Traceability analysis method and system for atmospheric pollutants

Abstract

The invention provides an atmospheric pollutant traceability analysis method and system. According to the method, a form classification standard is established through particle shape features, and atmospheric particles are divided into naturally raised dust and industrial emission particle swarms. A dual-wavelength laser technology is adopted for natural flying dust, specifically, the first wavelength is matched with the inherent frequency of mineral crystals, amorphous scattering interference is blocked through polarization beam splitting, and a reflection signal sequence is extracted to generate a mineral migration path; and the second wavelength screens heavy metal scattering signals of the industrial particles, and a heavy metal diffusion track is constructed by utilizing spatial intensity gradient. And superposing the two paths and aerosol three-dimensional distribution data to generate a thermal distribution diagram, and when the space density exceeds a threshold value, calculating a migration path deviation value and a trajectory offset, dynamically distributing pollution source contribution weights, and outputting a pollution source analysis result. According to the invention, dynamic and accurate analysis of natural and industrial pollution sources is realized, and the regional pollution traceability and management and control strategy targeting are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of pollution source analysis, and particularly relates to an atmospheric pollution source analysis method and system. BACKGROUND

[0002] In the mixed pollution scene of natural sources (such as dust and biomass burning) and anthropogenic sources (such as industrial emissions and vehicle exhaust), the complexity of pollution sources is significantly increased, and the traditional source analysis method is difficult to accurately distinguish the dynamic contribution ratio of the two types of pollution sources due to the dependence on single chemical component analysis or static source spectrum library. At present, it is urgent to build a high-resolution analysis system based on the multi-dimensional characteristics of particulate matter to support differentiated control strategies. At the same time, it is necessary to break through the bottleneck of morphology characterization technology of ultrafine particles and solve the problem of heterogeneous fusion of multi-modal data to improve the reliability and timeliness of the source tracing results.

[0003] The current mainstream scheme adopts a multi-modal sensing and deep learning collaborative architecture: the micro-morphology and element composition of particulate matter are extracted through a scanning electron microscope and energy spectrum analysis, and organic molecular fingerprints are captured by combining single-particle aerosol mass spectrometry. In the feature fusion layer, the morphological classification features in the scanning electron microscope image are analyzed based on the residual neural network optimized by transfer learning, and a morphology-chemistry correlation matrix is constructed using a graph convolution network to screen key discriminators of natural sources and anthropogenic sources. The dynamic analysis layer integrates a positive matrix decomposition model and an adaptive clustering algorithm, and simultaneously realizes real-time response of local pollution hotspots through edge computing nodes. SUMMARY

[0004] The present application provides an atmospheric pollution source analysis method and system to solve the problems of unclear classification of natural and industrial pollution sources, inaccurate migration path tracking and poor dynamic weight adaptation in the prior art.

[0005] In a first aspect, the present application provides an atmospheric pollution source analysis method, comprising:

[0006] Collecting shape feature parameters of atmospheric particulate matter, simultaneously establishing a morphological classification standard of the atmospheric particulate matter according to the shape feature parameters, and dividing the atmospheric particulate matter into a natural dust particle group and an industrial emission particle group according to the morphological classification standard;

[0007] Using a dual-wavelength laser technology on the natural dust particle group, wherein the inherent frequency of mineral crystals in the natural dust particle group is matched by a first wavelength laser to enhance the reflection signal, and the scattering interference of amorphous substances in the natural dust particle group is blocked by a polarization light splitting technology to extract a reflection signal sequence and obtain the time variation law of the reflection signal sequence, and a mineral migration path of the natural dust particle group is generated according to the time variation law;

[0008] Screening the scattering signal specific to the heavy metal element in the industrial emission particle group by the second wavelength laser, acquiring the spatial intensity gradient of the scattering signal, and generating the heavy metal diffusion track in the industrial pollution characteristic spectrum according to the spatial intensity gradient;

[0009] Superimposing the mineral migration path and the heavy metal diffusion track respectively with the aerosol three-dimensional distribution data acquired by the high-altitude detection device to generate the thermal distribution map of the pollution source;

[0010] When the spatial distribution density of the natural dust-raising particle group and the industrial emission particle group and the corresponding spatial distribution density in the thermal distribution map exceed a preset threshold, extracting the distribution deviation amount of the mineral migration path and the shift amount of the heavy metal diffusion track, re-distributing the pollution source contribution weight of the natural dust-raising particle group and the industrial emission particle group, and outputting the pollution source analysis result.

[0011] Optionally, the mineral migration path and the heavy metal diffusion track are superimposed respectively with the aerosol three-dimensional distribution data acquired by the high-altitude detection device to generate the thermal distribution map of the pollution source, including:

[0012] Acquiring the spatial coordinate point set of the mineral migration path and the spatial coordinate point set of the heavy metal diffusion track, simultaneously extracting the three-dimensional spatial coordinates and the corresponding aerosol concentration value in the aerosol three-dimensional distribution data collected by the high-altitude detection device, and aligning the spatial positions of the spatial coordinate points of the mineral migration path, the spatial coordinate points of the heavy metal diffusion track, and the spatial coordinate points corresponding to the aerosol concentration value;

[0013] Dividing the mineral migration path coordinate point set according to the spatial grid to count the density value of the mineral migration path in each spatial grid, simultaneously dividing the heavy metal diffusion track coordinate point set according to the same spatial grid to count the density value of the diffusion track in each spatial grid, and mapping the aerosol concentration value to the corresponding spatial grid;

[0014] Superimposing and calculating the density value of the mineral migration path, the density value of the heavy metal diffusion track, and the aerosol concentration value in each spatial grid to generate the pollution thermal value;

[0015] Connecting the pollution thermal values of all spatial grids according to the spatial coordinate sequence to construct the thermal distribution map of the pollution source in the three-dimensional space.

[0016] Optionally, the shape feature parameters of the atmospheric particulate matter are collected, the shape classification standard of the atmospheric particulate matter is established according to the shape feature parameters, and the atmospheric particulate matter is divided into the natural dust-raising particle group and the industrial emission particle group according to the shape classification standard, including:

[0017] The surface profile images of the atmospheric particulate matter are continuously acquired by a microscopic imaging device, and the edge sawtooth shape and surface concave-convex interval of the atmospheric particulate matter are extracted from the surface profile images as shape feature parameters;

[0018] The atmospheric particulate matter is subjected to energy spectrum analysis to detect the composition ratio of mineral elements and artificial synthetic elements inside the atmospheric particulate matter, and when the ratio of the mineral elements exceeds a set threshold value, it is marked as natural attribute, and when the ratio of the artificial synthetic elements exceeds a set threshold value, it is marked as industrial attribute;

[0019] The edge sawtooth shape in the shape feature parameters is matched with the natural attribute mark of the energy spectrum analysis, and the surface concave-convex interval is matched with the industrial attribute mark to generate a morphological classification standard;

[0020] According to the matching degree of the edge sawtooth shape and the natural attribute and the matching degree of the surface concave-convex interval and the industrial attribute in the morphological classification standard, the particulate matter is divided into a natural dust particulate group or an industrial emission particulate group.

[0021] Optionally, the atmospheric particulate matter is subjected to energy spectrum analysis to detect the composition ratio of mineral elements and artificial synthetic elements inside the atmospheric particulate matter, and when the ratio of the mineral elements exceeds a set threshold value, it is marked as natural attribute, and when the ratio of the artificial synthetic elements exceeds a set threshold value, it is marked as industrial attribute, comprising:

[0022] The atmospheric particulate matter is placed in the detection area of the energy spectrum analysis device, and the energy spectrum scanning probe of the energy spectrum analysis device is started to continuously irradiate the surface of the atmospheric particulate matter to collect characteristic energy spectrum signals of the elements inside the atmospheric particulate matter;

[0023] The characteristic energy spectrum signals are subjected to filtering and denoising processing, and the waveform of the characteristic energy spectrum signals is extracted, and the peak value corresponding to the mineral elements and the peak value corresponding to the artificial synthetic elements are extracted from the waveform to generate energy spectrum peak value signals of the atmospheric particulate matter;

[0024] The integral area of the peak value corresponding to the mineral elements in the energy spectrum peak value signals is compared with the total integral area of the peak values of all elements to calculate the ratio value of the mineral elements, and the integral area of the peak value corresponding to the artificial synthetic elements is compared with the total integral area to calculate the ratio value of the artificial synthetic elements;

[0025] The atmospheric particulate matter with a ratio value of the mineral elements exceeding a preset natural attribute threshold value is marked as natural attribute, and the atmospheric particulate matter with a ratio value of the artificial synthetic elements exceeding a preset industrial attribute threshold value is marked as industrial attribute.

[0026] Optionally, a double-wavelength laser technology is used for the natural dust particles, wherein a first wavelength laser is used to match the natural frequency of the mineral crystals in the natural dust particles to enhance the reflection signal, and a polarization light splitting technology is used to block the scattering interference of the amorphous substances in the natural dust particles to extract the reflection signal sequence and obtain the time variation of the reflection signal sequence, and the time variation is used to generate the mineral migration path of the natural dust particles, including:

[0027] A first wavelength laser beam in the double-wavelength laser technology is projected on the natural dust particles, and the frequency of the first wavelength laser beam is adjusted to the natural vibration frequency range of the mineral crystals to enhance the reflection signal intensity of the mineral crystals, and a polarization light splitting technology is used to block the scattering signal of the amorphous substance particles to separate the reflection signal of the mineral crystals;

[0028] The reflection signal is collected in time sequence to form a continuous time sequence of reflection signal sequences;

[0029] The time variation of the reflection signal intensity in the reflection signal sequence is extracted, and the wave peak and wave trough of the reflection signal intensity are obtained according to the time variation;

[0030] The wave peak and wave trough correspond to the position migration nodes of the natural dust particles in the airflow, and the time interval of the position migration nodes and the displacement direction of the particle group are obtained;

[0031] The time interval of the position migration nodes and the displacement direction of the particle group are integrated to generate the mineral migration path.

[0032] Optionally, a second wavelength laser is used to screen the scattering signal specific to the heavy metal elements in the industrial emission particles, and the spatial intensity gradient of the scattering signal is obtained, and the heavy metal diffusion track in the industrial pollution characteristic spectrum is generated according to the spatial intensity gradient, including:

[0033] The second wavelength laser is projected on the industrial emission particles, and the wavelength of the second wavelength laser is adjusted to the preset optical resonance frequency of the heavy metal elements to enhance the scattering signal intensity of the atmospheric particulate matter containing the heavy metal elements;

[0034] An optical detector is moved in the area covered by the second wavelength laser to collect the data of the scattering signal intensity at different spatial positions, and the spatial intensity gradient of the scattering signal is generated according to the spatial distribution difference of the data of the scattering signal intensity at different spatial positions;

[0035] The high-value area of the scattering signal intensity in the spatial intensity gradient is obtained, and the change direction of the re-scattering signal intensity of the adjacent high-value area is taken as the diffusion direction to generate the heavy metal diffusion track.

[0036] Optionally, when the spatial distribution density of the natural dust raising particle group and the industrial emission particle group and the corresponding spatial distribution density in the thermal distribution map exceed a preset threshold, the distribution deviation amount of the mineral migration path and the offset amount of the heavy metal diffusion trajectory are extracted, the pollution source contribution weight of the natural dust raising particle group and the industrial emission particle group is redistributed, and a pollution source analysis result is output, including:

[0037] The spatial distribution density of the natural dust raising particle group and the industrial emission particle group in the corresponding area of the thermal distribution map is obtained, and the difference between the spatial distribution density of the natural dust raising particle group and the industrial emission particle group in each area and the corresponding spatial distribution density in the thermal distribution map is calculated. When the difference exceeds a preset threshold, it is marked as an abnormal density area;

[0038] The mineral migration path of the natural dust raising particle group is extracted in the abnormal density area, and the distribution deviation amount of the mineral migration path is calculated. The heavy metal diffusion trajectory of the industrial emission particle group is extracted synchronously, and the offset amount of the heavy metal diffusion trajectory is calculated;

[0039] According to the proportional relationship of the distribution deviation amount and the offset amount, the pollution source contribution weight of the natural dust raising particle group and the industrial emission particle group in the thermal distribution map is adjusted. The proportion of the distribution deviation amount is inversely proportional to the weight of the natural dust raising particle group, and the proportion of the offset amount is inversely proportional to the weight of the industrial emission particle group;

[0040] The adjusted pollution source contribution weight is fed back to the generation process of the thermal distribution map, and the spatial distribution density of the particle group and the corresponding relationship with the thermal distribution map are recalculated according to the change of the weight of the natural dust raising particle group and the industrial emission particle group, to form a pollution source analysis result.

[0041] In a second aspect, the present application provides an atmospheric pollutant tracing analysis system, comprising:

[0042] The division module is used for collecting shape feature parameters of atmospheric particulate matters, establishing a shape classification standard of the atmospheric particulate matters according to the shape feature parameters, and dividing the atmospheric particulate matters into a natural dust raising particle group and an industrial emission particle group according to the shape classification standard;

[0043] The generation module is used for adopting a dual-wavelength laser technology on the natural dust raising particle group. The first wavelength laser is used to match the inherent frequency of mineral crystals in the natural dust raising particles to enhance the reflection signal, and the polarization light splitting technology is used to block the scattering interference of amorphous substances in the natural dust raising particle group, so as to extract a reflection signal sequence and obtain the time variation law of the reflection signal sequence. The mineral migration path of the natural dust raising particle group is generated according to the time variation law;

[0044] The screening module is configured to screen a scattering signal specific to a heavy metal element in the industrial emission particle group by using a second wavelength laser, acquire a spatial intensity gradient of the scattering signal, and generate a heavy metal diffusion track in an industrial pollution characteristic spectrum according to the spatial intensity gradient.

[0045] The superimposition module is configured to superimpose the mineral migration path and the heavy metal diffusion track with aerosol three-dimensional distribution data acquired by the high-altitude detection device to generate a thermal distribution map of the pollution source.

[0046] The distribution module is configured to extract a distribution deviation amount of the mineral migration path and a shift amount of the heavy metal diffusion track when the spatial distribution density of the natural dust raising particle group and the industrial emission particle group and the corresponding spatial distribution density in the thermal distribution map exceed a preset threshold, re-distribute pollution source contribution weights of the natural dust raising particle group and the industrial emission particle group, and output a pollution source analysis result.

[0047] In a third aspect, an embodiment of the present application provides a computing device, including a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the atmospheric pollutant source tracing analysis method in the first aspect.

[0048] In a fourth aspect, an embodiment of the present application provides a computer storage medium, which stores a computer program, and when the computer program is executed by a computer, the atmospheric pollutant source tracing analysis method in the first aspect is implemented.

[0049] In the technical solution of the present application, a full-link technical system for intelligent analysis of atmospheric pollution sources is constructed, and the visualization monitoring of particle source tracing and diffusion process is realized. Through the fusion classification of multi-dimensional morphological feature analysis and spectral element detection, the natural dust raising particle group and the industrial emission particle group are accurately distinguished. Based on the laser resonance and polarization spectroscopy technology, the optical signals of specific types of particles are enhanced, the mineral migration path is inverted, and the dynamic tracking of the heavy metal diffusion track is realized. Combined with the three-dimensional space superposition of high-altitude three-dimensional detection data, the multi-source pollution thermal distribution is generated, and the dimensional limitation of traditional plane monitoring is broken. Through the pollution contribution weight dynamic distribution mechanism triggered by the density threshold, the adaptive correction model of pollution source analysis is established, and the source tracing problem of multi-source pollution cross interference in a complex scene is effectively solved, thereby providing a high-precision, full-factor decision support scheme for atmospheric pollution prevention and control.

[0050] Further, the visualization modeling of three-dimensional spatial distribution of atmospheric pollution and the comprehensive quantitative analysis of multi-source pollution contribution are realized. Through multi-source spatial coordinate alignment technology, the mineral migration path, heavy metal diffusion trajectory and aerosol concentration data are integrated into a unified three-dimensional coordinate system, breaking through the data fragmentation limitation of traditional single-dimensional pollution monitoring. The spatial gridding density statistical method is used to convert the discrete migration path and diffusion trajectory into continuous spatial distribution characteristics, and the pollution thermal value model is constructed by combining the aerosol concentration superposition calculation. The three-dimensional thermal distribution map is generated based on the spatial connection technology of gridding data, realizing the stereoscopic presentation of pollution intensity and source type. This multi-parameter fusion spatial thermal modeling method effectively solves the problem of fuzzy spatial contribution of pollution sources in traditional two-dimensional pollution distribution map, and provides high-resolution spatial decision basis for atmospheric pollution source tracing and governance.

[0051] These aspects or other aspects of the present application will be more apparent in the following description of the embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0053] Figure 1 A flow chart of an atmospheric pollutant source tracing analysis method provided by the present application is shown;

[0054] Figure 2 A scene diagram of an atmospheric pollutant source tracing analysis system provided by the present application is shown;

[0055] Figure 3 A scene diagram of an atmospheric pollutant source tracing analysis system provided by the present application is shown;

[0056] Figure 4 A structural schematic diagram of an atmospheric pollutant source tracing analysis system provided by the present application is shown;

[0057] Figure 5 A structural schematic diagram of a computing device provided by the present application is shown. DETAILED DESCRIPTION

[0058] In order to make the person skilled in the art better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application.

[0059] In some of the flowcharts described in this specification and the accompanying drawings, and in the claims, operations can be described in a particular, sequential order, which should not be understood as a requirement that such operations be performed in the described order, or in sequential order at all. Rather, the operations can be performed in any order or simultaneously, unless otherwise specified. Additionally, some of the flowcharts can include more or fewer operations than those depicted and described, and these operations can be performed in serial, in parallel, or in some other manner. Further, the descriptions of the “first,” “second,” etc. messages, devices, modules, etc., are used for identification purposes only and are not meant to be taken literally. The “first” and “second” are not meant to be different types of messages, devices, modules, etc.

[0060] The research found that although the existing particle source analysis technology realizes the preliminary identification of pollution source characteristics through multi-modal sensing and deep learning, its core defects are: the dynamic correlation modeling of physical morphology and chemical composition is insufficient, and the time and space resolution of multi-source data fusion is limited. On the one hand, the single particle morphology classification relying on scanning electron microscope images is easily disturbed by the agglomeration effect of ultrafine particles, resulting in blurred morphological differentiation threshold between natural and artificial sources; on the other hand, the chemical component matching based on static source spectrum library is difficult to adapt to the dynamic evolution of source spectrum in mixed pollution scenarios, resulting in cumulative error in contribution weight allocation. In addition, the hotspot response mechanism realized by the existing method through edge computing is difficult to support high-resolution pollution migration path reconstruction due to the insufficient time and space alignment accuracy of aerosol three-dimensional distribution data and ground monitoring.

[0061] To solve the above problems, the application provides a particle source analysis method based on morphology-spectrum dynamic coupling, which has the innovation that: through the cooperation of particle shape characteristics and dual-wavelength laser enhanced detection technology, a full-chain analysis system of "topography classification-signal enhancement-trajectory inversion" is constructed. Specifically, based on the topography classification standard, the physical separation of natural dust and industrial emission particles is realized, the dual-wavelength laser is used to activate the mineral crystal inherent frequency reflection signal and the heavy metal scattering signal respectively, the polarization spectrometry technology is combined to block the non-target interference, and the mineral migration path of the natural dust and the heavy metal diffusion trajectory of the industrial emission are accurately extracted; further, the two types of trajectories are superimposed with the spatial thermal force of the high-altitude detection aerosol distribution data to dynamically correct the contribution weight of the pollution source. The method breaks through the inherent limitation of the traditional static source spectrum, realizes the real-time quantification of the distribution deviation of the mineral migration path and the displacement of the heavy metal trajectory, and analyzes the dynamic interaction mechanism of the natural source and the artificial source; at the same time, the signal enhancement technology based on the physical topography prior classification can significantly improve the distinguishability of ultrafine particles in mixed pollution, and solve the classification ambiguity problem caused by the resolution limitation of the scanning electron microscope image. In addition, the threshold linkage mechanism of the thermal distribution map and the spatial distribution density can dynamically optimize the spatial distribution accuracy of the source contribution weight, realize the technical leap from "component matching" to "topography-trajectory-space multi-dimensional coupling", and provide high timeliness and high reliability decision support for the differentiated management and control in the complex pollution scene.

[0062] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0063] For the above system, the embodiment of the application provides a method for tracing and analyzing atmospheric pollutants, Figure 1 For the method for tracing and analyzing atmospheric pollutants, the embodiment of the application provides a flowchart as shown in the figure, Figure 1 The method comprises the following steps.

[0064] 101, collect shape feature parameters of atmospheric particles, establish a topography classification standard of atmospheric particles according to the shape feature parameters, and divide the atmospheric particles into a natural dust particle group and an industrial emission particle group according to the topography classification standard.

[0065] Optionally, step 101 can specifically include the following steps:

[0066] 1011, continuously acquire surface contour images of atmospheric particles by a microscopic imaging device, and extract edge zigzag shape and surface concave-convex spacing of the atmospheric particles as shape feature parameters from the surface contour images.

[0067] 1012. Perform energy spectrum analysis on the atmospheric particulate matter to detect the composition ratio of mineral elements and artificial synthetic elements inside the atmospheric particulate matter, and mark as natural property when the ratio of mineral elements exceeds a set threshold, and mark as industrial property when the ratio of artificial synthetic elements exceeds a set threshold.

[0068] The step 1012 can specifically include the following process: placing the atmospheric particulate matter in the detection area of the energy spectrum analysis device, and simultaneously starting the energy spectrum scanning probe of the energy spectrum analysis device to continuously irradiate the surface of the atmospheric particulate matter, so as to collect the characteristic energy spectrum signal of the elements inside the atmospheric particulate matter; performing filtering and denoising processing on the characteristic energy spectrum signal, extracting the waveform of the characteristic energy spectrum signal, and extracting the peak value corresponding to the mineral elements and the peak value corresponding to the artificial synthetic elements from the waveform to generate the energy spectrum peak value signal of the atmospheric particulate matter; comparing the integral area of the peak value corresponding to the mineral elements in the energy spectrum peak value signal with the total integral area of the peak values of all elements to calculate the ratio value of the mineral elements, and simultaneously comparing the integral area of the peak value corresponding to the artificial synthetic elements with the total integral area to calculate the ratio value of the artificial synthetic elements; marking the atmospheric particulate matter with a ratio value of the mineral elements exceeding a preset natural property threshold as natural property, and marking the atmospheric particulate matter with a ratio value of the artificial synthetic elements exceeding a preset industrial property threshold as industrial property.

[0069] 1013. Match the edge sawtooth shape in the shape feature parameter with the natural property mark of the energy spectrum analysis, and match the surface concave-convex interval with the industrial property mark, to generate a morphological classification standard.

[0070] 1014. According to the matching degree of the edge sawtooth shape and the natural property, and the matching degree of the surface concave-convex interval and the industrial property in the morphological classification standard, divide the particulate matter into a natural dust particulate group or an industrial emission particulate group.

[0071] In the above scheme, atmospheric particulate matter is solid or liquid particulate matter suspended in the atmosphere. Shape feature parameters are quantitative indicators describing the geometric morphology of particulate matter. Morphological classification criteria are rules for classifying particulate matter based on shape and elemental composition. Natural dust particle group is a collection of particulate matter generated by natural activities. Industrial emission particle group is a collection of particulate matter emitted by industrial production. Microscopic imaging device is an optical device for capturing the microscopic morphology of the surface of particulate matter. Surface profile image is a microscopic image showing the relief features of the surface of particulate matter. Edge jagged shape is an irregular tooth-shaped geometric feature of the profile edge of particulate matter. Surface concave-convex spacing is the average distance between adjacent concave or convex parts of the surface of particulate matter. Energy spectrum analysis is a technique for identifying the composition of a substance by its characteristic X-ray spectrum. Mineral elements are inorganic elements present in rocks or soil in nature. Energy spectrum analysis device is a scientific instrument for detecting the elemental composition of a substance. Detection area is the physical space in the energy spectrum analyzer where the sample is placed for detection. Energy spectrum scanning probe is a detection component that emits and receives characteristic X-rays. Characteristic energy spectrum signal is the X-ray energy spectrum data emitted by an element when excited. Filtering and denoising processing is an algorithmic operation to remove interference waveforms in the energy spectrum signal. Waveform is the intensity distribution curve of the energy spectrum signal on the energy axis. Energy spectrum peak signal is the filtered elemental characteristic peak data. Integral area is the area value of a specific peak in the energy spectrum waveform. Mineral element ratio value is the percentage of the peak area of a mineral element to the total elemental peak area. Pre-set natural attribute threshold is a critical value of the proportion of mineral elements to determine whether particulate matter is of natural origin. Pre-set industrial attribute threshold is a critical value of the proportion of artificially synthesized elements to determine whether particulate matter is of industrial origin. Natural attribute label is a classification label added to particulate matter that meets the natural threshold. Natural dust particle group division is a process of classifying particulate matter that matches the natural attribute. Industrial emission particle group division is a process of classifying particulate matter that matches the industrial attribute.

[0072] In the embodiments of the present application, first, by step 1011, the surface profile image of the atmospheric particulate matter is continuously captured by the microscopic imaging device, and the edge jagged shape (such as irregular burrs or smooth arcs) and the surface concave-convex spacing (such as the average distance between convex and concave parts) of the particulate matter are extracted from the image by digital image processing algorithms (such as edge detection and contour extraction). For example, high-resolution microscopy is used to capture the details of the particle surface, and morphological analysis algorithms are used to quantify the complexity of the edge jaggedness and the distribution of the concave-convex spacing, generating a shape feature parameter dataset describing the morphology of the particulate matter.

[0073] Subsequently, through step 1012, the collected atmospheric particulate sample is placed in the detection area of the energy spectrum analysis device, a high-energy electron beam is emitted by the energy spectrum scanning probe to irradiate the surface of the particle, excite the internal elements to generate characteristic X-ray signals, and form characteristic energy spectrum signals. After the signal is filtered and denoised (such as wavelet transform or low-pass filtering) to remove background noise, the characteristic peak values of mineral elements (such as silicon and aluminum) and artificially synthesized elements (such as lead and cadmium) in the energy spectrum waveform are extracted. By calculating the proportion of the integral area of the mineral element peak value in the total integral area, the natural properties of the particulate matter are determined; similarly, when the integral area proportion of the artificially synthesized element peak value exceeds the threshold, it is marked as industrial properties.

[0074] Next, through step 1013, the edge sawtooth shape extracted in step 1011 is matched with the natural properties marked in step 1012, and the surface concave-convex spacing is matched with the industrial properties, to construct the morphological classification standard of the particulate matter. The matching process uses a pattern recognition algorithm (such as support vector machine or decision tree), trains the model to learn the characteristics of natural particulate matter (such as dust) with irregular sawtooth edges and industrial particulate matter (such as coal fly ash) with dense surface concave-convex spacing, and finally generates a classification rule based on the association of morphology and elements.

[0075] Finally, through step 1014, according to the matching degree in the morphological classification standard, the particulate matter is divided into a natural dust raising particle group or an industrial emission particle group. For example, if the edge sawtooth feature of a certain particulate matter conforms to the natural properties and the surface concave-convex spacing does not match the industrial properties, it is classified into the natural dust raising group; otherwise, if the concave-convex spacing is dense and the proportion of artificially synthesized elements is high, it is determined to be the industrial emission group. The classification results are stored and output through the database, providing data support for pollution source analysis.

[0076] In practical applications, at atmospheric monitoring stations surrounding industrial parks, a microscopic imaging device continuously images the collected particulate matter samples (step 1011), capturing surface contour images of two typical types of particulate matter: one type exhibits irregular serrated edges, while the other has densely distributed micro-pits on its surface. The system extracts the sharpness parameter of the serrated shape of the former and the pit spacing parameter of the latter as shape feature parameters. Simultaneously, an energy dispersive spectroscopy (EDS) analysis device is activated (step 1012). When scanning the elemental characteristic EDS signals inside the particulate matter, it is found that the serrated particulate matter shows significant integrated areas of mineral elements such as silicon and aluminum in its EDS peaks, while the pitted particulate matter shows prominent peaks of synthetic elements such as lead and cadmium. Based on preset thresholds, the former is labeled as a natural attribute, and the latter as an industrial attribute. Through matching analysis (step 1013), the edge serration shape parameter highly matches the natural attribute marker, and the surface pit spacing parameter is strongly correlated with the industrial attribute marker. This generates a morphological classification standard: serrated particles with a high proportion of mineral elements are classified as natural dust particles, while pitted particles with a high proportion of synthetic elements are classified as industrial emission particles (step 1014). Source tracing results show that natural dust particles mainly originate from exposed soil transport routes within the area, while industrial emission particles correspond to electroplating exhaust emissions from metal surface treatment workshops within the industrial park. This method provides a dual criterion of morphology and composition for accurately identifying pollution sources in industrial parks, supporting the formulation of targeted governance strategies.

[0077] The overall scheme in step 101 above achieves accurate identification and multi-dimensional classification of atmospheric particulate matter sources. By integrating microscopic imaging and energy dispersive spectroscopy (EDS) techniques, the system simultaneously captures the surface morphology features and elemental composition information of particulate matter, establishing a correlation classification standard between morphological parameters and elemental properties. Based on the matching mechanism between edge jaggedness and mineral element ratios, it effectively distinguishes between natural dust and industrial emission particle groups, overcoming the limitations of traditional single-parameter classification. Utilizing the correlation between surface unevenness spacing and artificially synthesized element concentrations enhances the identification accuracy of industrial particles. This technical solution solves the problem of separating morphological features and chemical composition analysis in particulate matter source determination, providing a dual-dimensional physicochemical classification basis for pollution source tracing.

[0078] 102. The natural dust particle group is subjected to dual-wavelength laser technology, wherein the first wavelength laser is used to match the inherent frequency of the mineral crystals in the natural dust particles to enhance the reflection signal, and polarization beam splitting technology is used to block the scattering interference of amorphous substances in the natural dust particle group to extract the reflection signal sequence and obtain the time variation law of the reflection signal sequence. The mineral migration path of the natural dust particle group is generated according to the time variation law.

[0079] Optionally, step 102 may specifically include the following steps:

[0080] 1021. Projecting a first wavelength laser beam in the dual-wavelength laser technology on the natural dust-raising particle group, and adjusting the frequency of the first wavelength laser beam to the inherent vibration frequency range of mineral crystals to enhance the reflection signal intensity of mineral crystals, while using polarization spectroscopy technology to block the scattering signal of non-crystalline substance particles to separate the reflection signal of mineral crystals;

[0081] 1022. Time-series collection of the reflection signal to form a continuous time sequence of reflection signal sequences.

[0082] 1023. Extracting the time variation law of the reflection signal intensity in the reflection signal sequence, and obtaining the wave peak and wave trough of the reflection signal intensity according to the time variation law.

[0083] 1024. Migrating the wave peak and wave trough to the position migration node of the natural dust-raising particle group in the airflow, and obtaining the time interval and particle group displacement direction of the position migration node.

[0084] 1025. Integrating the time interval and particle group displacement direction of the position migration node to generate a mineral migration path.

[0085] In the above scheme, the dual-wavelength laser technology is a technology of alternating two wavelengths of laser for detection and signal enhancement. The first wavelength laser beam is a specific wavelength laser matching the inherent frequency of mineral crystals. The inherent vibration frequency range of mineral crystals is the frequency interval of the resonance response of the molecular structure of the mineral. The reflection signal intensity is a quantitative indicator of the reflection light intensity after the particle is irradiated by laser. The polarization spectroscopy technology is an optical method for separating different scattering signals by using the polarization light characteristics. The non-crystalline substance particle is a solid particle with disordered internal structure. The scattering interference is a non-target scattering signal generated by non-crystalline substance to laser. The reflection signal of mineral crystals is the characteristic reflection light signal of crystal particles after being excited by laser. Time-series collection is an operation of recording the reflection signal intensity in time sequence. The reflection signal sequence is a data set of reflection signal intensity at continuous time points. The time variation law is the characteristic mode of the fluctuation of reflection signal intensity with time. The wave peak and wave trough are the highest point and lowest point of the change of reflection signal intensity. The position migration node is the key time point of the displacement of the particle group in the airflow movement. The time interval is the time difference between adjacent migration nodes. The particle group displacement direction is the orientation trend of the particle group movement in the airflow. The mineral migration path is a comprehensive description of the trajectory of the mineral particle group movement in the airflow.

[0086] In the embodiments of the present application, first, the classified natural dust particles are projected with a first wavelength laser beam in step 1021. The wavelength is adjusted to the inherent vibration frequency range of the target mineral crystal (such as quartz or feldspar) by laser frequency tuning technology, so that the reflection signal is significantly enhanced due to resonance effect. At the same time, the scattering interference of non-crystalline substances (such as glass or organic matter) in the particle group is blocked by polarization light splitting technology (such as a polarization prism-based light splitting device), and only the pure reflection signal of the mineral crystal is retained. For example, the mineral crystal produces strong reflection to the laser of a specific polarization direction due to the regularity of the crystal lattice structure, while the non-crystalline substance scatters light which is blocked by the polarization filter due to the disordered structure, thereby realizing signal separation.

[0087] Subsequently, the separated reflection signal is continuously sampled by a high-speed photoelectric detector in step 1022, the signal intensity change is recorded in time sequence, and a reflection signal sequence containing time stamp is generated. The time sequence acquisition process is controlled by a synchronous clock to ensure that the signal strictly corresponds to the time axis, thereby providing a data basis with continuous time dimension for subsequent analysis.

[0088] Then, the change rule of signal intensity with time is extracted from the reflection signal sequence in step 1023, and the peak (representing the moment when the particle group reflects the strongest) and the trough (representing the moment when the particle group reflects the weakest) of the signal intensity are identified by a peak detection algorithm (such as sliding window extreme value search). For example, the peak corresponds to the moment when the mineral crystal densely gathers in the air flow, and the trough corresponds to the moment when the particle group disperses or the movement direction deviates from the detection area.

[0089] Then, the peak and the trough are mapped to the position migration nodes of the particle group in the air flow in step 1024. The motion trajectory of the particle group is determined by time interval calculation (such as the time difference between the peaks) and displacement direction analysis (such as Doppler frequency shift or phase change). For example, the shortening of the time interval between adjacent peaks indicates that the migration speed of the particle group is accelerated, and the change of the frequency shift direction indicates that the movement direction of the particle group is deviated due to the influence of the wind.

[0090] Finally, the time interval and displacement direction data of all position migration nodes are integrated in step 1025, and a continuous mineral migration path is generated by a trajectory fitting algorithm (such as polynomial interpolation or Kalman filtering). The path shows the motion trajectory of the particle group in the monitoring area in the form of time-space coordinates, thereby providing a quantitative basis for analyzing the diffusion rule of natural dust and tracking the pollution source.

[0091] In practical applications, during atmospheric monitoring in farmland soil tillage season, to address the need for tracing the source of dust particle clusters, the system projects a dual-wavelength laser onto the classified natural dust particle clusters (step 1021). The frequency of the first wavelength laser beam is adjusted to the natural vibration band of the quartz crystal to enhance the reflected signal. At the same time, a polarizer is used to block scattering interference from amorphous substances such as mixed plant pollen, ensuring that the reflected signal originates from the mineral crystal. The laser probe continuously collects the reflected signal (step 1022), generating time-series data: the intensity of the reflected signal increases sharply during the midday strong wind period and decreases when the wind weakens at night. Peak and trough analysis (step 1023) shows the migration nodes of the dust particle clusters—the peak interval shortens when the northeast wind dominates in the morning, pointing to the diffusion path of farmland, while the peak interval lengthens when the wind turns to southwest in the afternoon, corresponding to the particle swirling and settling trend (step 1024). By integrating the time intervals and displacement directions of migration nodes throughout the day (step 1025), a migration path map of quartz particles is generated: the source area is farmland with tilled bare soil, and the particles spread to neighboring towns along a fan-shaped path with the wind direction, providing a spatial positioning basis for the deployment of windbreak and dust suppression nets, and effectively curbing the spread of dust pollution.

[0092] The overall scheme in step 102 above achieves dynamic tracking and migration path reconstruction of natural dust particles. Through the synergistic effect of dual-wavelength laser and polarization-based spectroscopy, the reflection signal of mineral crystals is selectively enhanced while background noise is suppressed, significantly improving the monitoring sensitivity of natural dust particle groups. Based on the mapping relationship between the temporal changes of the reflection signal and particle displacement, the signal fluctuation characteristics are transformed into spatial migration nodes, accurately reconstructing the trajectory of mineral particles in the airflow. This technology breaks through the spatiotemporal resolution limitations of traditional static observation, achieving dynamic analysis of the natural dust migration process throughout the entire time period, providing high-precision data support for the study of regional dust transport patterns.

[0093] 103. The unique scattering signals of heavy metal elements in the industrial emission particle group are screened out by the second wavelength laser, and the spatial intensity gradient of the scattering signal is obtained. The diffusion trajectory of heavy metals in the industrial pollution characteristic spectrum is generated according to the spatial intensity gradient.

[0094] Optionally, step 103 may specifically include the following steps:

[0095] 1031. Project a second wavelength laser onto the industrial emission particle group and adjust the wavelength of the second wavelength laser to the preset optical resonance frequency of the heavy metal element to enhance the scattering signal intensity of the atmospheric particles containing the heavy metal element.

[0096] 1032. Move the optical detector within the area covered by the second wavelength laser to collect data on the intensity of scattered signals at different spatial locations, and generate a spatial intensity gradient of the scattered signal based on the spatial distribution differences of the scattered signal intensity data at different spatial locations.

[0097] 1033、acquire the high-value area of the scattered signal intensity in the spatial intensity gradient, and take the change direction of the re-scattered signal intensity of adjacent high-value areas as the diffusion direction to generate a heavy metal diffusion track.

[0098] In the above scheme, the second wavelength laser is a specific wavelength laser beam set for the optical characteristics of heavy metal elements in the dual-wavelength technology. The preset optical resonance frequency is the resonance frequency at which the scattered signal of the heavy metal element is enhanced at a specific wavelength. The scattered signal intensity is the light intensity value of the reflected or scattered light after the laser excites the heavy metal particles. The optical detector is a photoelectric sensing device for collecting scattered signal intensity. Different spatial positions are coordinate points of the optical detector moving in the detection area. The scattered signal intensity data is a set of scattered light intensity values recorded at each spatial position. The spatial distribution difference is the change characteristics of the scattered signal intensity at different positions. The spatial intensity gradient is the rate and direction of the change of the scattered signal intensity with the spatial position. The high-value area is a local range in which the scattered signal intensity is significantly higher than the surrounding area in the spatial intensity gradient. The diffusion direction is the moving direction of the concentration change trend of the heavy metal particles in space. The heavy metal diffusion track is the path model of the heavy metal elements in the industrial emissions diffusing in the air.

[0099] In the embodiments of the present application, first, the classified industrial emission particle group is projected with the second wavelength laser through step 1031. The wavelength is adjusted to the preset optical resonance frequency of the target heavy metal element (such as lead and cadmium) through the laser wavelength tuning technology, and the scattered signal intensity of the heavy metal element in the particulate matter is enhanced by using the resonance effect. For example, the wavelength is adjusted to the characteristic absorption peak range of the lead element by using the tunable laser, so that the scattered signal of the lead-containing particles is significantly enhanced due to resonance amplification, and the interference signal of other non-heavy metal components is suppressed.

[0100] Subsequently, the scattered signal intensity data at different spatial positions is continuously collected by moving the detector position through the optical detector scanning technology in the monitoring area covered by the second wavelength laser through step 1032. The scanning process covers the target area by using a gridding path or a spiral track, and the intensity data of discrete points is converted into a continuous spatial intensity gradient (such as the distribution mode in which the intensity decreases from the center to the edge) by using a spatial interpolation algorithm, so as to reveal the concentration distribution characteristics of the heavy metal particles in space.

[0101] Finally, in step 1033, high-value regions of scattered signal intensity (such as local intensity peak regions) are identified from the spatial intensity gradient. The diffusion direction of heavy metal particles is determined by the direction of signal intensity change between adjacent high-value regions (such as a decrease in intensity from the factory chimney towards the southeast). For example, if the intensity of a high-value region gradually decreases eastward, the diffusion direction is determined to be eastward. By integrating all diffusion direction data, a continuous heavy metal diffusion trajectory is generated through a trajectory fitting algorithm, visually demonstrating the migration path of industrial pollution in the air and providing a basis for pollution source location and diffusion model construction.

[0102] In practical applications, during atmospheric monitoring in urban industrial areas, the system projects a second wavelength laser (step 1031) to classify industrial emission particles, adjusting its wavelength to the optical resonance frequency of lead to enhance the scattering signal of lead-containing particles. An optical detector moves downwind along the factory area (step 1032), collecting scattering signal intensity at different locations: the signal intensity is highest near the factory boundary, decreasing gradually with distance, with a lower signal attenuation rate on the northeast side than on the southwest side. The generated spatial intensity gradient is displayed (step 1032), showing high-value areas extending outwards along the chimney cluster of the smelting workshop, with the signal intensity change direction of adjacent high-value areas pointing northeast (step 1033). Combined with meteorological data, the lead particle diffusion trajectory is determined to be a northeastward migration influenced by the prevailing wind after being released from the smelting workshop chimneys, combining with halogen compounds in traffic exhaust to form stable aerosols along the way. This trajectory identifies the pollution source as lead dust emission from the smelting workshop, prompting the installation of high-efficiency dust removal facilities and adjustment of exhaust direction within the factory area, effectively blocking the lead pollution diffusion path.

[0103] The overall scheme in step 103 above achieves targeted detection and visualization of diffusion trajectories for industrial pollutant particles. Optical resonance frequency matching technology is used to enhance the scattering signal of heavy metal particles, and a scattering intensity distribution model is constructed through spatial gradient scanning. The main diffusion path is identified based on the direction of signal change in high-value regions, transforming discrete detection data into continuous diffusion trajectories. This technology solves the technical challenges of low detection sensitivity and ambiguous diffusion paths for heavy metal components in industrial emission particles, providing an innovative means for three-dimensional monitoring of industrial pollution diffusion through the dual correlation of spectral characteristics and spatial distribution.

[0104] 104. The mineral migration path and heavy metal diffusion trajectory are spatially superimposed with the aerosol three-dimensional distribution data obtained by the high-altitude detection equipment to generate a thermal distribution map of the pollution source.

[0105] Optionally, step 104 may specifically include the following steps:

[0106] 1041、Obtain the set of spatial coordinate points of the mineral migration path and the set of spatial coordinate points of the heavy metal diffusion trajectory, and extract the three-dimensional spatial coordinates and corresponding aerosol concentration values in the aerosol three-dimensional distribution data collected by the high-altitude detection equipment, and align the spatial positions of the spatial coordinate points of the mineral migration path, the spatial coordinate points of the heavy metal diffusion trajectory, and the spatial coordinate points corresponding to the aerosol concentration values.

[0107] 1042、Divide the set of mineral migration path coordinate points by spatial grid to count the density value of the mineral migration path in each spatial grid, and divide the set of heavy metal diffusion trajectory coordinate points by the same spatial grid to count the density value of the diffusion trajectory in each spatial grid, and map the aerosol concentration value to the corresponding spatial grid.

[0108] 1043、Superimpose the density value of the mineral migration path, the density value of the heavy metal diffusion trajectory, and the aerosol concentration value in each spatial grid to generate a pollution thermodynamic value.

[0109] 1044、Connect the pollution thermodynamic values of all spatial grids in order of spatial coordinates to construct a pollution source thermodynamic distribution map in three-dimensional space.

[0110] In the above scheme, the spatial distribution density is a quantitative indicator of the distribution concentration of natural dust or industrial particle groups in a specific three-dimensional space region. The preset threshold is the critical value for determining the abnormality of the spatial distribution density. The abnormal density area is the spatial area where the actual distribution density of the particle group deviates from the predicted density of the thermodynamic distribution map by more than the threshold value. The distribution deviation amount is a quantitative value of the deviation degree of the mineral migration path of the natural dust particle group from the expected trajectory. The offset amount is a quantitative value of the deviation degree of the heavy metal diffusion trajectory of the industrial emission particle group from the standard path. The pollution source contribution weight is the contribution proportion parameter of natural dust and industrial emission particle groups to regional pollution. The pollution source analysis result is the pollution source analysis conclusion output after comprehensive weight adjustment. The difference exceeding the preset threshold is the judgment condition that the actual distribution density and the thermodynamic distribution map density difference exceeds the allowed range. The distribution deviation amount of the mineral migration path is the spatial deviation amount of the natural dust path from the theoretical trajectory in the abnormal area. The offset amount of the heavy metal diffusion trajectory is the spatial deviation amount of the industrial emission trajectory from the standard path in the abnormal area. The proportional relationship is the numerical correspondence between the distribution deviation amount and the offset amount. The weight is inversely proportional to the correlation rule that the larger the distribution deviation amount, the lower the natural dust weight, and the larger the offset amount, the lower the industrial emission weight. The adjusted pollution source contribution weight is the distribution proportion of natural dust and industrial emission to pollution after correction. The corresponding relationship is the matching relevance of the particle group distribution density and the thermodynamic distribution map value. The pollution source analysis result is the quantitative conclusion of the final determination of the contribution proportion of natural dust and industrial emission to pollution.

[0111] In the embodiments of the present application, first, through step 1041, the spatial coordinate point set of the mineral migration path and the spatial coordinate point set of the heavy metal diffusion trajectory are obtained, and the three-dimensional spatial coordinates and the corresponding aerosol concentration values are extracted from the aerosol three-dimensional distribution data collected by the high-altitude detection device (such as a meteorological satellite or an unmanned aerial sensor). The mineral migration path coordinates, the heavy metal diffusion trajectory coordinates and the aerosol concentration coordinates are matched by a spatial position alignment algorithm, for example, a geographic coordinate system conversion technology (such as WGS84 to UTM projection) or a spatial interpolation method is used to ensure that the positions of data from different sources are strictly aligned in the three-dimensional space, forming a unified spatial reference datum.

[0112] Subsequently, through step 1042, the mineral migration path coordinate point set and the heavy metal diffusion trajectory coordinate point set are respectively regionally segmented according to the same spatial grid, and the number of path points or the trajectory density value (such as the trajectory length per unit volume) in each grid is counted. At the same time, the aerosol concentration value is mapped to the corresponding spatial grid according to the grid position, for example, the concentration value is averaged or weightedly distributed according to the grid center point, so as to assign the mineral migration density, the heavy metal diffusion density and the aerosol concentration to each grid as three types of data attributes.

[0113] Then, through step 1043, the mineral migration path density value, the heavy metal diffusion trajectory density value and the aerosol concentration value in each spatial grid are superimposed and calculated. For example, a multi-factor weighted fusion algorithm (such as the mineral density can account for 30%, the heavy metal density can account for 40%, and the aerosol concentration can account for 30%) is used to generate a pollution thermal value reflecting the comprehensive pollution intensity. The high and low of the thermal value directly represents the aggregation degree of the pollution source in the grid, and the high value area represents the dense or diffusion active area of the pollution source.

[0114] Finally, through step 1044, the pollution thermal values of all spatial grids are connected in order of three-dimensional spatial coordinates, and a three-dimensional pollution source thermal distribution map covering the monitoring area is constructed by three-dimensional visualization technology (such as volume rendering or isosurface generation). For example, the thermal value is rendered in the three-dimensional space through the color gradient (such as red representing high value and blue representing low value) and the transparency, which intuitively displays the spatial distribution characteristics and superposition effect of the natural dust and industrial emission pollution, and provides stereoscopic data support for pollution control decision-making.

[0115] In practical applications, in atmospheric monitoring around coal-fired power plants, the system superimposes the spatial coordinate point set of the mineral migration path of the natural dust particle group and the heavy metal diffusion track of the industrial emission particle group (step 1041) with the spatial alignment of the aerosol three-dimensional distribution data collected by the high-altitude detection equipment carried by the unmanned aerial vehicle. For example, the mineral migration path coordinate points of the dust raised by the ash storage on the west side of the power plant overlap with the high-value area of the high-altitude aerosol concentration, and the coordinate points of the heavy metal diffusion track in the downwind direction of the desulfurization tower coincide with the near-ground aerosol concentration peak zone. After being divided according to the spatial grid (step 1042), the mineral migration path density value, the heavy metal track density value and the aerosol concentration value in the west side grid are all significantly higher than those in other areas, and the pollution thermal value generated by superimposed calculation (step 1043) reveals that the grid is the pollution core area. The three-dimensional thermal distribution map (step 1044) shows that the pollution spreads in a feather shape along the dominant wind direction, and the core area thermal value extends northeast and gradually decays, which accurately corresponds to the open storage point of the ash storage and the exhaust port position of the desulfurization tower of the power plant, providing spatial targeting basis for closed ash storage reconstruction and flue gas purification upgrading, and realizing the precise matching of pollution source analysis and treatment measures.

[0116] The overall scheme of the above step 104 realizes the spatial fusion of multi-source pollution data and three-dimensional thermal modeling. By aligning the three-dimensional coordinates of the mineral migration path, the heavy metal track and the aerosol distribution, a multi-parameter spatial density superposition algorithm is established to convert discrete pollution characteristics into a continuous thermal field. Using the grid density mapping technology, the pollution contribution values from different sources are quantified as a unified thermal index to construct a three-dimensional distribution map containing pollution intensity and source characteristics. This technology breaks through the limitations of traditional two-dimensional pollution distribution expression, realizes the three-dimensional and multi-source spatial representation of atmospheric pollution, and provides visual decision support for pollution source spatial positioning.

[0117] 105、When the spatial distribution density of the natural dust particle group and the industrial emission particle group exceeds the corresponding spatial distribution density in the thermal distribution map by a preset threshold, the distribution deviation of the mineral migration path and the shift of the heavy metal diffusion track are extracted, and the pollution source contribution weight of the natural dust particle group and the industrial emission particle group is redistributed to output the pollution source analysis result.

[0118] Optionally, step 105 can specifically include the following steps:

[0119] 1051、Obtain the spatial distribution density of the natural dust particle group and the industrial emission particle group in the corresponding area of the thermal distribution map, and calculate the difference between the spatial distribution density of the natural dust particle group and the industrial emission particle group in each area and the corresponding spatial distribution density in the thermal distribution map. When the difference exceeds the preset threshold, it is marked as an abnormal density area.

[0120] 1052. Extract the mineral migration path of the natural dust particle group in the abnormal density area, and calculate the distribution deviation of the mineral migration path. Simultaneously extract the heavy metal diffusion trajectory of the industrial emission particle group, and calculate the offset of the heavy metal diffusion trajectory.

[0121] 1053. Based on the proportional relationship between the distribution deviation and the offset, adjust the pollution source contribution weights of natural dust particle groups and industrial emission particle groups in the thermal distribution map. The proportion of the distribution deviation is inversely proportional to the weight of the natural dust particle group, and the proportion of the offset is inversely proportional to the weight of the industrial emission particle group.

[0122] 1054. Feedback the adjusted pollution source contribution weights into the generation process of the heat map, and recalculate the correspondence between the spatial distribution density of the particle group and the heat map based on the changes in the weights of natural dust particle groups and industrial emission particle groups to form the pollution source analysis results.

[0123] In the above scheme, the preset threshold is the critical value for determining whether the difference between the actual distribution density and the predicted density on the thermal distribution map is abnormal. A difference exceeding the preset threshold is the criterion for determining that the difference between the spatial distribution density of natural dust or industrial emission particle groups and the corresponding density on the thermal distribution map exceeds the allowable range. Abnormal density areas are spatial areas where the difference exceeds the preset threshold. The distribution deviation of mineral migration paths is a quantitative value of the deviation between the actual migration path of natural dust particle groups and the theoretical trajectory in abnormal areas. The offset of heavy metal diffusion trajectories is a quantitative value of the deviation between the actual diffusion path of industrial emission particle groups and the standard trajectory in abnormal areas. The pollution source contribution weight is the proportional parameter for the allocation of the contribution of natural dust and industrial emission particle groups to regional pollution. The proportional relationship is the numerical correspondence between the distribution deviation and the offset. The inverse weighting is an association rule where the larger the distribution deviation, the lower the weight of natural dust particle groups; and the larger the offset, the lower the weight of industrial emission particle groups. The adjusted pollution source contribution weight is the corrected allocation ratio of the contribution of natural dust and industrial emissions to pollution. The pollution source analysis result is the quantitative analysis conclusion of pollution sources output after comprehensive weight adjustment.

[0124] In this embodiment, firstly, step 1051 obtains the spatial distribution density of natural dust particles and industrial emission particles in the corresponding areas of the thermal distribution map, and calculates the difference between the density and the density in the thermal distribution map. When the difference exceeds a preset threshold, the system marks such areas as abnormal density areas using a density comparison algorithm. For example, if the density of natural dust in a certain area is 1.5 times that of the density in the thermal distribution map, it is determined to be an abnormal area, triggering subsequent analysis.

[0125] Subsequently, by step 1052, the mineral migration path of the natural dust particle group in the abnormal density area is extracted, and the distribution deviation amount (such as the deviation distance of the actual path from the theoretical path) of the path is calculated by using a path analysis algorithm (such as least square fitting). At the same time, the heavy metal diffusion trajectory of the industrial emission particle group is extracted, and the deviation amount (such as the angle between the trajectory direction and the expected wind direction) of the trajectory is calculated by using a trajectory deviation detection algorithm (such as direction angle change analysis).

[0126] Then, by step 1053, according to the proportional relationship between the distribution deviation amount and the deviation amount, the pollution source contribution weight of the two types of particle groups is dynamically adjusted. For example, if the proportion of the distribution deviation amount of the natural dust increases, the weight is reduced; if the proportion of the deviation amount of the industrial emission increases, the weight is also reduced. The weight adjustment is realized by using an inverse proportional mapping algorithm, which ensures that the evaluation of the overall pollution contribution of the high deviation or deviation area is more accurate.

[0127] Finally, by step 1054, the adjusted pollution source contribution weight is fed back to the generation process of the thermal distribution map, and the spatial distribution density of the particle group and the corresponding relationship between the thermal value are re-integrated by using a density recalculation algorithm. For example, after reducing the weight of the natural dust, the influence of its density on the thermal value is weakened, and after adjusting the weight of the industrial emission, the contribution proportion of its diffusion trajectory to the thermal value changes accordingly. The updated thermal distribution map combines the weight correction result to output the final pollution source analysis result, which clearly shows the actual contribution degree of different pollution sources and provides dynamic data support for governance decision-making.

[0128] In actual application, in the atmospheric monitoring of the coastal port area, the system finds that the spatial distribution density of the natural dust particle group and the industrial emission particle group around the cargo terminal in the thermal distribution map both exceed the preset threshold (step 1051), which is marked as an abnormal density area. When the mineral migration path of this area is extracted, the path distribution deviation amount of the dust particle group pointing to the beach sand source significantly increases (step 1052), indicating that the dust particles abnormally diffuse inland; at the same time, the heavy metal diffusion trajectory deviation amount of the industrial emission particle group shows that the ship exhaust particles that should have migrated with the sea wind are retained in the port area. According to the proportional relationship between the deviation amount and the deviation amount (step 1053), the system reduces the weight of the natural dust particle group and simultaneously reduces the contribution proportion of the industrial emission particle group, and finds that the mixed pollution of the diesel particles of the loading and unloading machinery and the exhaust gas of the ship fuel is underestimated. The adjusted weight is fed back to the thermal distribution map (step 1054), and the re-calculated analysis result reveals that the nickel particles emitted by the loading and unloading machinery and the vanadium particles not burned by the ship fuel form a pollution core, which dominates the aerosol over-standard in the port area. Accordingly, the port starts the electric transformation of the loading and unloading machinery and the upgrade of the ship shore power system to accurately reduce the contribution amount of the mixed pollution source.

[0129] The overall scheme of the above step 105 realizes dynamic analysis of the contribution of pollution sources and self-optimization of the model. Based on the quantitative analysis of the spatial density deviation and the trajectory deviation, a dynamic adjustment mechanism of the pollution source weight is established, and through the inverse proportional relationship between the path deviation and the industrial trajectory deviation, the precise redistribution of the contribution values of natural and industrial pollution is realized. The weight feedback correction technology is adopted to form a closed loop optimization link of the thermal distribution map and the pollution source analysis. This technical scheme solves the problem of misjudgment of the contribution degree in the superposition scene of multi-source pollution, and through real-time iterative updating of the model parameters, the source analysis accuracy and reliability in complex pollution scenes are significantly improved.

[0130] The following is a complete embodiment for steps 101-105:

[0131] In the atmospheric monitoring of the high-speed road construction area, the particulate matter captured by the overpass microscopic imaging device shows two types of morphology: the surface of the edge broken angular particulate matter is covered with small sand bumps, and the smooth spherical particulate matter surface is attached with nanoscale pits. Through energy spectrum analysis, it is found that the proportion of mineral elements such as silicon and calcium in the angular particulate matter exceeds the threshold value, which is marked as natural attribute; the proportion of alloy elements such as vanadium and nickel in the spherical particulate matter is outstanding, which is marked as industrial attribute. The morphological classification standard is established accordingly, and the angular particulate group is classified as construction dust source, and the spherical particulate group is attributed to diesel vehicle exhaust emission source.

[0132] For the natural dust particulate group, the first wavelength of the dual-wavelength laser is adjusted to the quartz crystal resonance frequency band, and the polarization light filter is used to remove pollen interference. The reflection signal sequence shows that the signal wave peak is dense during morning and evening peak hours, which corresponds to the diffusion of dust caused by the frequent passing of slag cars. The mineral migration path is generated by integrating the wave peak interval and the wind direction data: the dust diffuses to the residential area along the northeast wind direction from the gap of the construction fence. For the industrial emission particulate group, the second wavelength laser resonance excites the vanadium element scattering signal in the exhaust particulate, and the spatial intensity gradient constructed by the mobile optical detector reveals the diffusion trajectory: the core high value area is concentrated in the tunnel exit, and presents a strip extending along the traffic direction.

[0133] The mineral migration path and the heavy metal diffusion trajectory are superimposed with the aerosol three-dimensional data space of the airborne laser radar to generate a thermal distribution map showing that the pollution thermal value of the grid on the northeast side of the construction area is abnormal. Analysis finds that the density of both natural dust and industrial emissions exceeds the threshold value: the dust path deviation points to the stockpile field that is not covered by the dustproof net, and the heavy metal trajectory deviation reveals that the exhaust particulate is retained due to poor ventilation in the tunnel. After redistributing the weight, the pollution source analysis result locks the dust contribution of the stockpile field, and the optimization of the tunnel exhaust scheme is synchronized to precisely suppress the diffusion of mixed pollution.

[0134] As shown in Figure 2 , Figure 3 , Figure 2A four-step process is used as the core: first, the classification standard is established by the morphological characteristics (such as shape parameters) of particulate matter, and the atmospheric particulate matter is divided into two types of natural dust and industrial emissions; second, the two types of particulate matter are differentiated by dual-wavelength laser technology - the first wavelength laser combined with polarization spectroscopy is used to extract the reflection signal and track the mineral migration path of the natural dust group, and the second wavelength laser is used to screen heavy metal elements in the industrial emission group, and the diffusion trajectory is generated according to the spatial intensity gradient; the third step is to superimpose the two types of migration paths and the three-dimensional distribution data of aerosol to form a visual atlas that integrates the thermal distribution characteristics; finally, the pollution source contribution degree analysis result is output by comparing the density value, extracting the offset and dynamically allocating the weight. Figure 3 On the basis of Figure 2 Supplement the key technical details: introduce the electric field and gravity settling physical separation technology in the particulate matter classification link, emphasize the mineral crystal frequency matching and amorphous interference blocking mechanism in the laser analysis stage, and highlight the three-dimensional space modeling process in the data superposition link. The two figures together build a closed-loop analysis system of "classification-detection-modeling-analysis", which realizes the accurate source tracing and contribution weight quantification of natural and industrial sources of pollutants through the synergistic effect of morphological screening, spectral feature recognition and spatial data fusion, and provides scientific decision-making basis for air pollution control.

[0135] Figure 4 A structural diagram of an atmospheric pollutant source tracing analysis system is provided for the embodiments of the present application, as Figure 2 shown, the system comprises:

[0136] A division module 41 is configured to collect shape feature parameters of atmospheric particulate matter, establish a morphological classification standard of the atmospheric particulate matter according to the shape feature parameters, and divide the atmospheric particulate matter into a natural dust particle group and an industrial emission particle group according to the morphological classification standard.

[0137] A generation module 42 is configured to use dual-wavelength laser technology on the natural dust particle group, wherein the inherent frequency of the mineral crystals in the natural dust particles is matched by the first wavelength laser to enhance the reflection signal, and the scattering interference of amorphous substances in the natural dust particle group is blocked by using polarization spectroscopy technology to extract the reflection signal sequence and obtain the time variation law of the reflection signal sequence, and the mineral migration path of the natural dust particle group is generated according to the time variation law.

[0138] A screening module 43 is configured to screen out the scattering signal characteristic of heavy metal elements in the industrial emission particle group by the second wavelength laser, obtain the spatial intensity gradient of the scattering signal, and generate the heavy metal diffusion trajectory in the industrial pollution characteristic spectrum according to the spatial intensity gradient.

[0139] The superposition module 44 is configured to perform spatial superposition of the mineral migration path and the heavy metal diffusion track with aerosol three-dimensional distribution data obtained by the high-altitude detection device, so as to generate a thermal distribution map of the pollution source.

[0140] The distribution module 45 is configured to extract a distribution deviation amount of the mineral migration path and a shift amount of the heavy metal diffusion track when the spatial distribution density of the natural dust raising particle group and the industrial emission particle group and the corresponding spatial distribution density in the thermal distribution map exceed a preset threshold value, and re-distribute the pollution source contribution weight of the natural dust raising particle group and the industrial emission particle group, so as to output a pollution source analysis result.

[0141] Figure 4 The atmospheric pollutant source tracing analysis system can perform Figure 1 The atmospheric pollutant source tracing analysis method of the embodiment has the same implementation principles and technical effects as the atmospheric pollutant source tracing analysis system. The specific implementation manners of the operations of each module and unit of the atmospheric pollutant source tracing analysis system in the embodiment have been described in detail in the embodiment of the method, and will not be described in detail here.

[0142] In one possible design, Figure 4 The atmospheric pollutant source tracing analysis system of the embodiment can be implemented as a computing device, such as a server or a personal computer. Figure 5 As shown in the figure, the computing device can include a storage component 51 and a processing component 52.

[0143] The storage component 51 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 52.

[0144] The processing component 52 is configured to perform the above Figure 1 The atmospheric pollutant source tracing analysis method of the embodiment.

[0145] The processing component 52 can include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component can also be one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components, for executing the above method.

[0146] The storage component 51 is configured to store various types of data to support the operation of the terminal. The storage component can be implemented by any type of volatile or nonvolatile storage devices or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic or optical disk.

[0147] Of course, the computing device can also necessarily include other components, such as an input / output interface, a display component, a communication component, etc.

[0148] The input / output interface provides an interface between the processing component and peripheral interface modules, which can be output devices, input devices, etc.

[0149] The communication component is configured to facilitate wired or wireless communication between the computing device and other devices, etc.

[0150] The computing device can be a physical device or an elastic computing host provided by a cloud computing platform, and the computing device can be a cloud server, and the processing component, the storage component, etc. can be a basic server resource rented or purchased from the cloud computing platform.

[0151] The embodiment of the application further provides a computer storage medium, which stores a computer program, and the computer program can realize the above-mentioned Figure 1 The embodiment shown provides a method for tracing and analyzing atmospheric pollutants.

[0152] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system, device and unit can refer to the corresponding process in the foregoing method embodiment, which will not be described here.

[0153] The device embodiment described above is only schematic, wherein the units described as separate components can or can not be physically separated, and the components displayed as units can or can not be physical units, that is, they can be located in one place, or distributed on multiple network units. According to actual needs, part or all of the modules can be selected to achieve the purpose of the embodiment. Those skilled in the art can understand and implement without creative labor.

[0154] Those skilled in the art can clearly understand the implementation of the various embodiments by means of software and necessary general hardware platforms through the description of the above embodiments, and of course, the embodiments can also be implemented by hardware. Based on such understanding, the above technical solutions, essentially or in other words, the part of the prior art that makes a contribution, can be embodied in the form of a software product. The computer software product can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, and the like, and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0155] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacement for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for tracing the source of air pollutants, characterized in that, include: The shape characteristic parameters of atmospheric particulate matter are collected, and a morphological classification standard for atmospheric particulate matter is established based on the shape characteristic parameters. The atmospheric particulate matter is then divided into natural dust particles and industrial emission particles according to the morphological classification standard. The natural dust particle group is treated with dual-wavelength laser technology. The first wavelength laser is used to match the inherent frequency of the mineral crystals in the natural dust particles to enhance the reflection signal. At the same time, polarization beam splitting technology is used to block the scattering interference of amorphous substances in the natural dust particle group to extract the reflection signal sequence and obtain the time variation law of the reflection signal sequence. The mineral migration path of the natural dust particle group is generated according to the time variation law. The unique scattering signals of heavy metal elements in the industrial emission particle group are screened out by a second wavelength laser. At the same time, the spatial intensity gradient of the scattering signal is obtained, and the diffusion trajectory of heavy metals in the industrial pollution characteristic spectrum is generated based on the spatial intensity gradient. The mineral migration paths and heavy metal diffusion trajectories are spatially superimposed with the aerosol three-dimensional distribution data obtained by the high-altitude detection equipment to generate a thermal distribution map of the pollution source. When the spatial distribution density of the natural dust particle group and the industrial emission particle group exceeds the corresponding spatial distribution density in the thermal distribution map by a preset threshold, the distribution deviation of the mineral migration path and the offset of the heavy metal diffusion trajectory are extracted, and the pollution source contribution weights of the natural dust particle group and the industrial emission particle group are reallocated to output the pollution source analysis results.

2. The method according to claim 1, characterized in that, The mineral migration paths and heavy metal diffusion trajectories are spatially overlaid with aerosol three-dimensional distribution data acquired by upper-air detection equipment to generate a thermal distribution map of the pollution source, including: The set of spatial coordinate points of the mineral migration path and the set of spatial coordinate points of the heavy metal diffusion trajectory are obtained. At the same time, the three-dimensional spatial coordinates and corresponding aerosol concentration values ​​are extracted from the aerosol three-dimensional distribution data collected by the high-altitude detection equipment. The spatial coordinate points of the mineral migration path, the spatial coordinate points of the heavy metal diffusion trajectory and the spatial coordinate points corresponding to the aerosol concentration values ​​are spatially aligned. The set of coordinate points of the mineral migration path is divided into spatial grids to count the density value of the mineral migration path in each spatial grid. At the same time, the set of coordinate points of the heavy metal diffusion trajectory is divided into the same spatial grids to count the density value of the diffusion trajectory in each spatial grid. The aerosol concentration value is then mapped to the corresponding spatial grid. The density values ​​of mineral migration paths, heavy metal diffusion trajectories, and aerosol concentrations within each spatial grid are superimposed to generate pollution thermal values. The pollution thermal values ​​of all spatial grids are connected in spatial coordinate order to construct a three-dimensional thermal distribution map of pollution sources.

3. The method according to claim 1, characterized in that, The shape characteristic parameters of atmospheric particulate matter are collected, and a morphological classification standard for atmospheric particulate matter is established based on these parameters. The atmospheric particulate matter is then classified into natural dust particles and industrial emission particles according to the morphological classification standard, including: The surface contour images of atmospheric particulate matter are continuously acquired by a microscopic imaging device, and the edge jagged shape and surface unevenness spacing of atmospheric particulate matter are extracted from the surface contour images as shape feature parameters. Energy dispersive spectroscopy (EDS) is performed on the atmospheric particulate matter to detect the composition ratio of mineral elements and synthetic elements within the particulate matter. When the proportion of mineral elements exceeds a set threshold, it is marked as a natural attribute, and when the proportion of synthetic elements exceeds a set threshold, it is marked as an industrial attribute. The edge jagged shape in the shape feature parameters is matched with the natural attribute markers of energy spectrum analysis, and the surface concave-convex distance is matched with the industrial attribute markers to generate a morphological classification standard. Based on the matching degree between the edge serrated shape and natural attributes, and the matching degree between the surface unevenness spacing and industrial attributes in the morphological classification criteria, particulate matter is divided into natural dust particle groups or industrial emission particle groups.

4. The method according to claim 3, characterized in that, Energy dispersive spectroscopy (EDS) is performed on the atmospheric particulate matter to detect the composition ratio of mineral elements and synthetic elements within the particulate matter. When the proportion of mineral elements exceeds a set threshold, it is marked as a natural attribute; when the proportion of synthetic elements exceeds a set threshold, it is marked as an industrial attribute. This includes: The atmospheric particulate matter is placed in the detection area of ​​the energy spectrum analyzer, and the energy spectrum scanning probe of the energy spectrum analyzer is activated to continuously irradiate the surface of the atmospheric particulate matter in order to collect the characteristic energy spectrum signals of the elements inside the atmospheric particulate matter. The characteristic energy spectrum signal is filtered and denoised, and the waveform of the characteristic energy spectrum signal is extracted. The peak values ​​corresponding to mineral elements and artificially synthesized elements are extracted from the waveform to generate the energy spectrum peak signal of atmospheric particulate matter. The integral area of ​​the peak corresponding to the mineral element in the energy spectrum peak signal is compared with the total integral area of ​​the peaks of all elements to calculate the proportion of the mineral element. At the same time, the integral area of ​​the peak corresponding to the artificially synthesized element is compared with the total integral area to calculate the proportion of the artificially synthesized element. Atmospheric particulate matter with a proportion of mineral elements exceeding a preset natural attribute threshold is marked as natural attribute, and atmospheric particulate matter with a proportion of artificially synthesized elements exceeding a preset industrial attribute threshold is marked as industrial attribute.

5. The method according to claim 1, characterized in that, A dual-wavelength laser technique is used on the natural dust particle group. The first wavelength laser is matched to the intrinsic frequency of the mineral crystals in the natural dust particles to enhance the reflection signal. Simultaneously, polarization-based beam splitting is used to block scattering interference from amorphous substances in the natural dust particle group, thereby extracting the reflection signal sequence and obtaining its temporal variation. Based on this temporal variation, the mineral migration path of the natural dust particle group is generated, including: The first wavelength laser beam in the dual-wavelength laser technology is projected onto the natural dust particle group, and the frequency of the first wavelength laser beam is adjusted to the natural vibration frequency range of the mineral crystal, so as to enhance the intensity of the reflected signal of the mineral crystal. At the same time, polarization beam splitting technology is used to block the scattering signal of amorphous material particles and separate the reflected signal of the mineral crystal. The reflected signals are acquired in a time sequence to form a continuous time sequence of reflected signals; Extract the temporal variation pattern of the reflected signal intensity in the reflected signal sequence, and obtain the peaks and troughs of the reflected signal intensity based on the temporal variation pattern; The peaks and troughs correspond to the position migration nodes of the natural dust particle group in the airflow, and the time interval of the position migration nodes and the displacement direction of the particle group are obtained. The time intervals of the location migration nodes are integrated with the particle group displacement direction to generate a mineral migration path.

6. The method according to claim 1, characterized in that, The unique scattering signals of heavy metal elements in the industrial emission particle group are screened using a second wavelength laser. Simultaneously, the spatial intensity gradient of the scattering signal is obtained, and the diffusion trajectory of heavy metals in the industrial pollution characteristic spectrum is generated based on the spatial intensity gradient, including: A second wavelength laser is projected onto the industrial emission particle group, and the wavelength of the second wavelength laser is adjusted to the preset optical resonance frequency of the heavy metal element to enhance the scattering signal intensity of the atmospheric particles containing heavy metal elements. An optical detector is moved within the area covered by the second wavelength laser to collect data on the intensity of scattered signals at different spatial locations, and a spatial intensity gradient of the scattered signal is generated based on the spatial distribution differences of the scattered signal intensity data at different spatial locations. The high-value regions of scattered signal intensity in the spatial intensity gradient are obtained, and the direction of change of rescattered signal intensity in adjacent high-value regions is taken as the diffusion direction to generate heavy metal diffusion trajectories.

7. The method according to claim 1, characterized in that, When the spatial distribution density of the natural dust particle group and the industrial emission particle group exceeds a preset threshold compared to the corresponding spatial distribution density in the thermal distribution map, the distribution deviation of the mineral migration path and the offset of the heavy metal diffusion trajectory are extracted. The pollution source contribution weights of the natural dust particle group and the industrial emission particle group are then reallocated to output the pollution source analysis results, including: Obtain the spatial distribution density of the natural dust particle group and the industrial emission particle group in the corresponding area of ​​the thermal distribution map, and calculate the difference between the spatial distribution density of the natural dust particle group and the industrial emission particle group in each area and the corresponding spatial distribution density in the thermal distribution map. When the difference exceeds a preset threshold, it is marked as an abnormal density area. Mineral migration paths of natural dust particle groups are extracted in the abnormal density region, and the distribution deviation of mineral migration paths is calculated. Simultaneously, heavy metal diffusion trajectories of industrial emission particle groups are extracted, and the offset of heavy metal diffusion trajectories is calculated. Based on the proportional relationship between the distribution deviation and the offset, the pollution source contribution weights of natural dust particle groups and industrial emission particle groups in the thermal distribution map are adjusted. The proportion of the distribution deviation is inversely proportional to the weight of the natural dust particle group, and the proportion of the offset is inversely proportional to the weight of the industrial emission particle group. The adjusted pollution source contribution weights are fed back into the generation process of the heat map, and the spatial distribution density of the particle group and the correspondence between the heat map are recalculated based on the changes in the weights of natural dust particle groups and industrial emission particle groups, thus forming the pollution source analysis results.

8. An air pollutant source tracing and analysis system, characterized in that, include: The classification module is used to collect shape characteristic parameters of atmospheric particulate matter, establish a morphological classification standard for atmospheric particulate matter based on the shape characteristic parameters, and classify the atmospheric particulate matter into natural dust particles and industrial emission particles according to the morphological classification standard. The generation module is used to apply dual-wavelength laser technology to the natural dust particle group. The first wavelength laser is used to match the inherent frequency of the mineral crystals in the natural dust particles to enhance the reflection signal. At the same time, polarization beam splitting technology is used to block the scattering interference of amorphous substances in the natural dust particle group to extract the reflection signal sequence and obtain the time variation law of the reflection signal sequence. The mineral migration path of the natural dust particle group is generated according to the time variation law. The screening module is used to screen out the unique scattering signals of heavy metal elements in the industrial emission particle group using a second wavelength laser, and at the same time obtain the spatial intensity gradient of the scattering signal, and generate the heavy metal diffusion trajectory in the industrial pollution characteristic spectrum based on the spatial intensity gradient. The overlay module is used to spatially overlay the mineral migration path and heavy metal diffusion trajectory with the aerosol three-dimensional distribution data obtained by the high-altitude detection equipment to generate a thermal distribution map of the pollution source. The allocation module is used to extract the distribution deviation of the mineral migration path and the offset of the heavy metal diffusion trajectory when the spatial distribution density of the natural dust particle group and the industrial emission particle group exceeds the corresponding spatial distribution density in the thermal distribution map. Then, it reallocates the pollution source contribution weights of the natural dust particle group and the industrial emission particle group to output the pollution source analysis results.

9. A computing device, characterized in that, It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement the atmospheric pollutant source tracing analysis method as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that, The device contains a computer program that, when executed by a computer, implements an atmospheric pollutant source tracing and analysis method as described in any one of claims 1 to 7.

Images