CN116050270A - PM2.5 concentration prediction method and system - Google Patents
PM2.5 concentration prediction method and system Download PDFInfo
- Publication number
- CN116050270A CN116050270A CN202310071237.8A CN202310071237A CN116050270A CN 116050270 A CN116050270 A CN 116050270A CN 202310071237 A CN202310071237 A CN 202310071237A CN 116050270 A CN116050270 A CN 116050270A
- Authority
- CN
- China
- Prior art keywords
- prediction
- concentration
- decomposition
- maximum
- monitoring data
- 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
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 37
- 238000013528 artificial neural network Methods 0.000 claims abstract description 11
- 238000012544 monitoring process Methods 0.000 claims description 25
- 238000004364 calculation method Methods 0.000 claims description 10
- 125000004122 cyclic group Chemical group 0.000 claims description 7
- 238000010586 diagram Methods 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000012935 Averaging Methods 0.000 claims description 3
- 230000036760 body temperature Effects 0.000 claims description 3
- 230000000306 recurrent effect Effects 0.000 claims description 3
- 230000004913 activation Effects 0.000 claims description 2
- 239000000809 air pollutant Substances 0.000 abstract description 9
- 231100001243 air pollutant Toxicity 0.000 abstract description 9
- 239000003344 environmental pollutant Substances 0.000 abstract description 3
- 231100000719 pollutant Toxicity 0.000 abstract description 3
- 239000003570 air Substances 0.000 description 23
- 241000282414 Homo sapiens Species 0.000 description 7
- 238000003915 air pollution Methods 0.000 description 3
- 230000036541 health Effects 0.000 description 3
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 230000008520 organization Effects 0.000 description 2
- 239000013618 particulate matter Substances 0.000 description 2
- 239000012080 ambient air Substances 0.000 description 1
- 210000000748 cardiovascular system Anatomy 0.000 description 1
- 239000007789 gas Substances 0.000 description 1
- 229910001385 heavy metal Inorganic materials 0.000 description 1
- 210000004185 liver Anatomy 0.000 description 1
- 210000004072 lung Anatomy 0.000 description 1
- 244000005700 microbiome Species 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 210000002345 respiratory system Anatomy 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 239000000725 suspension Substances 0.000 description 1
- 231100000331 toxic Toxicity 0.000 description 1
- 230000002588 toxic effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/25—Design optimisation, verification or simulation using particle-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/20—Finite element generation, e.g. wire-frame surface description, tesselation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Computer Graphics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a PM2.5 concentration prediction method and a PM2.5 concentration prediction system, which mainly relate to the technical field of air pollutant concentration prediction. After the history data are processed, the history data are subjected to feature selection, and corresponding threshold values are set according to feature values displayed among the features to perform feature selection. And then performing fully-adaptive noise set empirical mode decomposition on the concentration characteristics of the PM2.5 pollutants which are predicted by more than one. And finally, respectively putting the decomposed eigenmode functions and the selected characteristics into a gating circulation unit neural network to train to obtain corresponding prediction results, wherein the prediction results are that the eigenmode functions and the residual prediction results are added. The PM2.5 concentration prediction method is accurate in prediction result and can also calculate the amount.
Description
Technical Field
The invention relates to the field of air pollutant prediction, in particular to a PM2.5 concentration prediction method and system.
Background
A report published by the world health organization and the United nations environmental organization states that: "air pollution" has become an unexplained reality in urban residents throughout the world. "when the harmful gases and pollutants in the atmosphere reach a certain concentration, huge disasters are brought to human beings and the environment. The harm to human health, namely, the human body is damaged by three ways, namely, the inhalation of polluted air, the contact of surface skin with polluted air and the inhalation of food containing atmospheric pollutants, can cause diseases of respiratory tract and lung, can also cause harm to cardiovascular system, liver and the like, and can seriously deprive human life.
One of the most important indicators of air pollution is the concentration of PM2.5. Also known as fines, PM2.5. Fine particulate matter refers to particulate matter having an aerodynamic equivalent diameter of 2.5 microns or less in ambient air. It can be suspended in air for a longer time, and the higher the content concentration of the suspension in the air is, the more serious the air pollution is. Although PM2.5 is only a component of the earth's atmosphere that is very small in content, it has an important influence on air quality, visibility, and the like. Compared with coarser atmospheric particulates, PM2.5 has small particle size, large area, strong activity, easy attachment of toxic and harmful substances (such as heavy metals, microorganisms and the like), long residence time in the atmosphere and long conveying distance, thus having larger influence on human health and atmospheric environmental quality. Accurate prediction of PM2.5 concentration can enable people to know air quality in time, so that corresponding measures can be timely made, and probability of occurrence of diseases is reduced. The features related to the PM2.5 association degree are selected to predict the features together, so that the prediction effect obtained by the method is much better than that obtained by single-target prediction.
Disclosure of Invention
In view of the above, the invention provides a PM2.5 concentration prediction method and a PM2.5 concentration prediction system based on MIC-CEEMDAN-GRU, which overcome the problems that air quality characteristics and meteorological characteristics related to PM2.5 cannot be effectively selected, and the related factors among nonlinearities and the accurate prediction cannot be implemented.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
a PM2.5 concentration prediction method comprises the following specific steps:
acquiring air quality monitoring data and weather monitoring data of a target place, and correspondingly preprocessing the air quality monitoring data and the weather monitoring data;
all the preprocessed data are put into a maximum information coefficient model for calculation, so that a maximum information coefficient is obtained;
putting the PM2.5 concentration time sequence into a fully self-adaptive noise set empirical mode decomposition model to decompose;
setting a corresponding threshold according to the maximum information coefficient to select corresponding features;
putting the decomposed eigenmode functions, residual errors and selected features into a gated cyclic neural network for training and prediction;
and adding the eigenmode functions and the residual errors of each prediction to obtain a final prediction result.
Optionally, in the above method for predicting PM2.5 concentration, the specific steps of data preprocessing include:
filling the acquired air quality monitoring data and missing values in the meteorological monitoring data;
calculating air temperature, body temperature, air pressure, relative humidity, rainfall, WSPD wind speed, AQI, PM10 and NO 2 、SO 2 、O 3 And the correlation between CO and PM2.5 concentration sequences, and selecting factors with large correlation with PM2.5 as the characteristics of the input maximum information coefficient model.
Optionally, in the above method for predicting PM2.5 concentration, the specific steps of performing the maximum information coefficient calculation include:
given two variables x and y, MI is defined as:
where p (x, y) is the joint probability between the variables x and y;
given i and j, gridding a scatter diagram formed by XY in i columns and j rows, and solving the maximum mutual information value; normalizing the maximum mutual information value;
the maximum value of mutual information under different scales is selected as the MIC value, and the MIC is calculated by the following formula:
wherein a and B are the number of lattices divided in the x and y directions respectively, the lattice distribution is basically grid distribution, B is a variable, and the size of B is 0.6 th power of the data quantity.
Optionally, in the above method for predicting the concentration of PM2.5, the specific step of performing time-series decomposition of the concentration of PM2.5 includes:
adding Gaussian white noise to the signal y (t) to be decomposed to obtain a new signal y (t) +(-1) q εv j (t) EMD-decomposing the new signal to obtain a first-order eigenmode component C 1 ：
The first eigenmode component of the CEEMDAN decomposition is obtained by ensemble averaging the N modal components generated:
calculating a residual error after removing the first modal component:
at r 1 Adding positive and negative Gaussian white noise pairs into (t) to obtain a new signal, and carrying out EMD (empirical mode decomposition) by taking the new signal as a carrier to obtain a first-order modal component D 1 Thereby yielding a second eigenmode component of the CEEMDAN decomposition:
calculating a residual error after removing the second modal component:
repeating the steps until the obtained residual signal is a monotonic function, the decomposition cannot be continued, the algorithm is ended, at the moment, the number of the obtained intrinsic mode components is K, and then the original signal y (t) is decomposed into:
wherein E is i (. Cndot.) is the ith eigen mode component obtained by EMD decomposition, and the ith eigen mode component obtained by CEEMDAN decomposition is
v j In order to satisfy the j-th gaussian white noise signal of the standard normal distribution, j=1, 2,3 … … N is the number of times white noise is added, epsilon is the gaussian noise weight coefficient, and y (t) is the signal to be decomposed.
Optionally, in the method for predicting PM2.5 concentration, the specific steps of predicting the gated recurrent neural network include:
the cell structure of the gated recurrent neural network comprises only two "gates": an update gate and a reset gate;
the gate control unit is connected with the sigmoid activation function, the gate control unit is used for discarding or adding information to the cell state, the sigmoid layer output unit is used for outputting data between 0 and 1, and describing each partial throughput and whether the partial throughput is passed or not, wherein 0 indicates that all the amounts are not passed, and 1 indicates that all the amounts are passed.
Optionally, in the above method for predicting PM2.5 concentration, the data prediction result is added to obtain a prediction result:
wherein, the liquid crystal display device comprises a liquid crystal display device,y is the final prediction result, n is the test set decomposition
Number of (I) and (II)>
And r k ' (t) is the predicted outcome of the test set.
A PM2.5 concentration prediction system, comprising:
the acquisition module is used for acquiring air quality monitoring data and weather monitoring data of a target place and carrying out corresponding pretreatment on the air quality monitoring data and the weather monitoring data;
the computing module is used for placing all the preprocessed data into the maximum information coefficient model for computing to obtain the maximum information coefficient;
the decomposition module is used for putting the PM2.5 concentration time sequence into a completely self-adaptive noise set empirical mode decomposition model to decompose;
the setting module is used for setting a corresponding threshold value according to the maximum information coefficient to select corresponding characteristics;
the prediction module is used for putting each decomposed eigenmode function, residual errors and selected features into a gated cyclic neural network for training and prediction;
and the output module is used for adding the eigenmode function and the residual error of each prediction to obtain a final prediction result.
Optionally, in the above PM2.5 concentration prediction system, the calculation module includes: given two variables x and y, MI is defined as:
where p (x, y) is the joint probability between the variables x and y;
given i and j, gridding a scatter diagram formed by XY in i columns and j rows, and solving the maximum mutual information value; normalizing the maximum mutual information value;
the maximum value of mutual information under different scales is selected as the MIC value, and the MIC is calculated by the following formula:
wherein a and B are the number of lattices divided in the x and y directions respectively, the lattice distribution is basically grid distribution, B is a variable, and the size of B is 0.6 th power of the data quantity.
Compared with the prior art, the PM2.5 concentration prediction method and system provided by the invention not only consider the air quality factor, but also consider the influence of meteorological factors and other air pollutants on the PM2.5 concentration, and can be applied to various air pollutant concentration prediction tasks in different areas, so that the harm of high-concentration air pollutants to human beings is effectively prevented.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a diagram of the decomposition into eigenmode functions and residuals according to the present invention;
FIG. 3 is a schematic diagram of the present invention compared with other methods.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The embodiment of the invention discloses a PM2.5 concentration prediction method, which is shown in fig. 1 and comprises the following specific steps:
acquiring air quality monitoring data and weather monitoring data of a target place, and correspondingly preprocessing the air quality monitoring data and the weather monitoring data;
all the preprocessed data are put into a maximum information coefficient model for calculation, so that a maximum information coefficient is obtained;
putting the PM2.5 concentration time sequence into a fully self-adaptive noise set empirical mode decomposition model to decompose;
setting a corresponding threshold according to the maximum information coefficient to select corresponding features;
putting the decomposed eigenmode functions, residual errors and selected features into a gated cyclic neural network for training and prediction;
and adding the eigenmode functions and the residual errors of each prediction to obtain a final prediction result.
Data preparation:
the air pollutant and climate factor hour level data of 2016 years are intercepted from the data monitored by the air monitoring station, and are respectively air temperature, temperature sensing degree, air pressure, relative humidity, rainfall, WSPD wind speed, AQI, PM10 and NO 2 、SO 2 、O 3 Thirteen features of CO and PM2.5.
Data preprocessing:
among the collected data, more or less missing data can appear, and the data is filled by adopting a linear interpolation method, and some error data are deleted first and then are filled by taking an average value of two days before and after the error data are deleted. The linear interpolation satisfies the following formula:
F (x) ＝ax+b；
feature selection:
calculating air temperature, body temperature, air pressure, relative humidity, rainfall, WSPD wind speed, AQI, PM10 and NO 2 、SO 2 、O 3 Correlation of CO with PM2.5 concentration sequence, and factor with large correlation with PM2.5 was selected as the characteristic of the input MIC-CEEMDAN-GRU model.
Given i and j, gridding a scatter diagram formed by XY in i columns and j rows, and solving the maximum mutual information value; the calculation of MIC exploits MI.
Given two variables x and y, MI is defined as:
where p (x, y) is the joint probability between the variables x and y.
The main principle of MIC is to discretize the relationship between two variables in two-dimensional space using a scatter plot. The current two-dimensional space is divided into a certain number of intervals in the X and Y directions, and then the current scattered points fall into each grid, so that the problem of joint probability in mutual information is solved.
Normalizing the maximum mutual information value;
selecting the maximum value of mutual information under different scales as an MIC value; the MIC calculation is given by:
wherein a and b are the number of grids divided in the x and y directions respectively, and are distributed in a grid manner. B is a variable, and the size of B is about 0.6 th power of the data amount.
MIC values for each feature versus PM2.5 are shown in the table below:
threshold selection:
based on the MIC value calculation, the threshold was set to 0.15, and the final selected features were PM2.5, AQI, CO, PM10, SO 2 Relative humidity, O 3 And NO 2 。
Data decomposition:
the PM2.5 concentration sequence is subjected to CEEMDAN decomposition, and the decomposition principle comprises the following specific steps:
adding Gaussian white noise to the signal y (t) to be decomposed to obtain a new signal y (t) +(-1) q εv j (t), wherein q=12. EMD decomposition is carried out on the new signal to obtain a first-order eigenmode component C 1 ：
The first eigenmode component of the CEEMDAN decomposition is obtained by ensemble averaging the N modal components generated:
calculating a residual error after removing the first modal component:
at r 1 Adding positive and negative Gaussian white noise pairs into (t) to obtain a new signal, and carrying out EMD (empirical mode decomposition) by taking the new signal as a carrier to obtain a first-order modal component D 1 From this, the second eigenmode component of the CEEMDAN decomposition can be obtained:
calculating a residual error after removing the second modal component:
repeating the steps until the obtained residual signal is a monotonic function, and the decomposition cannot be continued, and ending the algorithm.
When the number of the eigenvalue components obtained at this time is K, the original signal y (t) is decomposed into:
wherein E is i (. Cndot.) is the ith eigen mode component obtained by EMD decomposition, and the ith eigen mode component obtained by CEEMDAN decomposition is
v j In order to satisfy the j-th gaussian white noise signal of the standard normal distribution, j=1, 2,3 … … N is the number of times white noise is added, epsilon is the gaussian noise weight coefficient, and y (t) is the signal to be decomposed.
The final PM2.5 concentration time is decomposed into twelve eigenmode functions and one residual, as shown in particular in fig. 2;
training data:
and respectively combining the selected data features with each eigenmode function and residual error decomposed by PM2.5, and putting the combined data features and the residual error into a gated cyclic neural network to train and predict.
Prediction result:
and adding the predicted results of the components, wherein the calculated final result is the predicted result, and the predicted result meets the following formula:
wherein y is the final prediction result, n is the decomposition of the test set
Number of (I) and (II)>
And r k ' (t) is the predicted outcome of the test set.
Prediction contrast:
as can be seen from FIG. 3, compared with the prior single model, the mixed model provided by the invention has higher precision than that of the single model, whether the mixed model is multivariable or univariable prediction, and the feasibility and the accuracy of the mixed model are verified.
Compared with the prior art, the invention has the advantages that: the method not only considers the air quality factor, but also considers the influence of meteorological factors and other air pollutants on the PM2.5 concentration, and can be applied to various air pollutant concentration prediction tasks in different areas, thereby effectively preventing the harm of high-concentration air pollutants to human beings.
In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (8)
1. The PM2.5 concentration prediction method is characterized by comprising the following specific steps of:
acquiring air quality monitoring data and weather monitoring data of a target place, and correspondingly preprocessing the air quality monitoring data and the weather monitoring data;
all the preprocessed data are put into a maximum information coefficient model for calculation, so that a maximum information coefficient is obtained;
putting the PM2.5 concentration time sequence into a fully self-adaptive noise set empirical mode decomposition model to decompose;
setting a corresponding threshold according to the maximum information coefficient to select corresponding features;
putting the decomposed eigenmode functions, residual errors and selected features into a gated cyclic neural network for training and prediction;
and adding the eigenmode functions and the residual errors of each prediction to obtain a final prediction result.
2. The PM2.5 concentration prediction method according to claim 1, characterized in that the data preprocessing specific step comprises:
filling the acquired air quality monitoring data and missing values in the meteorological monitoring data;
calculating air temperature, body temperature, air pressure, relative humidity, rainfall, WSPD wind speed, AQI, PM10 and NO 2 、SO 2 、O 3 And the correlation between CO and PM2.5 concentration sequences, and selecting factors with large correlation with PM2.5 as the characteristics of the input maximum information coefficient model.
3. The PM2.5 concentration prediction method according to claim 1, wherein the specific step of performing maximum information coefficient calculation comprises:
given two variables x and y, MI is defined as:
where p (x, y) is the joint probability between the variables x and y;
given i and j, gridding a scatter diagram formed by XY in i columns and j rows, and solving the maximum mutual information value; normalizing the maximum mutual information value;
the maximum value of mutual information under different scales is selected as the MIC value, and the MIC is calculated by the following formula:
wherein a and B are the number of lattices divided in the x and y directions respectively, the lattice distribution is basically grid distribution, B is a variable, and the size of B is 0.6 th power of the data quantity.
4. The PM2.5 concentration prediction method according to claim 1, wherein the specific step of performing time-series decomposition of the PM2.5 concentration comprises:
adding Gaussian white noise to the signal y (t) to be decomposed to obtain a new signal y (t) +(-1) q εv j (t) EMD-decomposing the new signal to obtain a first-order eigenmode component C 1 ：
E(y(t)+(-1) q εv j (t))＝C 1 j (t)+r j ；
The first eigenmode component of the CEEMDAN decomposition is obtained by ensemble averaging the N modal components generated:
calculating a residual error after removing the first modal component:
at r 1 Adding positive and negative Gaussian white noise pairs into (t) to obtain a new signal, and carrying out EMD (empirical mode decomposition) by taking the new signal as a carrier to obtain a first-order modal component D 1 Thereby yielding a second eigenmode component of the CEEMDAN decomposition:
calculating a residual error after removing the second modal component:
repeating the steps until the obtained residual signal is a monotonic function, the decomposition cannot be continued, the algorithm is ended, at the moment, the number of the obtained intrinsic mode components is K, and then the original signal y (t) is decomposed into:
wherein E is i (. Cndot.) is the ith eigen mode component obtained by EMD decomposition, and the ith eigen mode component obtained by CEEMDAN decomposition is
v j In order to satisfy the j-th gaussian white noise signal of the standard normal distribution, j=1, 2,3 … … N is the number of times white noise is added, epsilon is the gaussian noise weight coefficient, and y (t) is the signal to be decomposed.
5. The PM2.5 concentration prediction method according to claim 1, wherein the gating cyclic neural network performs the specific steps of:
the cell structure of the gated recurrent neural network comprises only two "gates": an update gate and a reset gate;
the gate control unit is connected with the sigmoid activation function, the gate control unit is used for discarding or adding information to the cell state, the sigmoid layer output unit is used for outputting data between 0 and 1, and describing each partial throughput and whether the partial throughput is passed or not, wherein 0 indicates that all the amounts are not passed, and 1 indicates that all the amounts are passed.
6. The PM2.5 concentration prediction method according to claim 1, wherein the data prediction results are added to obtain the prediction results:
wherein y is the final prediction result, n is the decomposition of the test set
Number of (I) and (II)>
And r k ' (t) is the predicted outcome of the test set.
7. A PM2.5 concentration prediction system, comprising:
the acquisition module is used for acquiring air quality monitoring data and weather monitoring data of a target place and carrying out corresponding pretreatment on the air quality monitoring data and the weather monitoring data;
the computing module is used for placing all the preprocessed data into the maximum information coefficient model for computing to obtain the maximum information coefficient;
the decomposition module is used for putting the PM2.5 concentration time sequence into the empirical mode of the complete self-adaptive noise set
Feeding into a decomposition model
Performing row decomposition;
the setting module is used for setting a corresponding threshold value according to the maximum information coefficient to select corresponding characteristics;
the prediction module is used for putting each decomposed eigenmode function, residual errors and selected features into a gated cyclic neural network for training and prediction;
and the output module is used for adding the eigenmode function and the residual error of each prediction to obtain a final prediction result.
8. The PM2.5 concentration prediction system according to claim 7, wherein the calculation module comprises: given two variables x and y, MI is defined as:
where p (x, y) is the joint probability between the variables x and y;
given i and j, gridding a scatter diagram formed by XY in i columns and j rows, and solving the maximum mutual information value; normalizing the maximum mutual information value;
the maximum value of mutual information under different scales is selected as the MIC value, and the MIC is calculated by the following formula:
wherein a and B are the number of lattices divided in the x and y directions respectively, the lattice distribution is basically grid distribution, B is a variable, and the size of B is 0.6 th power of the data quantity.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310071237.8A CN116050270A (en) | 2023-02-07 | 2023-02-07 | PM2.5 concentration prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310071237.8A CN116050270A (en) | 2023-02-07 | 2023-02-07 | PM2.5 concentration prediction method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116050270A true CN116050270A (en) | 2023-05-02 |
Family
ID=86114526
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310071237.8A Pending CN116050270A (en) | 2023-02-07 | 2023-02-07 | PM2.5 concentration prediction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116050270A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113409072A (en) * | 2021-05-31 | 2021-09-17 | 河北科技师范学院 | Empirical mode decomposition and distributed GRU neural network and price prediction method |
| CN114611808A (en) * | 2022-03-18 | 2022-06-10 | 河海大学 | Short-term offshore wind power prediction method based on CEEMDAN-SSA-BilSTM |
| CN114692998A (en) * | 2022-04-25 | 2022-07-01 | 陕西理工大学 | Comprehensive energy source stealing and leakage detection method based on probability density regression prediction |
| CN114897257A (en) * | 2022-05-31 | 2022-08-12 | 三峡大学 | Displacement step type reservoir landslide accumulated displacement prediction method based on EEMD-GRU model optimal time scale |
| CN114997510A (en) * | 2022-06-17 | 2022-09-02 | 合肥学院 | PM2.5 prediction method and system based on Copula entropy correlation analysis |
| CN115034473A (en) * | 2022-06-10 | 2022-09-09 | 国网山东省电力公司临沂供电公司 | Electricity price prediction method, system and device |
| CN115099499A (en) * | 2022-06-27 | 2022-09-23 | 合肥学院 | Method for predicting PM2.5 concentration based on EMD-LSTM of random forest |
-
2023
- 2023-02-07 CN CN202310071237.8A patent/CN116050270A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113409072A (en) * | 2021-05-31 | 2021-09-17 | 河北科技师范学院 | Empirical mode decomposition and distributed GRU neural network and price prediction method |
| CN114611808A (en) * | 2022-03-18 | 2022-06-10 | 河海大学 | Short-term offshore wind power prediction method based on CEEMDAN-SSA-BilSTM |
| CN114692998A (en) * | 2022-04-25 | 2022-07-01 | 陕西理工大学 | Comprehensive energy source stealing and leakage detection method based on probability density regression prediction |
| CN114897257A (en) * | 2022-05-31 | 2022-08-12 | 三峡大学 | Displacement step type reservoir landslide accumulated displacement prediction method based on EEMD-GRU model optimal time scale |
| CN115034473A (en) * | 2022-06-10 | 2022-09-09 | 国网山东省电力公司临沂供电公司 | Electricity price prediction method, system and device |
| CN114997510A (en) * | 2022-06-17 | 2022-09-02 | 合肥学院 | PM2.5 prediction method and system based on Copula entropy correlation analysis |
| CN115099499A (en) * | 2022-06-27 | 2022-09-23 | 合肥学院 | Method for predicting PM2.5 concentration based on EMD-LSTM of random forest |
Non-Patent Citations (1)
| Title |
|---|
| 丁子昂;乐曹伟;吴玲玲;付明磊;: "基于CEEMD-Pearson和深度LSTM混合模型的PM_(2.5)浓度预测方法", 计算机科学, no. 1, 15 June 2020 (2020-06-15), pages 454 - 459 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108426812B (en) | PM2.5 concentration value prediction method based on memory neural network | |
| CN108268935B (en) | PM2.5 concentration value prediction method and system based on time sequence recurrent neural network | |
| CN107194139B (en) | Atmospheric pollution source grading method and computing equipment | |
| CN110727717B (en) | Monitoring method, device, equipment and storage medium for gridding atmospheric pollution intensity | |
| CN108426815B (en) | Method for estimating concentration of components of near-surface fine particulate matters | |
| Zhou et al. | Effects of potential recirculation on air quality in coastal cities in the Yangtze River Delta | |
| CN110989041B (en) | Method and system for forecasting dust-haze and sand-dust weather | |
| CN112180472A (en) | Atmospheric visibility integrated forecasting method based on deep learning | |
| CN114662344B (en) | Atmospheric pollution source tracing prediction method and system based on continuous online observation data | |
| CN113011455B (en) | Air quality prediction SVM model construction method | |
| CN112230266B (en) | Radiation monitoring system | |
| CN112132336B (en) | Quarterly prediction method for PM2.5 concentration | |
| CN115438848A (en) | PM based on deep mixed graph neural network 2.5 Long-term concentration prediction method | |
| Schlünzen | On the validation of high-resolution atmospheric mesoscale models | |
| Garner et al. | Ensemble statistical post-processing of the national air quality forecast capability: Enhancing ozone forecasts in Baltimore, Maryland | |
| CN116050270A (en) | PM2.5 concentration prediction method and system | |
| CN113435068A (en) | Radionuclide assimilation prediction method based on logarithmic variational assimilation | |
| Lee et al. | Development of a deep neural network for predicting 6 h average PM 2.5 concentrations up to 2 subsequent days using various training data | |
| Wang et al. | Real-Time data driven simulation of air contaminant dispersion using particle filter and UAV sensory system | |
| Yu et al. | Performance and diagnostic evaluation of ozone predictions by the eta-community multiscale air quality forecast system during the 2002 New England Air Quality Study | |
| Miller | A comparison in the accuracy of mapping nuclear fallout patterns using HPAC, HYSPLIT, DELFIC FPT and an AFIT FORTRAN95 fallout deposition code | |
| Represa et al. | Daily concentrations of PM2. 5 in the Valencian community using random forest for the period 2008–2018 | |
| Liu et al. | Sensitivity of air quality simulation to smoke plume rise | |
| CN110751345A (en) | Method for predicting PM2.5 value of target location based on distance features | |
| Park | Development of a deep neural network for predicting 6-hour average PM2. 5 concentrations up to two subsequent days using various training data |
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 | ||
| CB02 | Change of applicant information |
Country or region after: China Address after: No.99 Jinxiu Avenue, Hefei Economic Development Zone, Anhui Province 230601 Applicant after: Hefei University Address before: No.99 Jinxiu Avenue, Hefei Economic Development Zone, Anhui Province 230601 Applicant before: HEFEI University Country or region before: China |
|
| CB02 | Change of applicant information |