CN108259099A - A kind of TD-SCDMA base station electromagnetic radiation Forecasting Methodology - Google Patents
A kind of TD-SCDMA base station electromagnetic radiation Forecasting Methodology Download PDFInfo
- Publication number
- CN108259099A CN108259099A CN201810071037.1A CN201810071037A CN108259099A CN 108259099 A CN108259099 A CN 108259099A CN 201810071037 A CN201810071037 A CN 201810071037A CN 108259099 A CN108259099 A CN 108259099A
- Authority
- CN
- China
- Prior art keywords
- value
- high frequency
- parameter
- low frequency
- frequency sequence
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/391—Modelling the propagation channel
- H04B17/3913—Predictive models, e.g. based on neural network models
Landscapes
- Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses a kind of TD SCDMA base stations electromagnetic radiation Forecasting Methodologies, and its step are as follows：TD SCDMA base station electromagnetic radiation historical datas are taken as training data, training data is decomposed by DB3 and COIF3 small echos, Decomposition order is 1, low frequency sequence R, high frequency series T are obtained by DB3 wavelet decompositions, low frequency sequence Q, high frequency series U are obtained by COIF3 wavelet decompositions, again by low frequency sequence R, high frequency series T, low frequency sequence Q, high frequency series U, input prediction model is trained respectively, then trained prediction model is done combined prediction to TD SCDMA base station electromagnetic radiation.The present invention analyzes the mode of TD SCDMA base stations electromagnetic radiation prediction, and this method accurately reliably can be predicted and be assessed to TD SCDMA base station electromagnetic radiation, have certain social benefit.
Description
Technical field
The present invention relates to a kind of TD-SCDMA base station electromagnetic radiation Forecasting Methodologies.
Background technology
When being predicted currently for the electromagnetic radiation around communication base station, often predicted with trained prediction model.
For example, application number 2018100095052 discloses a kind of GSM base stations electromagnetic radiation Forecasting Methodology, the electromagnetic radiation in base station
In prediction, first prediction model is trained, is then predicted using trained model, but in this prediction, due to
Model uses Individual forecast model, and the influence that Individual forecast model is easily disturbed during prediction is not sufficiently stable, together
When acutely be not easy to predict using the model due to TD-SCDMA base station electromagnetic radiation variation.
For the deficiencies in the prior art, the rule of this patent combination TD-SCDMA base station electromagnetic radiation sequence is special
Property, taking TD-SCDMA base station electromagnetic radiation historical data as training data is trained model, first passes through training data
DB3 and COIF3 small echos are decomposed, Decomposition order 1, are obtained low frequency sequence R, high frequency series T by DB3 wavelet decompositions, are led to
It crosses COIF3 wavelet decompositions and obtains low frequency sequence Q, high frequency series U, then by low frequency sequence R, high frequency series T, low frequency sequence Q, high frequency
Sequence U, then input prediction model is trained respectively, then trained prediction model does TD-SCDMA base station electromagnetic radiation
Combined prediction can realize that more accurate reliable electromagnetic radiation is predicted by method proposed by the present invention.
Invention content
In order to solve the above technical problem, the present invention provides a kind of TD-SCDMA base station electromagnetic radiation Forecasting Methodologies.
The present invention solves above-mentioned technical problem, and the technical scheme comprises the following steps：
1), take TD-SCDMA base station electromagnetic radiation historical data as training data, by training data by DB3 and
COIF3 small echos are decomposed, Decomposition order 1, are obtained low frequency sequence R, high frequency series T by DB3 wavelet decompositions, are passed through
COIF3 wavelet decompositions obtain low frequency sequence Q, high frequency series U；
2), low frequency sequence R, high frequency series T, low frequency sequence Q and the high frequency series U for obtaining step 1, input prediction mould
Type is trained, and obtains the parameter ω of prediction model1、b1、c1, ω2、b2、c2, ω3、b3、c3, ω4、b4、c4；
3), by input data when predicting, input step 2 respectively) trained prediction model carry out height frequency sequence
Prediction, expression formula is：
Wherein, ω1、b1、c1, ω2、b2、c2, ω3、b3、c3, ω4、b4、c4, to be obtained in claim 1 by step 2 training
The prediction model parameters obtained, xRiFor the low frequency sequence that the historical data before future position is obtained by DB3 wavelet decompositions, unit is
V/m, xTiFor the high frequency series that the historical data before future position is obtained by DB3 wavelet decompositions, unit V/m, xQiFor prediction
The low frequency sequence that historical data before point is obtained by COIF3 wavelet decompositions, unit V/m, xUiFor going through before future position
The high frequency series that history data are obtained by COIF3 wavelet decompositions, unit V/m, yRFor low frequency sequence xRiPredicted value, unit is
V/m, yTFor high frequency series xTiPredicted value, unit V/m, yQFor low frequency sequence xQiPredicted value, unit V/m, yUFor height
Frequency sequence xUiPredicted value, unit V/m；
4), the low-and high-frequency predicted value obtained according to step 3 training does combined prediction, and combined prediction expression formula is：
yDB3=yR+yT (5)
yCOIF3=yQ+yU (6)
Y=(yDB3+yCOIF3)/2 (7)
Wherein, yDB3For by DB3 wavelet decompositions, Decomposition order is 1 low frequency sequence prediction value yRIt is predicted with high frequency series
Value yTThe predicted value that summation obtains, unit V/m, yCOIF3For by COIF3 wavelet decompositions, Decomposition order is 1 low frequency sequence
Predicted value yQWith high frequency series predicted value yUSum obtained predicted value, unit V/m, y by proposition model combined prediction
Value, unit V/m.
A kind of above-mentioned TD-SCDMA base station electromagnetic radiation Forecasting Methodology, in the step 2), prediction model is following institute
Show：
In above formula (8), n historical data is passed through into DB3 wavelet decompositions, Decomposition order 1, the low frequency sequence R of acquisition
Input prediction model is trained and obtains optimum prediction model parameter ω1、b1、c1, by parameter ω1、b1It is assigned a value of between -1 to 1
Random number, then training data input training pattern is obtained into parameter c1Value, model parameter ω is determined with this1、b1、c1；
In above formula (9), n historical data is passed through into DB3 wavelet decompositions, Decomposition order 1, the high frequency series T of acquisition
Input prediction model is trained and obtains optimum prediction model parameter ω2、b2、c2, by parameter ω2、b2It is assigned a value of between -1 to 1
Random number, then training data input training pattern is obtained into parameter c2Value, model parameter ω is determined with this2、b2、c2；
In above formula (10), n historical data is passed through into COIF3 wavelet decompositions, Decomposition order 1, the low frequency sequence of acquisition
Q input prediction models are arranged, is trained and obtains optimum prediction model parameter ω3、b3、c3, by parameter ω3、b3Be assigned a value of -1 to 1 it
Between random number, then training data input training pattern is obtained into parameter c3Value, model parameter ω is determined with this3、b3、c3；
In above formula (11), n historical data is passed through into COIF3 wavelet decompositions, Decomposition order 1, the high frequency sequence of acquisition
U input prediction models are arranged, is trained and obtains optimum prediction model parameter ω4、b4、c4, by parameter ω4、b4Be assigned a value of -1 to 1 it
Between random number, then training data input training pattern is obtained into parameter c4Value, model parameter ω is determined with this4、b4、c4。
The beneficial effects of the present invention are：By Decomposition order it is 1 by TD-SCDMA base station electromagnetic radiation historical data
DB3 and COIF3 small echos resolve into low frequency sequence and high frequency series, and the base station electromagnetic radiation for making variation violent, which becomes flat, to be convenient for
Prediction, then obtained low frequency sequence will be decomposed and high frequency series are trained by prediction model, then by trained prediction mould
Type does combined prediction to TD-SCDMA base station electromagnetic radiation.The model established can carry out essence to TD-SCDMA base station electromagnetic radiation
Really reliable prediction, this method have base station construction and environmental protection pretty good reference value, have certain social benefit.
Description of the drawings
Fig. 1 is the flow chart of the present invention.
Specific embodiment
The present embodiment is carried out lower premised on the content of present invention, gives detailed implementation steps, but the guarantor of the present invention
Shield range is not limited to following embodiments.
This sample plot implemented is in spacious flat region, and objective for implementation is school teaching building roof base station, laboratory apparatus
For spectrum analyzer AT6024D, measurement object is TD-SCDMA base station.
The present invention is further illustrated with reference to the accompanying drawings and examples.
Fig. 1 is the block diagram of technical solution of the present invention, the specific steps are：
Step 1：TD-SCDMA base station electromagnetic radiation historical data is taken as training data, by training data by DB3 and
COIF3 small echos are decomposed, Decomposition order 1, are obtained low frequency sequence R, high frequency series T by DB3 wavelet decompositions, are passed through
COIF3 wavelet decompositions obtain low frequency sequence Q, high frequency series U.
In this embodiment, TD-SCDMA base station electromagnetic radiation historical data is taken to take 12 history as training data
Data are decomposed by DB3 the and COIF3 small echos that Decomposition order is 1 as training data, pass through the DB3 that Decomposition order is 1
Wavelet decomposition obtains 12 low frequency sequence R, 12 high frequency series T, and 12 are obtained by the COIF3 wavelet decompositions that Decomposition order is 1
A low frequency sequence Q, 12 high frequency series U.
Step 2：Low frequency sequence R, high frequency series T, low frequency sequence Q and the high frequency series U that step 1 is obtained, input are pre-
It surveys model to be trained, obtains the parameter ω of prediction model1、b1、c1, ω2、b2、c2, ω3、b3、c3, ω4、b4、c4。
In this embodiment, low frequency sequence R, high frequency series T, low frequency sequence Q and high frequency series that step 1 is taken to obtain
U is trained as training data, respectively input model, when low frequency sequence R is trained model, by R1, i=
[0.197623,0.202044 ..., 0.191881], i=1,2 ..., 12, yR=[0.188432], inputs with drag：
In above formula (12), by historical data by DB3 wavelet decompositions, Decomposition order 1, the low frequency sequence R of acquisition is defeated
Enter prediction model, while by parameter ω1And b1The random number being assigned a value of between -1 to 1, ω1It is assigned a value of 0.8129, b1It is assigned a value of
0.2119 training obtains prediction model parameters c1=-0.62911；
When high frequency series T is trained model, by T1, i=[- 0.00093, -0.00041 ..., 0.002599], i
=1,2 ..., 12, yT=[- 0.00058] is inputted with drag：
In above formula (13), by historical data by DB3 wavelet decompositions, Decomposition order 1, the high frequency series T of acquisition is defeated
Enter prediction model, while by parameter ω2And b2The random number being assigned a value of between -1 to 1, ω2It is assigned a value of 0.6213, b2It is assigned a value of
0.3175 training obtains prediction model parameters c2=-0.3248；
When low frequency sequence Q is trained model, by Q1, i=[0.196319,0.201446 ..., 0.192642], i
=1,2 ..., 12, yQ=[0.190339] is inputted with drag:
In above formula (14), historical data is passed through into COIF3 wavelet decompositions, Decomposition order 1, the low frequency sequence Q of acquisition
Input prediction model, while by parameter ω3And b3The random number being assigned a value of between -1 to 1, ω3It is assigned a value of -0.2315, b3Assignment
Prediction model parameters c is obtained for 0.2614 training3=0.717792；
When high frequency series U is trained model, by U1, i=[0.000371, -0.00012 ..., 0.001838], i
=1,2 ..., 12, yu=[- 0.00249] is inputted with drag：
In above formula (15), historical data is passed through into COIF3 wavelet decompositions, Decomposition order 1, the high frequency series U of acquisition
Input prediction model, while by parameter ω4And b4The random number being assigned a value of between -1 to 1, ω4It is assigned a value of 0.2755, b4It is assigned a value of
0.3145 training obtains prediction model parameters c4=-0.3125.
Step 3: by input data when predicting, input step 2 respectively) trained prediction model carry out low-and high-frequency
The prediction of sequence, expression formula are：
Wherein, ω1、b1、c1, ω2、b2、c2, ω3、b3、c3, ω4、b4、c4, to be obtained in claim 1 by step 2 training
The prediction model parameters obtained, xRiFor the low frequency sequence that the historical data before future position is obtained by DB3 wavelet decompositions, unit is
V/m, xTiFor the high frequency series that the historical data before future position is obtained by DB3 wavelet decompositions, unit V/m, xQiFor prediction
The low frequency sequence that historical data before point is obtained by COIF3 wavelet decompositions, unit V/m, xUiFor going through before future position
The high frequency series that history data are obtained by COIF3 wavelet decompositions, unit V/m, yRFor low frequency sequence xRiPredicted value, unit is
V/m, yTFor high frequency series xTiPredicted value, unit V/m, yQFor low frequency sequence xQiPredicted value, unit V/m, yUFor height
Frequency sequence xUiPredicted value, unit V/m.
In this embodiment, by the input data input step 2 when predicting) trained prediction model, it carries out
It is as shown in the table to obtain predicted value for the prediction of the height frequency sequence of 14 times：
The predicted value of 1 height frequency sequence of table
| yR | 0.21238 | 0.214961 | 0.21863 | 0.224595 | 0.220502 | 0.213321 | 0.213727 | 0.214849 | 0.21355 | 0.219635 | 0.213473 | 0.211362 | 0.212888 | 0.212662 |
| yT | 0.00064 | -0.00012 | -0.00022 | 0.000645 | -0.00079 | 8.89E-05 | 0.000333 | -0.00131 | 0.0018 | 0.000405 | -0.00146 | 0.000618 | 0.000132 | -0.00056 |
| yQ | 0.18359 1 | 0.188421 | 0.195196 | 0.196114 | 0.194289 | 0.189728 | 0.185898 | 0.189025 | 0.193145 | 0.191914 | 0.189213 | 0.187663 | 0.186847 | 0.18827 |
| yU | 0.00368 2 | 0.000574 | -0.00276 | 0.003121 | -0.00054 | -0.00219 | 0.002268 | -0.00134 | -0.00077 | 0.002209 | -0.00036 | -0.00121 | 0.00036 | 0.000988 |
Step 4：Combined prediction is done according to the low-and high-frequency predicted value that step 3 training obtains, combined prediction expression formula is：
yDB3=yR+yT (20)
yCOIF3=yQ+yU (21)
Y=(yDB3+yCOIF3)/2 (22)
Wherein, yDB3For by DB3 wavelet decompositions, Decomposition order is 1 low frequency sequence prediction value yRIt is predicted with high frequency series
Value yTThe predicted value that summation obtains, unit V/m, yCOIF3For by COIF3 wavelet decompositions, Decomposition order is 1 low frequency sequence
Predicted value yQWith high frequency series predicted value yUSum obtained predicted value, unit V/m, y by proposition model combined prediction
Value, unit V/m.
The low-and high-frequency predicted value that step 3 is acquired substitutes into formula (20) respectively, (21) acquire yDB3、yCOIF3Value, then will
yDB3、yCOIF3Value substitute into formula (22) and acquire the value of y, wherein yDB3、yCOIF3, the value of y is all put in following table：
2 y of tableDB3、yCOIF3, the value of y
| yDB3 | 0.21242 | 0.21484 | 0.218412 | 0.22524 | 0.219712 | 0.21341 | 0.21406 | 0.21354 | 0.21835 | 0.22004 | 0.21471 | 0.21198 | 0.21242 | 0.212 1 |
| yCOIF | 0.187273 | 0.188995 | 0.192436 | 0.199238 | 0.193749 | 0.187538 | 0.188266 | 0.187685 | 0.192405 | 0.194123 | 0.188853 | 0.186123 | 0.18720 7 | 0.189 25 |
| y | 0.200147 | 0.201914 | 0.205426 | 0.212239 | 0.206732 | 0.200474 | 0.201133 | 0.200613 | 0.205377 | 0.207082 | 0.201481 | 0.199067 | 0.20011 3 | 0.202 14 |
| It measures Value | 0.2020 | 0.2038 | 0.2073 | 0.2142 | 0.2086 | 0.2024 | 0.2024 | 0.2025 | 0.2073 | 0.2090 | 0.2037 | 0.2009 | 0.2020 | 0.204 1 |
From experimental result as can be seen that each predicted value and measured value all relatively, can be stablized different when
TD-SCDMA base station electromagnetic radiation is predicted, illustrate using the method can realize TD-SCDMA base station electromagnetic radiation essence
True reliable prediction, while the experiment show validity of method used herein.
Claims (2)
1. a kind of TD-SCDMA base station electromagnetic radiation Forecasting Methodology, which is characterized in that include the following steps：
1) TD-SCDMA base station electromagnetic radiation historical data, is taken as training data, training data is small by DB3 and COIF3
Wave is decomposed, Decomposition order 1, is obtained low frequency sequence R, high frequency series T by DB3 wavelet decompositions, is passed through the small wavelength-divisions of COIF3
Solution obtains low frequency sequence Q, high frequency series U；
2), low frequency sequence R, high frequency series T, low frequency sequence Q and the high frequency series U for obtaining step 1, input prediction model into
Row training obtains the parameter ω of prediction model1、b1、c1, ω2、b2、c2, ω3、b3、c3, ω4、b4、c4；
3), by input data when predicting, difference input step 2) trained prediction model progress height frequency sequence is pre-
It surveys, expression formula is：
Wherein, ω1、b1、c1, ω2、b2、c2, ω3、b3、c3, ω4、b4、c4, it is to be obtained in claim 1 by step 2 training
Prediction model parameters, xRiFor the low frequency sequence that the historical data before future position is obtained by DB3 wavelet decompositions, unit V/m,
xTiFor the high frequency series that the historical data before future position is obtained by DB3 wavelet decompositions, unit V/m, xQiFor future position it
The low frequency sequence that preceding historical data is obtained by COIF3 wavelet decompositions, unit V/m, xUiFor the history number before future position
According to the high frequency series obtained by COIF3 wavelet decompositions, unit V/m, yRFor low frequency sequence xRiPredicted value, unit V/m,
yTFor high frequency series xTiPredicted value, unit V/m, yQFor low frequency sequence xQiPredicted value, unit V/m, yUFor high frequency sequence
Arrange xUiPredicted value, unit V/m；
4), the low-and high-frequency predicted value obtained according to step 3 training does combined prediction, and combined prediction expression formula is：
yDB3=yR+yT (5)
yCOIF3=yQ+yU (6)
Y=(yDB3+yCOIF3)/2 (7)
Wherein, yDB3For by DB3 wavelet decompositions, Decomposition order is 1 low frequency sequence prediction value yRWith high frequency series predicted value yT
The predicted value that summation obtains, unit V/m, yCOIF3For by COIF3 wavelet decompositions, Decomposition order is 1 low frequency sequence prediction
Value yQWith high frequency series predicted value yUSum obtained predicted value, unit V/m, y be proposition model combined prediction value, singly
Position is V/m.
2. a kind of TD-SCDMA base station electromagnetic radiation Forecasting Methodology as described in claim 1, in the step 2), feature exists
In prediction model is as shown below：
In above formula (8), n historical data is passed through into DB3 wavelet decompositions, Decomposition order 1, the low frequency sequence R inputs of acquisition
Prediction model is trained and obtains optimum prediction model parameter ω1、b1、c1, by parameter ω1、b1Be assigned a value of between -1 to 1 with
Machine number, then training data input training pattern is obtained into parameter c1Value, model parameter ω is determined with this1、b1、c1；
In above formula (9), n historical data is passed through into DB3 wavelet decompositions, Decomposition order 1, the high frequency series T inputs of acquisition
Prediction model is trained and obtains optimum prediction model parameter ω2、b2、c2, by parameter ω2、b2Be assigned a value of between -1 to 1 with
Machine number, then training data input training pattern is obtained into parameter c2Value, model parameter ω is determined with this2、b2、c2；
In above formula (10), by n historical data by COIF3 wavelet decompositions, Decomposition order 1, the low frequency sequence Q of acquisition is defeated
Enter prediction model, be trained and obtain optimum prediction model parameter ω3、b3、c3, by parameter ω3、b3It is assigned a value of between -1 to 1
Random number, then training data input training pattern is obtained into parameter c3Value, model parameter ω is determined with this3、b3、c3；
In above formula (11), by n historical data by COIF3 wavelet decompositions, Decomposition order 1, the high frequency series U of acquisition is defeated
Enter prediction model, be trained and obtain optimum prediction model parameter ω4、b4、c4, by parameter ω4、b4It is assigned a value of between -1 to 1
Random number, then training data input training pattern is obtained into parameter c4Value, model parameter ω is determined with this4、b4、c4。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810071037.1A CN108259099B (en) | 2018-01-25 | 2018-01-25 | Electromagnetic radiation prediction method for TD-SCDMA base station |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810071037.1A CN108259099B (en) | 2018-01-25 | 2018-01-25 | Electromagnetic radiation prediction method for TD-SCDMA base station |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108259099A true CN108259099A (en) | 2018-07-06 |
| CN108259099B CN108259099B (en) | 2021-04-23 |
Family
ID=62742692
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810071037.1A Active CN108259099B (en) | 2018-01-25 | 2018-01-25 | Electromagnetic radiation prediction method for TD-SCDMA base station |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108259099B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101819253A (en) * | 2010-04-20 | 2010-09-01 | 湖南大学 | Probabilistic neural network-based tolerance-circuit fault diagnosis method |
| CN102056183A (en) * | 2010-12-10 | 2011-05-11 | 北京交通大学 | Network flow prediction method and device based on cognitive network |
| CN104408529A (en) * | 2014-11-21 | 2015-03-11 | 广东工业大学 | Short-term load predicting method of power grid |
| CN104992248A (en) * | 2015-07-07 | 2015-10-21 | 中山大学 | Microgrid photovoltaic power station generating capacity combined forecasting method |
| CN107292462A (en) * | 2017-08-25 | 2017-10-24 | 广东工业大学 | A kind of short-term load forecasting method, apparatus and system |
| CN107294623A (en) * | 2017-06-20 | 2017-10-24 | 湘潭大学 | A kind of Novel Communication base station electromagnetic radiation Forecasting Methodology |
| CN107545321A (en) * | 2017-07-25 | 2018-01-05 | 东南大学 | A kind of ARMA RBF by-product gas generating capacity combination forecasting methods based on wavelet transformation |
-
2018
- 2018-01-25 CN CN201810071037.1A patent/CN108259099B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101819253A (en) * | 2010-04-20 | 2010-09-01 | 湖南大学 | Probabilistic neural network-based tolerance-circuit fault diagnosis method |
| CN102056183A (en) * | 2010-12-10 | 2011-05-11 | 北京交通大学 | Network flow prediction method and device based on cognitive network |
| CN104408529A (en) * | 2014-11-21 | 2015-03-11 | 广东工业大学 | Short-term load predicting method of power grid |
| CN104992248A (en) * | 2015-07-07 | 2015-10-21 | 中山大学 | Microgrid photovoltaic power station generating capacity combined forecasting method |
| CN107294623A (en) * | 2017-06-20 | 2017-10-24 | 湘潭大学 | A kind of Novel Communication base station electromagnetic radiation Forecasting Methodology |
| CN107545321A (en) * | 2017-07-25 | 2018-01-05 | 东南大学 | A kind of ARMA RBF by-product gas generating capacity combination forecasting methods based on wavelet transformation |
| CN107292462A (en) * | 2017-08-25 | 2017-10-24 | 广东工业大学 | A kind of short-term load forecasting method, apparatus and system |
Non-Patent Citations (2)
| Title |
|---|
| 王东风等: "小波分解层数及其组合分量对短期风速多步预测的影响分析", 《电力系统保护与控制》 * |
| 田中大等: "基于小波变换的风电场短期风速组合预测", 《电工技术学报》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108259099B (en) | 2021-04-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Dombry et al. | Conditional simulation of max-stable processes | |
| CN112308008B (en) | Radar radiation source individual identification method based on working mode open set of transfer learning | |
| Miyoshi et al. | A multi-scale localization approach to an ensemble Kalman filter | |
| CN104899448B (en) | A kind of self-adapting compensation method of the static localization scheme of Ensemble Kalman Filter | |
| CN105404280A (en) | Industrial process fault detection method based on autoregression dynamic hidden variable model | |
| CN105046046B (en) | A kind of Ensemble Kalman Filter localization method | |
| CN112711083B (en) | Multi-source precipitation data dynamic fusion method and system based on adaptive weight characteristics | |
| CN106019257B (en) | The interpolation method of feature when based on high-frequency ground wave radar Current Observations result sky | |
| CN109344993B (en) | River channel flood peak water level forecasting method based on conditional probability distribution | |
| CN109145251A (en) | A kind of atmospheric parameter method for solving of modified simultaneous perturbation stochastic approximation algorithm | |
| CN109239653B (en) | Multi-radiation source passive direct time difference positioning method based on subspace decomposition | |
| CN104539340A (en) | Steady direction of arrival estimation method based on sparse representation and covariance fitting | |
| CN103364368B (en) | Rapid detection method for properties of mixed crude oil | |
| Braham et al. | Coverage mapping using spatial interpolation with field measurements | |
| CN105303051A (en) | Air pollutant concentration prediction method | |
| CN103117823B (en) | Short wave channel model building method | |
| CN108183754A (en) | A kind of GSM base stations electromagnetic radiation Forecasting Methodology | |
| CN106682782A (en) | Short-term photovoltaic power prediction method based on EWT-KMPMR (empirical wavelet transform and kernel minimax probability machine classification) | |
| CN108259099A (en) | A kind of TD-SCDMA base station electromagnetic radiation Forecasting Methodology | |
| CN108362951A (en) | A kind of base station electromagnetic radiation Interval evaluation method | |
| CN104008292B (en) | Broad-band antenna super-broadband electromagnetic impulse response prediction method | |
| Lin et al. | Short-Term Fine-Grained Regional MUF Prediction for HF Communication Based on Time Series Decomposition | |
| CN102651071B (en) | Support vector machine-based cabin interior path loss prediction method | |
| CN107528312B (en) | Power system state estimation method | |
| CN108199792A (en) | A kind of WCDMA base stations electromagnetic radiation Forecasting Methodology |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |