CN108761571A - Atmospheric visibility prediction technique based on neural network and system - Google Patents
Atmospheric visibility prediction technique based on neural network and system Download PDFInfo
- Publication number
- CN108761571A CN108761571A CN201810289058.0A CN201810289058A CN108761571A CN 108761571 A CN108761571 A CN 108761571A CN 201810289058 A CN201810289058 A CN 201810289058A CN 108761571 A CN108761571 A CN 108761571A
- Authority
- CN
- China
- Prior art keywords
- atmospheric visibility
- model
- atmospheric
- parameter
- prediction
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
- G01W1/10—Devices for predicting weather conditions
Abstract
The atmospheric visibility prediction technique and system, this method that the present invention relates to a kind of based on neural network include：Acquire moisture absorption growth factor, the optics microphysical property parameter of current environment；By the moisture absorption growth factor collected, optics microphysical property parameter, input atmospheric visibility prediction model trained in advance, output obtains prediction atmospheric visibility, and the atmospheric visibility prediction model is the model based on neural network.Inventive process avoids cannot react the situation of real atmosphere very well or completely because of experimental situation or in-site measurement, simultaneously based on god by the powerful nonlinear prediction ability of network, the Accurate Prediction for realizing visibility has certain directive significance for the monitoring and early-warning and predicting of carrying out low visibility.
Description
Technical field
The present invention relates to atmospheric sounding techniques field, more particularly to a kind of atmospheric visibility prediction side based on neural network
Method and system.
Background technology
The aerosol component part important as air-earth system, affects the earth-atmospheric radiation revenue and expenditure.Aerosol
The variation of characteristic affects numerous aspects of air, such as weather, environment, rainfall, visibility.And moisture-absorption characteristics is aerosol
One of main character, under certain relative humidity of atomsphere, due to moisture-absorption characteristics, particulate size may increase, this will
The Spectral structure and relevant optics and microphysical property of change particulate.In aerosol moisture absorption growth, moisture absorption increases
The factor is a crucial parameter, can be by the scattering coefficient under specific relative humidity and the scattering coefficient under reference humidity
Ratio calculation obtains, and is mainly used for describing dependence of the aerosol light scattering coefficient to relative humidity.Growth factor depends primarily on
Both the chemical composition and volume size distribution of aerosol, dependent on, when particulate absorbs moisture, particle size will
Increase, i.e., particle sectional area increases, to scatter more light, i.e., the variation of particle size will generate different refractive index and
Angle scatter properties.In general, moisture absorption growth factor can utilize turbidimetry, front and back difference under different relative humidities
Migration analysis instrument method, power balance method etc. measure.But these measurement methods be largely in laboratory environment into
Row or in-site measurement.Therefore, these measurement methods cannot very well or fully reflect the situation of real atmosphere.
Invention content
The purpose of the present invention is to provide a kind of atmospheric visibility prediction technique and system based on neural network, Neng Gouti
The accuracy that highly hygroscopic growth factor measures, and then improve the accuracy of atmospheric visibility prediction.
In order to achieve the above-mentioned object of the invention, an embodiment of the present invention provides following technical schemes：
On the one hand, the atmospheric visibility prediction technique based on neural network that an embodiment of the present invention provides a kind of, including with
Lower step：
Acquire moisture absorption growth factor, the optics microphysical property parameter of current environment；
By the moisture absorption growth factor collected, optics microphysical property parameter, input atmospheric visibility trained in advance
Prediction model, output obtain prediction atmospheric visibility, and the atmospheric visibility prediction model is the model based on neural network.
On the other hand, the atmospheric visibility forecasting system based on neural network that an embodiment of the present invention provides a kind of, including
With lower module：
Model training module, for being based on neural network, training obtains atmospheric visibility prediction model；
Atmospheric visibility prediction module, moisture absorption growth factor, optics speck for receiving the current environment collected
Characterisitic parameter is managed, and the data received are input to the atmospheric visibility prediction model, output, which obtains prediction air, to be seen
Degree.
In another aspect, an embodiment of the present invention provides a kind of electronic equipment, which includes：Memory stores journey
Sequence instructs；Processor is connected with the memory, executes the program instruction in memory, presses any embodiment of the present invention
The method predicts atmospheric visibility.
Compared with prior art, the present invention utilizes a kind of energy synchronizing detection relative humidity and aerosol optical and Microphysical
The multi-wavelength multi-parameter of characteristic polarizes Raman lidar, obtains aerosol moisture absorption growth factor, avoid because of experimental situation or
In-site measurement and the case where the situation of real atmosphere cannot be reacted very well or completely；Simultaneously based on powerful non-linear of neural network
Predictive ability, the non-linear relation between elements and visibility such as the research aerosol moisture absorption factor, pollution situation, meteorological condition,
Atmospheric visibility prediction model is established, realizes the Accurate Prediction of visibility, for carrying out monitoring and the early-warning and predicting of low visibility
With certain directive significance.
Description of the drawings
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached
Figure is briefly described, it should be understood that the following drawings illustrates only certain embodiments of the present invention, therefore is not construed as pair
The restriction of range for those of ordinary skill in the art without creative efforts, can also be according to this
A little attached drawings obtain other relevant attached drawings.
Fig. 1 is the flow chart of the atmospheric visibility prediction technique based on neural network described in the embodiment of the present invention.
Fig. 2 is the training flow chart of neural network described in the embodiment of the present invention.
Fig. 3 is three layers of BP neural network structure chart of the visibility prediction model described in the embodiment of the present invention.
Fig. 4 is the structural representation frame of the atmospheric visibility forecasting system based on neural network described in the embodiment of the present invention
Figure.
Fig. 5 is the structural schematic block diagram of the electronic equipment described in the embodiment of the present invention.
Specific implementation mode
Below in conjunction with attached drawing in the embodiment of the present invention, technical solution in the embodiment of the present invention carries out clear, complete
Ground describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.Usually exist
The component of the embodiment of the present invention described and illustrated in attached drawing can be arranged and be designed with a variety of different configurations herein.Cause
This, the detailed description of the embodiment of the present invention to providing in the accompanying drawings is not intended to limit claimed invention below
Range, but it is merely representative of the selected embodiment of the present invention.Based on the embodiment of the present invention, those skilled in the art are not doing
The every other embodiment obtained under the premise of going out creative work, shall fall within the protection scope of the present invention.
Referring to Fig. 1, a kind of atmospheric visibility prediction technique based on neural network is provided in the present embodiment, including with
Lower step：
S100 acquires moisture absorption growth factor, the optics microphysical property parameter of current environment.
In this step, moisture absorption growth factor obtains in the following manner：
Using backscattering coefficient of the laser radar under the certain relative humidity of detection at 532nm, after detecting
Moisture absorption growth factor is calculated using following formula (1) to scattering coefficient：
fβ(RH)=β (z, RH)/β (RHref) (1)
In formula, β (z, RH) and β (RHref) be respectively the backscattering coefficient of laser radar detection under certain relative humidity,
The backscattering coefficient of laser radar detection under reference state (dry state).
It is fitted using following one-parameter fitting formula (2) and two-parameter fitting formula (3)：
Wherein, RH is relative humidity, and γ is empirical parameter, represents aerosol moisture absorption growing ability, which can be preferably
A fitting part in ground increases without deliquescing the aerosol moisture absorption of phenomenon；A, b are empirical parameter, which can be preferably
Artificial source emission type aerosol and biomass combustion aerosol are fitted.It is molten to gas using the two formula in this method
The glue moisture absorption factor is fitted, and can obtain more accurate moisture absorption growth factor.
In this step, the optics microphysical property parameter of acquisition includes：PM2.5、PM10、SO2、O3、NO2, atmospheric temperature, wind
Some or all of in speed, air pressure, above-mentioned whole parameters are preferably comprised, because parameter is more, the result of prediction is more accurate.These
Supplemental characteristic can utilize heliograph, aerodynamic size spectrometer to be surveyed with, laser dust monitor and other environment detectors
Amount obtains.
S200, by the moisture absorption growth factor collected, optics microphysical property parameter, input air energy trained in advance
Degree of opinion prediction model, output obtain prediction atmospheric visibility, and the atmospheric visibility prediction model is the mould based on neural network
Type.
In this step, as shown in Fig. 2, during using neural metwork training atmospheric visibility prediction model, first
It determines the structure of atmospheric visibility prediction model, that is, determines the foundation structure of neural network, and initialization model parameter；Then defeated
Enter historical data, carry out parameter normalization processing, updates model parameter；Whether last training of judgement error or iterations reach
Given threshold if it is terminates to train, and obtains the atmospheric visibility prediction model, otherwise continues to input historical data progress
Model training, until training error or iterations reach given threshold.
More specifically, determining that the foundation structure of neural network is to determine input neuron number, node in hidden layer, defeated
Go out node layer number.As shown in figure 3, this neural network shares 10 input nodes, i.e. the neuron of input layer is 10, is respectively inhaled
Wet growth factor, PM2.5、PM10、SO2、O3、NO2, atmospheric temperature, wind speed, air pressure and humidity.The number of nodes of hidden layer is very little, net
Network can get that carry out the information of self-training just few, therefore network may not train not come out.Neuronal quantity can then increase too much
Trained time even results in larger error.According to the experience of forefathers, hidden layer neuron is carried out using following formula herein
Design：In formula：Node in hidden layer is indicated with n；Input number of nodes niIt indicates；n0It is saved for output
Points；Constant between 1-10 is indicated with a, final to determine that hidden layer neuron number is 13.Output layer is a node, i.e.,
The atmospheric visibility of prediction.
When establishing neural network, first with a large amount of (such as 800 groups) input data moisture absorption growth factors, PM2.5、PM10、
SO2、O3、NO2, atmospheric temperature, wind speed, air pressure, humidity and output data atmospheric visibility historical data, to neural network
Carry out repetition training, when training error or cycle (iteration) number reach requirement when terminate to train.
During building atmospheric visibility prediction model using neural network
(1) basic BP neural network is established
The input vector of BP neural network is x ∈ Rn, hidden layer has n1, export and bePower of the input layer to hidden layer
Value is wij, threshold value θj, the weights of hidden layer to output layer are w 'ij, threshold value is θ 'k, then each layer neuron, which exports, is：
Wherein, n is input layer dimension, m
For the dimension of output layer；x′jFor the output of hidden layer, ycFor the output of output layer.
(2) after the training data of BP networks and topological structure determine, since BP algorithm requires the excitation letter of neuron operation
Number is continuous guidable, therefore excitation function selects Sigmoid type functions
(3) during iteration, when updating weights, larger improved LM (Levenberg- can be selected
Marquardt) algorithm carries out gradient calculating, and compared with traditional BP algorithm, gradient declines will soon very much, and constringency performance is more
It is good.Levenberg-Marquardt (LM) algorithm is a kind of iterative method solving nonlinear least squares minimum, has become one kind
The basic skills of solving non-linear least square problem, and the combination of steepest descent method and Gauss-Newton method can be regarded as
Algorithm.
Objective function is：
Y in formulac(k) --- expected network output vector；ym(n) --- actual network output vector；E (k) --- when
Preceding error.
If ωkIt is ω according to the new weight vector of Newton's algorithm for the network weight vector of kth time iterationk+1, according to LM
Algorithm arrives：
In formula, ek--- error vector；Jk--- Jacobi matrix of the network error to weights derivative；I --- it is unit square
Battle array；λk--- scalar.
Wherein, λkValue determines that algorithm is realized according to Newton method or gradient method.Work as coefficient lambdakWhen being 0, above formula is
Newton method；Work as coefficient lambdakValue it is very big when, above formula becomes the smaller gradient descent method algorithm of step-length.In this approach, λkIt is also
Automatic adjusument.The specific steps are：
A gives initial weight vector ω at random0, setting target error ε, k=1.
B calculates output and the error vector E (k) of network.
C calculates error vector to the Grad of network weight and forms Jacobi matrixes.
E updates weights by formula：
F, if E (ωkThen algorithm terminates)≤ε, obtains the weight vector for meeting required precision；Otherwise, step b is turned to.
Alternatively embodiment, can also utilize and momentum term update weights are added, and the value increase of hidden layer is：
η is learning rate in formula, and α is factor of momentum.
It can be rewritten into
Δωij(n)=η δi(n)yi(n)+αΔωij(n-1)
When training data is added, above formula can be write as the time series using t as variable, and above formula is converted to Δ ωij(n)
Difference equation, solve Δ ωij(n) it obtains
When n-thWith n-1 jack per line, weighted sum value increases, therefore Δ ωij(t) bigger, therefore increase
The rate of change of ω；When n-thWhen with n-1 contrary sign, illustrate, in the presence of concussion, can thus make Δ ωij(t) subtract
It is small so that convergence steadily approach with it is optimal.So after introducing momentum term, learning rate is not only accelerated, can also effectively be kept away
Exempt from network and sinks into local minimum.
The method that the present invention can be used LM-BP neural networks or the BP neural network update weights of momentum term are added, can be with
So that convergence is reached most fast and network is avoided to be absorbed in local minimum.
In the historical data used in the training process, including input data and corresponding output data, input data
Including PM2.5、PM10、SO2、O3、NO2, atmospheric temperature, wind speed, air pressure, all or part of parameter in humidity, output data is then
Atmospheric visibility corresponding to these input datas when environment, only atmospheric visibility needs, which are calculated, obtains, that is, utilizes aerosol
Extinction coefficient inverting obtains the atmospheric extinction coefficient of horizontal journey, recycles the atmospheric extinction coefficient of the horizontal journey that level is calculated
The atmospheric visibility in direction.
Specifically, it using vibrating Raman signal inverting Aerosol Extinction, obtains
, in formula, λLFor laser emission wavelength, λRFor the vibrating Raman dispersion wavelength of nitrogen；aM(λL,z)、aA(λL, z) respectively
For the extinction coefficient of atmospheric molecule and aerosol；N (z) is the number density of molecule of nitrogen；Z is level height；P(z,λLλR) be
At height z, the vibrating Raman dispersion wavelength of laser emission wavelength and nitrogen is λ respectivelyL、λRWhen echo-signal reception power, aM
(λR, z) and the value of N (z) can be obtained by United States standard atmosphere model.
The atmospheric horizontal visibility RV and most sensitive 550nm wavelength atmospheric molecule Horizontal extinction coefficients a of human eyeA(λL,z)
And the relationship between the comparison threshold value ε of human eye is
Therefore the atmospheric molecule Horizontal extinction coefficient a obtained by invertingA(λL, z) and atmospheric horizontal visibility RV can be calculated.
After training obtains atmospheric visibility prediction model, collected moisture absorption growth factor, optics under current environment are inputted
Microphysical property parameter, you can export the atmospheric visibility of prediction.If input the moisture absorption growth factor under a certain specific environment,
Optics microphysical property parameter also can be predicted and obtain the atmospheric visibility under the specific environment.
Laser radar technique is used for the measurement of aerosol moisture absorption rising characteristic by the present invention, is avoided because of experimental situation or is showed
Field measurement and the case where the situation of real atmosphere cannot be reacted very well or completely；Simultaneously based on powerful non-linear pre- of neural network
Survey ability, the non-linear relation between elements and visibility such as the research aerosol moisture absorption factor, pollution situation, meteorological condition, builds
Vertical atmospheric visibility prediction model, realizes the Accurate Prediction of atmospheric visibility, pre- for the monitoring and early warning of carrying out low visibility
Report has certain directive significance.
Referring to Fig. 4, being based on identical inventive concept, the embodiment of the present invention additionally provides a kind of based on the big of neural network
Gas visibility forecasting system, comprises the following modules：
Model training module 41, for being based on neural network, training obtains atmospheric visibility prediction model；
Atmospheric visibility prediction module 42, moisture absorption growth factor, the optics for receiving the current environment collected are micro-
Physical characteristic parameter, and the data received are input to the atmospheric visibility prediction model, output obtains prediction air energy
Degree of opinion.
In model training module 41, moisture absorption growth factor, optics microphysical property parameter are being built using neural network
When atmospheric visibility prediction model between atmospheric visibility, in an iterative process, network weight is updated in the following way：
Wherein, ωkNetwork weight for kth time iteration is vectorial,
ωk+1For updated network weight vector, ekFor error vector, JkIt is network error to the Jacobi matrixes of weights derivative, I is
Unit matrix, λkFor scalar.
As other embodiment, in an iterative process, network weight is updated in the following way：
Wherein Δ ωij(n) weights for being hidden layer node n
Increment, η are learning rate, and a is factor of momentum, and E (k) is error vector.
The optics microphysical property parameter may include：PM2.5、PM10、SO2、O3、NO2, atmospheric temperature, wind speed, air pressure,
Humidity, these data can utilize the monitoring of heliograph, aerodynamic size spectrometer and laser dust monitor to obtain.
As shown in figure 5, the present embodiment provides a kind of electronic equipment simultaneously, which may include 51 He of processor
Memory 52, wherein memory 52 are coupled to processor 51.It is worth noting that, the figure is exemplary, it can also be used
The structure of his type carrys out the supplement or alternative structure, realize task perception and receive, task priority judgement, alternative satellite function
Whether judgement, task switching, communication or other functions are had.
As shown in figure 5, the electronic equipment can also include：Input unit 53, display unit 54 and power supply 55.It is worth noting
, which is also not necessary to include all components shown in Fig. 5.In addition, electronic equipment can also include
The component being not shown in Fig. 5 can refer to the prior art.
Processor 51 is sometimes referred to as controller or operational controls, may include microprocessor or other processor devices and/
Or logic device, the processor 51 receive the operation of all parts of input and control electronics.
Wherein, memory 52 for example can be buffer, flash memory, hard disk driver, removable medium, volatile memory, it is non-easily
It is one or more in the property lost memory or other appropriate devices, configuration information, the processor 51 of above-mentioned processor 51 can be stored
The information such as the instruction of execution.Processor 51 can execute the program of the storage of memory 52, to realize information storage or processing etc..?
Further include buffer storage in memory 52 in one embodiment, i.e. buffer, to store average information.
Input unit 53 for example can be used for providing historical data or current measurement data to processor 51.Display unit
54 for showing prediction result or the structure of prediction model, which can be for example LCD display, but the present invention is not
It is limited to this.Power supply 55 is used to provide electric power for electronic equipment.
The embodiment of the present invention also provides a kind of computer-readable instruction, wherein when executing described instruction in the electronic device
When, described program makes electronic equipment execute the atmospheric visibility prediction technique based on neural network as shown in Figure 1 and be included
Operating procedure.
The embodiment of the present invention also provides a kind of storage medium being stored with computer-readable instruction, wherein the computer can
Reading instruction makes electronic equipment execute the operation that the atmospheric visibility prediction technique based on neural network as shown in Figure 1 is included
Step.
Those of ordinary skill in the art may realize that lists described in conjunction with the examples disclosed in the embodiments of the present disclosure
Member and algorithm steps, can be realized with electronic hardware, computer software, or a combination of the two, in order to clearly demonstrate hardware
With the interchangeability of software, each exemplary composition and step are generally described according to function in the above description.This
A little functions are implemented in hardware or software actually, depend on the specific application and design constraint of technical solution.Specially
Industry technical staff can use different methods to achieve the described function each specific application, but this realization is not
It is considered as beyond the scope of this invention.
It is apparent to those skilled in the art that for convenience of description and succinctly, foregoing description is
The specific work process of system and unit, can refer to corresponding processes in the foregoing method embodiment, details are not described herein.
In several embodiments provided herein, it should be understood that disclosed system and method can pass through it
Its mode is realized.For example, the apparatus embodiments described above are merely exemplary, for example, the division of the unit, only
Only a kind of division of logic function, formula that in actual implementation, there may be another division manner, such as multiple units or component can be tied
Another system is closed or is desirably integrated into, or some features can be ignored or not executed.In addition, shown or discussed phase
Coupling, direct-coupling or communication connection between mutually can be INDIRECT COUPLING or the communication by some interfaces, device or unit
Connection can also be electricity, mechanical or other form connections.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can easily think of the change or the replacement, and should all contain
Lid is within protection scope of the present invention.
Claims (7)
1. a kind of atmospheric visibility prediction technique based on neural network, which is characterized in that include the following steps：
Acquire moisture absorption growth factor, the optics microphysical property parameter of current environment；
By the moisture absorption growth factor collected, optics microphysical property parameter, input atmospheric visibility prediction trained in advance
Model, output obtain prediction atmospheric visibility, and the atmospheric visibility prediction model is the model based on neural network.
2. according to the method described in claim 1, it is characterized in that, the optics microphysical property parameter includes：PM2.5、PM10、
SO2、O3、NO2, atmospheric temperature, wind speed, air pressure, all or part of parameter in humidity.
3. according to the method described in claim 2, it is characterized in that, the atmospheric visibility prediction model is instructed in the following manner
It gets：
Determine the structure of atmospheric visibility prediction model, and initialization model parameter；
It inputs historical data and carries out model training, update model parameter；The historical data includes moisture absorption growth factor, PM2.5、
PM10、SO2、O3、NO2, atmospheric temperature, wind speed, air pressure, all or part of parameter in humidity and parameter correspond under visibility
Data；
Whether training of judgement error or iterations reach given threshold, if it is terminate to train, and obtaining the air can see
Prediction model is spent, otherwise continues to input historical data progress model training, until training error or iterations reach setting threshold
Value.
4. according to the method described in claim 1, it is characterized in that, the moisture absorption growth factor acquires in the following manner
It arrives：
It is backward scattered using what is detected using backscattering coefficient of the laser radar under the certain relative humidity of detection at 532nm
It penetrates coefficient and calculates moisture absorption growth factor：
fβ(RH)=β (z, RH)/β (RHref)
In formula, β (z, RH) is backscattering coefficient, the β (RH of laser radar detection under certain relative humidityref) it is reference state
Under laser radar detection backscattering coefficient；
Moisture absorption growth factor is fitted using following two fitting formulas：
Wherein, RH is relative humidity, and γ is empirical parameter, represents aerosol moisture absorption growing ability, and a, b are empirical parameter.
5. a kind of atmospheric visibility forecasting system based on neural network, which is characterized in that comprise the following modules：
Model training module, for being based on neural network, training obtains atmospheric visibility prediction model；
Atmospheric visibility prediction module, moisture absorption growth factor, the optics Microphysical for receiving the current environment collected are special
Property parameter, and the data received are input to the atmospheric visibility prediction model, output obtains prediction atmospheric visibility.
6. system according to claim 5, which is characterized in that the model training module is trained especially by following manner
Obtain the atmospheric visibility prediction model：
Determine the structure of visibility prediction model, and initialization model parameter；
It inputs historical data and carries out model training, update model parameter；
Whether training of judgement error or iterations reach given threshold, if it is terminate to train, and obtaining the air can see
Prediction model is spent, otherwise continues to input historical data progress model training, until training error or iterations reach setting threshold
Value.
7. a kind of electronic equipment, which is characterized in that the electronic equipment includes：
Memory stores program instruction；
Processor is connected with the memory, execute memory in program instruction, by claim 1-4 it is any described in
Method predicts atmospheric visibility.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810289058.0A CN108761571A (en) | 2018-04-03 | 2018-04-03 | Atmospheric visibility prediction technique based on neural network and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810289058.0A CN108761571A (en) | 2018-04-03 | 2018-04-03 | Atmospheric visibility prediction technique based on neural network and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108761571A true CN108761571A (en) | 2018-11-06 |
Family
ID=63980854
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810289058.0A Pending CN108761571A (en) | 2018-04-03 | 2018-04-03 | Atmospheric visibility prediction technique based on neural network and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108761571A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109374488A (en) * | 2018-11-14 | 2019-02-22 | 国网安徽省电力有限公司电力科学研究院 | Prediction technique for photovoltaic power station haze weather visibility |
| KR102002158B1 (en) * | 2018-11-15 | 2019-07-19 | 주식회사 포디솔루션 | System for prediction of visibility |
| CN110531444A (en) * | 2019-08-29 | 2019-12-03 | 中国气象局广州热带海洋气象研究所（广东省气象科学研究所） | The error source of numerical weather prediction model determines method and device |
| CN111060926A (en) * | 2019-12-25 | 2020-04-24 | 新奇点企业管理集团有限公司 | Rainfall calculation system and method |
| CN111398109A (en) * | 2020-03-10 | 2020-07-10 | 上海眼控科技股份有限公司 | Atmospheric visibility measuring method, sensor module, system and storage medium |
| CN112461799A (en) * | 2020-11-25 | 2021-03-09 | 北京心中有数科技有限公司 | Method and device for acquiring visibility of highway foggy |
| CN113466181A (en) * | 2021-09-02 | 2021-10-01 | 成都信息工程大学 | Atmospheric visibility data processing method, system and application |
| CN113625369A (en) * | 2021-08-10 | 2021-11-09 | 中国科学院大气物理研究所 | Miniaturized intelligent measuring system and method for atmospheric visibility |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106326520A (en) * | 2015-07-10 | 2017-01-11 | 中国科学院遥感与数字地球研究所 | Moisture absorption and reconstruction simulation method and system of black carbon cluster particles based on fractal theory |
| CN106709903A (en) * | 2016-11-22 | 2017-05-24 | 南京理工大学 | PM2.5 concentration prediction method based on image quality |
| CN107664628A (en) * | 2017-08-29 | 2018-02-06 | 中国计量大学 | It is a kind of predicting the method for atmospheric visibility |
-
2018
- 2018-04-03 CN CN201810289058.0A patent/CN108761571A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106326520A (en) * | 2015-07-10 | 2017-01-11 | 中国科学院遥感与数字地球研究所 | Moisture absorption and reconstruction simulation method and system of black carbon cluster particles based on fractal theory |
| CN106709903A (en) * | 2016-11-22 | 2017-05-24 | 南京理工大学 | PM2.5 concentration prediction method based on image quality |
| CN107664628A (en) * | 2017-08-29 | 2018-02-06 | 中国计量大学 | It is a kind of predicting the method for atmospheric visibility |
Non-Patent Citations (4)
| Title |
|---|
| 张芷言: "广州地区PM_1质量浓度对能见度的影响以及气溶胶吸湿增长因子", 《中国科学院大学学报》 * |
| 熊兴隆: "基于消光系数的机场PM2.5质量浓度神经网络预测模型", 《科学技术与工程》 * |
| 王小川: "《MATLAB神经网络43个案例分析》", 31 August 2013 * |
| 郭庆春: "《人工神经网络应用研究》", 31 October 2016 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109374488A (en) * | 2018-11-14 | 2019-02-22 | 国网安徽省电力有限公司电力科学研究院 | Prediction technique for photovoltaic power station haze weather visibility |
| CN109374488B (en) * | 2018-11-14 | 2021-06-29 | 国网安徽省电力有限公司电力科学研究院 | Prediction method for visibility of haze weather of distributed photovoltaic power station |
| KR102002158B1 (en) * | 2018-11-15 | 2019-07-19 | 주식회사 포디솔루션 | System for prediction of visibility |
| CN110531444A (en) * | 2019-08-29 | 2019-12-03 | 中国气象局广州热带海洋气象研究所（广东省气象科学研究所） | The error source of numerical weather prediction model determines method and device |
| CN110531444B (en) * | 2019-08-29 | 2021-10-08 | 中国气象局广州热带海洋气象研究所（广东省气象科学研究所） | Error source determination method and device for numerical weather forecast mode |
| CN111060926A (en) * | 2019-12-25 | 2020-04-24 | 新奇点企业管理集团有限公司 | Rainfall calculation system and method |
| CN111060926B (en) * | 2019-12-25 | 2022-07-12 | 新奇点智能科技集团有限公司 | Rainfall calculation system and method |
| CN111398109A (en) * | 2020-03-10 | 2020-07-10 | 上海眼控科技股份有限公司 | Atmospheric visibility measuring method, sensor module, system and storage medium |
| CN112461799A (en) * | 2020-11-25 | 2021-03-09 | 北京心中有数科技有限公司 | Method and device for acquiring visibility of highway foggy |
| CN112461799B (en) * | 2020-11-25 | 2023-08-18 | 北京心中有数科技有限公司 | Method and device for obtaining visibility of fog on expressway |
| CN113625369A (en) * | 2021-08-10 | 2021-11-09 | 中国科学院大气物理研究所 | Miniaturized intelligent measuring system and method for atmospheric visibility |
| CN113466181A (en) * | 2021-09-02 | 2021-10-01 | 成都信息工程大学 | Atmospheric visibility data processing method, system and application |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108761571A (en) | Atmospheric visibility prediction technique based on neural network and system | |
| Qiu et al. | Atmospheric dispersion prediction and source estimation of hazardous gas using artificial neural network, particle swarm optimization and expectation maximization | |
| Boutahar et al. | Development and validation of a fully modular platform for numerical modelling of air pollution: POLAIR | |
| Sausen et al. | Coupled ocean-atmosphere models with flux correction | |
| CN111158401B (en) | Distributed unmanned aerial vehicle path planning system and method for encouraging space-time data exploration | |
| BR102018010426A2 (en) | SATELLITE SCHEDULING METHOD AND PROCESSING SYSTEM FOR EARTH OBSERVATION SYSTEMS, INCLUDING ONE OR MORE REMOTE SENSING SATELLITES | |
| Darus et al. | Parametric modelling of a twin rotor system using genetic algorithms | |
| CN107885933B (en) | Pavement structure fatigue cracking numerical simulation method based on extended finite element | |
| CN109828578A (en) | A kind of instrument crusing robot optimal route planing method based on YOLOv3 | |
| CN105784624A (en) | Retrieval method and device for water vapor profile | |
| Zhou et al. | Combined estimation of fire perimeters and fuel adjustment factors in FARSITE for forecasting wildland fire propagation | |
| Chung et al. | Examination of situation-dependent background error covariances at the convective scale in the context of the ensemble Kalman filter | |
| Jiao et al. | Damage identification of bridge based on Chebyshev polynomial fitting and fuzzy logic without considering baseline model parameters | |
| Tang et al. | A power transfer model-based method for lithium-ion battery discharge time prediction of electric rotatory-wing UAV | |
| CN111488709A (en) | Part flatness error analysis method, error analysis system and computer | |
| CN114999117B (en) | Airport pavement ice and snow condition monitoring and early warning method, system, terminal and medium | |
| Zheng et al. | An effective genetic algorithm to VDA with discontinuous “on-off” switches | |
| CN115859789A (en) | Method for improving inversion accuracy of polar atmosphere temperature profile | |
| Rasp | Statistical methods and machine learning in weather and climate modeling | |
| CN115270239A (en) | Bridge reliability prediction method based on dynamic characteristics and intelligent algorithm response surface method | |
| Politis et al. | Bayesian updating of atmospheric dispersion after a nuclear accident | |
| Wu et al. | Correlation between PM concentrations and aerosol optical depth in eastern China based on BP neural networks | |
| Xi et al. | Adaptive compensation for measurement error in remote sensing of mobile source emissions | |
| Elbern et al. | Inverse modelling and combined state-source estimation for chemical weather | |
| Wang et al. | Efficient climate simulation via machine learning method |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181106 |
|
| RJ01 | Rejection of invention patent application after publication |