AU2021105563A4 - Method for Traceability of Air Pollutants Based on Coupled Machine Learning and Correlation Analysis - Google Patents
Method for Traceability of Air Pollutants Based on Coupled Machine Learning and Correlation Analysis Download PDFInfo
- Publication number
- AU2021105563A4 AU2021105563A4 AU2021105563A AU2021105563A AU2021105563A4 AU 2021105563 A4 AU2021105563 A4 AU 2021105563A4 AU 2021105563 A AU2021105563 A AU 2021105563A AU 2021105563 A AU2021105563 A AU 2021105563A AU 2021105563 A4 AU2021105563 A4 AU 2021105563A4
- Authority
- AU
- Australia
- Prior art keywords
- machine learning
- spatio
- traceability
- model
- pollution
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/30—Services specially adapted for particular environments, situations or purposes
- H04W4/38—Services specially adapted for particular environments, situations or purposes for collecting sensor information
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/022—Knowledge engineering; Knowledge acquisition
- G06N5/025—Extracting rules from data
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Evolutionary Computation (AREA)
- Computational Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Signal Processing (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Networks & Wireless Communication (AREA)
- Evolutionary Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computing Systems (AREA)
- Automation & Control Theory (AREA)
- Probability & Statistics with Applications (AREA)
- Operations Research (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention discloses a method for traceability of air pollutants based on
coupled machine learning and correlation analysis. The method specifically comprises the
following steps: establishing a transient model for pollutant concentration distribution based
on spatio-temporal pollutant concentration data of a regional grid source; extracting features
from a pollutant concentration correlation coefficient matrix for a grid source by Gaussian
regression, and coupling machine learning algorithms for intelligent recognition of pollutant
transport channels and pollutant source regions. According to the present invention, features
are extracted from the spatio-temporal correlation matrix of pollutant concentrations by
Gaussian regression and correlation analysis based on the transient model established based on
spatio-temporal pollutant concentration data of a regional grid source, so as to solve the
problems of delayed response and uncertainty in time window of pollutant concentrations. In
addition, model training data are constantly updated by using machine learning algorithms,
thereby ensuring the continuous and effective improvement of the accuracy of traceability
algorithms.
1/1
FIGURES OF THE SPECIFICATION
d itorica -Data
Preproces
n
Data sing the
source data in Possiblc
New pollution
data source
election Model
model Optimization
Inew
data
Results
of model
applicati
on
FIG. 1
Description
1/1
d itorica -Data
n Preproces Data sing the source data in Possiblc New pollution data source election Model model Optimization Inew data
Results of model applicati on
FIG. 1
Method for Traceability of Air Pollutants Based on Coupled Machine Learning
and Correlation Analysis
The present invention relates to the technical field of pollutant traceability, in
particular to a method for traceability of air pollutants based on coupled machine
learning and correlation analysis.
With the rapid development of economy, the acceleration of industrialization and
urbanization, and the increase of energy consumption in China, a series of
atmospheric environmental problems have emerged. Compared with pollutants in the
environment such as pollutants in waters and soil, air pollutants have the
characteristics of easy diffusion, easy mixing and unclear pollution paths, and may be
affected by emission sources, pollution processes, meteorological conditions and the
like. Among them, emission sources are internal factors, meteorological conditions
are external factors, and pollution processes are motivation factors. Since motivation
factors and external factors are mainly influenced by the objective laws of nature, it is
difficult for human beings to control these factors. Therefore, the control of internal
factors is the most effective method for air pollution control and environmental
management, the core of which is to identify pollution sources, clarify causes of
pollution, achieve targeted governance and improve control efficiency.
Identification of the sources of air pollution can be divided into two categories:
pollution traceability, focusing on the traceability of emission sources in terms of
spatio-temporal distribution; and analysis of emission sources, focusing on the
analysis of composition and industry of emission sources. As the main means for
precise control and scientific control of ambient air quality, atmospheric fine grid system is widely used. Grid-based environmental statistical data analysis can achieve rough traceability of air pollution, but the response time is long. As a result, researchers improve the response time by traceability of air pollution based on model software and machine learning algorithms. However, the existing methods have the following disadvantages in achieving traceability of air pollution: (1) backward trajectory: backward trajectory is an integrated model system for calculating and analyzing airflow motion, deposition and diffusion trajectories, the core of which is to calculate and describe air mass motion through wind direction and wind speed in a three-dimensional meteorological field, and thus localize pollution sources through air mass trajectories. However, the method is highly dependent on wind field data and is limited by the input field of multiple meteorological elements. The current research mainly focuses on the long-range transport on a short time scale and the identification of external pollution sources, which can provide theoretical reference in dealing with overseas pollution sources and coordinated inter-regional prevention and control, but is not applicable in dealing with the traceability of small-scale regional endogenous pollution for the time being. (2) Probabilistic method: probabilistic method is a method for traceability of pollution mainly developed for the complexity of physical and chemical processes of air pollution and the discreteness of numerical models. The main principle is to combine available concentration observations with priori information, and analyze and mine the uncertainty and confidence intervals of posterior parameters based on a large number of historical data. In the application, a large amount of supporting data and priori information on known pollution sources are required, which is difficult to achieve in atmospheric emergency response. (3)
Analytical method for sources of particulate matter: the method is developed for
qualitative identification of pollution sources by analyzing the physicochemical properties of particulate matters in ambient air and samples from the pollution sources. Meanwhile, the contribution rate of pollution sources can be calculated quantitatively based on mathematical statistics and numerical model simulation.
However, the method focuses on the analysis of the composition and industry of
emission sources, and cannot obtain the lock-in and contribution rate of pollution
sources in geographical space. Therefore, it is difficult for the method to meet the
requirements for accurate traceability of air pollution, targeted governance and
efficient control of air pollution.
At this stage, traceability analysis based on models is mostly conducted from the
perspective of the influence of boundary conditions on pollutant diffusion, such as
wind direction, wind power and other factors. Such models are not universal and
cannot achieve rapid deployment in different regional scenarios, and personnel with
certain knowledge background are required to adjust localization parameters. In
addition, changes in real-time data during pollutant transport are not considered,
therefore, such models are used for steady-state modelling, and it is impossible to
correct dynamic response of the models according to transient spatio-temporal
pollutant concentration data. Moreover, the problems of delay effects during pollutant
concentration transport and uncertainty in valid time window of pollutant events
cannot be effectively considered for the existing models.
The purpose of the present invention is to provide a method for traceability of air
pollutants based on coupled machine learning and correlation analysis. According to
the method, intelligent recognition is performed on pollutant transport channels and
pollutant source regions by using coupled machine learning and correlation analysis
based on spatio-temporal pollutant concentration data of a regional grid source.
In order to achieve the purpose, the present invention provides a method for
traceability of air pollutants based on coupled machine learning and correlation
analysis, comprising the following steps:
Si. acquiring real-time spatio-temporal data and historical spatio-temporal data
for each grid source site in a target region;
S2. constructing a database based on the real-time spatio-temporal data and the
historical spatio-temporal data; and extracting historical spatio-temporal data over a
period of time from the database;
S3. constructing a transient model for pollutant concentration distribution based
on the historical spatio-temporal data over a period of time;
S4. extracting features from the transient model for pollutant concentration
distribution by Gaussian regression, and standardizing the extracted features;
S5. constructing a possible pollution source selection model by using machine
learning algorithms, and training the possible pollution source selection model with
the extracted historical spatio-temporal data over a period of time as a training set;
then inputting the extracted features into the trained possible pollution source
selection model, and outputting the result of whether the grid source is on a transport
path;
S6. repeating S3-S4 for the real-time spatio-temporal data to obtain features R2 ,
and 6, and standardizing the features to obtain preprocessed new data; and
S7. taking the preprocessed new data as an input into the trained possible
pollution source selection model, and outputting whether the grid source is on the
transport path; and adding the output result to the training set for optimization and
continual learning of the possible pollution source selection model.
Preferably, both the real-time spatio-temporal data and the historical spatio-temporal data include geographic location information, pollutant concentration information, sampling time and meteorological information.
Preferably, the S3 specifically comprises:
extracting pollutant concentration information from the historical
spatio-temporal data, then obtaining a valid time window i and a transport response
delay j according to set pollution events based on a hierarchical tree structure
constructed by a transport channel grid source k, and constructing a matrix to be
compared step by step and a correlation coefficient matrix in real time, i.e., a transient
model for pollutant concentration distribution.
Preferably, the S4 specifically comprises:
converting the correlation coefficient matrix into a correlation coefficient vector,
and extracting features from the correlation coefficient vector by Gaussian regression
to obtain Gaussian regression eigenvalues R2 k, pk and 6k.
Preferably, the possible pollution source selection model has an expression as
follows:
yk fk(Rk, pk,Sk)
where ykE[0,1], 0 means that the grid source k is not on the transport path and 1
means that the grid source k is on the transport path;fk represents the possible pollution
source selection model, and each site is analyzed based onfk to record the grid source
k with yk=1.
Preferably, the machine learning algorithms include random forest, decision tree,
clustering, Bayesian classification, support vector machine, EM and Adaboost.
Preferably, the S7 specifically comprises:
reassigning the pollution event time and pollution event concentration vector,
repeating S1-S5, then marking step by step until the correlation coefficient is the lowest, and outputting the marking result, i.e., the transport channel and possible pollution source regions, and ending iterative computation to realize continual learning and optimization of the model.
Preferably, the method for constructing the transient model for pollutant
concentration distribution specifically comprises:
Step 1. setting trigger conditions for a pollution event according to national
standards, and automatically marking the pollution event time t;
Step 2. constructing a pollution event concentration vector Xi and a vector Yif to
be compared step by step by the valid time window i and the transport response delay
j of the set pollution event based on Step 1:
Step 3. constructing a matrix Zijk to be compared step by step according to the
vector Yijk to be compared step by step; and
Step 4. constructing a correlation coefficient matrix Rif for the grid source k
based on the matrix Zij to be compared step by step, i.e., a transient model for
pollutant concentration distribution.
The following is the advantageous technical effect of the present invention as
compared with the prior art:
Since most of existing traceability models are steady-state models considering
the influence of boundaries on pollutant diffusion, without considering the problems
of delay effects during pollutant concentration transport and uncertainty in valid time
window of pollutant events, as well as the problem of inability to continuously
improve the stability, universality and accuracy of models. According to the present
invention, features are extracted from the spatio-temporal correlation matrix of
pollutant concentrations by Gaussian regression and correlation analysis based on the
transient model established based on spatio-temporal pollutant concentration data of a regional grid source, so as to solve the problems of delayed response and uncertainty in time window of pollutant concentrations. In addition, model training data are constantly updated by using machine learning algorithms, thereby ensuring the continuous and effective improvement of the accuracy of traceability algorithms.
In order to explain the technical solutions in the embodiments of the present
invention or the prior art more clearly, the drawings used in the embodiments will be
briefly introduced below. Obviously, the drawings in the following description are
some embodiments of the present invention. For those of ordinary skill in the art,
other drawings can be obtained based on these drawings without paying creative
labor.
FIG. 1 is a flowchart showing the method of the present invention.
The technical solutions in the embodiments of the present invention will be
described clearly and completely with reference to the accompanying drawings in the
embodiments of the present invention. Apparently, the described embodiments are
only a part of the embodiments of the present invention, not all of the embodiments.
Based on the embodiments of the present invention, all other embodiments obtained
by those of ordinary skill in the art without creative work should fall within the
protection scope of the present invention.
The present invention will be further described in detail with reference to
accompanying drawings and preferred embodiments for clear understanding of the
above purpose, features and advantages of the present invention.
Embodiment 1
Referring to FIG. 1, the present invention provides a method for traceability of air pollutants based on coupled machine learning and correlation analysis, specifically comprising the following steps:
Si. acquiring real-time spatio-temporal data and historical spatio-temporal data
for each grid source site in a target region;
where both the real-time spatio-temporal data and the historical spatio-temporal
data include geographic location information (latitude and longitude), sampling time,
meteorological information (e.g., wind power and wind direction) and pollutant
concentration information;
S2. constructing a database based on the real-time spatio-temporal data and the
historical spatio-temporal data; and extracting historical spatio-temporal data over a
period of time from the database;
S3. constructing a transient model for pollutant concentration distribution based
on the historical spatio-temporal data over a period of time to improve the dynamic
response speed of the model;
extracting pollutant concentration information from the historical
spatio-temporal data, then obtaining a valid time window i and a transport response
delay j according to set pollution events based on a hierarchical tree structure
constructed by a transport channel grid source k, and constructing a matrix to be
compared step by step and a correlation coefficient matrix, i.e., a transient model for
pollutant concentration distribution, which specifically comprises:
S3.1. setting trigger conditions for a pollution event according to national
standards, and automatically marking the pollution event time t;
S3.2. constructing a pollution event concentration vector Xi, as shown in formula
(1): where xt represents the pollutant concentration of the grid source at which a standard event occurs at time t; i represents the valid time window of a set pollutant event, iE[3, 1];
I = -(2)
where I represents the upper limit of i; the operator [ indicates rounding up;
T is the antecedent duration of the pollution event; and AT is the data monitoring
cycle of the grid source;
S3.3. constructing a vector Yijk to be compared step by step, as shown in
formulas (3)-(5):
Yi/l=(yt-jk,ytr-+1*,...,yt-j+,k) (3)
J [(d/)] (4) AT
rn-kn- kd d = 1,n1dmn (5) Ck
where Y#X, Yif represents the pollutant concentration of a site k at time t-j; j
represents the set pollutant transport response delay, jE[1,J]; J represents the upper
limit ofj; v is the wind speed; a is the angle between wind direction and two points in
space; d is the average distance between any two grid sources; m, nEk, k is the total
number of grid sources; d, n is the distance between grid sources m and n; and C2is
the number of combinations of any two grid sources from k grid sources;
S3.4. constructing a matrix Zijk to be compared step by step according to the
vector Yijk to be compared step by step, as shown in formula (6):
11 1,2 " 6j
Z Y2 1 22 .. 2,j6
S3.5. constructing a correlation coefficient matrix Ri,; of the grid source k based
on the matrix Zi- to be compared step by step, as shown in formula (7):
k k k ri,1 ri,2 --- rij k k k Rk- r2 1 r 2 --- ~j (7) k k k ri-2,1 ri-2,2 --- ri-2,j
k Cov(,XiYi 2 j) Ti-2,j = a~j-ary j (8) Var(Xi)-VarY2j
S4. extracting features from a pollutant concentration correlation coefficient
matrix for a grid source by Gaussian regression, and standardizing the extracted
2 features to obtain Gaussian regression eigenvalues R k, k,6;
converting the correlation coefficient matrix RiJ into a vectorRk
r = (I - 2) -j (9)
then extracting features from the vector Rrk by Gaussian regression to obtain
Gaussian regression eigenvalues R 2 , pk and 6k; where R 2 represents the fitting effect
of Gaussian regression, p represents the mean value of the correlation coefficient, and
6 represents the variance of the correlation coefficient;
S5. constructing a possible pollution source selection model by using machine
learning algorithms, and training the possible pollution source selection model with
the extracted historical spatio-temporal data over a period of time as a training set;
then inputting the extracted features into the trained possible pollution source
selection model, outputting the result of whether the grid source is on a transport path,
and performing intelligent recognition on pollutant transport channels and pollution
source regions to reduce the difficulty in manual analysis and improve the universality
of the model, which specifically comprises: establishing the possible pollution source selection model based on the transient model fkfor pollutant concentration distribution by using a random forest machine learning algorithm, as shown in formula (10):
Y - fk(R , pk,S) (10)
where ykE[0,1], 0 means that the grid source k is not on the transport path and 1
means that the grid source k is on the transport path; Here, whether the grid source site
is on the pollution transport path is manually marked by a professional (0 means that
the grid source is not on the transport path, and 1 means that the grid source is on the
transport path).
Each site is analyzed based onf to record the grid source k with yk=-.
The machine learning algorithms include random forest, decision tree, clustering,
Bayesian classification, support vector machine, EM and Adaboost. Other machine
learning algorithms do not include but are not limited to random forest; a variety of
machine learning algorithms are used for modeling, and a champion model is selected
as the possible pollution source selection model;
S6. extracting new geographic location information (longitude and latitude),
pollutant concentration information, sampling time and meteorological information
(e.g., wind power and wind direction) when a new pollution path grid source (new
data) appears, i.e., real-time spatio-temporal data; repeating S3-S4 for dynamic
correlation analysis to obtain features R 2, p and 6, and standardizing the features to
obtain preprocessed new data; and
S7. inputting the preprocessed new data into the possible pollution source
selection model to obtain the result of whether the grid source is on the transport path;
and adding the result to the training set based on application feedback of the possible
pollution source selection model to realize continual learning and optimization of the model based on new data, thus improving the judgment accuracy, which specifically comprises: reassigning by formulas (11) and (12), repeating S1-S5, then marking k. step by step until the correlation coefficient R,/ is the lowest, and outputting k., i.e., the transport channel and possible pollution source regions, and ending iterative computation to realize continual learning and optimization of the model, where w indicates the wth cycle.
t =t-j (11)
X, = Yk; (12)
The preferred embodiments described herein are only for illustration purpose,
and are not intended to limit the present invention. Various modifications and
improvements on the technical solution of the present invention made by those of
ordinary skill in the art without departing from the design spirit of the present
invention shall fall within the protection scope set forth in claims of the present
invention.
Claims (8)
1. A method for traceability of air pollutants based on coupled machine learning
and correlation analysis, comprising the following steps:
Si. acquiring real-time spatio-temporal data and historical spatio-temporal data
for each grid source site in a target region;
S2. constructing a database based on the real-time spatio-temporal data and the
historical spatio-temporal data; and extracting historical spatio-temporal data over a
period of time from the database;
S3. constructing a transient model for pollutant concentration distribution based
on the historical spatio-temporal data over a period of time;
S4. extracting features from the transient model for pollutant concentration
distribution by Gaussian regression, and standardizing the extracted features;
S5. constructing a possible pollution source selection model by using machine
learning algorithms, and training the possible pollution source selection model with
the extracted historical spatio-temporal data over a period of time as a training set;
then inputting the extracted features into the trained possible pollution source
selection model, and outputting the result of whether the grid source is on a transport
path;
S6. repeating S3-S4 for the real-time spatio-temporal data to obtain features R2 ,
and 6, and standardizing the features to obtain preprocessed new data; and
S7. taking the preprocessed new data as an input into the trained possible
pollution source selection model, and outputting whether the grid source is on the
transport path; and adding the output result to the training set for optimization and
continual learning of the possible pollution source selection model.
2. The method for traceability of air pollutants based on coupled machine learning and correlation analysis according to claim 1, characterized in that both the real-time spatio-temporal data and the historical spatio-temporal data include geographic location information, pollutant concentration information, sampling time and meteorological information.
3. The method for traceability of air pollutants based on coupled machine
learning and correlation analysis according to claim 2, characterized in that the S3
specificallycomprises:
extracting pollutant concentration information from the historical
spatio-temporal data, then obtaining a valid time window i and a transport response
delay j according to set pollution events based on a hierarchical tree structure
constructed by a transport channel grid source k, and constructing a matrix to be
compared step by step and a correlation coefficient matrix in real time, i.e., a transient
model for pollutant concentration distribution.
4. The method for traceability of air pollutants based on coupled machine
learning and correlation analysis according to claim 3, characterized in that the S4
specifically comprises:
converting the correlation coefficient matrix into a correlation coefficient vector,
and extracting features from the correlation coefficient vector by Gaussian regression
to obtain Gaussian regression eigenvalues R2 k, Pk and 6k.
5. The method for traceability of air pollutants based on coupled machine
learning and correlation analysis according to claim 1, characterized in that the
possible pollution source selection model has an expression as follows: 2 k F( Y - fk(k , Pk, 5k) (10)
where ykE[0,1], 0means that the grid source k is not on the transport path and 1
means that the grid source k is on the transport path;fk represents the possible pollution source selection model, and each site is analyzed based onf to record the grid source k with yk=1.
6. The method for traceability of air pollutants based on coupled machine
learning and correlation analysis according to claim 5, characterized in that the
machine learning algorithms include random forest, decision tree, clustering,
Bayesian classification, support vector machine, EM and Adaboost.
7. The method for traceability of air pollutants based on coupled machine
learning and correlation analysis according to claim 5, characterized in that the S7
specifically comprises:
reassigning the pollution event time and pollution event concentration vector,
repeating S1-S5, then marking step by step until the correlation coefficient is the
lowest, and outputting the marking result, i.e., the transport channel and possible
pollution source regions, and ending iterative computation to realize continual
learning and optimization of the model.
8. The method for traceability of air pollutants based on coupled machine
learning and correlation analysis according to claim 3, characterized in that the
method for constructing the transient model for pollutant concentration distribution
specifically comprises:
Step 1. setting trigger conditions for a pollution event according to national
standards, and automatically marking the pollution event time t;
Step 2: constructing a pollution event concentration vector Xi and a vector Yij to
be compared step by step by the valid time window i and the transport response delay
j of the set pollution event based on Step 1: Step 3. constructing a matrix Zijk to be compared step by step according to the
vector Yijk to be compared step by step; and
Step 4. constructing a correlation coefficient matrix Rif for the grid source k
based on the matrix Zij to be compared step by step, i.e., a transient model for
pollutant concentration distribution.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
AU2021105563A AU2021105563A4 (en) | 2021-08-16 | 2021-08-16 | Method for Traceability of Air Pollutants Based on Coupled Machine Learning and Correlation Analysis |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
AU2021105563A AU2021105563A4 (en) | 2021-08-16 | 2021-08-16 | Method for Traceability of Air Pollutants Based on Coupled Machine Learning and Correlation Analysis |
Publications (1)
Publication Number | Publication Date |
---|---|
AU2021105563A4 true AU2021105563A4 (en) | 2021-10-14 |
Family
ID=78007596
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
AU2021105563A Active AU2021105563A4 (en) | 2021-08-16 | 2021-08-16 | Method for Traceability of Air Pollutants Based on Coupled Machine Learning and Correlation Analysis |
Country Status (1)
Country | Link |
---|---|
AU (1) | AU2021105563A4 (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116011317A (en) * | 2022-11-29 | 2023-04-25 | 北京工业大学 | Small-scale near-real-time atmospheric pollution tracing method based on multi-method fusion |
-
2021
- 2021-08-16 AU AU2021105563A patent/AU2021105563A4/en active Active
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116011317A (en) * | 2022-11-29 | 2023-04-25 | 北京工业大学 | Small-scale near-real-time atmospheric pollution tracing method based on multi-method fusion |
CN116011317B (en) * | 2022-11-29 | 2023-12-08 | 北京工业大学 | Small-scale near-real-time atmospheric pollution tracing method based on multi-method fusion |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110363347B (en) | Method for predicting air quality based on neural network of decision tree index | |
Chen et al. | Time series forecasting of temperatures using SARIMA: An example from Nanjing | |
CN113610243B (en) | Atmospheric pollutant tracing method based on coupled machine learning and correlation analysis | |
CN115358332A (en) | Atmospheric pollution tracing method for multi-source data | |
CN111814111A (en) | Industrial park atmospheric pollutant tracing method | |
CN115759488A (en) | Carbon emission monitoring and early warning analysis system and method based on edge calculation | |
CN104091216A (en) | Traffic information predication method based on fruit fly optimization least-squares support vector machine | |
Qin et al. | Prediction of air quality based on KNN-LSTM | |
AU2021105563A4 (en) | Method for Traceability of Air Pollutants Based on Coupled Machine Learning and Correlation Analysis | |
CN116957838A (en) | Crop growth environment monitoring method based on knowledge graph representation learning | |
Yamac | Reference evapotranspiration estimation with kNN and ANN models using different climate input combinations in the semi-arid environment | |
CN115578227A (en) | Method for determining atmospheric particulate pollution key area based on multi-source data | |
Zhang et al. | Spatiotemporal prediction of O3 concentration based on the KNN-Prophet-LSTM model | |
CN113935228A (en) | L-band rough sea surface radiation brightness and temperature simulation method based on machine learning | |
Bramantoro et al. | Precision agriculture through weather forecasting | |
Saha et al. | Recurrent neural network based prediction of Indian summer monsoon using global climatic predictors | |
CN107977727B (en) | Method for predicting blocking probability of optical cable network based on social development and climate factors | |
Ahmadi et al. | Identification of dominant sources of sea level pressure for precipitation forecasting over Wales | |
AlSkaif et al. | On the interdependence and importance of meteorological variables for photovoltaic output power estimation | |
CN111582567B (en) | Wind power probability prediction method based on hierarchical integration | |
CN110019167A (en) | Long-term new forms of energy resource data base construction method and system in one kind | |
CN111382147A (en) | Meteorological data missing interpolation method and system | |
Xiang et al. | GA-ARIMA model-based analysis of arrival time at bus stop | |
CN117932232B (en) | Wind speed prediction system based on state identification RIME-DELM multivariable time sequence prediction | |
Patino et al. | Assessment of Odour Dispersion Modelling Tools through the Comparison with Field Inspections and Complaints Records |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
FGI | Letters patent sealed or granted (innovation patent) |