A VOCs gas dissipation traceability accurate identification method, system, medium and device
Patent Information
- Application Number
- CN202611111677.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-24
- Publication Date
- 2026-10-27
AI Technical Summary
[0004]然而现有技术难以满足工业园区VOCs逸散溯源对快速、精准、大范围的核心需求，并且存在显著的缺陷和不足之处：
1、本发明采用地面标定空间域校正方法，依托高精度实测标定点位浓度数据计算残差、构建全域残差场模型，对遥感观测浓度数据逐像素校正，提升数据可靠性；将校正后的浓度场图像与实时气象数据耦合输入高斯扩散模型，结合遗传算法等迭代优化算法，最小化模拟场与观测场误差，精准反演单一潜在排放源的空间位置与排放强度。同时基于实时温度梯度、大气稳定度及风速数据，依据帕斯奎尔稳定度分级标准动态修正模型扩散参数，让模拟过程贴合实际大气环境条件。
Smart Images
Figure CN122990168A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of gas emission tracing technology, and in particular relates to a method, system, medium and device for accurate identification of VOCs gas emission sources. Background Technology
[0002] Volatile organic compounds (VOCs) are among the major air pollutants originating from various anthropogenic and natural sources, including industrial production and transportation. Their control has become a focal point of current environmental management because VOCs are not only key precursors to ozone and fine particulate matter (PM2.5), exacerbating photochemical smog and haze pollution, but also pose direct health risks, potentially causing respiratory irritation, nervous system damage, and even cancer.
[0003] Industrial parks are concentrated emission areas for VOCs, and effective source tracing and detection of their emitted gases are prerequisites for implementing precise supervision and pollution control. In this technological field, developing rapid, accurate, and reliable methods for tracing and detecting VOC emissions is of great significance for improving environmental management efficiency and protecting public health.
[0004] However, existing technologies are insufficient to meet the core requirements of rapid, accurate, and large-scale VOCs emission tracing in industrial parks, and have significant defects and shortcomings: (1) Sensor technology has poor selectivity and is susceptible to interference: Sensors such as photoionization detectors generally have limited ability to identify VOCs components and are susceptible to cross-interference from environmental humidity, temperature and other coexisting gases, resulting in low accuracy of measurement data and difficulty in conducting reliable source tracing analysis in complex industrial environments.
[0005] (2) Laboratory analysis techniques have long cycles and cannot respond in real time: Although methods such as GC-MS are accurate, the process is cumbersome and takes a long time from sampling to issuing results. They cannot capture the transient process of VOCs emission and provide real-time early warning, thus missing the best opportunity for source tracing.
[0006] (3) Online monitoring technology is costly and has limited coverage: Fixed online monitoring systems are expensive to operate and maintain, and their monitoring results only represent local information from a single point. Due to the vast area of the park, the limited number of monitoring points makes it difficult to fully reflect the spatial distribution of pollutants and trace their escape paths, resulting in monitoring blind spots. Summary of the Invention
[0007] In view of the shortcomings of the existing technology, the present invention provides a method, system, medium and device for accurate identification of VOCs gas emission sources.
[0008] Firstly, the method includes, Within the industrial park, calibration points are set up based on potential pollution areas and prevailing wind directions; Infrared remote sensing equipment is used to acquire infrared spectral image sequences at different altitudes and orientations, and these sequences are combined to construct three-dimensional spectral data. Based on the measured VOCs concentration data at the calibration points, the observed concentration data in the constructed three-dimensional spectral data are optimized to obtain a concentration dataset. Based on the corrected concentration dataset, the VOCs concentration field of the park is reconstructed using a three-dimensional interpolation algorithm to generate a VOCs concentration distribution image; By combining VOCs concentration distribution images with real-time meteorological data and using diffusion models for inversion calculations, the escape paths of VOCs gases are deduced, and the locations of potential emission sources are identified as selected points. Based on the model inversion results, extract the multidimensional feature vectors of selected points; The extracted multidimensional feature vectors are used to divide the pollution impact clusters, identify independent emission sources, and generate judgment results.
[0009] Furthermore, the construction of the three-dimensional spectral data specifically includes unifying all infrared spectral image sequences at different altitudes and orientations into a three-dimensional geographic coordinate system, performing spatiotemporal matching and synchronization based on high-precision timestamps, and using a three-dimensional spatial interpolation algorithm to reconstruct all discrete spatial concentration points into three-dimensional spectral data with a regular grid.
[0010] Furthermore, the acquisition of the concentration dataset specifically includes, Extract the observed concentration value of the pixel corresponding to the coordinates of each calibration point in the three-dimensional spectral data, compare it with the measured concentration value of the same point at the same time, and calculate the concentration residual. Using the spatial coordinates and residual values of all calibration points as known data, a residual field model covering the entire scanning area is constructed through a spatial interpolation algorithm; This residual field is used for three-dimensional quality control to generate a corrected concentration dataset.
[0011] Furthermore, the generation of the VOCs concentration distribution image specifically includes, Using the corrected concentration dataset as input, a three-dimensional interpolation algorithm is used to interpolate the entire target space to generate a three-dimensional VOCs concentration field; A VOCs concentration distribution image is generated by visualizing and rendering the three-dimensional VOCs concentration field.
[0012] Furthermore, the location of the potential emission source as the selected point specifically includes, The VOCs concentration distribution image is combined with real-time acquired meteorological data and input into a Gaussian diffusion model that has been dynamically corrected by meteorological parameters. By using an iterative optimization algorithm, the hypothetical potential emission source locations and source strengths are adjusted to minimize the difference between the predicted concentration field simulated by the Gaussian diffusion model and the observed concentration field. The grid centers where the optimized emission intensity is significantly higher than the background value are identified as potential emission source locations.
[0013] Secondly, the system includes: a punctuation layout module, a spectral data construction module, a concentration data acquisition module, a concentration image generation module, an inversion calculation module, a feature extraction module, and a result output module; The marker placement module is used to place marker points within the industrial park based on potential pollution areas and prevailing wind direction. The spectral data construction module is used to acquire infrared spectral image sequences at different altitudes and orientations through infrared remote sensing equipment, and combine them to construct three-dimensional spectral data. The concentration data acquisition module is used to optimize the observed concentration data in the constructed three-dimensional spectral data based on the measured VOCs concentration data at the calibration points, and to obtain a concentration dataset. The concentration image generation module is used to reconstruct the VOCs concentration field of the park using a three-dimensional interpolation algorithm based on the corrected concentration dataset, and generate a VOCs concentration distribution image. The inversion calculation module is used to combine VOCs concentration distribution images with real-time meteorological data, and use diffusion models to perform inversion calculations, deduce the emission paths of VOCs gases, and locate potential emission source locations as selected points. The feature extraction module is used to extract multidimensional feature vectors of selected points based on the model inversion results; The result output module is used to divide the extracted multidimensional feature vector into pollution impact clusters using a clustering algorithm, identify independent emission sources, and generate judgment results.
[0014] Furthermore, the spectral data construction module is specifically used to unify all infrared spectral image sequences at different altitudes and orientations into a three-dimensional geographic coordinate system, perform spatiotemporal matching and synchronization based on high-precision timestamps, and reconstruct all discrete spatial concentration points into three-dimensional spectral data with a regular grid using a three-dimensional spatial interpolation algorithm.
[0015] Furthermore, the concentration data acquisition module is specifically used for: Extract the observed concentration value of the pixel corresponding to the coordinates of each calibration point in the three-dimensional spectral data, compare it with the measured concentration value of the same point at the same time, and calculate the concentration residual. Using the spatial coordinates and residual values of all calibration points as known data, a residual field model covering the entire scanning area is constructed through a spatial interpolation algorithm; This residual field is used for three-dimensional quality control to generate a corrected concentration dataset.
[0016] Furthermore, the concentration image generation module is specifically used for, Using the corrected concentration dataset as input, a three-dimensional interpolation algorithm is used to interpolate the entire target space to generate a three-dimensional VOCs concentration field; A VOCs concentration distribution image is generated by visualizing and rendering the three-dimensional VOCs concentration field.
[0017] Furthermore, the inversion calculation module is specifically used for, The VOCs concentration distribution image is combined with real-time acquired meteorological data and input into a Gaussian diffusion model that has been dynamically corrected by meteorological parameters. By using an iterative optimization algorithm, the hypothetical potential emission source locations and source strengths are adjusted to minimize the difference between the predicted concentration field simulated by the Gaussian diffusion model and the observed concentration field. The grid centers where the optimized emission intensity is significantly higher than the background value are identified as potential emission source locations.
[0018] Thirdly, a computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements any of the steps of the VOCs gas emission tracing and detection method based on high-point infrared spectroscopy of iron towers described above.
[0019] Fourthly, an electronic device includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus. Memory, used to store computer programs; When the processor executes the program stored in the memory, it implements any of the steps of the VOCs gas emission tracing and detection method based on infrared spectroscopy at high points of iron towers described above.
[0020] Compared with the prior art, the present invention has the following advantages: 1. This invention employs a ground-based calibration spatial domain correction method. It relies on high-precision measured concentration data from calibration points to calculate residuals and construct a global residual field model. This model performs pixel-by-pixel correction on remotely sensed concentration data, improving data reliability. The corrected concentration field image is coupled with real-time meteorological data and input into a Gaussian diffusion model. Iterative optimization algorithms, such as genetic algorithms, minimize the error between the simulated and observed fields, accurately retrieving the spatial location and emission intensity of a single potential emission source. Simultaneously, based on real-time temperature gradient, atmospheric stability, and wind speed data, the model's diffusion parameters are dynamically corrected according to the Pasqual stability classification standard, ensuring the simulation process closely matches actual atmospheric environmental conditions.
[0021] 2. This invention relies on a high-tower platform to conduct multi-height layer horizontal rotation scanning and vertical elevation scanning fusion detection, collect full-domain observation data and uniformly map it to a three-dimensional geographic coordinate system, and generate regular grid three-dimensional spectral data through spatiotemporal matching and three-dimensional interpolation processing, providing complete and standardized basic data support for subsequent three-dimensional spatial source tracing analysis of pollution sources.
[0022] 3. This invention extracts multi-dimensional features such as temporal changes and spatial distribution of alarm points, key points, and sensitive points from the pollution concentration field, constructs feature vectors, and automatically divides pollution impact clusters using clustering algorithms such as K-means++, intelligently identifying the number and distribution patterns of emission sources. For complex scenarios with multiple emission sources in industrial parks, spatial grid discretization is adopted, setting emission intensity variables for each grid. Through multi-variable synchronous iterative optimization, accurate analysis and differentiation of the location and intensity of multiple independent emission sources are achieved. Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description
[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0024] Figure 1 The diagram shows a schematic flow chart of a VOCs gas emission tracing and detection method based on infrared spectroscopy at a high point on a steel tower according to the present invention.
[0025] Figure 2 The diagram shows a schematic of a VOCs gas emission tracing and detection system module based on infrared spectroscopy at high points on iron towers according to the present invention. Detailed Implementation
[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0027] like Figure 1 As shown, this invention provides a method for tracing and detecting VOCs gas emissions based on infrared spectroscopy at high points on iron towers. This method achieves blind-spot-free, rapid, and accurate source tracing and location of VOCs emissions from industrial parks, providing an efficient technical means for environmental supervision. The steps include: S1. Within the industrial park, benchmark points are set up based on potential pollution areas and prevailing wind directions.
[0028] In this embodiment, within the industrial park, based on pollution source survey data and prevailing wind direction (wind rose diagram), the present invention uses a combination of functional zone layout method and grid layout method to plan and deploy multiple calibration points with precise spatial coordinates (obtained through GPS module) in the downwind direction of major emission facilities and environmentally sensitive points; wherein, high-precision VOCs sensors will be deployed at the calibration points and synchronized with infrared remote sensing equipment in time.
[0029] S2. Using infrared remote sensing equipment, acquire infrared spectral image sequences at different altitudes and orientations, and combine them to construct three-dimensional spectral data.
[0030] In this embodiment, the present invention performs two modes of scanning by using an infrared remote sensing device deployed on a high tower platform in the park, including: (1) horizontal scanning: performing 360-degree rotation scanning at multiple preset height layers to obtain horizontal cross-sectional concentration distribution data; (2) vertical scanning: performing elevation scanning on key areas to obtain vertical cross-sectional concentration distribution data.
[0031] In this embodiment, the present invention unifies all infrared spectral image sequences from different heights and orientations into a three-dimensional geographic coordinate system, and performs spatiotemporal matching and synchronization based on high-precision timestamps. Finally, through a three-dimensional spatial interpolation algorithm, all discrete spatial concentration points are reconstructed into three-dimensional spectral data of a regular grid. Specifically, a three-dimensional spatial coordinate system is established with the tower as the origin or directly using the geodetic coordinate system. Using the azimuth, elevation, equipment height, and equipment geometric calibration parameters recorded in each frame of data, the precise three-dimensional coordinates of each data point in the unified coordinate system are calculated through a spatial forward intersection algorithm. Based on high-precision timestamps, data from different height layers and different scanning modes are time-synchronized and matched. Finally, all spatiotemporally matched three-dimensional spatial points and their concentration values are used as discrete samples, and a three-dimensional spatial interpolation algorithm is used to reconstruct them into three-dimensional spectral data of a regular grid, so that each voxel contains spatial coordinates and spectral concentration attributes.
[0032] S3. Based on the measured VOCs concentration data at the calibration points, optimize the observed concentration data in the constructed three-dimensional spectral data to obtain a concentration dataset.
[0033] In this embodiment, the present invention extracts the observed concentration value of the pixel corresponding to the coordinates of each calibration point in the three-dimensional spectral data, and compares it with the measured concentration value of the point at the same time to calculate the concentration residual; then, using the spatial coordinates and residual values of all calibration points as known data, a residual field model covering the entire scanning area is constructed by a spatial interpolation algorithm (such as the inverse distance weighting method); and this residual field is used for three-dimensional quality control (such as outlier removal) to generate a spatially consistent corrected concentration dataset.
[0034] Alternatively, when constructing the residual field or three-dimensional concentration field, other spatial interpolation algorithms such as Kriging interpolation, radial basis function interpolation, and natural neighbor interpolation can be used to replace the inverse distance weighted method.
[0035] S4. Based on the corrected concentration dataset, reconstruct the VOCs concentration field of the park using a three-dimensional interpolation algorithm to generate a VOCs concentration distribution image.
[0036] In this embodiment, the present invention takes the corrected concentration dataset as input and uses a three-dimensional interpolation algorithm (such as natural neighbor interpolation) to perform interpolation calculations on the entire target space to generate a three-dimensional raster model (i.e., a three-dimensional VOCs concentration field) composed of regular voxels; and generates a VOCs concentration distribution image that can intuitively and accurately reflect the three-dimensional spatial distribution characteristics of pollutants by visually rendering the three-dimensional raster model (such as using isosurface rendering technology).
[0037] S5. Combine the VOCs concentration distribution image with real-time meteorological data, and use the diffusion model to perform inversion calculations to deduce the VOCs gas emission path and locate the potential emission source as the selected point.
[0038] In this embodiment, the present invention combines VOCs concentration distribution images with real-time acquired meteorological data (wind speed, wind direction, temperature gradient, atmospheric stability) and inputs them into a Gaussian diffusion model that has been dynamically corrected by meteorological parameters. Through iterative optimization algorithms, the hypothetical potential emission source locations and source intensities are adjusted to minimize the difference between the predicted concentration field simulated by the model and the observed concentration field. Finally, the grid center where the optimized emission intensity is significantly higher than the background value is located as the potential emission source location.
[0039] Alternatively, the iterative optimization algorithm may employ other global or local optimization algorithms such as genetic algorithms, particle swarm optimization, simulated annealing, Levenberg-Marquardt algorithms, etc.
[0040] Optionally, the core source tracing model of this invention is not limited to the Gaussian diffusion model; a Lagrange particle diffusion model or a computational fluid dynamics model can also be used as the engine for reverse source tracing, with the corrected three-dimensional concentration field and meteorological data as input, and the location of the pollution source can be located through inversion calculation.
[0041] S6. Based on the model inversion results, extract the multidimensional feature vectors of the selected points.
[0042] In this embodiment, based on the model inversion results, the present invention extracts the features of three types of key points from the VOCs concentration distribution image: (1) alarm points: points that continuously exceed the standard; (2) key points: sampling points on the emission path; (3) sensitive points: preset environmental sensitive points.
[0043] Optionally, for the selected points mentioned above, a concentration time series is constructed, and multidimensional features such as concentration change trend (e.g., average concentration, change slope), spatial distribution characteristics (e.g., three-dimensional coordinates, distance from emission sources) and time series response characteristics (e.g., variance, correlation with wind direction) are extracted from it.
[0044] S7. For the extracted multidimensional feature vectors, a clustering algorithm is used to divide the pollution impact clusters, identify independent emission sources, and generate judgment results.
[0045] In this embodiment, the present invention employs clustering algorithms (such as K-means++, DBSCAN, hierarchical clustering, Gaussian mixture model, and other unsupervised clustering algorithms) to perform cluster analysis on the multidimensional feature vectors of all selected points extracted in step S6, automatically dividing them into different pollution impact clusters. By calculating the spatial centroid of each cluster and analyzing the synchronicity of concentration changes within the cluster, the invention intelligently identifies whether there is a single emission source or multiple independent emission sources within the park, and outputs the classification results of their quantity and spatial distribution, providing direct basis for regulatory decisions.
[0046] like Figure 2 As shown, this invention also proposes a VOCs gas emission tracing and detection system based on infrared spectroscopy at high points on iron towers, which includes a marker layout module, a spectral data construction module, a concentration data acquisition module, a concentration image generation module, an inversion calculation module, a feature extraction module, and a result output module.
[0047] 1. The marker placement module is used to place marker points within an industrial park based on potential pollution areas and prevailing wind direction.
[0048] In this embodiment, within the industrial park, the marker deployment module, based on pollution source survey data and prevailing wind direction (wind rose diagram), uses a combination of functional zone deployment and grid deployment methods to plan and deploy multiple calibration points with precise spatial coordinates (obtained via GPS module) downwind of major emission facilities and environmentally sensitive points. Among these, high-precision VOCs sensors will be deployed at the calibration points and synchronized with infrared remote sensing equipment in time.
[0049] 2. The spectral data construction module is used to acquire infrared spectral image sequences at different altitudes and orientations through infrared remote sensing equipment, and combine them to construct three-dimensional spectral data.
[0050] In this embodiment, the spectral data construction module performs two scanning modes using infrared remote sensing equipment deployed on a high tower platform within the park, including: (1) Horizontal scanning: performing 360-degree rotational scanning at multiple preset height layers to obtain horizontal cross-sectional concentration distribution data; (2) Vertical scanning: performing elevation scanning on key areas to obtain vertical profile concentration distribution data.
[0051] In this embodiment, the spectral data construction module unifies all scan data into a three-dimensional geographic coordinate system and performs spatiotemporal matching and synchronization based on high-precision timestamps. Finally, through a three-dimensional spatial interpolation algorithm, all discrete spatial concentration points are reconstructed into three-dimensional spectral data with a regular grid. Specifically, a three-dimensional spatial coordinate system is established with the tower as the origin or directly using the geodetic coordinate system. Using the azimuth, elevation, equipment height, and equipment geometric calibration parameters recorded in each frame of data, the precise three-dimensional coordinates of each data point in the unified coordinate system are calculated through a spatial forward intersection algorithm. Based on high-precision timestamps, data from different height layers and different scanning modes are time-synchronized and matched. Finally, all spatiotemporally matched three-dimensional spatial points and their concentration values are used as discrete samples and reconstructed into three-dimensional spectral data with a regular grid using a three-dimensional spatial interpolation algorithm, so that each voxel contains spatial coordinates and spectral concentration attributes.
[0052] 3. The concentration data acquisition module is used to optimize the observed concentration data in the constructed three-dimensional spectral data based on the measured VOCs concentration data at the calibration points, and to obtain the concentration dataset.
[0053] In this embodiment, the concentration data acquisition module extracts the pixel observation concentration value corresponding to the coordinates of each calibration point in the three-dimensional spectral data, and compares it with the measured concentration value of the same point at the same time to calculate the concentration residual. Then, using the spatial coordinates and residual values of all calibration points as known data, a residual field model covering the entire scanning area is constructed through a spatial interpolation algorithm (such as the inverse distance weighting method). This residual field is then used for three-dimensional quality control (such as outlier removal) to generate a spatially consistent corrected concentration dataset.
[0054] Alternatively, when constructing the residual field or three-dimensional concentration field, other spatial interpolation algorithms such as Kriging interpolation, radial basis function interpolation, and natural neighbor interpolation can be used to replace the inverse distance weighted method.
[0055] 4. The concentration image generation module is used to reconstruct the VOCs concentration field of the park through a three-dimensional interpolation algorithm based on the corrected concentration dataset, and generate a VOCs concentration distribution image.
[0056] In this embodiment, the concentration image generation module takes the corrected concentration dataset as input, uses a three-dimensional interpolation algorithm (such as natural neighbor interpolation) to perform interpolation calculations on the entire target space, and generates a three-dimensional raster model (i.e., a three-dimensional VOCs concentration field) composed of regular voxels; and generates a VOCs concentration distribution image that can intuitively and accurately reflect the three-dimensional spatial distribution characteristics of pollutants by visually rendering the three-dimensional raster model (such as using isosurface rendering technology).
[0057] 5. The inversion calculation module is used to combine VOCs concentration distribution images with real-time meteorological data, and use diffusion models to perform inversion calculations to deduce the emission paths of VOCs gases and locate potential emission source locations as selected points.
[0058] In this embodiment, the inversion calculation module combines the VOCs concentration distribution image with real-time acquired meteorological data (wind speed, wind direction, temperature gradient, atmospheric stability) and inputs it into a Gaussian diffusion model that has been dynamically corrected by meteorological parameters. Through iterative optimization algorithms, the assumed potential emission source locations and source intensities are adjusted to minimize the difference between the predicted concentration field simulated by the model and the observed concentration field. Finally, the grid center where the optimized emission intensity is significantly higher than the background value is located as the potential emission source location.
[0059] Alternatively, the iterative optimization algorithm may employ other global or local optimization algorithms such as genetic algorithms, particle swarm optimization, simulated annealing, Levenberg-Marquardt algorithms, etc.
[0060] Optionally, the core source tracing model of the inversion calculation module is not limited to the Gaussian diffusion model; a Lagrange particle diffusion model or a computational fluid dynamics model can also be used as the engine for reverse source tracing, with the corrected three-dimensional concentration field and meteorological data as input, and the pollution source location can be located through inversion calculation.
[0061] 6. The feature extraction module is used to extract multidimensional feature vectors of selected points based on the inversion results.
[0062] In this embodiment, the feature extraction module extracts features of three types of key points from the VOCs concentration distribution image based on the model inversion results: (1) alarm points: points that continuously exceed the standard; (2) key points: sampling points on the emission path; (3) sensitive points: preset environmental sensitive points.
[0063] Optionally, for the selected points mentioned above, a concentration time series is constructed, and multidimensional features such as concentration change trend (e.g., average concentration, change slope), spatial distribution characteristics (e.g., three-dimensional coordinates, distance from emission sources) and time series response characteristics (e.g., variance, correlation with wind direction) are extracted from it.
[0064] 7. The results generation module performs cluster analysis on the extracted multidimensional feature vectors, uses clustering algorithms to divide pollution impact clusters, identifies independent emission sources, and generates judgment results.
[0065] In this embodiment, the result generation module uses clustering algorithms (such as K-means++, DBSCAN, hierarchical clustering, Gaussian mixture model, and other unsupervised clustering algorithms) to perform cluster analysis on the multidimensional feature vectors of all selected points extracted in step S6, automatically dividing them into different pollution impact clusters. By calculating the spatial centroid of each cluster and analyzing the synchronicity of concentration changes within the cluster, the module intelligently identifies whether there is a single emission source or multiple independent emission sources in the park, and outputs the classification results of their quantity and spatial distribution, providing a direct basis for regulatory decisions.
[0066] The foregoing description and accompanying drawings fully illustrate embodiments of the invention to enable those skilled in the art to practice them. Other embodiments may include structural and other changes. The embodiments represent only possible variations. Individual components and functions are optional unless explicitly required, and the order of operation may vary. Some portions and features of some embodiments may be included or substituted for portions and features of other embodiments. Embodiments of the invention are not limited to the structures described above and shown in the accompanying drawings, and various modifications and changes may be made without departing from their scope. The scope of the invention is limited only by the appended claims.
Claims
1. A method for tracing and detecting VOCs gas emission sources based on infrared spectroscopy at high points on iron towers, characterized in that, The method includes, Within the industrial park, calibration points are set up based on potential pollution areas and prevailing wind directions; Infrared remote sensing equipment is used to acquire infrared spectral image sequences at different altitudes and orientations, and these sequences are combined to construct three-dimensional spectral data. Based on the measured VOCs concentration data at the calibration points, the observed concentration data in the constructed three-dimensional spectral data are optimized to obtain a concentration dataset. Based on the corrected concentration dataset, the VOCs concentration field of the park is reconstructed using a three-dimensional interpolation algorithm to generate a VOCs concentration distribution image; By combining VOCs concentration distribution images with real-time meteorological data and using diffusion models for inversion calculations, the escape paths of VOCs gases are deduced, and the locations of potential emission sources are identified as selected points. Based on the model inversion results, extract the multidimensional feature vectors of selected points; The extracted multidimensional feature vectors are used to divide the pollution impact clusters, identify independent emission sources, and generate judgment results.
2. The method for tracing and detecting VOCs gas emission based on high-point infrared spectroscopy of a steel tower according to claim 1, characterized in that, The construction of the three-dimensional spectral data specifically includes unifying all infrared spectral image sequences at different altitudes and orientations into a three-dimensional geographic coordinate system, performing spatiotemporal matching and synchronization based on high-precision timestamps, and using a three-dimensional spatial interpolation algorithm to reconstruct all discrete spatial concentration points into three-dimensional spectral data with a regular grid.
3. The method for tracing and detecting VOCs gas emission based on high-point infrared spectroscopy of a steel tower according to claim 1, characterized in that, The acquisition of the concentration dataset specifically includes, Extract the observed concentration value of the pixel corresponding to the coordinates of each calibration point in the three-dimensional spectral data, compare it with the measured concentration value of the same point at the same time, and calculate the concentration residual. Using the spatial coordinates and residual values of all calibration points as known data, a residual field model covering the entire scanning area is constructed through a spatial interpolation algorithm; This residual field is used for three-dimensional quality control to generate a corrected concentration dataset.
4. The method for tracing and detecting VOCs gas emission based on high-point infrared spectroscopy of a steel tower according to claim 1, characterized in that, The generation of the VOCs concentration distribution image specifically includes, Using the corrected concentration dataset as input, a three-dimensional interpolation algorithm is used to interpolate the entire target space to generate a three-dimensional VOCs concentration field; A VOCs concentration distribution image is generated by visualizing and rendering the three-dimensional VOCs concentration field.
5. The method for tracing and detecting VOCs gas emission based on high-point infrared spectroscopy of a steel tower according to claim 1, characterized in that, The location of the potential emission source is selected as the point, specifically including: The VOCs concentration distribution image is combined with real-time acquired meteorological data and input into a Gaussian diffusion model that has been dynamically corrected by meteorological parameters. By using an iterative optimization algorithm, the hypothetical potential emission source locations and source strengths are adjusted to minimize the difference between the predicted concentration field simulated by the Gaussian diffusion model and the observed concentration field. The grid centers where the optimized emission intensity is significantly higher than the background value are identified as potential emission source locations.
6. A VOCs gas emission tracing and detection system based on infrared spectroscopy at high points on iron towers, characterized in that, The system includes: a punctuation layout module, a spectral data construction module, a concentration data acquisition module, a concentration image generation module, an inversion calculation module, a feature extraction module, and a result output module; The marker placement module is used to place marker points within the industrial park based on potential pollution areas and prevailing wind direction. The spectral data construction module is used to acquire infrared spectral image sequences at different altitudes and orientations through infrared remote sensing equipment, and combine them to construct three-dimensional spectral data. The concentration data acquisition module is used to optimize the observed concentration data in the constructed three-dimensional spectral data based on the measured VOCs concentration data at the calibration points, and to obtain a concentration dataset. The concentration image generation module is used to reconstruct the VOCs concentration field of the park using a three-dimensional interpolation algorithm based on the corrected concentration dataset, and generate a VOCs concentration distribution image. The inversion calculation module is used to combine VOCs concentration distribution images with real-time meteorological data, and use diffusion models to perform inversion calculations, deduce the emission paths of VOCs gases, and locate potential emission source locations as selected points. The feature extraction module is used to extract multidimensional feature vectors of selected points based on the model inversion results; The result output module is used to divide the extracted multidimensional feature vector into pollution impact clusters using a clustering algorithm, identify independent emission sources, and generate judgment results.
7. The VOCs gas emission tracing and detection system based on high-point infrared spectroscopy of iron towers according to claim 6, characterized in that, The spectral data construction module is specifically used to unify all infrared spectral image sequences at different altitudes and orientations into a three-dimensional geographic coordinate system, perform spatiotemporal matching and synchronization based on high-precision timestamps, and reconstruct all discrete spatial concentration points into regular grid three-dimensional spectral data using a three-dimensional spatial interpolation algorithm.
8. The VOCs gas emission tracing and detection system based on high-point infrared spectroscopy of iron towers according to claim 6, characterized in that, The concentration data acquisition module is specifically used for, Extract the observed concentration value of the pixel corresponding to the coordinates of each calibration point in the three-dimensional spectral data, compare it with the measured concentration value of the same point at the same time, and calculate the concentration residual. Using the spatial coordinates and residual values of all calibration points as known data, a residual field model covering the entire scanning area is constructed through a spatial interpolation algorithm; This residual field is used for three-dimensional quality control to generate a corrected concentration dataset.
9. The VOCs gas emission tracing and detection system based on high-point infrared spectroscopy of iron towers according to claim 6, characterized in that, The concentration image generation module is specifically used for, Using the corrected concentration dataset as input, a three-dimensional interpolation algorithm is used to interpolate the entire target space to generate a three-dimensional VOCs concentration field; A VOCs concentration distribution image is generated by visualizing and rendering the three-dimensional VOCs concentration field.
10. The VOCs gas emission tracing and detection system based on high-point infrared spectroscopy of iron towers according to claim 6, characterized in that, The inversion calculation module is specifically used for, The VOCs concentration distribution image is combined with real-time acquired meteorological data and input into a Gaussian diffusion model that has been dynamically corrected by meteorological parameters. By using an iterative optimization algorithm, the hypothetical potential emission source locations and source strengths are adjusted to minimize the difference between the predicted concentration field simulated by the Gaussian diffusion model and the observed concentration field. The grid centers where the optimized emission intensity is significantly higher than the background value are identified as potential emission source locations.
11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the VOCs gas emission tracing and detection method based on high-point infrared spectroscopy of a steel tower as described in any one of claims 1-5.
12. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; When the processor executes the program stored in the memory, it implements the steps of the VOCs gas emission tracing and detection method based on the infrared spectrum of the high point of the iron tower as described in any one of claims 1-5.