CN106056210A - PM 2.5 concentration value prediction method based on hybrid neural network - Google Patents
PM 2.5 concentration value prediction method based on hybrid neural network Download PDFInfo
- Publication number
- CN106056210A CN106056210A CN201610401881.7A CN201610401881A CN106056210A CN 106056210 A CN106056210 A CN 106056210A CN 201610401881 A CN201610401881 A CN 201610401881A CN 106056210 A CN106056210 A CN 106056210A
- Authority
- CN
- China
- Prior art keywords
- network
- data
- concentration value
- training
- layer
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a PM 2.5 concentration value prediction method based on a hybrid neural network. The method comprises the following steps: 1) acquisition of four types of sample data, the four types of sample data comprising PM 2.5 concentration historical data, PM 2.5 concentration value related index historical data, weather historical data and PM 2.5 composition analysis data; 2) collecting an initially-forecasted PM 2.5 concentration value of a first neural network; 3) collecting a secondary-prediction PM 2.5 concentration value of a second neural network; and 4) collecting a final-prediction PM 2.5 concentration value of a third neural network, and outputting the final PM 2.5 concentration predication value. Besides the three kinds of data of PM 2.5 concentration value historical data, the PM 2.5 concentration value related index historical data and the weather historical data, the PM 2.5 composition analysis data is also introduced, so that PM 2.5 concentration value change and development rules can be accurately described and prediction accuracy is improved.
Description
Technical field
The present invention relates to the electric powder prediction of air particle PM2.5 concentration value, particularly relate to a kind of based on mixing god
Forecasting Methodology through the PM2.5 concentration value of network.
Background technology
Nowadays air pollution has become as focus of concern.In densely populated city, air pollution is the tightest
Heavily have influence on the Health and Living of people.In air pollution regulations, the PM2.5 (diameter granule less than or equal to 2.5 microns
Thing) concentration value have become as weigh air quality significant Testing index.Dense to future time section PM2.5 according to historical data
The prediction of angle value has become as has studying a question of stronger academic significance and using value.
In order to solve the problems referred to above, Shi Xuhua et al. is in patent " a kind of regional air PM2.5 concentration prediction method ", logical
Cross set up Support vector regression model carry out PM2.5 concentration value prediction.Wang Shuqiang et al. is in patent " a kind of air quality
PM2.5 Forecasting Methodology and system " in, support that the concentration value that tensor regression model carries out PM2.5 is predicted by setting up.Ma Tiancheng etc.
People, in paper " fuzzy neural network PM2.5 concentration prediction based on modified model PSO ", uses a kind of modified model PSO to optimize
The concentration value of fuzzy neural network prediction PM2.5.Zhang Yiwen et al. is at paper " PM2.5 forecast model based on neutral net "
In, by setting up the prediction of neural fusion PM2.5 concentration value." the controlled genetic algorithm of t-distribution is excellent at paper for Jing Tao et al.
Change the PM2.5 mass concentration prediction of BP neutral net " in, by setting up based on t-distribution controlled genetic algorithm optimization BP nerve net
Network model, is predicted PM2.5 mass concentration.Sun Rongji et al. is " a kind of based on BP improved neural network algorithm at paper
PM2.5 Forecasting Methodology " in, by combining principal component analysis, in advance termination coaching method and the BP neutral net mould concentration to PM2.5
Value is predicted.Chen Qiang et al. is in paper " Zhengzhou City PM2.5 concentration spatial-temporal distribution characteristic and forecast model research ", by building
The concentration value of locality, Zhengzhou City PM2.5 is predicted by vertical BP neutral net.
Through document investigation and analysis, the PM2.5 concentration value Forecasting Methodology having pointed out at present is required to substantial amounts of historical data and makees
Sample is inputted for system.These historical datas can be divided into three classifications.Primary sources are PM2.5 concentration value historical datas.
Secondary sources are PM2.5 concentration value index of correlation historical data (such as AQI, PM10, NO2、CO、SO2、O3).3rd class data
It it is meteorological historical data (such as temperature, relative humidity, air pressure, wind speed, precipitation etc.).But, through literature survey, existing
PM2.5 concentration value Forecasting Methodology does not accounts for using the 4th class data, i.e. PM2.5 component resolving data (possess by such as motor vehicles
Amount, industrial gas emission amount, power consumption).For different cities or area, PM2.5 Crack cause has bigger difference.As
Fruit only considers first three class data, then designed PM2.5 concentration value prognoses system is difficult to accurate simulation locality PM2.5 concentration value
Change and Development rule.
Summary of the invention
In order to overcome existing PM2.5 concentration value prediction mode cannot describe the Change and Development rule of PM2.5 concentration value, prediction
The deficiency that precision is relatively low, the present invention use PM2.5 concentration value historical data, PM2.5 concentration value index of correlation historical data and
Outside the data of meteorological historical data these three classification, also introduce PM2.5 component resolving data, it is provided that a kind of accurate description
The Change and Development rule of PM2.5 concentration value, the PM2.5 concentration value prediction side based on hybrid neural networks of raising precision of prediction
Method.
The technical solution adopted for the present invention to solve the technical problems is:
A kind of PM2.5 concentration value Forecasting Methodology based on hybrid neural networks, described method comprises the steps:
Step 1, four class sample datas gather.Described four class sample datas include PM2.5 concentration value historical data, PM2.5
Concentration value index of correlation historical data, meteorological historical data and PM2.5 component resolving data.
Step 2, employing first nerves network initial forecast PM2.5 concentration value, process is as follows:
Step 2.1, one four layers of neutral net comprising input layer, hidden layer, articulamentum and output layer of establishment, set hidden
Containing layer and the nodes of output layer.The node number of described hidden layer uses empirical equation to provide estimated value, described empirical equation
As follows:
In above formula, m and n is respectively the neuron number of input layer and output layer, and a is the constant between [0,10].
Step 2.2, set training function, connectivity function and the output function of hidden layer, articulamentum and output layer respectively, if
Determine anticipation error minima, maximum iteration time and the learning rate of network.
Step 2.3, using described PM2.5 concentration value historical data as first nerves network input data, and by first god
It is divided into first nerves network training data and first nerves network testing data, by described first nerves net through network input data
Network training data normalized, normalization formula is as follows:
In above formula, x be described in treat normalized data, xminAnd xmaxIt is respectively the minima in described new sample data
And maximum, y is the data after normalization, is distributed in [0.1,0.9] interval.
Step 2.4, described first nerves network training data are input in the first nerves network created training the
One neutral net, calculates error, and according to error transfer factor first nerves network weight.
Step 2.5, judge whether described first nerves network restrains, when error is less than anticipation error minima, algorithm
Convergence.Terminate algorithm when reaching maximum iteration time, described first nerves network training completes.
Step 2.6, the first nerves network that will train described in the input of described first nerves network testing data, obtain
The initial predicted value of PM2.5 concentration value.
Step 2.7, by the initial predicted value of described PM2.5 concentration value and described PM2.5 concentration value index of correlation history
Data and meteorological historical data are incorporated as nervus opticus network input data.
Step 3, employing nervus opticus network re prediction PM2.5 concentration value, process is as follows:
Step 3.1, create a three-layer neural network comprising input layer, hidden layer and output layer, set hidden layer and
The node number of output layer.
Step 3.2, set the training function of described nervus opticus network, the anticipation error minima of setting network, maximum
Iterations and learning rate.
Step 3.3, by described nervus opticus network input data be divided into nervus opticus network training data and nervus opticus
Described nervus opticus network training data are normalized by network testing data.
Step 3.4, the nervus opticus network training data after described normalization are input to the nervus opticus net that created
Network, and train nervus opticus network, calculate error, and according to error transfer factor nervus opticus network weight.
Step 3.5, judge whether described nervus opticus network restrains, when error is less than anticipation error minima, algorithm
Convergence.Terminate algorithm when reaching maximum iteration time, described nervus opticus network training completes.
Step 3.6, described nervus opticus network testing data is input in the nervus opticus network that described training completes,
The re prediction value of output PM2.5 concentration value, by re prediction value and the PM2.5 component resolving data of described PM2.5 concentration value
It is incorporated as third nerve network input data.
Step 4, employing third nerve network finally predict PM2.5 concentration value, and process is as follows:
Step 4.1, create a three-layer neural network comprising input layer, hidden layer and output layer, set hidden layer and
The node number of output layer.
Step 4.2, set the training function of described third nerve network, the anticipation error minima of setting network, maximum
Iterations and learning rate.
Step 4.3, described third nerve network data is divided into third nerve network training data and third nerve network
Described third nerve network training data are normalized by test data.
Step 4.4, the third nerve network training data after described normalization are input to the third nerve net that created
Network, and train third nerve network, calculate error, and according to error transfer factor third nerve network weight.
Step 4.5, judge whether described third nerve network restrains, when error is less than anticipation error minima, algorithm
Convergence.Terminate algorithm when reaching maximum iteration time, described third nerve network training completes.
Step 4.6, described third nerve network testing data is input in the third nerve network that described training completes,
The final predictive value of output PM2.5 concentration value.
Further, in described step 1, described PM2.5 concentration value index of correlation historical data includes that (air quality refers to AQI
Number), PM10 (Particulate Matter 10), SO2(sulfur dioxide), CO (carbon monoxide), CO2(carbon dioxide), O3(smelly
Oxygen), described meteorological historical data includes temperature on average, dew point, relative humidity, pressure, wind speed, precipitation, described PM2.5 composition
Resolve data and include vehicle guaranteeding organic quantity, industrial gas emission amount and power consumption etc..
The technology of the present invention is contemplated that: using PM2.5 concentration value historical data, going through of PM2.5 concentration value index of correlation
History data (AQI, PM10, NO2、CO、SO2、O3) and meteorological historical data (temperature, relative humidity, air pressure, wind speed, precipitation etc.)
Outside these three classes data, emphasis introduces PM2.5 component resolving data (vehicle guaranteeding organic quantity, industrial gas emission amount, electricity consumption
Amount) these the 4th class data carry out the prediction of PM2.5 concentration value.Further, these four classes data are divided by three neutral nets
Stage forecast PM2.5 concentration value.
Beneficial effects of the present invention is mainly manifested in: it is dense that technical scheme can simulate local PM2.5 exactly
The Changing Pattern of angle value, effective precision of prediction that must improve current PM2.5 concentration value, it is achieved " localization " and " compartmentalization " is accurate
Prediction.
Accompanying drawing explanation
Fig. 1 is a kind of PM2.5 concentration value Forecasting Methodology schematic diagram based on hybrid neural networks.
Fig. 2 is the training flow chart of ELman neutral net.
Fig. 3 is the training flow chart of BP neutral net.
Detailed description of the invention
The invention will be further described below in conjunction with the accompanying drawings.
With reference to Fig. 1~Fig. 3, a kind of PM2.5 concentration value Forecasting Methodology based on hybrid neural networks, described method include as
Lower step:
Step 1, four class sample datas gather.Described four class sample datas include PM2.5 (Particulate Matter
2.5) concentration value historical data, PM2.5 concentration value index of correlation historical data, meteorological historical data and PM2.5 component resolving number
According to.Further, described PM2.5 concentration value index of correlation historical data includes AQI (air quality index), PM10
(Particulate Matter 10)、SO2(sulfur dioxide), CO (carbon monoxide), CO2(carbon dioxide), O3(ozone), described
Meteorological historical data includes temperature on average, dew point, relative humidity, pressure, wind speed, precipitation, described PM2.5 component resolving data
Including vehicle guaranteeding organic quantity, industrial gas emission amount and power consumption.
The present invention gathers four class data of Hangzhou.The Hangzhou AQI of 2015 (air quality index), PM2.5
(Particulate Matter 2.5)、PM10(Particulate Matter 10)、SO2(sulfur dioxide), a CO (oxidation
Carbon), CO2(carbon dioxide), O3(ozone) is collected in China air quality on-line monitoring analysis platform website, 2015, Hangzhou
Temperature on average, dew point, relative humidity, pressure, wind speed, precipitation WEATHER UNDERGROUND website collect, Hangzhou
Vehicle guaranteeding organic quantity, industrial gas emission amount and the power consumption of 2014 is collected in Information Statistics net website, Hangzhou.
In the present invention, the four class sample datas such as table 1 collected:
Table 1
Step 2, an Elman neutral net initial forecast PM2.5 concentration value, process is as follows:
Step 2.1, one four layers of neutral net comprising input layer, hidden layer, articulamentum and output layer of establishment, set hidden
Nodes containing layer and output layer is respectively 6 and 1.The node number of described hidden layer uses empirical equation to provide estimated value, institute
State empirical equation as follows:
In formula: m and n is respectively the neuron number of input layer and output layer, a is the constant between [0,10];
Step 2.2, set training function, connectivity function and the output function of hidden layer, articulamentum and output layer respectively, institute
Stating training function is Sigmoid function, and articulamentum and output layer function are purelin function.The anticipation error of setting network is
Little value is 10-5, maximum iteration time is 104It is 0.1 with learning rate;
Step 2.3, described PM2.5 concentration value historical data is divided into first as an Elman Neural Network Data
Elman neural metwork training data and Elman neutral net test data two parts, by a described Elman nerve net
Network training data normalized, normalization formula is as follows:
In formula: x be described in treat normalized data, xminAnd xmaxBe respectively the minima in described new sample data and
Maximum, y is the data after normalization, is distributed in [0.1,0.9] interval;
Step 2.4, described Elman neural metwork training data are input to the Elman nerve net that creates
Network is trained an Elman neutral net, calculates error, according to error transfer factor weights;
Step 2.5, judge whether a described Elman neutral net restrains, when error is less than anticipation error minima,
Algorithmic statement;Or reaching to terminate during maximum iteration time algorithm, a described Elman neural metwork training completes;
Step 2.6, the Elman nerve that will train described in a described Elman neutral net test data input
Network, obtains the predictive value of preliminary PM2.5 concentration value;
Step 2.7, by the predictive value of described preliminary PM2.5 concentration value and described PM2.5 concentration value index of correlation history
Data and meteorological historical data are incorporated as the 2nd BP Neural Network Data;
Step 3, the 2nd BP neutral net re prediction PM2.5 concentration value, process is as follows:
Step 3.1, one three-layer neural network comprising input layer, hidden layer and output layer of establishment, set implicit respectively
The node number of layer and output layer is 12 and 1.
Step 3.2, set the training function of described 2nd BP neutral net as Sigmoid function, the expectation of setting network
Error minima is 10-5, maximum iteration time is 104It is 0.1 with learning rate;
Step 3.3, described 2nd BP Neural Network Data is divided into the 2nd BP neural metwork training data and the 2nd BP god
Through network testing data, described 2nd BP neural metwork training data are normalized;
Step 3.4, the 2nd BP being input to the 2nd BP neural metwork training data after told normalization create are refreshing
In network, train the 2nd BP neutral net, calculate error, according to error transfer factor weights;
Step 3.5, judge whether described 2nd BP neutral net restrains, when error is less than anticipation error minima, calculate
Method restrains;Or reaching to terminate during maximum iteration time algorithm, described 2nd BP neural metwork training completes;
Step 3.6, by described 2nd BP neutral net test data be input to the 2nd BP nerve net that described training completes
In network, the re prediction value of output PM2.5 concentration value, by re prediction value and the PM2.5 component resolving of described PM2.5 concentration value
Data are incorporated as the 3rd BP Neural Network Data;
Step 4, the 3rd BP neutral net finally predict PM2.5 concentration value, and process is as follows:
Step 4.1, one three-layer neural network comprising input layer, hidden layer and output layer of establishment, set implicit respectively
The node number of layer and output layer is 4 and 1.
Step 4.2, set the training function of described 3rd BP neutral net as Sigmoid function, the expectation of setting network
Error minima is 10-5, maximum iteration time is 104It is 0.1 with learning rate;
Step 4.3, described 3rd BP Neural Network Data is divided into the 3rd BP neural metwork training data and the 3rd BP god
Through network testing data, described 3rd BP neural metwork training data are normalized;
Step 4.4, the 3rd BP being input to the 3rd BP neural metwork training data after told normalization create are refreshing
In network, train the 3rd BP neutral net, calculate error, according to error transfer factor weights;
Step 4.5, judge whether described 3rd BP neutral net restrains, when error is less than anticipation error minima, calculate
Method restrains;Or reaching to terminate during maximum iteration time algorithm, described 3rd BP neural metwork training completes;
Step 4.6, described third nerve network testing data is input to the 3rd BP neutral net that described training completes
In, the final predictive value of output PM2.5 concentration value.
The predictive value of PM2.5 concentration value is obtained, such as table 2 according to above operation:
Table 2.
Claims (2)
1. a PM2.5 concentration value Forecasting Methodology based on hybrid neural networks, it is characterised in that: described method includes walking as follows
Rapid:
Step 1, four class sample datas gather, and described four class sample datas include PM2.5 concentration value historical data, PM2.5 concentration
Value index of correlation historical data, meteorological historical data and PM2.5 component resolving data;
Step 2, employing first nerves network initial forecast PM2.5 concentration value.Process is as follows:
Step 2.1, one four layers of neutral net comprising input layer, hidden layer, articulamentum and output layer of establishment, set hidden layer
With the nodes of output layer, the node number of described hidden layer uses empirical equation to provide estimated value, and described empirical equation is as follows:
In above formula, m and n is respectively the neuron number of input layer and output layer, and a is the constant between [0,10];
Step 2.2, set training function, connectivity function and the output function of hidden layer, articulamentum and output layer respectively, set net
Anticipation error minima, maximum iteration time and the learning rate of network;
Step 2.3, described PM2.5 concentration value historical data is inputted data as first nerves network, and by first nerves net
Network input data are divided into first nerves network training data and first nerves network testing data, are instructed by described first nerves network
Practicing data normalization to process, normalization formula is as follows:
In above formula, x be described in treat normalized data, xminAnd xmaxIt is respectively the minima in described new sample data and
Big value, y is the data after normalization, is distributed in [0.1,0.9] interval;
Step 2.4, described first nerves network training data are input in the first nerves network created training first god
Through network, calculate error, and according to error transfer factor first nerves network weight;
Step 2.5, judge whether described first nerves network restrains, when error is less than anticipation error minima, algorithmic statement.
Terminate algorithm when reaching maximum iteration time, described first nerves network training completes;
Step 2.6, the first nerves network that will train described in the input of described first nerves network testing data, obtain PM2.5
The initial predicted value of concentration value;
Step 2.7, by the initial predicted value of described PM2.5 concentration value and described PM2.5 concentration value index of correlation historical data
It is incorporated as nervus opticus network input data with meteorological historical data;
Step 3, employing nervus opticus network re prediction PM2.5 concentration value, process is as follows:
Step 3.1, one three-layer neural network comprising input layer, hidden layer and output layer of establishment, set hidden layer and output
The node number of layer;
Step 3.2, set the training function of described nervus opticus network, the anticipation error minima of setting network, greatest iteration
Number of times and learning rate;
Step 3.3, by described nervus opticus network input data be divided into nervus opticus network training data and nervus opticus network
Described nervus opticus network training data are normalized by test data;
Step 3.4, the nervus opticus network training data after described normalization are input to the nervus opticus network that created, and
And training nervus opticus network, calculate error, and according to error transfer factor nervus opticus network weight;
Step 3.5, judge whether described nervus opticus network restrains, when error is less than anticipation error minima, algorithmic statement.
Terminate algorithm when reaching maximum iteration time, described nervus opticus network training completes;
Step 3.6, described nervus opticus network testing data is input in the nervus opticus network that described training completes, output
Re prediction value and the PM2.5 component resolving data of described PM2.5 concentration value are merged by the re prediction value of PM2.5 concentration value
Data are inputted as third nerve network;
Step 4, employing third nerve network finally predict PM2.5 concentration value, and process is as follows:
Step 4.1, one three-layer neural network comprising input layer, hidden layer and output layer of establishment, set hidden layer and output
The node number of layer;
Step 4.2, set the training function of described third nerve network, the anticipation error minima of setting network, greatest iteration
Number of times and learning rate;
Step 4.3, described third nerve network data is divided into third nerve network training data and third nerve network test
Described third nerve network training data are normalized by data;
Step 4.4, the third nerve network training data after described normalization are input to the third nerve network that created, and
Training third nerve network, calculates error, and according to error transfer factor third nerve network weight;
Step 4.5, judge whether described third nerve network restrains, when error is less than anticipation error minima, algorithmic statement;
Terminate algorithm when reaching maximum iteration time, described third nerve network training completes;
Step 4.6, described third nerve network testing data is input in the third nerve network that described training completes, output
The final predictive value of PM2.5 concentration value.
A kind of PM2.5 concentration value Forecasting Methodology based on hybrid neural networks, it is characterised in that:
In described step 1, described PM2.5 concentration value index of correlation historical data includes air quality index AQI, PM10, SO2、CO、CO2
And O3, described meteorological historical data includes temperature on average, dew point, relative humidity, pressure, wind speed and precipitation, and described PM2.5 becomes
Decompose analysis data and include vehicle guaranteeding organic quantity, industrial gas emission amount and power consumption.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610401881.7A CN106056210B (en) | 2016-06-07 | 2016-06-07 | A kind of PM2.5 concentration value Forecasting Methodologies based on hybrid neural networks |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610401881.7A CN106056210B (en) | 2016-06-07 | 2016-06-07 | A kind of PM2.5 concentration value Forecasting Methodologies based on hybrid neural networks |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106056210A true CN106056210A (en) | 2016-10-26 |
CN106056210B CN106056210B (en) | 2018-06-01 |
Family
ID=57171107
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610401881.7A Active CN106056210B (en) | 2016-06-07 | 2016-06-07 | A kind of PM2.5 concentration value Forecasting Methodologies based on hybrid neural networks |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106056210B (en) |
Cited By (25)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106529746A (en) * | 2016-12-29 | 2017-03-22 | 南京恩瑞特实业有限公司 | Method for dynamically fusing, counting and forecasting air quality based on dynamic and thermal factors |
CN106529081A (en) * | 2016-12-03 | 2017-03-22 | 安徽新华学院 | PM2.5 real-time level prediction method and system based on neural net |
CN106599520A (en) * | 2016-12-31 | 2017-04-26 | 中国科学技术大学 | LSTM-RNN model-based air pollutant concentration forecast method |
CN106611090A (en) * | 2016-12-31 | 2017-05-03 | 中国科学技术大学 | Roadside air pollutant concentration prediction method based on reconstruction deep learning |
CN106650825A (en) * | 2016-12-31 | 2017-05-10 | 中国科学技术大学 | Automotive exhaust emission data fusion system |
CN107192690A (en) * | 2017-05-19 | 2017-09-22 | 重庆大学 | Near infrared spectrum Noninvasive Blood Glucose Detection Methods and its detection network model training method |
CN107368894A (en) * | 2017-07-28 | 2017-11-21 | 国网河南省电力公司电力科学研究院 | The prevention and control of air pollution electricity consumption data analysis platform shared based on big data |
CN107545122A (en) * | 2017-09-27 | 2018-01-05 | 重庆长安汽车股份有限公司 | A kind of simulation system of the vehicle gaseous effluent based on neutral net |
CN108399470A (en) * | 2018-02-09 | 2018-08-14 | 苏州科技大学 | A kind of indoor PM2.5 prediction techniques based on more example genetic neural networks |
CN108426812A (en) * | 2018-04-08 | 2018-08-21 | 浙江工业大学 | A kind of PM2.5 concentration value prediction techniques based on Memory Neural Networks |
WO2018220566A1 (en) * | 2017-06-01 | 2018-12-06 | International Business Machines Corporation | Neural network classification |
CN109376903A (en) * | 2018-09-10 | 2019-02-22 | 浙江工业大学 | A kind of PM2.5 concentration value prediction technique based on game neural network |
CN109670646A (en) * | 2018-12-21 | 2019-04-23 | 浙江工业大学 | A kind of PM2.5 concentration value prediction technique based on mixing threshold neural network |
TWI662422B (en) * | 2018-04-23 | 2019-06-11 | 國家中山科學研究院 | Air quality prediction method based on machine learning model |
CN109916788A (en) * | 2019-01-14 | 2019-06-21 | 南京大学 | A kind of differentiation different zones discharge variation and meteorological condition variation are to PM2.5The method that concentration influences |
CN110046771A (en) * | 2019-04-25 | 2019-07-23 | 河南工业大学 | A kind of PM2.5 concentration prediction method and apparatus |
CN110796284A (en) * | 2019-09-20 | 2020-02-14 | 平安科技(深圳)有限公司 | Method and device for predicting pollution level of fine particulate matters and computer equipment |
US10599977B2 (en) | 2016-08-23 | 2020-03-24 | International Business Machines Corporation | Cascaded neural networks using test ouput from the first neural network to train the second neural network |
CN111174824A (en) * | 2019-12-27 | 2020-05-19 | 北京首钢自动化信息技术有限公司 | Control platform that acid mist discharged |
US10713783B2 (en) | 2017-06-01 | 2020-07-14 | International Business Machines Corporation | Neural network classification |
CN112529344A (en) * | 2019-09-18 | 2021-03-19 | 中国科学院沈阳计算技术研究所有限公司 | Algorithm for optimizing air quality value based on Elman neural network |
WO2021099427A1 (en) * | 2019-11-22 | 2021-05-27 | Elichens | Method for estimating the concentration of analyte in air close to a route travelled by means of transport |
CN113011660A (en) * | 2021-03-23 | 2021-06-22 | 上海应用技术大学 | Air quality prediction method, system and storage medium |
CN113281229A (en) * | 2021-02-09 | 2021-08-20 | 北京工业大学 | Multi-model self-adaptive atmosphere PM based on small samples2.5Concentration prediction method |
CN114814092A (en) * | 2022-04-12 | 2022-07-29 | 上海应用技术大学 | IP index measuring method based on BP neural network |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104535465A (en) * | 2015-01-07 | 2015-04-22 | 东北大学 | PM2.5 concentration detection method and device based on neural network |
CN104729965A (en) * | 2015-01-28 | 2015-06-24 | 东北大学 | PM2.5 concentration detection method based on interzone radial basis function nerve network |
-
2016
- 2016-06-07 CN CN201610401881.7A patent/CN106056210B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104535465A (en) * | 2015-01-07 | 2015-04-22 | 东北大学 | PM2.5 concentration detection method and device based on neural network |
CN104729965A (en) * | 2015-01-28 | 2015-06-24 | 东北大学 | PM2.5 concentration detection method based on interzone radial basis function nerve network |
Non-Patent Citations (1)
Title |
---|
刘敏等: "《基于混合神经网络的臭氧浓度软测量》", 《计算机测量与控制》 * |
Cited By (39)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10599977B2 (en) | 2016-08-23 | 2020-03-24 | International Business Machines Corporation | Cascaded neural networks using test ouput from the first neural network to train the second neural network |
CN106529081A (en) * | 2016-12-03 | 2017-03-22 | 安徽新华学院 | PM2.5 real-time level prediction method and system based on neural net |
CN106529746A (en) * | 2016-12-29 | 2017-03-22 | 南京恩瑞特实业有限公司 | Method for dynamically fusing, counting and forecasting air quality based on dynamic and thermal factors |
CN106611090B (en) * | 2016-12-31 | 2018-04-10 | 中国科学技术大学 | A kind of road side air pollutant concentration Forecasting Methodology based on reconstruct deep learning |
CN106650825A (en) * | 2016-12-31 | 2017-05-10 | 中国科学技术大学 | Automotive exhaust emission data fusion system |
CN106599520A (en) * | 2016-12-31 | 2017-04-26 | 中国科学技术大学 | LSTM-RNN model-based air pollutant concentration forecast method |
CN106650825B (en) * | 2016-12-31 | 2020-05-12 | 中国科学技术大学 | Motor vehicle exhaust emission data fusion system |
CN106611090A (en) * | 2016-12-31 | 2017-05-03 | 中国科学技术大学 | Roadside air pollutant concentration prediction method based on reconstruction deep learning |
CN107192690A (en) * | 2017-05-19 | 2017-09-22 | 重庆大学 | Near infrared spectrum Noninvasive Blood Glucose Detection Methods and its detection network model training method |
CN107192690B (en) * | 2017-05-19 | 2019-04-23 | 重庆大学 | Near infrared spectrum Noninvasive Blood Glucose Detection Methods and its detection network model training method |
US11138724B2 (en) | 2017-06-01 | 2021-10-05 | International Business Machines Corporation | Neural network classification |
WO2018220566A1 (en) * | 2017-06-01 | 2018-12-06 | International Business Machines Corporation | Neural network classification |
CN110622175B (en) * | 2017-06-01 | 2023-09-19 | 国际商业机器公司 | Neural network classification |
US11935233B2 (en) | 2017-06-01 | 2024-03-19 | International Business Machines Corporation | Neural network classification |
GB2577017A (en) * | 2017-06-01 | 2020-03-11 | Ibm | Neural network classification |
US10713783B2 (en) | 2017-06-01 | 2020-07-14 | International Business Machines Corporation | Neural network classification |
CN110622175A (en) * | 2017-06-01 | 2019-12-27 | 国际商业机器公司 | Neural network classification |
CN107368894A (en) * | 2017-07-28 | 2017-11-21 | 国网河南省电力公司电力科学研究院 | The prevention and control of air pollution electricity consumption data analysis platform shared based on big data |
CN107545122A (en) * | 2017-09-27 | 2018-01-05 | 重庆长安汽车股份有限公司 | A kind of simulation system of the vehicle gaseous effluent based on neutral net |
CN108399470A (en) * | 2018-02-09 | 2018-08-14 | 苏州科技大学 | A kind of indoor PM2.5 prediction techniques based on more example genetic neural networks |
CN108399470B (en) * | 2018-02-09 | 2021-10-08 | 苏州科技大学 | Indoor PM2.5 prediction method based on multi-example genetic neural network |
CN108426812B (en) * | 2018-04-08 | 2020-07-31 | 浙江工业大学 | PM2.5 concentration value prediction method based on memory neural network |
CN108426812A (en) * | 2018-04-08 | 2018-08-21 | 浙江工业大学 | A kind of PM2.5 concentration value prediction techniques based on Memory Neural Networks |
TWI662422B (en) * | 2018-04-23 | 2019-06-11 | 國家中山科學研究院 | Air quality prediction method based on machine learning model |
CN109376903A (en) * | 2018-09-10 | 2019-02-22 | 浙江工业大学 | A kind of PM2.5 concentration value prediction technique based on game neural network |
CN109670646A (en) * | 2018-12-21 | 2019-04-23 | 浙江工业大学 | A kind of PM2.5 concentration value prediction technique based on mixing threshold neural network |
CN109916788B (en) * | 2019-01-14 | 2020-05-19 | 南京大学 | PM pair for distinguishing emission change and meteorological condition change of different areas2.5Method for influencing concentration |
CN109916788A (en) * | 2019-01-14 | 2019-06-21 | 南京大学 | A kind of differentiation different zones discharge variation and meteorological condition variation are to PM2.5The method that concentration influences |
CN110046771A (en) * | 2019-04-25 | 2019-07-23 | 河南工业大学 | A kind of PM2.5 concentration prediction method and apparatus |
CN112529344A (en) * | 2019-09-18 | 2021-03-19 | 中国科学院沈阳计算技术研究所有限公司 | Algorithm for optimizing air quality value based on Elman neural network |
CN112529344B (en) * | 2019-09-18 | 2023-09-05 | 中国科学院沈阳计算技术研究所有限公司 | Algorithm for optimizing air quality value based on Elman neural network |
CN110796284A (en) * | 2019-09-20 | 2020-02-14 | 平安科技(深圳)有限公司 | Method and device for predicting pollution level of fine particulate matters and computer equipment |
WO2021099427A1 (en) * | 2019-11-22 | 2021-05-27 | Elichens | Method for estimating the concentration of analyte in air close to a route travelled by means of transport |
CN111174824A (en) * | 2019-12-27 | 2020-05-19 | 北京首钢自动化信息技术有限公司 | Control platform that acid mist discharged |
CN111174824B (en) * | 2019-12-27 | 2022-04-12 | 北京首钢自动化信息技术有限公司 | Control platform that acid mist discharged |
CN113281229A (en) * | 2021-02-09 | 2021-08-20 | 北京工业大学 | Multi-model self-adaptive atmosphere PM based on small samples2.5Concentration prediction method |
CN113281229B (en) * | 2021-02-09 | 2022-11-29 | 北京工业大学 | Multi-model self-adaptive atmosphere PM based on small samples 2.5 Concentration prediction method |
CN113011660A (en) * | 2021-03-23 | 2021-06-22 | 上海应用技术大学 | Air quality prediction method, system and storage medium |
CN114814092A (en) * | 2022-04-12 | 2022-07-29 | 上海应用技术大学 | IP index measuring method based on BP neural network |
Also Published As
Publication number | Publication date |
---|---|
CN106056210B (en) | 2018-06-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106056210A (en) | PM 2.5 concentration value prediction method based on hybrid neural network | |
Krishan et al. | Air quality modelling using long short-term memory (LSTM) over NCT-Delhi, India | |
CN108426812B (en) | PM2.5 concentration value prediction method based on memory neural network | |
Zhou et al. | Industrial structural upgrading and spatial optimization based on water environment carrying capacity | |
Chen et al. | Predict the effect of meteorological factors on haze using BP neural network | |
CN106611090B (en) | A kind of road side air pollutant concentration Forecasting Methodology based on reconstruct deep learning | |
CN106779151B (en) | A kind of line of high-speed railway wind speed multi-point multi-layer coupling prediction method | |
CN108268935B (en) | PM2.5 concentration value prediction method and system based on time sequence recurrent neural network | |
Cai et al. | Prediction of hourly air pollutant concentrations near urban arterials using artificial neural network approach | |
He et al. | Modeling the urban landscape dynamics in a megalopolitan cluster area by incorporating a gravitational field model with cellular automata | |
CN103226741B (en) | Public supply mains tube explosion prediction method | |
Kang et al. | Application of BP neural network optimized by genetic simulated annealing algorithm to prediction of air quality index in Lanzhou | |
CN114280695A (en) | Air pollutant monitoring and early warning method and cloud platform | |
Asghari et al. | Predicting air pollution in Tehran: Genetic algorithm and back propagation neural network | |
CN106781489A (en) | A kind of road network trend prediction method based on recurrent neural network | |
Gadhavi et al. | Student final grade prediction based on linear regression | |
CN109785618A (en) | Short-term traffic flow prediction method based on combinational logic | |
CN103942398A (en) | Traffic simulation correction method based on genetic algorithm and generalized recurrent nerve network | |
Abdul-Wahab et al. | Prediction of optimum sampling rates of air quality monitoring stations using hierarchical fuzzy logic control system | |
CN113011455B (en) | Air quality prediction SVM model construction method | |
Shen et al. | Prediction of entering percentage into expressway service areas based on wavelet neural networks and genetic algorithms | |
CN111077048A (en) | Opportunistic group intelligent air quality monitoring and evaluating method based on mobile equipment | |
Al_Janabi et al. | Pragmatic method based on intelligent big data analytics to prediction air pollution | |
Kadiyala et al. | Application of MATLAB to select an optimum performing genetic algorithm for predicting in-vehicle pollutant concentrations | |
Das et al. | Prediction of air pollutants for air quality using deep learning methods in a metropolitan city |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |