Atmospheric environment particulate pollution monitoring method and system based on multi-source data fusion

Abstract

The application relates to the technical field of atmospheric pollution monitoring, and provides a method and system for monitoring atmospheric environmental particulate matter pollution based on multi-source data fusion. The method comprises the following steps: accessing an industrial management database to establish a pollution source dataset; configuring a sensing grid in a target area to perform multi-source monitoring and identification of atmospheric pollution and establish a pollution identifier; configuring a flight task of a UAV according to the pollution identifier, performing flight operation, and establishing an additional dataset; reconstructing the identifier according to the additional dataset, and performing correlation matching in combination with the pollution source dataset; performing source tracing evolution authentication according to the matching result, and reporting a pollution monitoring result, so that the technical problem that the dynamic source tracing accuracy of pollution sources is low and the pollution evolution process cannot be comprehensively and accurately reflected due to insufficient integration of multi-source heterogeneous data in existing atmospheric pollution monitoring is solved, the technical effect that accurate monitoring and positioning of atmospheric environmental particulate matter pollution are realized through multi-dimensional data fusion and dynamic monitoring of a UAV is achieved, and the source tracing capability is improved.

Description

Technical Field

[0001] This application relates to the field of atmospheric pollution monitoring technology, specifically to a method and system for monitoring atmospheric particulate matter pollution based on multi-source data fusion. Background Technology

[0002] Currently, with the rapid advancement of industrialization and urbanization, particulate matter pollution in the atmospheric environment (such as PM2.5 and PM10) has become a major challenge affecting public health and ecological balance. Traditional pollution monitoring methods mostly rely on fixed ground monitoring stations or periodic manual sampling. Although these methods can obtain pollution data for local areas, they are limited by spatial coverage, temporal resolution, and the ability to capture dynamic changes in pollution sources, making it difficult to achieve accurate source tracing and real-time analysis of diffusion paths for complex pollution events. Especially in densely industrialized areas, the types of pollution sources are diverse, and emission patterns are complex. Significant differences exist in the process characteristics, pollutant composition, and emission patterns among different enterprises. This leads to bottlenecks in the integration of multi-source heterogeneous data, pollution cause correlation analysis, and the accuracy of dynamic source tracing in existing monitoring systems, making it difficult to meet the needs of refined environmental governance. Summary of the Invention

[0003] This application provides a method and system for monitoring particulate matter pollution in the atmospheric environment based on multi-source data fusion, aiming to solve the technical problems in existing atmospheric pollution monitoring, such as low accuracy of dynamic source tracing of pollution sources and inability to comprehensively and accurately reflect the pollution evolution process due to insufficient integration of multi-source heterogeneous data.

[0004] The first aspect disclosed in this application provides a method for monitoring atmospheric particulate matter pollution based on multi-source data fusion. The method includes: accessing an industrial management database; extracting pollution source data from the industrial management database to establish a pollution source fusion dataset; configuring a basic sensing grid in a target area and using the basic sensing grid to perform multi-source monitoring and identification of atmospheric pollution to establish a regional pollution identifier; configuring a UAV flight mission using the regional pollution identifier, controlling the UAV to initiate flight operations, and establishing an additional dataset; reconstructing the identifier of the regional pollution identifier based on the additional dataset, and performing association matching between the identifier reconstruction result and the pollution source fusion dataset, wherein the association matching includes explicit association verification matching and implicit association identification matching; performing source tracing evolution authentication using the association matching result, and reporting pollution monitoring results based on the source tracing evolution authentication result.

[0005] Another aspect of this application discloses an atmospheric particulate matter pollution monitoring system based on multi-source data fusion. The system includes: a pollution source data extraction module: accessing an industrial management database, extracting pollution source data from the database, and establishing a pollution source fusion dataset; a multi-source monitoring and identification module: configuring a basic sensing grid in a target area, and performing multi-source monitoring and identification of atmospheric pollution through the basic sensing grid to establish a regional pollution identifier; a flight operation module: configuring a UAV's flight mission using the regional pollution identifier, controlling the UAV to initiate flight operations, and establishing an additional dataset; an association matching module: reconstructing the identifier of the regional pollution identifier based on the additional dataset, and performing association matching between the identifier reconstruction result and the pollution source fusion dataset, wherein the association matching includes explicit association verification matching and implicit association identification matching; and a source tracing evolution authentication module: performing source tracing evolution authentication using the association matching results, and reporting pollution monitoring results based on the source tracing evolution authentication results.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The aforementioned method for monitoring atmospheric particulate matter pollution based on multi-source data fusion first accesses an industrial management database to extract pollution source data and construct a fused dataset. Then, a basic sensing grid is deployed in the target area to monitor multi-source air pollution data and generate regional pollution labels. Next, drone flight missions are scheduled based on the pollution labels, and the drones are controlled to perform these missions and collect additional data. Then, the pollution labels are reconstructed by analyzing the additional data, and the reconstructed labels are matched with the fused pollution source dataset to verify explicit and implicit correlations. Finally, source tracing and evolution verification are performed based on the matching results to obtain accurate pollution monitoring results, aiding in the identification and tracking of pollution sources.

[0008] The above description is merely an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1This is a flowchart illustrating an atmospheric particulate matter pollution monitoring method based on multi-source data fusion in one embodiment.

[0011] Figure 2 This is an architecture diagram of an atmospheric particulate matter pollution monitoring system based on multi-source data fusion in one embodiment.

[0012] Figure labeling: 11 Pollution source data extraction module, 12 Multi-source monitoring and identification module, 13 Flight operation module, 14 Association matching module, 15 Source tracing evolution authentication module. Detailed Implementation

[0013] This application provides a method and system for monitoring particulate matter pollution in the atmospheric environment based on multi-source data fusion, which solves the technical problems in existing atmospheric pollution monitoring, such as low accuracy of dynamic source tracing of pollution sources and inability to comprehensively and accurately reflect the pollution evolution process due to insufficient integration of multi-source heterogeneous data.

[0014] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0015] It should be noted that the terms “comprising” and “having”, and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such process, method, product, or device.

[0016] Example 1, as Figure 1 As shown, this application provides a method for monitoring atmospheric particulate matter pollution based on multi-source data fusion, the method comprising:

[0017] Access the industrial management database, extract pollution source data based on the industrial management database, and establish a pollution source fusion dataset.

[0018] In this embodiment, the system first connects to an industrial management database via a reserved interface. This database contains a large amount of data related to industrial activities, such as enterprise emission data, production processes, and emission standards. After connecting to the industrial management database, relevant pollution source data is extracted through a data interface or API. This data includes information such as the types of pollutants emitted by the enterprise, emission amounts, emission times, and emission locations. This data may involve different pollution sources, such as industrial plants, power plants, and chemical enterprises. During the data extraction process, the data obtained from the industrial management database is screened, cleaned, and preprocessed. Specifically, all industrial enterprise data located within the target area boundary are screened, invalid records with zero emissions or exceeding the equipment's processing capacity threshold are removed, data with format errors is corrected (e.g., time format is standardized to UTC+8), pollutant names are mapped to a unified coding system (e.g., GB / T 16157-1996), and emission units are converted to standard units (e.g., kg / h). After these preprocessing steps, the cleaned data is categorized and organized according to pollution source type, ultimately forming a pollution source fusion dataset. This dataset integrates data from multiple different sources, including enterprise emission data and environmental monitoring data. By establishing this pollution source fusion dataset, a comprehensive understanding of the emissions from various pollution sources can be achieved, providing an accurate data foundation for subsequent air pollution monitoring, pollution source tracing, and pollution assessment.

[0019] A basic sensing grid is configured in the target area, and multi-source monitoring and identification of air pollution are carried out through the basic sensing grid to establish regional pollution identification.

[0020] In one embodiment, a basic sensing grid is first deployed within the target area. This grid is a network of multiple environmental sensors and monitoring devices distributed across different locations within the area, collecting data according to a defined sensing cycle. The monitoring devices can include various types such as air quality monitors, PM2.5 / PM10 sensors, temperature and humidity sensors, and anemometers, enabling real-time monitoring of air pollutant concentrations, meteorological data, and environmental changes. Through these sensing devices, the basic sensing grid collects real-time monitoring data from the target area, such as pollutant concentrations, types, wind speeds, temperatures, and humidity. After collecting the real-time monitoring data, it is identified into the corresponding grid cells, establishing regional pollution identification and providing crucial support for subsequent pollution source tracing and pollution prevention.

[0021] Furthermore, this application provides the aforementioned method for multi-source monitoring and identification of air pollution through the basic sensing grid, and the establishment of regional pollution identification, including:

[0022] After configuring the sensing cycle, the monitoring sensors of the basic sensing grid are activated at the cycle nodes of the sensing cycle to perform sensing monitoring and establish pollution identification for the area.

[0023] Preferably, when performing source monitoring and identification, a sensing cycle is first configured. This sensing cycle refers to the data acquisition cycle of each monitoring sensor in the basic sensing grid. The configuration of the sensing cycle should be set according to the pollution status of the target area, monitoring needs, and actual environmental conditions to ensure the timeliness and accuracy of the monitoring data. After configuring the sensing cycle, the monitoring sensors are activated at the cycle node of each sensing cycle. Whenever a new cycle node is entered, all monitoring sensors in the sensing grid are automatically activated to perform sensing and monitoring tasks and collect air quality data in the target area in real time. By encapsulating the collected air quality data and identifying it in the corresponding grid, a regional pollution identification system for the target area is established, providing an accurate basis for subsequent pollution source tracing, pollution control, and UAV flight mission configuration, ensuring that the monitoring work has a high degree of real-time performance and targeting.

[0024] After configuring the drone's flight mission using the aforementioned area pollution markers, control the drone to initiate flight operations and establish an additional dataset.

[0025] In one embodiment, after obtaining regional pollution markers, the markers are first deconstructed to obtain data such as the types and concentration distribution of pollutants, the location of pollution sources, and the diffusion trend of pollution. Then, this deconstructed data is combined with actual monitoring points, and the drone's flight path is rationally planned through gray-scale area identification to construct the drone's current flight mission, ensuring that the drone can accurately reach pollution hotspots or potential pollution sources. The flight mission is then sent to the drone's control terminal, which controls the drone's flight according to the mission. During flight, the drone uses its onboard sensors to collect environmental data in real time, such as air quality, particulate matter concentration, and meteorological data. This data is continuously transmitted to the system for real-time recording. The data collected by the drone can access data from areas that are inaccessible or difficult to obtain by ground sensors, especially for areas where pollution sources are hidden or difficult to access, providing drones with more flexible monitoring capabilities. By continuously updating and storing the data collected by drones, an additional dataset can be established. This additional dataset includes real-time environmental data during flight, changes in pollutant concentrations, the specific locations of pollutants, and surrounding meteorological data. This data can not only further supplement the information on regional pollution identification, but also provide detailed data support for subsequent pollution source tracing, pollution assessment, and prevention and control measures.

[0026] Furthermore, this application provides a flight mission for configuring a drone using the aforementioned area pollution marker, including:

[0027] The pollution level is deconstructed based on the regional pollution markers to establish a deconstruction result, which includes pollutant types, concentration ranges, and diffusion trend data. The actual distribution of monitoring points in the region is obtained based on the regional pollution markers. The importance of gray areas is identified based on the actual monitoring point distribution and the deconstruction result, and regional monitoring points are configured using the importance identification result. Flight path optimization and fitting are performed on the regional monitoring points to establish a flight mission.

[0028] Preferably, the pollution level of the target area is first deconstructed based on the generated regional pollution labels. During this process, information such as pollutant types, concentration ranges, and diffusion trend data are extracted from the regional pollution labels according to pre-set key names. Then, each grid is sorted in descending order according to the concentration range, so that grids with higher pollution levels are placed earlier, thus generating the deconstruction results. The pollutant types reflect the main pollutants in the target area, such as PM2.5 and PM10; the concentration ranges reflect the real-time concentration data of the main pollutants; and the diffusion trend data reflects the diffusion trend of pollutants, such as diffusion speed. Subsequently, based on the regional pollution labels, the actual distribution of monitoring points within each grid area is obtained, i.e., the actual spatial location and coverage radius of the monitoring equipment within the grid (e.g., a 500m circular coverage). The actual monitoring point distribution is then combined with the deconstruction results to identify the importance of gray-scale areas. Specifically, the importance of each grid is calculated by weighting the normalized concentration level (the normalized result of the maximum value minus the minimum value of the concentration interval), normalized diffusion rate, and normalized coverage radius. The weight of each parameter is an empirical weight, pre-set by domain experts. Grids with importance exceeding the importance threshold are marked as priority monitoring areas, and UAV regional monitoring points are prioritized for these areas. Then, the regional monitoring points and their importance are loaded into a node set and bound to the UAV take-off and landing coordinates. The upper limit of UAV endurance time, the minimum monitoring duration per point, and the safe flight altitude range are set as constraints. Next, an optimization model is constructed, where the objective function is to maximize the total importance score while minimizing the total flight distance. The distance between nodes is converted into Euclidean cost, and the node importance serves as a heuristic factor in the path weight design. After the optimization model is established, an ant colony algorithm is used for iterative solution. During this process, the ant colony is initialized, and starting nodes are randomly assigned. Each ant selects the next node based on probability, calculated as the product of node importance and the reciprocal of distance. After each selection, pheromone levels are updated, increasing the pheromone concentration for the optimal path. Through multiple iterations, the optimal node sequence after convergence is finally extracted (e.g., starting point → monitoring point 3 → monitoring point 1 → monitoring point 2 → endpoint). Finally, the optimal path is converted into a waypoint instruction set, including coordinates, hovering time, and sensor mode, and a JSON-formatted flight mission is generated. This ensures coverage of each monitoring point and effective collection of necessary environmental data, thereby improving the timeliness, accuracy, and coverage of pollution monitoring and providing strong data support for subsequent pollution source tracing and remediation.

[0029] After the regional pollution identifiers are reconstructed based on the additional dataset, the reconstructed identifiers are used to perform association matching with the pollution source fusion dataset. The association matching includes explicit association verification matching and implicit association identification matching.

[0030] In one embodiment, after obtaining the supplementary dataset, the data from each grid in the supplementary dataset is merged into the regional pollution label for label reconstruction. During this process, data from the same grid are weighted and fused to form a more timely and accurate label reconstruction result. Subsequently, the label reconstruction result is combined with the pollution source fusion dataset, and association matching is performed from two aspects: explicit association verification and implicit association identification. Explicit association verification matching is used to verify the direct relationship between pollution sources and pollution labels. Specifically, by comparing the reported data in the pollution source fusion dataset with the concentration and distribution of corresponding pollutants in the label reconstruction result, it can be directly verified whether the pollution source is indeed related to the diffusion of the pollutant. For example, if a reported data in the pollution source dataset shows a large amount of sulfide emissions, and the label reconstruction result shows significant sulfide pollution in the area, then it indicates that there is a direct association between the pollution source and the diffusion of the pollutant. Implicit association identification and matching focuses on potential connections between pollution sources and pollutants that lack explicit associations. For example, a declaration might show very little or no sulfide emissions, but the identification reconstruction results show significant sulfide pollution. In this case, it's necessary to conduct in-depth analysis of the process flow, material usage, and other possible indirect emission pathways to determine whether a potential association exists between the pollution source and sulfide pollution. After the association matching analysis is completed, the explicit association verification matching results and the implicit association identification matching results are summarized into an association matching result, providing data support for subsequent source control and pollution source tracing.

[0031] Furthermore, this application provides the method of using the identifier reconstruction results and the pollution source fusion dataset for association matching, wherein the association matching includes explicit association verification matching and implicit association identification matching, including:

[0032] Extract the declared emission data from the pollution source fusion dataset; use the declared emission data and the identification reconstruction results to verify the consistency of pollution components, pollution time windows, and pollution spatial proximity indicators; configure explicit association scores based on the consistency verification results to complete explicit association verification matching.

[0033] Preferably, the process begins by extracting declared emission data from factories, enterprises, or other pollution sources from the pollution source fusion dataset. This declared emission data is typically reported periodically by the pollution source entities according to environmental regulations, and includes information such as pollutant type, emission quantity, emission time, and emission location. Subsequently, consistency verification is performed between the declared emission data and the reconstructed labeling results from three aspects: pollutant composition, pollution time, and pollution spatial distribution. For pollutant type, the declared emission data and reconstructed labeling results are input into a composition evaluation sub-channel to calculate a composition consistency score. For pollution time window, the declared emission data and reconstructed labeling results are input into a time window evaluation sub-channel to calculate a time consistency score. For pollution spatial proximity, the declared emission data and reconstructed labeling results are input into a spatial proximity evaluation sub-channel to calculate a spatial proximity index. These consistency verification results are then weighted and summed to calculate an explicit correlation score. After configuring the explicit correlation score, the explicit correlation verification matching results can be used, providing a reliable basis for subsequent pollution control and traceability.

[0034] Furthermore, this application provides consistency verification of pollutant components, pollution time windows, and pollution spatial proximity indicators using the declared emission data and the label reconstruction results, including:

[0035] A consistency evaluation channel is established, comprising a component evaluation sub-channel, a time window evaluation sub-channel, and a spatial proximity evaluation sub-channel. After synchronizing the declared emission data and the label reconstruction results to the consistency evaluation channel, the component evaluation sub-channel is used to score the component consistency of the declared emission data and the label reconstruction results, establishing a first verification result. The time window evaluation sub-channel is used to analyze the time window intersection ratio of the declared emission data and the label reconstruction results, establishing a second verification result. The spatial proximity evaluation sub-channel is used to evaluate the spatial proximity index of the declared emission data and the label reconstruction results, establishing a third verification result. Consistency verification is completed based on the first, second, and third verification results.

[0036] Optionally, before conducting consistency verification, a consistency evaluation channel is constructed. This channel includes a component evaluation sub-channel, a time window evaluation sub-channel, and a spatial proximity evaluation sub-channel. The component evaluation sub-channel verifies whether the components emitted by the pollution source are consistent with the components of the pollutants in the regional pollution label. The time window evaluation sub-channel verifies whether the time window of the pollution source emissions matches the time window of the regional pollution. The spatial proximity evaluation sub-channel verifies the spatial proximity of the geographical location of the pollution source emissions to the polluted area in the pollution label. When the consistency evaluation channel receives the synchronized declared emission data and label reconstruction results, it copies this data and transmits it to the component evaluation sub-channel, the time window evaluation sub-channel, and the spatial proximity evaluation sub-channel, respectively. In the component evaluation sub-channel, the total number of pollutant types is counted from the declared emission data and label reconstruction results. Then, the number of pollutant types coexisting in the declared emission data and label reconstruction results is counted. By dividing the number of coexisting pollutant types by the total number of pollutant types, a component consistency score is obtained, and this score is used to establish the first verification result. In the time window evaluation sub-channel, the intersection of the declared time window and the actual pollution time window is calculated to obtain the overlap duration. Dividing the overlap duration by the declared time window yields a time consistency score, which serves as the second verification result. In the spatial proximity evaluation sub-channel, Euclidean distance is used to calculate the distance between the coordinates of the declared emission outlet in the declared emission data and the coordinates of the pollution center in the reconstructed labeling results. Dividing the calculated distance by the tolerance deviation distance yields a spatial proximity index, which serves as the third verification result. Finally, the first, second, and third verification results are merged into a single set for output, completing the consistency verification of the declared emission data and the reconstructed labeling results, providing a basis for subsequent pollution source tracing and pollution control.

[0037] Furthermore, this application provides the method of using the identifier reconstruction results and the pollution source fusion dataset for association matching, wherein the association matching includes explicit association verification matching and implicit association identification matching, and further includes:

[0038] Enterprise retrospective analysis is performed based on the pollution source fusion dataset to obtain a retrospective dataset, which includes industry category, process type, and material usage dataset. A process emission fingerprint database is established based on the retrospective dataset. The process emission fingerprint database is used to perform implicit emission association matching of the identifier reconstruction results to complete implicit association identification matching. After performing joint matching verification of explicit association verification matching results and implicit association identification matching results, the association matching results are constructed.

[0039] Optionally, when performing implicit association identification and matching, a retrospective analysis of enterprises is first conducted based on the pollution source fusion dataset. That is, based on the enterprise names in the pollution source fusion dataset, historical data of all relevant enterprises is extracted to form a retrospective dataset. This retrospective dataset includes industry category, process type, and material usage dataset. Among them, the industry category indicates which industry the enterprise belongs to (such as chemical, metallurgical, manufacturing, etc.), and different industries have different pollution emission characteristics and types of emissions; the process type indicates the type of production process adopted by the enterprise (such as high-temperature refining, drying, chemical synthesis, etc.), and different processes will lead to different types of pollutant emissions; the material usage dataset includes data on raw materials, chemicals, and fuels used by the enterprise in the production process. This data helps to analyze the types and quantities of potential pollutants. Subsequently, based on the industry category, process type, and material usage data in the retrospective dataset, a process emission fingerprint database is established. This database categorizes and summarizes pollutants that may be generated during production under different industries, processes, and material usage methods. Each process fingerprint includes a process emission pattern and a latent emission identifier. The process emission pattern refers to the types of pollutants that may be emitted based on different process types. For example, some chemical synthesis processes may latently emit solvents and volatile organic compounds (VOCs). The latent emission identifier refers to the pollutants that may be transformed from certain raw materials or intermediate products during production, as determined by material usage data. Then, by comparing the pollutant data in the identifier reconstruction results with the potential latent emission information in the process emission fingerprint database, potential unreported pollutant emissions are identified. For example, if the identifier reconstruction results show a high concentration of a certain pollutant (such as benzene or sulfur dioxide) in a certain area, but this pollutant does not appear in the company's reported data, the process emission fingerprint database will be used to analyze the company's processes and material usage to determine whether the pollutant may be a latent emission. By comparing process emission patterns and latent emission identifiers in the process fingerprint database, potential latent pollution sources and pollutant types are further identified, forming latent association identification and matching results. Finally, the latent association identification and matching results are combined with the previously obtained explicit association verification and matching results to verify whether latent emissions match explicit emissions, ensuring the comprehensiveness and accuracy of pollution source identification. By summarizing the verification results, latent association identification and matching results, and explicit association verification and matching results, the final association matching result is constructed. This association matching result will be used for pollution source tracing, pollutant source analysis, etc., providing more comprehensive data support for pollution monitoring and control.

[0040] The source tracing and evolution authentication is performed using the correlation matching results, and the pollution monitoring results are reported based on the source tracing and evolution authentication results.

[0041] In one embodiment, after obtaining the correlation matching results, information such as the occurrence time, location, and concentration changes of pollutants are first extracted based on the previously obtained identifier reconstruction results. A pollution propagation path map is then constructed based on this information to simulate the diffusion path of pollutants within the region, revealing the propagation process from the source to various monitoring points. Subsequently, a pollution causal chain is established using the correlation matching results. This chain connects the causal relationship between pollution sources and pollution events, helping to analyze the source and propagation path of pollutants. Then, by combining the pollution causal chain with the pollution propagation path map, the consistency of pollutant diffusion is verified to ensure the accuracy of pollution source identification and pollutant propagation. Finally, the consistency verification result is output as the source evolution certification result, generating a pollution monitoring report. This report includes the location of the pollution source, the pollution diffusion path, and the spatiotemporal distribution of pollution concentration, providing a scientific basis for subsequent pollution control and decision support.

[0042] Furthermore, this application provides the aforementioned method of using correlation matching results for source tracing and evolution authentication, and reporting pollution monitoring results based on the source tracing and evolution authentication results, including:

[0043] Based on the identification reconstruction results, the spatiotemporal features of the pollution event are extracted, and a three-dimensional feature tensor of pollution time-space-concentration is constructed. Based on the three-dimensional feature tensor, a pollution propagation path map is constructed. The pollution causal chain is established using the correlation matching results, and the diffusion consistency source tracing verification of the pollution propagation path map is performed through the pollution causal chain. The diffusion consistency source tracing verification result is output as the source tracing evolution authentication result.

[0044] Preferably, the spatiotemporal features of the pollution event are first extracted from the identification reconstruction results. These features include pollution time characteristics, pollution space characteristics, and pollution concentration characteristics. Pollution time characteristics record the time points of pollutant concentration changes, which can be used to determine the occurrence time of the pollution event and the peak period of pollutant concentration. Pollution space characteristics record the spatial distribution of pollutants, which can be used to identify the location of pollution sources and the diffusion range of pollutants within a region. Pollution concentration characteristics record the numerical data of pollutants at specific times and locations, helping to reveal the severity of pollution and its changing trends. By combining these spatiotemporal features, a three-dimensional feature tensor of pollution time-space-concentration can be constructed. The dimensions of this three-dimensional feature tensor correspond to time, spatial location, and concentration value, respectively, where each data point represents the pollutant concentration at a specific time and location. Subsequently, a diffusion model (such as a Gaussian plume model or a CALPUFF model) is set up, and three-dimensional feature tensors and meteorological data (wind direction, wind speed, temperature, and humidity) are input into the model. In the model, key locations such as pollution source locations, concentration peak points, and monitoring points are marked as graph nodes, and the edges connecting the nodes represent the direction of pollutant diffusion (e.g., "A→B: diffusion in the southeast direction") and diffusion intensity (e.g., "edge weight = wind speed × diffusion time"), thereby constructing a pollution propagation path graph. Then, the pollution sources and emission behaviors (explicit / implicit associations) in the association matching results, as well as the diffusion paths in the pollution propagation path graph and the pollution events in the identification reconstruction results are linked together to construct a pollution causal chain. Then, the pollution causal chain and pollution propagation path map are subjected to diffusion consistency source tracing verification. This process compares the relative errors between the predicted pollution concentration in the pollution propagation path map and the actual monitored concentration in the pollution causal chain, determining consistency based on the relative error (whether it meets the error tolerance threshold). It also checks whether the diffusion path covers a sufficient number of actual polluted areas (whether the error between the number of covered pollution grids and the actual number of pollution grids is within the tolerance range). Finally, it compares the delay between the predicted pollution arrival time and the actual monitoring time to see if it is within the delay tolerance. If all are consistent, the diffusion consistency source tracing verification is considered successful. At this point, data such as pollution events, 3D feature tensors, propagation path maps, and pollution causal chains are integrated to generate a diffusion consistency source tracing verification result. This result is then output as the source tracing evolution certification result, providing a scientific basis for pollution source identification and pollution control.

[0045] Furthermore, this application provides that after reporting pollution monitoring results based on source tracing and evolution certification results, the following steps are included:

[0046] Obtain the pollution level and pollutants from pollution monitoring results; construct a regional map of the target area, and configure visual identifiers based on the pollution level and pollutants; embed the visual identifiers into the regional map, and then execute a visual early warning display.

[0047] Preferably, to better display pollution monitoring results, pollution levels and pollutants are extracted from the monitoring data. Pollution levels are categorized based on pollutant concentration data within the region, such as light pollution, moderate pollution, and heavy pollution. Subsequently, a regional map of the target area is constructed using a Geographic Information System (GIS), displaying the geographical boundaries, main roads, industrial areas, and other information as the foundation for pollution information visualization. Next, the locations of all monitoring points are marked on the regional map, and visual identifiers for different pollution levels and pollutants are constructed using colors and symbols. For example, light pollution might be represented by green, moderate pollution by yellow, and heavy pollution by red. The higher the pollutant concentration, the darker the color. Pollutant types can also be distinguished by different colors or symbols, such as PM2.5, PM10, NOx, and SO2. The size or shape of the symbols may be adjusted according to the pollutant concentration to highlight the type and concentration of pollutants. Finally, the visual identifiers are combined with the regional map to generate a complete visualized pollution map. This visualized pollution map can be displayed on a console, webpage, or mobile device for relevant personnel to view. This visualized pollution map can automatically generate pollution warnings. For example, if the pollution level in a certain area exceeds a predetermined threshold, relevant personnel will be alerted through flashing or highlighted visual markers on the map. Warning displays can be made through color changes, warning signs, and notification messages, helping relevant personnel understand the pollution situation and take appropriate action.

[0048] Furthermore, this application provides that the pollution monitoring results reported based on the source tracing and evolution certification results also include:

[0049] Determine whether the reconstruction result contains passive authentication data; if passive authentication data exists, start redundant observation nodes based on the contaminated nodes of the passive authentication data, use the redundant observation nodes to perform reconstruction observation, and establish reconstruction observation results; re-authenticate and manage the passive authentication data based on the reconstruction observation results.

[0050] Preferably, after obtaining the identification reconstruction results, the system checks whether passive authentication data exists in the reconstruction results. Passive authentication data refers to pollution events without clear pollution source data support. This data may be data that cannot be directly correlated with known pollution sources in pollution monitoring, such as areas with high pollutant concentrations but no corresponding pollution source emission records. If passive authentication data exists, the system will activate redundant observation nodes based on the location information of the pollution nodes that detected the passive authentication data. These redundant observation nodes can cover the blind spots of the original monitoring network, ensuring more accurate pollution data is obtained in the key areas where pollution events occur. After the redundant observation nodes are activated, reconstruction observations are performed through these nodes to collect real-time pollutant data, including pollutant concentrations, types, meteorological parameters, etc., forming reconstruction observation results. Then, based on the reconstruction observation results of the redundant observation nodes, the passive authentication data is re-authenticated. If the error between the reconstruction observation results and the passive authentication data is within the tolerance range, it indicates that they are consistent and the pollution source data is valid. Conversely, if the reconstruction observation results deviate from the tolerance range, it indicates that the passive authentication data is abnormal data. In this case, it will be corrected (replaced with the reconstruction observation results) or excluded. This series of processes ensures the accuracy and integrity of pollution monitoring data. In particular, for cases where it is difficult to directly identify pollution sources, the deployment of redundant nodes and reconstruction of observations can greatly improve the credibility of pollution source tracing.

[0051] In summary, the embodiments of this application have at least the following technical effects:

[0052] This application first accesses an industrial management database, extracts pollution source data based on the database, and establishes a pollution source fusion dataset. Then, a basic sensing grid is configured in the target area, and multi-source monitoring and identification of air pollution are performed through the basic sensing grid to establish regional pollution identifiers. Next, the flight missions of drones are configured using the regional pollution identifiers, and the drones are controlled to initiate flight operations, establishing an additional dataset. Then, the identifiers of the regional pollution identifiers are reconstructed based on the additional dataset, and the reconstructed identifiers are correlated with the pollution source fusion dataset. This correlation matching includes explicit correlation verification matching and implicit correlation identification matching. Finally, the correlation matching results are used for source tracing evolution authentication, and the pollution monitoring results are reported based on the source tracing evolution authentication results. These technical effects collectively solve the technical problems of low accuracy in dynamic source tracing of pollution sources and inability to comprehensively and accurately reflect the pollution evolution process caused by insufficient integration of multi-source heterogeneous data in existing air pollution monitoring. This achieves the technical effect of accurately monitoring and locating particulate matter pollution in the atmospheric environment through multi-dimensional data fusion and dynamic drone monitoring, thereby improving the ability to trace pollution sources.

[0053] Example 2 is based on the same inventive concept as the atmospheric particulate matter pollution monitoring method based on multi-source data fusion in the previous examples, such as... Figure 2 As shown, this application provides an atmospheric particulate matter pollution monitoring system based on multi-source data fusion. The system includes: a pollution source data extraction module 11: accessing an industrial management database, extracting pollution source data based on the industrial management database, and establishing a pollution source fusion dataset; a multi-source monitoring and identification module 12: configuring a basic sensing grid in the target area, and performing multi-source monitoring and identification of atmospheric pollution through the basic sensing grid to establish a regional pollution label; a flight operation module 13: configuring a UAV flight mission using the regional pollution label, controlling the UAV to start and execute flight operations, and establishing an additional dataset; an association matching module 14: reconstructing the label of the regional pollution label based on the additional dataset, and performing association matching between the label reconstruction result and the pollution source fusion dataset, wherein the association matching includes explicit association verification matching and implicit association identification matching; and a source tracing evolution authentication module 15: performing source tracing evolution authentication using the association matching result, and reporting the pollution monitoring result based on the source tracing evolution authentication result.

[0054] Furthermore, the multi-source monitoring and identification module 12 is also used to perform the following method:

[0055] After configuring the sensing cycle, the monitoring sensors of the basic sensing grid are activated at the cycle nodes of the sensing cycle to perform sensing monitoring and establish pollution identification for the area.

[0056] Furthermore, the flight operation module 13 is also used to perform the following methods:

[0057] The pollution level is deconstructed based on the regional pollution markers to establish a deconstruction result, which includes pollutant types, concentration ranges, and diffusion trend data. The actual distribution of monitoring points in the region is obtained based on the regional pollution markers. The importance of gray areas is identified based on the actual monitoring point distribution and the deconstruction result, and regional monitoring points are configured using the importance identification result. Flight path optimization and fitting are performed on the regional monitoring points to establish a flight mission.

[0058] Furthermore, the association matching module 14 is also used to perform the following method:

[0059] Extract the declared emission data from the pollution source fusion dataset; use the declared emission data and the identification reconstruction results to verify the consistency of pollution components, pollution time windows, and pollution spatial proximity indicators; configure explicit association scores based on the consistency verification results to complete explicit association verification matching.

[0060] Furthermore, the association matching module 14 is also used to perform the following method:

[0061] Enterprise retrospective analysis is performed based on the pollution source fusion dataset to obtain a retrospective dataset, which includes industry category, process type, and material usage dataset. A process emission fingerprint database is established based on the retrospective dataset. The process emission fingerprint database is used to perform implicit emission association matching of the identifier reconstruction results to complete implicit association identification matching. After performing joint matching verification of explicit association verification matching results and implicit association identification matching results, the association matching results are constructed.

[0062] Furthermore, the association matching module 14 is also used to perform the following method:

[0063] A consistency evaluation channel is established, comprising a component evaluation sub-channel, a time window evaluation sub-channel, and a spatial proximity evaluation sub-channel. After synchronizing the declared emission data and the label reconstruction results to the consistency evaluation channel, the component evaluation sub-channel is used to score the component consistency of the declared emission data and the label reconstruction results, establishing a first verification result. The time window evaluation sub-channel is used to analyze the time window intersection ratio of the declared emission data and the label reconstruction results, establishing a second verification result. The spatial proximity evaluation sub-channel is used to evaluate the spatial proximity index of the declared emission data and the label reconstruction results, establishing a third verification result. Consistency verification is completed based on the first, second, and third verification results.

[0064] Furthermore, the source tracing and evolution authentication module 15 is also used to perform the following methods:

[0065] Based on the identification reconstruction results, the spatiotemporal features of the pollution event are extracted, and a three-dimensional feature tensor of pollution time-space-concentration is constructed. Based on the three-dimensional feature tensor, a pollution propagation path map is constructed. The pollution causal chain is established using the correlation matching results, and the diffusion consistency source tracing verification of the pollution propagation path map is performed through the pollution causal chain. The diffusion consistency source tracing verification result is output as the source tracing evolution authentication result.

[0066] Furthermore, the source tracing and evolution authentication module 15 is also used to perform the following methods:

[0067] Obtain the pollution level and pollutants from pollution monitoring results; construct a regional map of the target area, and configure visual identifiers based on the pollution level and pollutants; embed the visual identifiers into the regional map, and then execute a visual early warning display.

[0068] Furthermore, the source tracing and evolution authentication module 15 is also used to perform the following methods:

[0069] Obtain the pollution level and pollutants from pollution monitoring results; construct a regional map of the target area, and configure visual identifiers based on the pollution level and pollutants; embed the visual identifiers into the regional map, and then execute a visual early warning display.

[0070] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0071] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0072] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. An atmospheric environment particulate pollution monitoring method based on multi-source data fusion, characterized in that, The method comprises: Accessing an industrial management database, extracting pollution source data from the industrial management database, and establishing a pollution source fusion data set; Configuring a basic perception grid in a target area, and performing multi-source monitoring and identification of atmospheric pollution through the basic perception grid to establish a regional pollution identification; After configuring the flight task of the unmanned aerial vehicle using the regional pollution identification, controlling the unmanned aerial vehicle to start and execute flight operations, and establishing an additional data set; After reconstructing the identification of the regional pollution identification according to the additional data set, performing association matching using the identification reconstruction result and the pollution source fusion data set, which comprises explicit association verification matching and implicit association identification matching; Using the association matching result to perform traceability evolution authentication, and reporting a pollution monitoring result according to the traceability evolution authentication result; The association matching using the identification reconstruction result and the pollution source fusion data set comprises explicit association verification matching and implicit association identification matching, which comprises: Extracting declared emission data in the pollution source fusion data set; Using the declared emission data and the identification reconstruction result to perform consistency verification of pollution components, pollution time windows, and pollution spatial proximity indexes; Configuring an explicit association score according to the consistency verification result to complete the explicit association verification matching; The association matching using the identification reconstruction result and the pollution source fusion data set comprises explicit association verification matching and implicit association identification matching, and further comprises: Performing enterprise traceability analysis according to the pollution source fusion data set to obtain a traceability data set, which comprises an industry category, a process type, and a material usage data set; Establishing a process emission fingerprint library according to the traceability data set; Using the process emission fingerprint library to perform implicit emission association matching of the identification reconstruction result to complete the implicit association identification matching; After performing joint matching verification of the explicit association verification matching result and the implicit association identification matching result, constructing an association matching result; The traceability evolution authentication using the association matching result and the reporting of the pollution monitoring result according to the traceability evolution authentication result comprise: Extracting the spatiotemporal characteristics of a pollution event according to the identification reconstruction result, and constructing a three-dimensional feature tensor of pollution time-space-concentration; Constructing a pollution propagation path map based on the three-dimensional feature tensor; Using the association matching result to establish a pollution causal chain, performing diffusion consistency traceability verification of the pollution propagation path map through the pollution causal chain, and outputting the diffusion consistency traceability verification result as the traceability evolution authentication result. 2.The atmospheric environment particulate pollution monitoring method based on multi-source data fusion according to claim 1, wherein, The consistency verification of pollution components, pollution time windows, and pollution spatial proximity indexes using the declared emission data and the identification reconstruction result comprises: Establishing a consistency evaluation channel, which comprises a component evaluation sub-channel, a time window evaluation sub-channel, and a spatial proximity evaluation sub-channel; After synchronizing the declared emission data and the identification reconstruction result to the consistency evaluation channel, performing component consistency scoring of the declared emission data and the identification reconstruction result using the component evaluation sub-channel to establish a first verification result; The time window evaluation sub-channel is used to evaluate the time window intersection of the declared emission data and the identification reconstruction result, and a second verification result is established; The spatial proximity evaluation sub-channel is used to evaluate the spatial proximity index of the declared emission data and the identification reconstruction result, and a third verification result is established; The consistency verification is completed according to the first verification result, the second verification result, and the third verification result. 3.The atmospheric environment particulate pollution monitoring method based on multi-source data fusion according to claim 1, wherein, The flight task of the unmanned aerial vehicle is configured according to the regional pollution identification, including: The pollution level is deconstructed according to the regional pollution identification, and a deconstruction result is established, which includes the type of pollutant, the concentration interval, and the diffusion trend data; The actual monitoring point distribution of the regional monitoring is obtained based on the regional pollution identification; The importance of the gray area is identified according to the actual monitoring point distribution and the deconstruction result, and the regional monitoring point is configured using the importance identification result; The flight path optimization fitting of the regional monitoring point is performed, and the flight task is established. 4.The atmospheric environment particulate pollution monitoring method based on multi-source data fusion according to claim 1, wherein, After the pollution monitoring result is reported according to the traceability evolution authentication result, including: The pollution level and the pollutant of the pollution monitoring result are obtained; After the regional map of the target area is constructed, the visualization identification is configured based on the pollution level and the pollutant; After the visualization identification is embedded in the regional map, the visualization early warning display is executed. 5.The atmospheric environment particulate pollution monitoring method based on multi-source data fusion according to claim 1, wherein, After the pollution monitoring result is reported according to the traceability evolution authentication result, including: It is judged whether there is sourceless authentication data in the identification reconstruction result; If there is sourceless authentication data, a redundant observation node is started according to the pollution node of the sourceless authentication data, and the reconstruction observation is performed using the redundant observation node to establish a reconstruction observation result; The sourceless authentication data is re-authenticated based on the reconstruction observation result. 6.The atmospheric environment particulate pollution monitoring method based on multi-source data fusion according to claim 1, wherein, The regional pollution identification is established by performing multi-source monitoring identification of atmospheric pollution through the basic perception grid, including: After configuring the perception period, the monitoring sensor of the basic perception grid is activated at the cycle node of the perception period to perform perception monitoring and establish the regional pollution identification.

7. The atmospheric environment particulate pollution monitoring system based on multi-source data fusion, characterized in that, The system is used to execute the atmospheric environment particulate matter pollution monitoring method based on multi-source data fusion according to any one of claims 1-6, and the system includes: A pollution source data extraction module: accessing an industrial management database, extracting pollution source data from the industrial management database, and establishing a pollution source fusion data set; A multi-source monitoring identification module: configuring a basic perception grid in a target area, and performing multi-source monitoring identification of atmospheric pollution through the basic perception grid to establish a regional pollution identification; A flight operation module: after configuring the flight task of the unmanned aerial vehicle using the regional pollution identification, controlling the unmanned aerial vehicle to start and execute the flight operation to establish an additional data set; An association matching module: after reconstructing the identification of the regional pollution identification according to the additional data set, using the identification reconstruction result and the pollution source fusion data set for association matching, the association matching includes explicit association verification matching and implicit association identification matching; A traceability evolution authentication module: using the association matching result to perform traceability evolution authentication, and reporting the pollution monitoring result according to the traceability evolution authentication result.

Images