperature monitoring, and leak detection of the gas storage tanks.

[0092] The risk management path traverses multiple risk management zones, which are typically divided based on factors such as equipment type, process flow, and historical accident records. Each sub-risk zone has its own specific sub-risk coefficient, which reflects the likelihood of a risk event occurring in that zone and the severity of its potential consequences. Along the risk management path, corresponding control measures need to be developed based on the sub-risk coefficient of each sub-risk zone. For example, in high-risk areas, it is necessary to increase the number of monitoring points, increase the frequency of inspections, and even install additional safety equipment.

[0093] Optionally, within a natural gas processing station, the storage tank area is designated as a high-risk sub-risk area because it stores a large amount of high-pressure natural gas, and a leak there would have very serious consequences. While the risk is relatively low in the processing equipment area, close monitoring of equipment operation is still necessary to prevent natural gas leaks caused by equipment malfunctions; therefore, this area is designated as a medium-risk sub-risk area. In the risk management path, corresponding control measures will be formulated based on the sub-risk coefficients of these sub-risk areas. For example, in the storage tank area, the frequency of leak detection will be increased, and emergency shut-off valves will be installed to prevent the spread of leaks; in the processing equipment area, equipment will be regularly maintained and inspected to ensure its normal operation.

[0094] In terms of risk management, the corresponding path segment needs to be adjusted according to the sub-risk coefficient of each sub-risk area. This usually means that different management measures are taken in different risk areas to reflect different risk levels. For example, in high-risk areas, higher-specification pipe materials are needed to increase the wall thickness and strength of the pipes to resist potential explosion and leakage risks. In medium-risk areas, conventional pipe materials are selected, but inspection and maintenance are strengthened. In low-risk areas, simpler management measures are taken, such as setting up basic monitoring points and alarm systems.

[0095] Optionally, in natural gas processing plants, for the outlet pipelines of gas storage tanks in high-risk areas, higher-specification pipeline materials will be selected, and emergency shut-off valves and leak detection devices will be installed. This way, once a leak or abnormality is detected, the gas supply will be immediately cut off to prevent the situation from deteriorating further. For the connecting pipelines between processing equipment in medium-risk areas, conventional pipeline materials will be selected, but inspection and maintenance will be strengthened to ensure that there are no leaks or damages at the pipeline connections. For the inlet and outlet pipelines of auxiliary equipment (such as pumps, compressors, etc.) in low-risk areas, simpler control measures will be adopted, such as setting up basic pressure monitoring and alarm systems so that alarms can be issued in a timely manner and corresponding countermeasures can be taken when equipment failure occurs.

[0096] Therefore, risk control paths are collected, and multiple risk control nodes are marked based on these paths. Sub-risk control measures are determined according to the location of each risk control node, the risk event corresponding to that node, and the actual location of the natural gas station. At this point, each risk control node is matched with one sub-risk control measure. Risk control measures for the risk control area are triggered based on the location of each node and its corresponding sub-risk control measure. This approach considers the overall factors of the location of each risk control node, the risk event corresponding to that node, and the actual location of the natural gas station, ensuring the accuracy of the corresponding sub-risk control measures. Simultaneously, risk control areas are introduced, enabling precise control of risk control paths. This facilitates the implementation of corresponding risk control measures within the risk control areas, further improving the risk detection and control effectiveness of the natural gas station. At this point, based on the established risk control path, multiple key risk control nodes are precisely marked. Subsequently, for each node, specific sub-risk control measures are formulated according to its geographical location, the risk events involved, and the actual layout of the natural gas station. Importantly, each risk control node will be matched with a specific sub-risk control measure, and these measures will be automatically triggered according to the location of the node to ensure the safety of the risk control area.

[0097] Once the risk management path is identified, several key risk management nodes will be marked on these paths. These nodes are typically critical points in the process flow, equipment connections, potential risk points, or areas with frequent historical accidents. Each node will be managed hierarchically based on its importance, risk level, and potential consequences. In natural gas processing plants, risk management nodes will be marked at locations such as the inlet and outlet of gas storage tanks, the inlet and outlet of processing equipment, the operation points of key valves, and the monitoring points of output pipelines. These nodes will serve as an important basis for the subsequent development of sub-risk management measures.

[0098] For each risk management node, specific sub-risk management measures will be developed based on its geographical location, the risk events involved, and the actual layout of the natural gas station. These measures include regular inspections, equipment maintenance, leak detection, installation of emergency shut-off valves, installation of fire alarm systems, and emergency response plan development. Importantly, each node will be matched with a specific sub-risk management measure to ensure rapid response and effective control in the event of a risk event. For example, in natural gas processing stations, emergency shut-off valves and leak detection devices will be installed at the inlet and outlet nodes of gas storage tanks to immediately shut off the gas supply and issue an alarm when a leak is detected. For the inlet and outlet nodes of processing equipment, equipment maintenance and inspection will be strengthened to ensure normal operation and prevent malfunctions. For monitoring points of output pipelines, pressure monitoring and fire alarm systems will be installed to promptly detect and take appropriate measures in the event of abnormal pipeline pressure or fire.

[0099] Finally, an automatic triggering mechanism will be set up based on the location of each risk control node and its corresponding sub-risk control measures. This means that once a risk event is detected or predicted, the corresponding risk control measures will be triggered immediately to ensure the safety of the risk control area. These triggering mechanisms include sensors, alarm systems, emergency shut-off valve controllers, and remote monitoring systems. Simultaneously, sensors and alarm systems will be installed at each risk control node in the natural gas processing plant. Once these systems detect anomalies or potential risks (such as leaks, abnormal pressure, etc.), they will immediately trigger corresponding sub-risk control measures (such as closing emergency shut-off valves, activating leak detection devices, etc.). At the same time, these systems will also send alarm information to the remote monitoring center so that managers can respond quickly and take further countermeasures. Through step S153, key risk control nodes in the natural gas plant can be accurately identified, and specific sub-risk control measures can be developed for each node. These measures will be automatically triggered based on the node's location and risk level to ensure the safe operation of the plant.

[0100] In some embodiments of this application, a sub-risk control measure matching table is collected, as shown in Table 3: Table 3: Matching Table of Sub-risk Management Measures Risk control node number Location Corresponding risk event Sub-risk control measures N001 Gas tank inlet valve Leakage Regularly check valve sealing, install leak detection device N002 Processing device A inlet Overpressure Set up pressure monitoring alarm system, regularly maintain equipment N003 Output pipeline monitoring point Pressure anomaly / leakage Install pressure sensor and leak detection device, remote monitoring Please see Figure 7 , Figure 7 This is a schematic diagram of the structural composition of a vision-based risk detection system for natural gas stations according to an embodiment of the present invention; the vision-based risk detection system for natural gas stations includes: Acquisition module 21 is used to acquire distribution maps of natural gas stations; Risk monitoring point module 22 is used to determine multiple risk monitoring points based on the distribution map of natural gas stations, and to configure corresponding visual detectors at multiple risk monitoring points; Risk event module 23 is used to determine the corresponding risk features based on the detection of images captured by each visual detector, and to predict multiple risk events of the natural gas station based on the shape of each risk feature, its current location and the environmental type of the natural gas station. Risk management area module 24 is used to determine the risk management area of ​​the natural gas station based on multiple risk events and the overall shape of the natural gas station, and to mark the risk coefficient of the risk management area; The risk management path module 25 is used to determine the corresponding risk management path based on the regional location, risk coefficient and risk management logic of the risk management area and to carry out corresponding risk management measures along the risk management path in the risk management area.

[0101] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A risk detection method for natural gas stations based on visual inspection, characterized in that, include: Collect a distribution map of natural gas stations; Based on the distribution map of natural gas stations, multiple risk monitoring points were identified, and corresponding visual detectors were configured at these points. Based on the detection of images captured by each visual detector, the corresponding risk characteristics are determined, and multiple risk events of the natural gas station are predicted according to the shape of each risk characteristic, its current location, and the environmental type of the natural gas station. The risk control area of ​​the natural gas station is determined based on multiple risk events and the overall shape of the natural gas station, and the risk coefficient of the risk control area is marked. Based on the regional location, risk coefficient, and risk management logic of the risk management area, the corresponding risk management path is determined, and corresponding risk management measures are implemented in the risk management area along the risk management path.

2. The risk detection method for natural gas stations based on visual inspection according to claim 1, characterized in that, The distribution map of the natural gas collection stations includes: The system collects the names of natural gas stations, determines the data space of natural gas stations based on the names and town databases, and determines the location of natural gas stations by traversing the data space. At this time, the drone detects the location of the natural gas station to collect its current status. The preset form of the natural gas station is determined by traversing the data space of the natural gas station. The consistency of the natural gas station is determined by matching the preset form and the current form of the natural gas station. At this time, when the preset form and the current form of the natural gas station are consistent, the distribution map of the natural gas station is determined by detecting the data space of the natural gas station.

3. The risk detection method for natural gas stations based on visual inspection according to claim 1, characterized in that, The step of determining multiple risk monitoring points based on the distribution map of natural gas stations and configuring corresponding visual detectors at these points includes: Based on the division of the distribution map of natural gas stations, multiple risk monitoring areas are determined. In each risk monitoring area, corresponding risk monitoring points are determined based on the location and current state of the risk monitoring area, so as to collect multiple risk monitoring points. In multiple risk monitoring points, the location of each risk monitoring point is collected, and the monitoring range of each risk monitoring point is determined based on the location of each risk monitoring point and the distance between two adjacent risk monitoring points. Based on the monitoring range and surrounding environment of each risk monitoring point, the corresponding visual detectors for each risk monitoring point are determined, so that multiple risk monitoring points are configured with corresponding visual detectors.

4. The risk detection method for natural gas stations based on visual inspection according to claim 1, characterized in that, The risk characteristics are determined based on the detection of images captured by each visual detector. Multiple risk events at the natural gas station are then predicted based on the morphology of each risk characteristic, its current location, and the environmental type of the natural gas station. These include: The system monitors multiple visual detectors in real time, collects images captured by each visual detector, identifies risk areas based on the recognition of multiple images, and determines corresponding risk features based on the synchronous detection of each risk area, thereby collecting multiple risk features. The location of the natural gas station is marked, and multiple environmental parameters are collected based on environmental monitoring of the location of the natural gas station. The environmental type of the natural gas station is determined based on the multiple environmental parameters, their location, and the shape of the natural gas station.

5. The risk detection method for natural gas stations based on visual inspection according to claim 4, characterized in that, The method of determining corresponding risk features based on the detection of images captured by each visual detector, and predicting multiple risk events at the natural gas station based on the morphology of each risk feature, its current location, and the environmental type of the natural gas station, also includes: A first risk set is determined based on the form of each risk characteristic and the environmental type of the natural gas station. A second risk set is determined based on the current location of each risk characteristic and the environmental type of the natural gas station. Multiple risk events of the natural gas station are predicted based on the synthesis of the first and second risk sets. These multiple risk events include weather risk events, collision risk events, vibration risk events, and indoor risk events.

6. The risk detection method for natural gas stations based on visual inspection according to claim 1, characterized in that, The process of determining the risk control area of ​​a natural gas station based on multiple risk events and the overall shape of the station, and marking the risk coefficient of the risk control area, includes: Collect the overall shape of the natural gas station and mark the locations of multiple risk events. Based on the overall shape of the natural gas station and the locations of multiple risk events, determine the three-dimensional risk model of the natural gas station.

7. The risk detection method for natural gas stations based on visual inspection according to claim 6, characterized in that, The method of determining the risk control area of ​​a natural gas station based on multiple risk events and the overall shape of the natural gas station, and marking the risk coefficient of the risk control area, also includes: The risk control areas of the natural gas station are determined based on the risk three-dimensional model of the station; within each risk control area, multiple sub-risk areas are determined based on the risk detection of each risk control area. Based on the risk identification of multiple sub-risk areas, the corresponding risk categories are determined. The sub-risk coefficients of the sub-risk areas are determined according to the regional environment and risk categories of the multiple sub-risk areas. The risk coefficients of the risk control area are determined based on the sub-risk coefficients and the risk coefficient mapping relationship of the multiple sub-risk areas.

8. The risk detection method for natural gas stations based on visual inspection according to claim 1, characterized in that, The process involves determining the corresponding risk control path based on the regional location, risk coefficient, and risk control logic of the risk control area, and implementing corresponding risk control measures along the risk control path in the risk control area, including: Collect the risk management logic of natural gas stations, determine the first risk path based on the regional location of the risk management area and the risk management logic of the natural gas stations, and determine the second risk path based on the risk coefficient of the risk management area and the risk management logic of the natural gas stations; In the risk management logic of natural gas stations, the corresponding risk management path is determined based on the first risk path, the second risk path and the risk management mapping relationship. At this time, the risk management path passes through multiple sub-risk areas of the risk management area in sequence, and the corresponding path segments are adjusted according to the different sub-risk coefficients of the sub-risk areas.

9. The risk detection method for natural gas stations based on visual inspection according to claim 8, characterized in that, The process of determining the corresponding risk control path based on the regional location, risk coefficient, and risk control logic of the risk control area, and implementing corresponding risk control measures along the risk control path, also includes: Collect risk control paths, mark multiple risk control nodes based on the risk control paths, and determine corresponding sub-risk control measures according to the location of each risk control node, the risk event corresponding to the risk control node, and the actual location of the natural gas station. At this time, each risk control node is matched with a sub-risk control measure, and the risk control measures of the risk control area are triggered according to the location of each risk control node and the corresponding sub-risk control measure.

10. A risk detection system for natural gas stations based on visual inspection, characterized in that, The vision-based risk detection system for natural gas stations is applied to the vision-based risk detection method for natural gas stations as described in any one of claims 1-9, wherein the vision-based risk detection system for natural gas stations comprises: The data acquisition module is used to collect distribution maps of natural gas stations; The risk monitoring point module is used to determine multiple risk monitoring points based on the distribution map of natural gas stations, and to configure corresponding visual detectors at multiple risk monitoring points; The risk event module is used to determine the corresponding risk features based on the detection of images captured by each visual detector, and to predict multiple risk events of the natural gas station based on the shape of each risk feature, its current location, and the environmental type of the natural gas station. The risk management area module is used to determine the risk management area of ​​a natural gas station based on multiple risk events and the overall shape of the natural gas station, and to mark the risk coefficient of the risk management area. The risk management path module is used to determine the corresponding risk management path based on the regional location, risk coefficient, and risk management logic of the risk management area, and to implement corresponding risk management measures along the risk management path in the risk management area.

Images