Gas leakage early warning method and system based on infrared technology

Abstract

The embodiment of the invention relates to the field of gas leakage detection, in particular to a gas leakage early warning method and system based on the infrared technology, and the method comprises the steps: obtaining a plurality of pieces of video monitoring data of a monitoring region, and real-time infrared thermal imaging data of the monitoring region in a historical preset time period; according to the method, visual video, infrared thermal imaging and dangerous gas leakage on-line sensing technologies are combined, data processing and analysis are carried out through a series of algorithms in combination with environment information such as the current time and the sun height of a monitored area, and the processed data can more accurately reflect whether the actual situation of the monitored area is related to gas leakage or not; and efficient and accurate early warning of gas leakage is realized.

Description

Technical Field

[0001] The embodiments of this specification relate to the field of gas leakage detection, and particularly to a gas leakage early warning method and system based on infrared technology. Background Art

[0002] Most of the existing gas leakage detection systems rely on manual detection. For example, touching, seeing, smelling, and listening by ear. There are also systems that use a single detection method, such as a separate gas sensor or thermal imaging device. These systems have problems such as insufficient sensitivity, high false alarm rate, and limited coverage. When dangerous gas leakage occurs, traditional systems may not be able to detect and give an early warning in a timely and accurate manner, resulting in potential safety hazards.

[0003] Therefore, it is necessary to provide a gas leakage early warning method and system based on infrared technology to better detect gas leakage. Summary of the Invention

[0004] The object of the present invention is to provide a gas leakage early warning method and system based on infrared technology. This method can combine visual video, infrared thermal imaging, and online sensing technology for dangerous gas leakage. Through intelligent algorithms for data processing and analysis, it can achieve efficient and accurate early warning of gas leakage.

[0005] One aspect of the embodiments of this specification provides a gas leakage early warning method based on infrared technology. The method includes:

[0006] Obtain multiple video surveillance data of the monitoring area. According to the current time, the current solar altitude, and the positions of the monitoring devices corresponding to each of the multiple video surveillance data, calculate the weight of each video surveillance data, and sort the multiple video surveillance data according to the weight size to obtain a surveillance data sequence Va = [V1, V2, V3,..., Va];

[0007] Obtain the real-time infrared thermal imaging data of the monitoring area, and calculate a real-time heat map based on the real-time infrared thermal imaging data;

[0008] Obtain the historical infrared thermal imaging data of the monitoring area within a historical preset time period, and calculate the historical heat map of the monitoring area;

[0009] Based on the real-time heat map and the historical heat map, determine multiple temperature anomaly positions in the monitoring area; where the temperature anomaly position is a position where the heat value difference at the same position in the real-time heat map and the historical heat map is greater than a preset value;

[0010] Sort the multiple temperature anomaly positions according to the size of the heat value difference to obtain an anomaly position sequence Wa = [W1, W2, W3,..., Wa];

[0011] Based on the monitored data sequence Va and the abnormal position sequence Wa, a matching sequence Pa is calculated through the following formula:

[0012]

[0013] where ∈ is a preset threshold; when the value of Match(Va, Wa) is 1, it means that the monitored data Va matches the abnormal position Wa. When the monitored data Va matches the abnormal position Wa, (Va, Wa) is added to the matching sequence Pa; when the value is 0, it means that the monitored data Va does not match the abnormal position Wa;

[0014] Based on the weights corresponding to each element in the matching sequence Pa = [P1, P2, P3,... Pa] and the actual monitored data corresponding to the element, screening is performed to obtain a target monitored data sequence;

[0015] Obtain the gas sensing data of the monitoring area;

[0016] Based on the target monitored data sequence and the gas sensing data, determine whether there is a gas leak;

[0017] In response to determining that there is a gas leak, generate a warning message.

[0018] Further, the screening based on the weights corresponding to each element in the matching sequence Pa = [P1, P2, P3,... Pa] and the actual monitored data corresponding to it to obtain a target monitored data sequence includes:

[0019] For each element,

[0020] Obtain the actual position of the monitoring device corresponding to the monitored data Va;

[0021] Calculate the physical distance between the actual position and the abnormal position Wa;

[0022] When the physical distance is less than a first preset value, determine whether the weight corresponding to the monitored data Va is greater than a second preset value;

[0023] In response to the weight being greater than the second preset value, add this element to the target monitored data sequence.

[0024] Further, the determination of whether there is a gas leak based on the target monitored data sequence and the gas sensing data includes:

[0025] Input the target monitored data sequence and the gas sensing data into a machine learning model that has been pre-trained, and obtain the output result of the machine learning model;

[0026] Based on the output result of the machine learning model, determine whether there is a gas leak.

[0027] Further, the inputting the target monitoring data sequence and the gas sensing data into a pre-trained machine learning model includes:

[0028] Construct graph data based on the target monitoring data sequence and the gas sensing data;

[0029] Input the graph data into the machine learning model.

[0030] Further, the constructing graph data based on the target monitoring data sequence and the gas sensing data includes:

[0031] Use the geographical locations of the target monitoring data and the gas sensing data as the nodes of the graph data;

[0032] Use the target monitoring data and the gas sensing data as the node features of the graph data nodes;

[0033] Determine the edge directions between the nodes of the graph data based on the wind direction between adjacent nodes of the graph data;

[0034] Use the wind force magnitude between the graph data nodes as the feature of the edge.

[0035] Further, the generating a warning message in response to determining that there is a gas leak includes:

[0036] Generate a real-time dashboard of the monitoring area based on the graph data;

[0037] Give a warning in a way of differentially displaying the nodes corresponding to the areas where gas leaks are determined on the real-time dashboard.

[0038] Further, the calculating the weight of each of the multiple video monitoring data according to the current time, the current solar altitude, and the position of the monitoring device corresponding to each of the multiple video monitoring data includes:

[0039] Obtain the installation direction of the monitoring device corresponding to each of the multiple video monitoring data;

[0040] Determine the solar altitude and the light angle corresponding to the monitoring device according to the position of the monitoring device corresponding to each of the multiple video monitoring data;

[0041] Determine the weight corresponding to the video monitoring data based on the solar altitude, the light angle, and the installation direction of the monitoring device.

[0042] Another aspect of the embodiments of this specification provides a gas leak warning system based on infrared technology, and the system includes:

[0043] A monitoring data acquisition module, which is used to acquire multiple video monitoring data of a monitoring area, calculate the weight of each video monitoring data according to the current time, the current solar altitude, and the location of each monitoring device corresponding to each of the multiple video monitoring data, sort the multiple video monitoring data according to the magnitude of the weights, and obtain a monitoring data sequence Va = [V1, V2, V3, ……, Va];

[0044] A heat map calculation module, which is used to acquire the real-time infrared thermal imaging data of the monitoring area and calculate a real-time heat map according to the real-time infrared thermal imaging data; acquire the historical infrared thermal imaging data of the monitoring area within a historical preset time period and calculate the historical heat map of the monitoring area;

[0045] A sequence determination module, which is used to determine multiple temperature anomaly positions in the monitoring area based on the real-time heat map and the historical heat map; wherein, the temperature anomaly position is a position where the difference in heat values at the same position between the real-time heat map and the historical heat map is greater than a preset value;

[0046] Sort the multiple temperature anomaly positions according to the magnitude of the difference in heat values to obtain an anomaly position sequence Wa = [W1, W2, W3, …… Wa];

[0047] Based on the monitoring data sequence Va and the anomaly position sequence Wa, calculate and obtain a matching sequence Pa through the following formula:

[0048]

[0049] Wherein, when the value is 1, it indicates that the monitoring data Va matches the anomaly position Wa. When the monitoring data Va matches the anomaly position Wa, add (Va, Wa) to the matching sequence Pa; when the value is 0, it indicates that the monitoring data Va does not match the anomaly position Wa;

[0050] Based on the weights corresponding to each element in the matching sequence Pa = [P1, P2, P3, …… Pa] and its corresponding actual monitoring data, perform screening to obtain a target monitoring data sequence;

[0051] A gas leakage determination module, which is used to acquire gas sensing data of the monitoring area;

[0052] Based on the target monitoring data sequence and the gas sensing data, determine whether there is gas leakage;

[0053] An early warning module, which is used to generate an early warning message in response to determining that there is gas leakage.

[0054] The gas leakage warning method based on infrared technology provided by some embodiments of this specification combines visual video, infrared thermal imaging, and online sensing technology for dangerous gas leakage. Through a series of algorithms, it processes and analyzes data by combining environmental information such as the current time and solar altitude in the monitored area. The processed data can more accurately reflect whether the actual situation in the monitored area is related to gas leakage, achieving efficient and accurate warning of gas leakage. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The accompanying drawings herein are incorporated into and constitute a part of this specification, showing embodiments consistent with the present invention, and are used together with the specification to explain the principles of the present invention.

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0057] Figure 1 is an exemplary flowchart of the gas leakage warning method based on infrared technology according to some embodiments of this specification;

[0058] Figure 2 is an exemplary flowchart for determining weights according to some embodiments of this specification. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0059] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the following will clearly and completely describe the technical solutions in the embodiments of this application with reference to the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are some, but not all, of the embodiments of this application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of this application without creative efforts fall within the scope of protection of this application.

[0060] Figure 1 is an exemplary flowchart of the gas leakage warning method based on infrared technology according to some embodiments of this specification. In some embodiments, Figure 1 the shown process 100 can be executed by a processing device (e.g., a processor). As Figure 1 shown, process 100 can include the following operations.

[0061] Step 101, obtain multiple video surveillance data of the monitoring area, calculate the weight of each video surveillance data according to the current time, the current sun altitude and the monitoring device position corresponding to each of the multiple video surveillance data, sort the multiple video surveillance data according to the weight, and obtain the monitoring data sequence Va=[V1, V2, V3,..., Va].

[0062] The weight reflects the importance of the video surveillance data when it is used to determine whether there is a gas leak. The more important the video surveillance data is, the greater its corresponding weight is.

[0063] The size of the weight is related to the current time, the current sun altitude, and the location of the monitoring device when the video surveillance data is obtained. For example, the closer the current time is to noon and the higher the current sun altitude is, the better the environmental conditions of the monitoring device when obtaining the video surveillance data, and the higher the clarity of the obtained video surveillance data, and the higher the accuracy of the subsequent identification of whether there is a gas leak.

[0064] Similarly, the location of the monitoring device also affects the video acquisition field of view of the monitoring device. The better the geographical location of the monitoring device, or in other words, the location of the monitoring device determines whether it can capture areas that require key monitoring. The more key monitoring areas in the captured image, the greater the weight of the video monitoring data.

[0065] In the monitoring data sequence, the video monitoring data with a greater weight is ranked higher.

[0066] In this embodiment, by calculating the weight of the video monitoring data, it is possible to screen out video monitoring data that is more consistent with the conditions for the current gas leakage calculation, thereby improving the accuracy of gas leakage early warning monitoring.

[0067] Step 102: acquiring real-time infrared thermal imaging data of the monitoring area, and calculating a real-time thermal map based on the real-time infrared thermal imaging data.

[0068] The thermal map can reflect the temperature of each location in the monitoring area. The higher the temperature, the more obvious the representation in the thermal map. For example, the higher the temperature, the darker the color.

[0069] In some embodiments, the processor may directly convert the temperature information in the infrared thermal imaging data into thermal values ​​to obtain a thermal map.

[0070] Step 103, obtaining historical infrared thermal imaging data of the monitoring area within a historical preset time period, and calculating a historical thermal map of the monitoring area.

[0071] The historical preset time period refers to a preset time period in the past based on the current time, which can be half an hour, 1 hour, or 2 hours.

[0072] The calculation method of the historical heat map can be the same as that of the real-time heat map, the difference lies in the different data used. For example, the real-time heat map is calculated based on the real-time infrared thermal imaging data, and the historical heat map is calculated based on the historical infrared thermal imaging data.

[0073] Step 104: determining a plurality of temperature anomaly locations within the monitoring area based on the real-time thermal map and the historical thermal map.

[0074] The temperature anomaly location is the location where the difference in thermal value between the real-time thermal map and the same location in the historical thermal map is greater than the preset value.

[0075] The preset value refers to a preset temperature difference, for example, 3 degrees, 5 degrees, etc.

[0076] Step 105 , sorting the plurality of temperature anomaly positions according to the difference in thermal values ​​to obtain an abnormal position sequence Wa=[W1, W2, W3, ... Wa].

[0077] In some embodiments, the sorting order is from large to small according to the difference in thermal value.

[0078] The greater the difference in thermal values, the greater the temperature change in the area. Gas leakage is usually accompanied by fire, explosion, etc., which will cause rapid temperature changes. Therefore, when sorting, the temperature anomaly configuration with a greater difference in thermal values ​​is ranked higher, indicating that the temperature anomaly location is more important and has a higher priority in the subsequent processing process.

[0079] Step 106: Based on the monitoring data sequence Va and the abnormal position sequence Wa, a matching sequence Pa is calculated by the following formula (1).

[0080]

[0081] Among them, ∈ is a preset threshold, which can be 0.1, 0.5 or 0.8, etc.; when the value of Match(Va, Wa) is 1, it means that the monitoring data Va matches the abnormal position Wa. When the monitoring data Va matches the abnormal position Wa, (Va, Wa) is added to the matching sequence Pa; when the value is 0, it means that the monitoring data Va does not match the abnormal position Wa.

[0082] Step 107 , screening is performed based on the weight corresponding to each element in the matching sequence Pa=[P1, P2, P3, ...Pa] and the corresponding actual monitoring data to obtain a target monitoring data sequence.

[0083] It is understandable that the purpose of screening in step 106 is to select those with relatively large weights and having a matching relationship with the temperature anomaly positions with relatively high temperature anomalies to a certain extent. However, in fact, the matching results are not necessarily accurate. Therefore, the matching sequence Pa needs to be further screened in a certain way.

[0084] Exemplarily, the matching sequence can be screened in the manner shown in the following embodiments.

[0085] In some embodiments, for each element in the matching sequence, that is, each combination of (Va, Wa), the following operations are performed for screening.

[0086] S1, Obtain the actual position of the monitoring device corresponding to the monitoring data Va.

[0087] The actual position is the actual geographical location of the monitoring device, which can be obtained through the positioning information of the monitoring device or by obtaining the pre-stored installation position information.

[0088] S2, Calculate the physical distance between the actual position and the anomaly position Wa.

[0089] The physical distance refers to the geographical straight-line distance between the actual position and the anomaly position.

[0090] In some embodiments, the physical distance can be directly calculated based on the geographical coordinates of the actual position and the anomaly position.

[0091] S3, When the physical distance is less than the first preset value, determine whether the weight corresponding to the monitoring data Va is greater than the second preset value.

[0092] The first preset value is the set value of the physical distance, and the unit can be meters / m, for example, 100m, 500m, 1000m, etc.

[0093] It can be understood that for the finally screened target monitoring data, not only does the physical distance need to meet the conditions (if the distance value is too large, it means that the position of the video monitoring data may not be accurate enough), but also the weight needs to be greater than the second preset value.

[0094] The second preset value can be 0.5, 0.6, 0.8, etc. The weight reflects the importance of the video monitoring data.

[0095] In this embodiment, through the double-screening mechanism, the accuracy and importance of the screened video monitoring data can be ensured, providing accurate data for the subsequent analysis and determination of gas leakage.

[0096] S4, In response to the weight being greater than the second preset value, add this element to the target monitoring data sequence.

[0097] Step 108: Obtain gas sensing data of the monitored area.

[0098] Gas sensing data refers to various gas data collected by gas sensing monitoring devices. The gases to be monitored can include various toxic or harmful gases.

[0099] In some embodiments, gas sensing data can be obtained by gas sensors arranged in the detection area.

[0100] Step 109: Determine whether there is a gas leak based on the target monitoring data sequence and the gas sensing data.

[0101] In some embodiments, comprehensive analysis can be performed on the target monitoring data sequence and the gas sensing data to determine whether there is a gas leak.

[0102] For example, the target monitoring data sequence and the gas sensing data can be displayed on a display screen for the monitoring personnel to determine whether there is a gas leak.

[0103] For another example, intelligent algorithms can be used to analyze the target monitoring data sequence and the gas sensing data to determine whether there is a gas leak.

[0104] Exemplarily, the intelligent algorithms shown in the following embodiments can be used to analyze the target monitoring data sequence and the gas sensing data to determine whether there is a gas leak.

[0105] In some embodiments, the target monitoring data sequence and the gas sensing data can be input into a pre-trained machine learning model to obtain the output result of the machine learning model; and, based on the output result of the machine learning model, determine whether there is a gas leak.

[0106] The output result of the machine learning model can be a binary classification result. For example, its output value can be 1 or 0, where 1 indicates a gas leak and 0 indicates no gas leak. The output result can include the prediction result of whether there is a gas leak for each target monitoring data.

[0107] When monitoring multiple locations simultaneously, the output result of the machine learning model can be a vector, and each element of the vector represents the prediction result of whether there is a gas leak at a location.

[0108] The machine learning model shown in the embodiments of this specification can be obtained by pre-training using training samples.

[0109] In some embodiments, the machine learning model can be a deep learning model or a graph neural network model.

[0110] When the machine learning model is a graph neural network model, inputting the target monitoring data sequence and the gas sensing data into the pre-trained machine learning model requires first constructing graph data. Graph data usually includes nodes and edges. Nodes represent entities that make up the graph data, and edges represent the relationships between entities.

[0111] In some embodiments, constructing graph data may include the following operations.

[0112] S21, using the geographical locations of the target monitoring data and the gas sensing data as the nodes of the graph data.

[0113] S22, using the target monitoring data and the gas sensing data as the node features of the nodes of the graph data.

[0114] S23, determining the edge directions between the nodes of the graph data based on the wind direction between adjacent nodes of the graph data.

[0115] S24, using the wind force magnitude between the nodes of the graph data as the feature of the edge.

[0116] Step 110, in response to determining that there is a gas leak, generating a warning message.

[0117] When it is determined that there is a gas leak, a corresponding warning message can be generated. For example, the warning message can include the gas leak location, the gas leak time, the gas leak degree, and the risk level of the gas leak, etc.

[0118] In some embodiments, after generating the warning message, the system can automatically issue an alarm. The alarm methods can include pop-up alarms, siren alarms, and automatically calling and sending text messages to supervisors.

[0119] In some embodiments, the system can also generate a real-time dashboard for the monitoring area, and display the monitoring locations in the form of a graph structure on the real-time dashboard, and highlight the locations where gas leaks occur.

[0120] For example, the real-time dashboard for the monitoring area can be generated based on the graph data; warnings can be given by differentiating the display of the nodes corresponding to the areas where gas leaks are determined on the real-time dashboard.

[0121] The ways of differentiating the display include highlighting, displaying in different colors, flashing alarms, etc.

[0122] In some embodiments, the real-time dashboard can also display the diffusion direction of the leaked gas in combination with the features of the edges of the graph data, so that the monitoring personnel can formulate reasonable gas leak control measures in a timely manner.

[0123] In some embodiments of the present specification, by obtaining visible video data, infrared temperature data, and gas leakage data of a monitoring area, and making scientific and reasonable predictions on the data by combining algorithms such as machine learning models, the influence of human subjective factors can be reduced, making the monitoring of gas leakage more objective and improving the accuracy of gas leakage detection. At the same time, the method of obtaining data in real time and analyzing whether there is gas leakage can also improve the real-time performance of gas leakage detection. When gas leakage is detected, control measures can be formulated in a timely and effective manner to avoid major impacts caused by out-of-control gas leakage.

[0124] Figure 2 is an exemplary flowchart of a method for determining weights according to some embodiments of the present specification. In some embodiments, Figure 2 the process 200 shown can be executed by a processing device (e.g., a processor). As Figure 2 shown, the process 200 may include the following operations.

[0125] Step 201, obtain the installation direction of the monitoring device corresponding to each of the multiple video monitoring data.

[0126] The installation direction refers to the orientation of the lens of the monitoring device.

[0127] In some embodiments, the installation direction can be obtained by querying the installation data of the monitoring device.

[0128] Step 202, determine the solar altitude and light angle corresponding to the monitoring device according to the position of the monitoring device corresponding to each of the multiple video monitoring data.

[0129] In some embodiments, the solar altitude and light angle corresponding to the monitoring device can be determined by a satellite device.

[0130] Among them, the light angle can be calculated from the solar altitude and the azimuth angle between the sun and the position of the monitoring device.

[0131] Step 203, determine the weight corresponding to the video monitoring data based on the solar altitude, light angle, and the installation direction of the monitoring device.

[0132] In some embodiments, the higher the solar altitude and the closer the light angle is to 90° with the position area of the monitoring device, the greater the weight of the video monitoring data.

[0133] Specifically, the weight of the video monitoring data can be calculated by the following formula (2).

[0134] λ = (1 - a)h + at (2)

[0135] Among them, λ represents the calculated weight, a is a preset coefficient, h is the solar altitude, and t is the light angle.

[0136] If the calculated weight is greater than 1, the weight can be normalized.

[0137] The beneficial effects that the embodiments of this specification may bring include but are not limited to: (1) converting the traditional manual detection methods of touching, seeing, smelling, and listening into an automatic technology based on multiple monitoring schemes, reducing the dependence on manual detection and saving labor costs; (2) integrating the visualization video technology, infrared thermal imaging technology, and online sensing technology for dangerous gas leakage, analyzing and processing data through intelligent algorithms, achieving day and night visualization, real-time detection, intelligent early warning, and remote platform inspection, and improving the real-time performance and accuracy of gas leakage detection.

[0138] It should be noted that in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, method, article or device including the said element.

[0139] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A gas leakage warning method based on infrared technology, characterized in that The method includes: Obtain multiple video surveillance data of the monitoring area, calculate the weight of each video surveillance data according to the current time, the current solar altitude, and the location of the monitoring device corresponding to each of the multiple video surveillance data, sort the multiple video surveillance data according to the magnitude of the weights, and obtain a surveillance data sequence Va = [V1, 2, V3, ……, Va]; Obtain the real-time infrared thermal imaging data of the monitoring area, and calculate a real-time thermal map according to the real-time infrared thermal imaging data; Obtain the historical infrared thermal imaging data of the monitoring area within a historical preset time period, and calculate the historical thermal map of the monitoring area; Based on the real-time thermal map and the historical thermal map, determine multiple temperature anomaly locations within the monitoring area; wherein, the temperature anomaly location is a location where the thermal value difference at the same location in the real-time thermal map and the historical thermal map is greater than a preset value; Sort the multiple temperature anomaly locations according to the magnitude of the thermal value difference, and obtain an anomaly location sequence Wa = [W1, W2, W3, …… Wa]; Based on the surveillance data sequence Va and the anomaly location sequence Wa, calculate and obtain a matching sequence Pa through the following formula: wherein, ∈ is a preset threshold; when the value of Match(Va, Wa) is 1, it indicates that the surveillance data Va matches the anomaly location Wa, and when the surveillance data Va matches the anomaly location Wa, (Va, Wa) is added to the matching sequence Pa; when the value is 0, it indicates that the surveillance data Va does not match the anomaly location Wa; Based on the weights corresponding to each element in the matching sequence Pa = [P1, P2, P3, …… Pa] and their corresponding actual surveillance data, perform screening to obtain a target surveillance data sequence; Obtain the gas sensing data of the monitoring area; Based on the target surveillance data sequence and the gas sensing data, determine whether there is a gas leak; In response to determining that there is a gas leak, generate a warning message.

2. The method according to claim 1, wherein The performing screening based on the weights corresponding to each element in the matching sequence Pa = [P1, P2, P3, …… Pa] and the actual surveillance data corresponding to the element to obtain a target surveillance data sequence includes: For each element, Obtain the actual location of the monitoring device corresponding to the surveillance data Va; Calculate the physical distance between the actual location and the anomaly location Wa; When the physical distance is less than a first preset value, determine whether the weight corresponding to the surveillance data Va is greater than a second preset value; In response to the weight being greater than the second preset value, add this element to the target surveillance data sequence.

3. The method according to claim 1, wherein The determining whether there is a gas leak based on the target surveillance data sequence and the gas sensing data includes: Input the target surveillance data sequence and the gas sensing data into a pre-trained machine learning model, and obtain the output result of the machine learning model; Based on the output result of the machine learning model, determine whether there is a gas leak.

4. The method according to claim 3, characterized in that, The inputting the target surveillance data sequence and the gas sensing data into a pre-trained machine learning model includes: Construct graph data based on the target surveillance data sequence and the gas sensing data; Input the graph data into the machine learning model.

5. The method according to claim 4, wherein Constructing the graph data based on the target monitoring data sequence and the gas sensing data includes: Using the geographical locations of the target monitoring data and the gas sensing data as the nodes of the graph data; Taking the target monitoring data and the gas sensing data as the node features of the nodes of the graph data; Determining the edge directions between the nodes of the graph data based on the wind direction between adjacent nodes of the graph data; Taking the wind force magnitude between the nodes of the graph data as the feature of the edge.

6. The method according to claim 5, wherein Generating a warning message in response to determining a gas leak, including: Generating a real-time dashboard of the monitoring area based on the graph data; Issuing a warning by distinguishing and displaying the nodes corresponding to the areas where gas leaks are determined on the real-time dashboard.

7. The method according to claim 1, characterized in that, Calculating the weight of each of the multiple video monitoring data according to the current time, the current solar altitude, and the location of the monitoring device corresponding to each of the multiple video monitoring data, including: Obtaining the installation direction of the monitoring device corresponding to each of the multiple video monitoring data; Determining the solar altitude and the light angle corresponding to the monitoring device based on the location of the monitoring device corresponding to each of the multiple video monitoring data; Determining the weight of the video monitoring data based on the solar altitude, the light angle, and the installation direction of the monitoring device.

8. A gas leakage warning system based on infrared technology, characterized in that, The system includes: A monitoring data acquisition module, configured to acquire multiple video monitoring data of a monitoring area, calculate the weight of each video monitoring data according to the current time, the current solar altitude, and the location of the monitoring device corresponding to each of the multiple video monitoring data, sort the multiple video monitoring data according to the weight size, and obtain a monitoring data sequence Va = [V1, V2, V3, ……, Va]; A heat map calculation module, configured to acquire real-time infrared thermal imaging data of the monitoring area and calculate a real-time heat map according to the real-time infrared thermal imaging data; acquire historical infrared thermal imaging data of the monitoring area within a historical preset time period and calculate the historical heat map of the monitoring area; A sequence determination module, configured to determine multiple temperature anomaly positions in the monitoring area based on the real-time heat map and the historical heat map; wherein, the temperature anomaly position is a position where the heat value difference at the same position in the real-time heat map and the historical heat map is greater than a preset value; Sorting the multiple temperature anomaly positions according to the magnitude of the heat value difference to obtain an anomaly position sequence Wa = [W1, W2, W3, …… Wa]; Calculating a matching sequence Pa based on the monitoring data sequence Va and the anomaly position sequence Wa through the following formula: where, when the value is 1, it indicates that the monitoring data Va matches the anomaly position Wa, and when the monitoring data Va matches the anomaly position Wa, (Va, Wa) is added to the matching sequence Pa; when the value is 0, it indicates that the monitoring data Va does not match the anomaly position Wa; Screening based on the weight corresponding to each element in the matching sequence Pa = [P1, P2, P3, …… Pa] and its corresponding actual monitoring data to obtain a target monitoring data sequence; A gas leakage determination module, which is used to obtain gas sensing data of a monitoring area; Based on the target monitoring data sequence and the gas sensing data, determine whether there is gas leakage; An early warning module, which is used to generate an early warning message in response to determining that there is gas leakage.

Images