CN112613675A - Analyzing pollution source and meteorological factor to PM of different degrees2.5Machine learning model of pollution impact contributions and effects - Google Patents
Analyzing pollution source and meteorological factor to PM of different degrees2.5Machine learning model of pollution impact contributions and effects Download PDFInfo
- Publication number
- CN112613675A CN112613675A CN202011592008.3A CN202011592008A CN112613675A CN 112613675 A CN112613675 A CN 112613675A CN 202011592008 A CN202011592008 A CN 202011592008A CN 112613675 A CN112613675 A CN 112613675A
- Authority
- CN
- China
- Prior art keywords
- pollution
- learning model
- machine learning
- source
- factor
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000000694 effects Effects 0.000 title claims abstract description 23
- 238000010801 machine learning Methods 0.000 claims abstract description 35
- 238000000034 method Methods 0.000 claims abstract description 18
- 238000004458 analytical method Methods 0.000 claims abstract description 17
- 238000011109 contamination Methods 0.000 claims abstract description 10
- 239000013618 particulate matter Substances 0.000 claims description 48
- 239000011159 matrix material Substances 0.000 claims description 25
- 239000000126 substance Substances 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 12
- 238000012544 monitoring process Methods 0.000 claims description 10
- 238000001228 spectrum Methods 0.000 claims description 9
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims description 6
- 238000004422 calculation algorithm Methods 0.000 claims description 6
- 229910052799 carbon Inorganic materials 0.000 claims description 6
- 150000002500 ions Chemical class 0.000 claims description 6
- 239000003344 environmental pollutant Substances 0.000 claims description 5
- 231100000719 pollutant Toxicity 0.000 claims description 5
- 238000007637 random forest analysis Methods 0.000 claims description 5
- 230000002195 synergetic effect Effects 0.000 claims description 5
- 230000001419 dependent effect Effects 0.000 claims description 4
- 238000010206 sensitivity analysis Methods 0.000 claims description 4
- 230000005764 inhibitory process Effects 0.000 claims description 3
- 229910052742 iron Inorganic materials 0.000 claims description 3
- 229910052748 manganese Inorganic materials 0.000 claims description 3
- 229910052759 nickel Inorganic materials 0.000 claims description 3
- 229910052700 potassium Inorganic materials 0.000 claims description 3
- 229910052725 zinc Inorganic materials 0.000 claims description 3
- 229910052787 antimony Inorganic materials 0.000 claims description 2
- 229910052785 arsenic Inorganic materials 0.000 claims description 2
- 238000013473 artificial intelligence Methods 0.000 claims description 2
- 229910052797 bismuth Inorganic materials 0.000 claims description 2
- 229910052793 cadmium Inorganic materials 0.000 claims description 2
- 229910052804 chromium Inorganic materials 0.000 claims description 2
- 229910052802 copper Inorganic materials 0.000 claims description 2
- 238000000354 decomposition reaction Methods 0.000 claims description 2
- 239000000284 extract Substances 0.000 claims description 2
- 229910052733 gallium Inorganic materials 0.000 claims description 2
- 229910001385 heavy metal Inorganic materials 0.000 claims description 2
- 229910052745 lead Inorganic materials 0.000 claims description 2
- 239000000203 mixture Substances 0.000 claims description 2
- 229910052718 tin Inorganic materials 0.000 claims description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 2
- 238000005315 distribution function Methods 0.000 claims 1
- OYFJQPXVCSSHAI-QFPUQLAESA-N enalapril maleate Chemical compound OC(=O)\C=C/C(O)=O.C([C@@H](C(=O)OCC)N[C@@H](C)C(=O)N1[C@@H](CCC1)C(O)=O)CC1=CC=CC=C1 OYFJQPXVCSSHAI-QFPUQLAESA-N 0.000 claims 1
- 238000005457 optimization Methods 0.000 claims 1
- 239000002245 particle Substances 0.000 claims 1
- 230000002265 prevention Effects 0.000 abstract description 5
- 238000012417 linear regression Methods 0.000 description 6
- 229910002651 NO3 Inorganic materials 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 239000003245 coal Substances 0.000 description 3
- 239000010419 fine particle Substances 0.000 description 3
- 239000002028 Biomass Substances 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- NHNBFGGVMKEFGY-UHFFFAOYSA-N nitrate group Chemical group [N+](=O)([O-])[O-] NHNBFGGVMKEFGY-UHFFFAOYSA-N 0.000 description 2
- 238000011002 quantification Methods 0.000 description 2
- 238000004445 quantitative analysis Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- QAOWNCQODCNURD-UHFFFAOYSA-N sulfuric acid group Chemical group S(O)(O)(=O)=O QAOWNCQODCNURD-UHFFFAOYSA-N 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 208000024172 Cardiovascular disease Diseases 0.000 description 1
- GRYLNZFGIOXLOG-UHFFFAOYSA-N Nitric acid Chemical group O[N+]([O-])=O GRYLNZFGIOXLOG-UHFFFAOYSA-N 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000003915 air pollution Methods 0.000 description 1
- 239000008277 atmospheric particulate matter Substances 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 229910052791 calcium Inorganic materials 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000002485 combustion reaction Methods 0.000 description 1
- 239000000356 contaminant Substances 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000002526 effect on cardiovascular system Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 239000010881 fly ash Substances 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 229910052749 magnesium Inorganic materials 0.000 description 1
- 229910017604 nitric acid Inorganic materials 0.000 description 1
- 230000002028 premature Effects 0.000 description 1
- 238000012950 reanalysis Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 208000023504 respiratory system disease Diseases 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 229910052708 sodium Inorganic materials 0.000 description 1
- QAOWNCQODCNURD-UHFFFAOYSA-L sulfate group Chemical group S(=O)(=O)([O-])[O-] QAOWNCQODCNURD-UHFFFAOYSA-L 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 229910052719 titanium Inorganic materials 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N15/00—Investigating characteristics of particles; Investigating permeability, pore-volume or surface-area of porous materials
- G01N15/06—Investigating concentration of particle suspensions
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N30/00—Investigating or analysing materials by separation into components using adsorption, absorption or similar phenomena or using ion-exchange, e.g. chromatography or field flow fractionation
- G01N30/96—Investigating or analysing materials by separation into components using adsorption, absorption or similar phenomena or using ion-exchange, e.g. chromatography or field flow fractionation using ion-exchange
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A50/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE in human health protection, e.g. against extreme weather
- Y02A50/20—Air quality improvement or preservation, e.g. vehicle emission control or emission reduction by using catalytic converters
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Theoretical Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Business, Economics & Management (AREA)
- Pathology (AREA)
- Immunology (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Artificial Intelligence (AREA)
- Economics (AREA)
- Evolutionary Computation (AREA)
- Combustion & Propulsion (AREA)
- Game Theory and Decision Science (AREA)
- Food Science & Technology (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Dispersion Chemistry (AREA)
- Computational Linguistics (AREA)
- Development Economics (AREA)
- Biophysics (AREA)
- Medicinal Chemistry (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
Abstract
Analyzing pollution source and meteorological factor to PM of different degrees2.5A machine learning model of pollution influence contribution and effect relates to the field of prevention and control of atmospheric particulate pollution, and comprises the following steps: step 1, calculating PM2.5 pollution source analysis based on online multi-component data; step 2, building a machine learning model; calculating the influence degree of each factor on PM2.5 by using the sharp value; step 4, analyzing pollution source and meteorological phenomenaFactor pairs of different degrees of PM2.5The effect of contamination; analyzing the influence effect of the primary pollution source and meteorological factors on the concentration generated by the secondary pollution source; the machine learning model built based on data driving can quickly identify PM with different degrees2.5Dominant factors in pollution, quantifying single and multiple factors versus different levels of PM2.5The influence effect of pollution is analyzed sensitively, and the method has good application prospect in the prevention, control and control of atmospheric pollution.
Description
Technical Field
The invention relates to the field of prevention and control of atmospheric particulate pollution, in particular to data-driven identification and quantification P.M2.5Machine learning models of pollution impact factors, contributions, and effects.
Background
Fine Particulate Matter (PM)2.5) As a main pollutant in the atmospheric environment, the pollution-free environment-friendly type air conditioner not only causes low environmental visibility and changes climate conditions, but also has great harm to human health, and is easy to cause cardiovascular and respiratory diseases and even premature death of human bodies. PM of China2.5Pollution is mainly synergistic by emission, weather, atmospheric chemical reaction, transmission and other factors, the pollution process is very complex, and different PM matters exist2.5During the contamination process, the contaminant chemistry reaction path and rate are different. Identifying and quantitatively evaluating fine Particulate Matter (PM) in various regions2.5) The influencing factors in pollution are important research problems in air pollution prevention and control.
At present, most of researches use source emission lists, chemical component data and reanalysis data, cause analysis is carried out on pollution events in one area by combining a chemical mass transmission model, an air quality prediction mode, a backward track model and the like, single heavy pollution events in one area are often focused, the limited range is narrow, quantitative analysis of multiple events with different pollution degrees and the same pollution degree on high-time classification is lacked, meanwhile, the analysis process is delayed, and PM with different degrees cannot be rapidly and accurately identified and quantified2.5And (3) evaluating the main factors in the pollution and the influence of each factor on the pollution.
With the requirement of refined source analysis, rapid, time-by-time and accurate quantitative analysis of pollution causes of fine particles is to be continuously and deeply researched. The machine learning method driven by the slave data has the characteristics of high speed, high accuracy, strong nonlinear characterization capability and the like, and has unique advantages on the method for quantifying the influence factors of the fine particles.
Disclosure of Invention
The purpose of the invention is to solve the problem of different degrees of PM2.5In pollution, the recognition efficiency of the leading factors is low, and the influence effect of each factor on the pollution is unclear; the method is used for identifying the PM with different degrees based on the machine learning model established by various online data (conventional pollutant concentration data, chemical component data, meteorological data, source analysis result and the like)2.5Leading factors in pollution and ranking the importance, and calculating the PM pair of each factor by using a xiapril value2.5Influence degree is combined with partial dependence calculation to respectively quantify pollution sources and meteorological factors on PM with different degrees2.5And carrying out sensitivity analysis on the single-factor and multi-factor synergistic influence of the pollution, and quantifying the influence contribution of the primary pollution source and the meteorological factors on the secondary pollution source generated concentration.
The invention provides a method for analyzing pollution sources and meteorological factors for PM with different degrees2.5The machine learning model of pollution influence contribution and effect adopts the following technical scheme:
monitoring the concentration of particulate matters and the concentration of chemical components by using an online monitoring instrument, constructing a multi-component online data set, inputting the data set into a PMF model, carrying out preliminary inspection on the data, carrying out basic model operation by setting parameters such as factor number, operation times and the like, analyzing to obtain a factor spectrum matrix and a factor contribution matrix, optimizing the factor spectrum matrix and the factor contribution matrix by rotating calculation, identifying the factors into different source types according to chemical identification components in the factor spectrum matrix, and carrying out multiple linear regression calculation on the factor contribution matrix and the concentration of the particulate matters to calculate the PM source by using the pollution source2.5Contribution of contamination.
based on a python3.9.0 platform, a machine learning model is built by using a random forest algorithm, a data set (conventional pollutant concentration data, chemical component data, meteorological data and source analysis result data) is divided into a training set and a testing set by a 10-fold cross validation method, and internal parameters of the machine learning model are continuously adjusted and optimized according to the size (accuracy, F1-Score and the like) of model evaluation indexes to form an optimal machine learning model;
step 3, calculating each factor pair PM2.5The degree of influence;
calculating the value of the Charapril, drawing a shape graph, and quantifying the influence factors on the PM2.5The extent of promotion or inhibition of concentration;
step 4, analyzing pollution sources and meteorological factors to PM with different degrees2.5The effect of contamination;
identifying different degrees of PM using an optimized machine learning model2.5The leading factors in the pollution are ranked in importance, and meanwhile, the pollution source and the meteorological factors are respectively quantized to PM with different degrees by combining partial dependence calculation and a xiapril value2.5Carrying out sensitivity analysis on the single-factor and multi-factor synergistic influence of pollution;
analyzing the influence effect of the primary pollution source and meteorological factors on the concentration generated by the secondary pollution source;
and quantifying the influence contribution of the primary pollution source and meteorological factors on the secondary pollution source generated concentration by using the optimized machine learning model.
The invention has the advantages and beneficial effects that:
the invention provides a machine learning algorithm-based method for analyzing PM with different degrees of pollution sources and meteorological factors2.5The influence contribution and effect of pollution can realize the rapid identification of PM with different degrees2.5Pollution leading factors, quantification of all pollution sources and meteorological factors, primary pollution sources and meteorological factors on different degrees of PM2.5The pollution is partially dependent on analysis and the Charpy value analysis, and the influence contribution of the primary pollution source and meteorological factors on the generation concentration of the secondary pollution source is quantified, so that scientific basis and guidance are provided for atmosphere pollution prevention and control, and the method has good popularization and application prospects.
Drawings
FIG. 1 is a schematic view showing the analysis of a contamination sourceDifferent degree of PM from meteorological factor2.5Pollution impact contribution and effect flow diagrams.
FIG. 2 shows the atmospheric Particulate Matter (PM) of the present invention2.5) And (5) source analysis result graph.
FIG. 3 shows the pollution source and meteorological factors of the present invention versus PM2.5The degree of influence of (c).
FIG. 4 is a graph of the present invention for each factor vs. PM2.5The degree of influence of (c).
FIG. 5 shows different levels of PM according to the present invention2.5The importance of each factor in contamination is ranked.
FIG. 6 shows the secondary nitrate versus different levels of PM according to the present invention2.5Susceptibility analysis of contamination.
FIG. 7 is a three-dimensional partial dependence calculation of vehicle, coal and secondary pollution source generation concentrations.
Detailed Description
Example 1
The embodiment utilizes online monitoring data and a machine learning algorithm to analyze the influence contribution and the effect of pollution sources and meteorological factors on PM2.5 pollution with different degrees, and comprises the following specific steps:
1. computing PM based on online multi-component data2.5Analyzing a pollution source;
inputting online multi-component data into PMF model input data, wherein the online multi-component data is multi-component data formed by online monitoring data of particulate matter concentration and chemical components thereof monitored by different instruments and comprises particulate matter concentration, water-soluble ions, carbon components and element concentration data; the particulate matter concentration refers to PM measured by a particulate matter on-line monitoring instrument2.5Concentration; the water soluble ions are measured by an on-line ion chromatograph, comprising NH4 +、Na+、K+、Ca2+、Mg2+、SO4 2-、NO3 -And Cl-(ii) a Carbon composition was measured by a semi-continuous OC/EC instrument, including OC and EC; elements are monitored by a heavy metal on-line analyzer and comprise K, Ca, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, As, Se, Ag, Cd, Sn, Sb, Ba, Au, Hg, Tl, Pb and Bi components. The PMF model input parameters are set up,decomposing an on-line monitoring receptor data matrix (X) into a factor spectrum matrix (F) and a source contribution matrix (G), identifying the factors into different source classes according to chemical identification components in the factor spectrum matrix (F), and performing multiple linear regression by using the source contribution matrix (G) and the concentration of particulate matters to calculate the PM of a pollution source2.5Contribution of contamination. The PMF model in the invention extracts a plurality of factors from a receptor matrix by utilizing multi-component data through positive definite matrix decomposition to obtain a factor spectrum matrix (X) and a source contribution matrix (G), identifies the factors into different source classes according to chemical identification components in the factor spectrum matrix (X), and calculates the PM source by utilizing the source contribution matrix (G) and the concentration of particulate matters to carry out multiple linear regression2.5A model of pollution contribution. The multivariate linear regression refers to regression comprising two or more independent variables, and the multivariate linear regression is carried out by setting the concentration of the particulate matters as dependent variables and setting the source contribution matrix as independent variables.
Continuously sampling from 1 month and 9 months in 2017 to 30 months in 2018, wherein the time resolution of the monitored data is 1 hour, so that 4742 pieces of receptor data containing K, Ca, Na, Mg, Zn, Fe, Mn, Ti, Ni and NO are obtained3 -、SO4 2-、NH4 +And the OC and the EC have 14 components. The input model parameters are as follows: the number of rows 4742, columns 14, and factor 6, is repeated 20 times. Identification component (dust-Ca, secondary sulfuric acid source-SO) according to the identified source class4 2-,NH4 +Secondary nitric acid source-NO3 -,NH4 +Second organic carbon/Biomass Combustion-OC, K+Coal sources-OC, EC, SO4 2-Motor vehicles-OC, EC), the most reasonable result in the iterative calculations is selected. Associating the selected source contribution matrix with the PM2.5And performing multiple linear regression to obtain the source concentration and the source contribution, wherein the source analysis result is shown in figure 2.
2. Establishing a machine learning model based on the source contribution data and the time-by-time meteorological data;
the machine learning open source library lightgbm based on python is used for constructing a machine learning model by using a random forest algorithm. Source contribution result data (Dust-fly ash, SS-secondary sulfate, SOC/BIO-secondary organic carbon and biomass combustion SN-secondary nitrate Vehicle-Coal source) and meteorological data (T-temperature, RH-relative humidity P-pressure WS-wind speed WD-wind direction) and PM2.5 data are included into a machine learning model for training. The machine learning model is a branch of artificial intelligence, and is an implementation method. The invention relates to a random forest, in particular to an integrated learning method for distinguishing and classifying data by utilizing a plurality of classification trees.
3. PM pair calculation method based on machine learning model and pollution source and meteorological factors2.5Degree of influence of
Calculating the value of the Charapril, drawing a shape graph, and quantifying the influence factors on the PM2.5The degree of concentration promotion or inhibition; the inventive value of the charapril is a compromise utility allocation scheme that comprehensively considers conflicting party requirements. The results are shown in FIGS. 3 and 4.
4. Analysis of pollution sources and meteorological factors for different levels of PM2.5Effect of contamination
Calculating PM of each input variable based on random forest algorithm2.5The importance of the pollution is calculated by the sum of the information gain or the reduction of the kini coefficient brought by each feature in the process of calculating the branches of the decision tree. Meanwhile, depending calculation and a xiapril value of machine learning are combined, pollution sources and meteorological factors are quantized respectively to PM with different degrees2.5Carrying out sensitivity analysis on the single-factor and multi-factor synergistic influence of pollution; the present invention relies in part on calculations to interpret the relationship of an independent variable to a dependent variable, reflecting how the predicted variable affects the prediction of the model, and the xiapril values used to calculate the contribution of each individual feature to the model output. The results are shown in FIGS. 5 and 6.
5. Analyzing the influence effect of the primary pollution source and meteorological factors on the generation concentration of the secondary pollutants;
and analyzing the influence and action degree of the meteorological factors and the change of the primary pollution source on the secondarily generated fine particles through machine learning three-dimensional partial dependence calculation. The results are shown in FIG. 7.
Claims (6)
1. Analyzing pollution source and meteorological factor to PM of different degrees2.5A machine learning model of pollution impact contributions and effects, comprising:
computing PM based on online multi-component data2.5Analyzing pollution sources, namely monitoring the concentration of the particulate matters and the concentration of chemical components by using an online monitoring instrument, constructing a multi-component online data set, inputting the data set into a PMF (particle size distribution function) model for analyzing the sources of the atmospheric particulate matters, analyzing to obtain the number of pollution sources and the PM (particulate matter) source of the pollution sources2.5A contribution of contamination;
building a machine learning model, building the machine learning model by using a random forest algorithm, and continuously adjusting parameters of the optimization model based on various data sets such as conventional pollutant concentration data, chemical component data, meteorological data and source analysis result data to form an optimal machine learning model;
analysis of pollution sources and meteorological factors on PM2.5Calculating the influence degree of the influence factor, calculating the value of the summer pril, drawing a shape graph, and quantifying the influence factor on the PM2.5The extent of promotion or inhibition of concentration;
analysis of pollution sources and meteorological factors for different levels of PM2.5Impact effect of pollution, identifying different degrees of PM by using optimized machine learning model2.5The leading factors in the pollution are ranked in importance, and meanwhile, the pollution source and the meteorological factors are respectively quantized to PM with different degrees by combining partial dependence calculation and a xiapril value2.5Carrying out sensitivity analysis on the single-factor and multi-factor synergistic influence of pollution;
analyzing the influence effect of the primary pollution source and the meteorological factors on the secondary pollution source generated concentration, and quantifying the influence contribution of the primary pollution source and the meteorological factors on the secondary pollution source generated concentration by utilizing an optimized machine learning model.
2. The method of claim 1, wherein the different levels of PM are analyzed for pollution sources and meteorological factors2.5A machine learning model of pollution impact contributions and effects characterized by: the online multi-component data is formed by online monitoring data of the concentration of the particulate matters and the chemical components thereof monitored by different instruments, and comprises the data of the concentration of the particulate matters, water-soluble ions, carbon components and element concentrations; the particulate matter concentration refers to PM measured by a particulate matter on-line monitoring instrument2.5Concentration; the water soluble ions are measured by an on-line ion chromatograph, comprising NH4 +、Na+、K+、Ca2+、Mg2+、SO4 2-、NO3 -And Cl-(ii) a Carbon composition was measured by a semi-continuous OC/EC instrument, including OC and EC; elements are monitored by a heavy metal on-line analyzer and comprise K, Ca, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, As, Se, Ag, Cd, Sn, Sb, Ba, Au, Hg, Tl, Pb and Bi components.
3. The method of claim 1, wherein the different levels of PM are analyzed for pollution sources and meteorological factors2.5A machine learning model of pollution impact contributions and effects characterized by: the PMF model is a model which utilizes multi-component data, extracts a plurality of factors from a receptor matrix through positive definite matrix decomposition to obtain factor spectrum matrix information and factor contribution matrix information, identifies the factors into different source classes according to chemical identification components in the factor spectrum matrix, and calculates the contribution of each source class by utilizing the factor contribution matrix.
4. The method of claim 1, wherein the different levels of PM are analyzed for pollution sources and meteorological factors2.5A machine learning model of pollution impact contributions and effects characterized by: analysis of pollution sources and meteorological factors on PM2.5The contribution of each individual feature to the model output is calculated.
5. As set forth in claim 1Pollution source and meteorological factor pair PM of different degrees2.5A machine learning model of pollution impact contributions and effects characterized by: the machine learning model refers to a branch of artificial intelligence, and is also an implementation method, and the model is used for predicting and deciding data according to the data learning model of the sample.
6. The method of claim 1, wherein the different levels of PM are analyzed for pollution sources and meteorological factors2.5A machine learning model of pollution impact contributions and effects characterized by: the partial dependence calculation is used for explaining the relation between an independent variable and a dependent variable and reflecting how a predictive variable influences the prediction of the model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011592008.3A CN112613675A (en) | 2020-12-29 | 2020-12-29 | Analyzing pollution source and meteorological factor to PM of different degrees2.5Machine learning model of pollution impact contributions and effects |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011592008.3A CN112613675A (en) | 2020-12-29 | 2020-12-29 | Analyzing pollution source and meteorological factor to PM of different degrees2.5Machine learning model of pollution impact contributions and effects |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112613675A true CN112613675A (en) | 2021-04-06 |
Family
ID=75248802
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011592008.3A Pending CN112613675A (en) | 2020-12-29 | 2020-12-29 | Analyzing pollution source and meteorological factor to PM of different degrees2.5Machine learning model of pollution impact contributions and effects |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112613675A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113295826A (en) * | 2021-05-24 | 2021-08-24 | 暨南大学 | Motor vehicle exhaust emission test management system based on Internet of vehicles and intelligent diagnosis method |
CN114295749A (en) * | 2021-12-30 | 2022-04-08 | 南京大学 | Intelligent tracing method and system for organic pollution of water body |
CN115453064A (en) * | 2022-09-22 | 2022-12-09 | 山东大学 | Fine particle air pollution cause analysis method and system |
CN117877647A (en) * | 2024-03-13 | 2024-04-12 | 苏州创腾软件有限公司 | Recipe generation method and device based on machine learning |
CN117875576A (en) * | 2024-03-13 | 2024-04-12 | 四川国蓝中天环境科技集团有限公司 | Urban atmosphere pollution analysis method based on structured case library |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107239613A (en) * | 2017-06-05 | 2017-10-10 | 南开大学 | A kind of intelligent source class recognition methods based on online data and Factor Analysis Model |
CN108956881A (en) * | 2018-07-16 | 2018-12-07 | 南开大学 | SDABB Source Apportionment, system, terminal device and computer readable storage medium |
US20200393434A1 (en) * | 2019-06-13 | 2020-12-17 | Southern University Of Science And Technology | Evaluation method for impact of emission control on air quality, device and storage medium thereof |
-
2020
- 2020-12-29 CN CN202011592008.3A patent/CN112613675A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107239613A (en) * | 2017-06-05 | 2017-10-10 | 南开大学 | A kind of intelligent source class recognition methods based on online data and Factor Analysis Model |
CN108956881A (en) * | 2018-07-16 | 2018-12-07 | 南开大学 | SDABB Source Apportionment, system, terminal device and computer readable storage medium |
US20200393434A1 (en) * | 2019-06-13 | 2020-12-17 | Southern University Of Science And Technology | Evaluation method for impact of emission control on air quality, device and storage medium thereof |
Non-Patent Citations (1)
Title |
---|
王雨晨: "基于随机森林的上海市PM2.5质量浓度预测研究", 《中国优秀博硕士学位论文全文数据库(硕士)工程科技Ⅰ辑》, no. 01, 15 January 2018 (2018-01-15) * |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113295826A (en) * | 2021-05-24 | 2021-08-24 | 暨南大学 | Motor vehicle exhaust emission test management system based on Internet of vehicles and intelligent diagnosis method |
CN114295749A (en) * | 2021-12-30 | 2022-04-08 | 南京大学 | Intelligent tracing method and system for organic pollution of water body |
CN114295749B (en) * | 2021-12-30 | 2022-10-25 | 南京大学 | Intelligent tracing method and system for organic pollution of water body |
WO2023024463A1 (en) * | 2021-12-30 | 2023-03-02 | 南京大学 | Intelligent tracing method and system for organic pollution of water body |
US11965871B2 (en) | 2021-12-30 | 2024-04-23 | Nanjing University | Method and system for intelligent source tracing of organic pollution of water body |
CN115453064A (en) * | 2022-09-22 | 2022-12-09 | 山东大学 | Fine particle air pollution cause analysis method and system |
CN115453064B (en) * | 2022-09-22 | 2023-09-05 | 山东大学 | Fine particulate matter air pollution cause analysis method and system |
CN117877647A (en) * | 2024-03-13 | 2024-04-12 | 苏州创腾软件有限公司 | Recipe generation method and device based on machine learning |
CN117875576A (en) * | 2024-03-13 | 2024-04-12 | 四川国蓝中天环境科技集团有限公司 | Urban atmosphere pollution analysis method based on structured case library |
CN117875576B (en) * | 2024-03-13 | 2024-05-24 | 四川国蓝中天环境科技集团有限公司 | Urban atmosphere pollution analysis method based on structured case library |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112613675A (en) | Analyzing pollution source and meteorological factor to PM of different degrees2.5Machine learning model of pollution impact contributions and effects | |
CN111814111B (en) | Industrial park atmospheric pollutant tracing method | |
CN110057983B (en) | Ozone source analysis method based on observation data and chemical mechanism | |
CN108268935B (en) | PM2.5 concentration value prediction method and system based on time sequence recurrent neural network | |
CN110489785B (en) | Online source analysis method and system for atmospheric pollutants | |
CN105631203A (en) | Method for recognizing heavy metal pollution source in soil | |
CN110044779B (en) | method for real-time online monitoring and source analysis combined use of atmospheric heavy metal particles | |
CN109060617B (en) | Simulation and analysis method for influences of different agricultural ammonia emission situations on PM2.5 concentration | |
CN107239613A (en) | A kind of intelligent source class recognition methods based on online data and Factor Analysis Model | |
Jiao et al. | Prediction of air quality index based on LSTM | |
CN112183709B (en) | Method for predicting and early warning excessive dioxin in waste incineration gas | |
CN108802262B (en) | Target factor conversion-PMF source analysis method based on online multi-component data | |
CN112967764B (en) | Multi-technology coupled pollutant source analysis method and device | |
CN115453064B (en) | Fine particulate matter air pollution cause analysis method and system | |
CN109377440B (en) | PM2.5 and O3 concentration collaborative prediction method based on multitask integrated learner | |
CN108446531A (en) | Receptor Si and Al reconstructing method based on actual measurement derived components spectrum and Source apportionment | |
CN114912343A (en) | LSTM neural network-based air quality secondary prediction model construction method | |
CN109376903B (en) | PM2.5 concentration value prediction method based on game neural network | |
CN114117893A (en) | Method for analyzing atmospheric dust-fall pollution source and evaluating dust-fall marginal effect of pollution source | |
CN109085291B (en) | Missing component iterative inversion calibration nesting-PMF source analysis algorithm | |
CN113192569A (en) | Harmful gas monitoring method based on improved particle swarm and error feedback neural network | |
CN116930423A (en) | Automatic verification and evaluation method and system for air quality model simulation effect | |
CN112711911B (en) | Rapid pollution tracing method applied to boundary observation based on pollution source spectrum library | |
CN114140296A (en) | Intelligent tracing method for atmospheric pollutants by decomposing and restricting pollution source component spectrum by ICA algorithm | |
Roy et al. | Development of PM2. 5 concentration prediction model using machine learning approaches |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |