CN109902874A - A kind of micro-capacitance sensor photovoltaic power generation short term prediction method based on deep learning - Google Patents
A kind of micro-capacitance sensor photovoltaic power generation short term prediction method based on deep learning Download PDFInfo
- Publication number
- CN109902874A CN109902874A CN201910151971.9A CN201910151971A CN109902874A CN 109902874 A CN109902874 A CN 109902874A CN 201910151971 A CN201910151971 A CN 201910151971A CN 109902874 A CN109902874 A CN 109902874A
- Authority
- CN
- China
- Prior art keywords
- data
- prediction
- power generation
- photovoltaic power
- lstm
- 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
- 238000010248 power generation Methods 0.000 title claims abstract description 42
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000013135 deep learning Methods 0.000 title claims abstract description 15
- 238000012549 training Methods 0.000 claims abstract description 23
- 230000001932 seasonal effect Effects 0.000 claims abstract description 6
- 239000013589 supplement Substances 0.000 claims abstract description 6
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 5
- 230000006870 function Effects 0.000 claims description 6
- 230000007423 decrease Effects 0.000 claims description 5
- 241001269238 Data Species 0.000 claims description 4
- 238000012360 testing method Methods 0.000 claims description 3
- 239000002421 finishing Substances 0.000 claims description 2
- 238000012423 maintenance Methods 0.000 claims description 2
- 238000010606 normalization Methods 0.000 claims description 2
- 238000012545 processing Methods 0.000 claims description 2
- 238000011160 research Methods 0.000 abstract description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 238000004146 energy storage Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 description 1
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000003466 anti-cipated effect Effects 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 229910052799 carbon Inorganic materials 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000012567 pattern recognition method Methods 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A30/00—Adapting or protecting infrastructure or their operation
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Photovoltaic Devices (AREA)
Abstract
The present invention relates to micro-capacitance sensor photovoltaic generating system research fields, more particularly, to a kind of micro-capacitance sensor photovoltaic power generation short term prediction method based on deep learning, it the steps include: 1, data prediction, supplement amendment is carried out to bad data, the missing data in historical data, is then normalized to form training sample data；2, processed sample data is divided into the sample set of Various Seasonal, different broad sense weather patterns, inputs LSTM prediction model as training sample；3, LSTM Short-term Forecasting Model is established according to weather pattern and complete to train；4, the season according to belonging to prediction day and weather pattern forecast information select corresponding LSTM to predict submodel, export the photovoltaic power generation short-term forecast value of future time instance.The present invention is deeply excavated the power producing characteristics of photovoltaic power generation and is established prediction model using deep learning algorithm, is not needed valuable meteorologic survey equipment and instrument, is reduced costs, and has the wider scope of application.
Description
Technical field
The present invention relates to micro-capacitance sensor photovoltaic generating system research field, specifically a kind of micro-capacitance sensor light based on deep learning
Volt power generation short term prediction method.
Background technique
Solar energy is a kind of inexhaustible, zero carbon emission new energy, and photovoltaic power generation is mainly utilized as it
One of mode has obtained sufficient development.With continuous soaring, the stochastic volatility of photovoltaic power generation of Photovoltaic generation installed capacity
And its large-scale grid integration make dispatching of power netwoks need to match energy storage device or using other adjustable units to its into
Row peak regulation is just able to achieve the control to photovoltaic power generation, otherwise can only abandon light.Therefore how to guarantee power grid security economical operation
Under the conditions of dissolve photovoltaic generation power to the maximum extent and be particularly important.Accurate photovoltaic power generation power prediction technology is to improve
It dissolves horizontal one of effective way.
Micro-capacitance sensor is mainly integrated by distributed generation unit, energy conversion unit, energy-storage units and subscriber unit etc., can be real
The power-balance in current situation portion and optimization.The dispersion of micro-capacitance sensor photovoltaic generating system accesses power distribution network, is for dispatching of power netwoks personnel
" invisible ", function is equivalent to load for power transmission network.The economic load dispatching operation of micro-capacitance sensor needs synthesis can be again
Various information such as energy forecast information, load prediction information and future market are given birth to make a plan, and traditional power grid is dispatched
Method only considered load prediction, to be unable to satisfy the safe and economical operation requirement of power grid.Therefore, accurate photovoltaic power generation
Prediction can provide richer reliable foundation, be the key technology for improving power grid consumption distributed photovoltaic power generation.Carrying out light
Before volt power generation prediction, also needs to analyse in depth the output characteristics that photovoltaic is contributed, accurately hold photovoltaic output power
Uncertainty, for microgrid energy management and control technical support is provided.
Summary of the invention
The present invention proposes a kind of micro-capacitance sensor photovoltaic power generation short term prediction method based on LSTM network, establishes and is suitable for moving
The prediction model of state timing.Photovoltaic is analyzed first to contribute itself periodicity and temporal characteristics；Describe energy in deep learning
Handle the working principle of the LSTM network of long Dependency Specification；And the input/output variable and hidden layer of LSTM prediction model is determined
Number of nodes；Training sample data are pre-processed using the repairing of transverse and longitudinal similar day and method for normalizing；Finally based on A,
B, tetra- class broad sense weather pattern of C, D establishes prediction submodel, realizes to future time instance micro-capacitance sensor photovoltaic power generation short-term forecast；Emulation
The result shows that LSTM prediction model precision of prediction with higher.
Above-mentioned technical problem of the invention is mainly addressed by following technical proposals:
A kind of micro-capacitance sensor photovoltaic power generation short term prediction method based on deep learning, comprising the following steps:
Step 1, data prediction carries out supplement amendment to bad data, the missing data in historical data, then carries out
Normalized forms training sample data；
Step 2, processed sample data is divided into the sample set of Various Seasonal, different broad sense weather patterns, input
LSTM prediction model is as training sample；
Step 3, LSTM Short-term Forecasting Model is established according to weather pattern and complete to train；
Step 4, the season according to belonging to prediction day and weather pattern forecast information select corresponding LSTM to predict submodel, defeated
The photovoltaic power generation short-term forecast value of future time instance out.
Supplement amendment is carried out to bad data, the missing data in historical data in the step 1, is then normalized
The process that processing forms training sample data is as follows:
Remove improper data as because interim accident or gadget maintenance caused by generated energy be zero historical data；Hair
The extraordinary data of electricity；The data being lost.It is handled using across comparison method: when photovoltaic power generation data exception some day,
Searching and the consistent similar day of its weather pattern, go out force data with the photovoltaic of similar day and carry out generation in a period of time near it
It replaces.
It is handled using mean-standard deviation (Z-score) method for normalizing, function formula are as follows:
In formula: x is the input data value before normalizing, and μ is the mean value of all input datas, and σ is all input datas
Standard deviation, x be normalization after input data.
Processed sample data is divided into the sample set of Various Seasonal, different broad sense weather patterns by the step 2, defeated
Enter process of the LSTM prediction model as training sample are as follows:
Using state of weather mode identification method, by 33 kinds of weather pattern classified finishings of China Meteorological Administration's subdivision to A, B,
C, D4 kind broad sense weather pattern, and prediction submodel is established respectively.
The step 3 establishes LSTM Short-term Forecasting Model according to weather pattern and to complete trained process as follows:
Step 3.1, parameter such as input layer, hidden layer, the output layer number of nodes for determining LSTM prediction model.Hidden layer node
Number chooses the complexity that excessively will increase network model, and the training time is elongated, and over-fitting is caused to close some non-features
System remembers and learns；Number of nodes chose the connection between few then excessively extensive input/output variable, can not find sample data
Rule.Rule of thumb formula obtains hidden layer node initial setting to the present invention, selects by test of many times optimal.Commonly
Empirical equation has:
M=log2n (3)
In formula: m is the number of hidden nodes, and n is input layer number, and l is output layer number of nodes, and α is normal between 1~10
Number.Increase and decrease after acquisition node initial value and on this basis, after to taking the network of different number of nodes to be trained emulation, choosing
Take network output error minimum and frequency of training it is few be used as optimum network model.
Step 3.2, random initializtion network weight and biasing determine the number of iterations N in network training process and every time
The weight of iteration adjusts ratio；
Step 3.3, according to error function (model output value otWith real output value ytFunction) calculate prediction error E, adopt
Gradient of the accumulated error of whole network relative to weighting parameter U, V, W, root are calculated with the error backpropagation algorithm along the time
It updates its weight according to gradient decline adjustment to predict that error is minimum, expression formula is as follows:
Et(o, y)=ot-yt (5)
Network weight iterates until error reaches setting value ε or the number of iterations reaches maximum value.
The step 4 season according to belonging to prediction day and weather pattern forecast information select corresponding LSTM to predict submodule
Type, the process for exporting future time instance photovoltaic power generation short-term forecast value are as follows:
The present invention passes through trained LSTM after obtaining training sample weather pattern classification results using direct forecast methods
Prediction model directly calculates future time instance photovoltaic power generation output power.
The advantages and positive effects of the present invention are: it is pre- in short term to propose a kind of micro-capacitance sensor photovoltaic power generation based on LSTM network
Survey method establishes the prediction model suitable for dynamic time sequence.Training sample is divided by 4 class broad sense days using weather typing method
Gas type reduces the model training time.Since photovoltaic power generation data are typical Nonlinear Time Series, and there is dynamic
The output of characteristic, that is, photovoltaic generating system subsequent time is not only with current time input value in relation to also and pervious input value has one
Fixed connection, and LSTM network is added to the memory unit for storing long period information, can learn into time series long
The information that distance relies on, therefore be highly suitable for predicting fluctuation biggish photovoltaic power generation power output sequence.The present invention uses
Deep learning algorithm deeply excavates the power producing characteristics of photovoltaic power generation and establishes prediction model, does not need valuable meteorologic survey equipment
And instrument, it reduces costs, there is the wider scope of application.
Detailed description of the invention
Attached drawing 1 is short term prediction method flow chart of the invention.
Attached drawing 2 is photovoltaic power curve figure under broad sense weather pattern.
Attached drawing 3 is A class weather pattern prediction contrast curve chart.
Attached drawing 4 is B class weather pattern prediction contrast curve chart.
Attached drawing 5 is C class weather pattern prediction contrast curve chart.
Attached drawing 6 is D class weather pattern prediction contrast curve chart.
Specific embodiment
Below by example, and in conjunction with attached drawing, the technical solutions of the present invention will be further described.
Example:
A kind of micro-capacitance sensor photovoltaic power generation short term prediction method based on deep learning, as shown in Figure 1, comprising the following steps:
Step 1: data prediction carries out supplement amendment to bad data, the missing data in historical data, then carries out
Normalized forms training sample data.
Step 2: processed sample data is divided into the sample set of Various Seasonal, different broad sense weather patterns, input
LSTM prediction model is as training sample.
In this example, using 2010 of Wuhan micro-capacitance sensor photovoltaic experiment porch record, 2015~2016 years light
Volt power output real data is sample, and the installed capacity of the photo-voltaic power generation station is 10kWp, the historical data of photovoltaic test platform record
Time interval is 5min.Since photovoltaic power generation has intermittence, photovoltaic power output is zero during night, therefore herein only to the phase in daytime
Between photovoltaic power output predicted.Since photovoltaic battery panel power output is mostly zero before 6 points of Wuhan Area and after 18 points, herein
Only the photovoltaic power output of 6:00~18:00 is predicted.Basic weather pattern and broad sense are obtained according to weather pattern recognition methods
The corresponding relationship of weather pattern is as shown in table 1, and the photovoltaic power curve figure of each broad sense weather pattern is as shown in Figure 2:
1 weather pattern classification results of table
Step 3: establishing LSTM Short-term Forecasting Model according to weather pattern and complete to train.
(1) parameter such as input layer, hidden layer, the output layer number of nodes of LSTM prediction model are determined；
The present invention chooses the time series for being used to carry out photovoltaic power generation prediction as input variable, chooses prediction day at per moment
Input variable of preceding 8 sample points as model.According to the difference of output layer number of nodes, there are mainly two types of network knots for prediction model
Structure: (a) single argument exports network, and prediction only exports a variate-value every time, establishes 48 prediction models and obtains one day 48 point
Output variable；(b) multivariable exports network, and the power prediction value that output layer number of nodes exports as needed determines, pre- every time
Survey can export multiple predicted values.The present invention predicts that output valve be to predict the photovoltaic power generating value of one day day, 48 points, selection above-mentioned the
Two kinds of multivariables export network, and output layer number of nodes is 4, and output valve is to predict the photovoltaic hair of each hour of day interval 15min
Electricity trains altogether 12 different LSTM models；It is 18 by repeatedly trying to gather more last determining node in hidden layer.
(2) random initializtion network weight and biasing determine the number of iterations N in network training process and each iteration
Weight adjusts ratio；
(3) according to error function (model output value otWith real output value ytFunction) calculate prediction error E, using edge
The error backpropagation algorithm of time calculates gradient of the accumulated error of whole network relative to weighting parameter U, V, W, according to ladder
Degree decline adjustment updates its weight and to predict that error is minimum.
Step 4: select corresponding LSTM to predict submodel according to season belonging to prediction day and weather pattern forecast information, it is defeated
The photovoltaic power generation short-term forecast value of future time instance out.
On the basis of step 1 and step 2, historical data is divided into A (fine day), B (cloudy), C (rainy day) and D (yin
It) four major class are predicted respectively, present invention selection is spaced the intensity of illumination of prediction day 6:00 to 18:00 period
The short-term forecast of 15min chooses root-mean-square error (RMSE) here and mean absolute percentage error comments prediction effect
Valence:
In formula: PmeasFor measured value；PpredFor predicted value；I is time series；PcapIt is total for the photovoltaic plant booting of i period
Capacity；N is number of samples.
For the advantage and disadvantage for more fully analyzing LSTM deep learning network, increases BP neural network prediction model and carry out pair
Compare model.Fig. 3 is that the photovoltaic power generation of on November 5th, 2016 (A class weather pattern) is practical, predicts contrast curve chart；Fig. 4 is 2016
The photovoltaic power generation on November 24, in (B class weather pattern) is practical, predicts contrast curve chart；Fig. 5 is (C class on November 16th, 2016
Weather pattern) photovoltaic power generation is practical, prediction contrast curve chart；Fig. 6 is the photovoltaic of on November 28th, 2016 (D class weather pattern)
Power generation is practical, predicts contrast curve chart.Table 2 is the prediction error of prediction model.
2 LSTM of table predicts error MAPE/RMSE fiducial value
The mentioned LSTM prediction model prediction effect of the prediction result display present invention is integrally better than traditional BP nerve in this example
Network Prediction Model.Under A class weather pattern the photovoltaic power generation output power predicted value and variation tendency of two kinds of prediction models with
Measured value height of curve is coincide, though there is small-scale fluctuation error smaller, the average absolute percentage of two model of BP and LSTM
Error substantially remains within 10%；The stochastic volatility that photovoltaic is contributed under B class weather pattern is most strong, the mistake of peak value, valley
Forecast increases so that precision of prediction declines, and LSTM prediction model can be better anticipated continuous as caused by cloud cover compared with BP model
Photovoltaic in time, which is contributed, to be changed, because LSTM network has complicated memory unit, can remember the data information of historical juncture, cloud
The power output decaying of photovoltaic caused by layer has timing dependence and duration, therefore LSTM neural network forecast accuracy is higher；C, D class day
LSTM network can preferably learn compared with BP network to the intermittent spy with minor swing of photovoltaic power output under the weather pattern under gas type
Sign.
Embodiment belonging to the present invention be it is illustrative, without being restrictive, because the invention is not limited to be embodied
Embodiment described in mode, all other embodiment party obtained according to the technique and scheme of the present invention by those skilled in the art
Formula also belongs to the scope of protection of the invention.
Claims (5)
1. a kind of micro-capacitance sensor photovoltaic power generation short term prediction method based on deep learning, comprising the following steps:
Step 1, data prediction carries out supplement amendment to bad data, the missing data in historical data, then carries out normalizing
Change processing and forms training sample data；
Step 2, processed sample data is divided into the sample set of Various Seasonal, different broad sense weather patterns, inputs LSTM
Prediction model is as training sample；
Step 3, LSTM Short-term Forecasting Model is established according to weather pattern and complete to train；
Step 4, the season according to belonging to prediction day and weather pattern forecast information select corresponding LSTM to predict submodel, and output is not
Carry out the photovoltaic power generation short-term forecast value at moment.
2. the micro-capacitance sensor photovoltaic power generation short term prediction method according to claim 1 based on deep learning, it is characterised in that:
The step 1 carries out supplement amendment to bad data, the missing data in historical data, is then normalized to form instruction
Practice sample data, method particularly includes: removal improper data as because interim accident or gadget maintenance caused by generated energy be
Zero historical data；The extraordinary data of generated energy；The data being lost；It is handled using across comparison method: when photovoltaic some day is sent out
When electric data exception, searching and the consistent similar day of its weather pattern in a period of time in its vicinity, with the photovoltaic of similar day
Force data is replaced out；It is handled using mean-standard deviation (Z-score) method for normalizing, function formula are as follows:
In formula: x is the input data value before normalizing, and μ is the mean value of all input datas, and σ is the mark of all input datas
Quasi- poor, x is the input data after normalization.
3. the micro-capacitance sensor photovoltaic power generation short term prediction method according to claim 1 based on deep learning, it is characterised in that:
Processed sample data is divided into the sample set of Various Seasonal, different broad sense weather patterns by the step 2, and input LSTM is pre-
Model is surveyed as training sample；
Using state of weather mode identification method, by 33 kinds of weather pattern classified finishings of China Meteorological Administration's subdivision to A, B, C, D4
Kind broad sense weather pattern, and prediction submodel is established respectively.
4. the micro-capacitance sensor photovoltaic power generation short term prediction method according to claim 1 based on deep learning, it is characterised in that:
The step 3 establishes LSTM Short-term Forecasting Model according to weather pattern and completes to train, comprising the following steps:
Step 3.1, parameter such as input layer, hidden layer, the output layer number of nodes for determining LSTM prediction model；Node in hidden layer choosing
The complexity that excessively will increase network model is taken, the training time is elongated, and over-fitting is caused to remember some non-characteristic relations
Firmly and learn；Number of nodes chose the connection between few then excessively extensive input/output variable, can not find the rule of sample data；
Rule of thumb formula obtains hidden layer node initial setting to the present invention, selects by test of many times optimal；Common experience
Formula has:
M=log2n (3)
In formula: m is the number of hidden nodes, and n is input layer number, and l is output layer number of nodes, and α is the constant between 1~10；It obtains
Increase and decrease after node initial value and on this basis, after to taking the network of different number of nodes to be trained emulation, selection net
Network output error minimum and frequency of training it is few be used as optimum network model；
Step 3.2, random initializtion network weight and biasing determine the number of iterations N in network training process and each iteration
Weight adjust ratio；
Step 3.3, according to error function (model output value otWith real output value ytFunction) calculate prediction error E, using edge
The error backpropagation algorithm of time calculates gradient of the accumulated error of whole network relative to weighting parameter U, V, W, according to ladder
Degree decline adjustment updates its weight and to predict that error is minimum, and expression formula is as follows:
Et(o, y)=ot-yt (5)
Network weight iterates until error reaches setting value ε or the number of iterations reaches maximum value.
5. the micro-capacitance sensor photovoltaic power generation short term prediction method according to claim 1 based on deep learning, it is characterised in that:
The step 4 season according to belonging to prediction day and weather pattern forecast information select corresponding LSTM to predict submodel, and output is not
Carry out moment photovoltaic power generation short-term forecast value；The present invention uses direct forecast methods, after obtaining training sample weather pattern classification results,
Future time instance photovoltaic power generation output power is directly calculated by trained LSTM prediction model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910151971.9A CN109902874A (en) | 2019-02-28 | 2019-02-28 | A kind of micro-capacitance sensor photovoltaic power generation short term prediction method based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910151971.9A CN109902874A (en) | 2019-02-28 | 2019-02-28 | A kind of micro-capacitance sensor photovoltaic power generation short term prediction method based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109902874A true CN109902874A (en) | 2019-06-18 |
Family
ID=66945954
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910151971.9A Pending CN109902874A (en) | 2019-02-28 | 2019-02-28 | A kind of micro-capacitance sensor photovoltaic power generation short term prediction method based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109902874A (en) |
Cited By (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110377447A (en) * | 2019-07-17 | 2019-10-25 | 腾讯科技（深圳）有限公司 | A kind of abnormal deviation data examination method, device and server |
| CN110543988A (en) * | 2019-08-28 | 2019-12-06 | 上海电力大学 | Photovoltaic short-term output prediction system and method based on XGboost algorithm |
| CN110738363A (en) * | 2019-09-30 | 2020-01-31 | 华中科技大学 | photovoltaic power generation power prediction model and construction method and application thereof |
| CN110766212A (en) * | 2019-10-15 | 2020-02-07 | 哈尔滨工程大学 | Ultra-short-term photovoltaic power prediction method for historical data missing electric field |
| CN110991689A (en) * | 2019-10-17 | 2020-04-10 | 国网河南省电力公司鹤壁供电公司 | Distributed photovoltaic power generation system short-term prediction method based on LSTM-Morlet model |
| CN111092454A (en) * | 2019-11-04 | 2020-05-01 | 重庆大学 | Unit combination rapid calculation method based on characteristic scheduling points |
| CN111461297A (en) * | 2020-03-25 | 2020-07-28 | 河海大学 | Solar irradiation quantity optimization prediction algorithm based on MPC and E L M neural network |
| CN111461890A (en) * | 2020-04-21 | 2020-07-28 | 上海积成能源科技有限公司 | Distributed power generation transaction system and method based on multiple prediction and block chain technologies |
| CN111626506A (en) * | 2020-05-27 | 2020-09-04 | 华北电力大学 | Regional photovoltaic power probability prediction method based on federal learning and cooperative regulation and control system thereof |
| CN112561189A (en) * | 2020-12-23 | 2021-03-26 | 宁夏中科嘉业新能源研究院（有限公司） | Method suitable for predicting power generation capacity of photovoltaic power station |
| CN112686443A (en) * | 2020-12-30 | 2021-04-20 | 南京工程学院 | Photovoltaic power generation prediction method based on artificial intelligence |
| CN112734073A (en) * | 2019-10-28 | 2021-04-30 | 国网河北省电力有限公司 | Photovoltaic power generation short-term prediction method based on long and short-term memory network |
| CN112749904A (en) * | 2021-01-14 | 2021-05-04 | 国网湖南省电力有限公司 | Power distribution network fault risk early warning method and system based on deep learning |
| CN112862630A (en) * | 2021-03-08 | 2021-05-28 | 海南省电力学校(海南省电力技工学校) | Weather type index interval-based photovoltaic power prediction method, terminal and medium |
| CN112949210A (en) * | 2021-03-31 | 2021-06-11 | 上海电机学院 | Micro-grid renewable energy source prediction method based on UT (user input) transformation and width learning |
| CN112989706A (en) * | 2021-04-01 | 2021-06-18 | 东北大学 | Tunnel lamp illumination attenuation prediction method |
| CN113392972A (en) * | 2021-06-23 | 2021-09-14 | 广东工业大学 | Photovoltaic short-term power prediction model training method, prediction method and device |
| CN113763184A (en) * | 2021-08-26 | 2021-12-07 | 甘肃同兴智能科技发展有限责任公司 | Carbon asset assessment method |
| CN113887809A (en) * | 2021-10-11 | 2022-01-04 | 国网新疆电力有限公司巴州供电公司 | Power distribution network supply and demand balance method, system, medium and computing equipment under double-carbon target |
| CN113962441A (en) * | 2021-09-26 | 2022-01-21 | 华能澜沧江水电股份有限公司 | Short-term irradiance prediction method and prediction system based on historical data analysis |
| CN115034484A (en) * | 2022-06-17 | 2022-09-09 | 国网江苏省电力有限公司经济技术研究院 | New energy output prediction method based on deep learning |
| CN115310657A (en) * | 2022-06-22 | 2022-11-08 | 江苏润和软件股份有限公司 | Online distributed photovoltaic power generation prediction method based on lstm neural network |
| CN116245259A (en) * | 2023-05-11 | 2023-06-09 | 华能山东发电有限公司众泰电厂 | Photovoltaic power generation prediction method and device based on depth feature selection and electronic equipment |
| CN116307291A (en) * | 2023-05-22 | 2023-06-23 | 国网信息通信产业集团有限公司 | Distributed photovoltaic power generation prediction method and prediction terminal based on wavelet decomposition |
| CN117713238A (en) * | 2023-12-14 | 2024-03-15 | 国网青海省电力公司清洁能源发展研究院 | Random optimization operation strategy combining photovoltaic power generation and energy storage micro-grid |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105005825A (en) * | 2015-07-03 | 2015-10-28 | 深圳供电局有限公司 | Method and system for predicting photovoltaic power based on dynamic neural network |
| CN105426956A (en) * | 2015-11-06 | 2016-03-23 | 国家电网公司 | Ultra-short-period photovoltaic prediction method |
| CN108205717A (en) * | 2017-12-30 | 2018-06-26 | 国网江苏省电力公司无锡供电公司 | A kind of photovoltaic generation power Multiple Time Scales Forecasting Methodology |
| CN108280551A (en) * | 2018-02-02 | 2018-07-13 | 华北电力大学 | A kind of photovoltaic power generation power prediction method using shot and long term memory network |
| CN108734331A (en) * | 2018-03-23 | 2018-11-02 | 武汉理工大学 | Short-term photovoltaic power generation power prediction method based on LSTM and system |
| CN108985521A (en) * | 2018-08-02 | 2018-12-11 | 广东工业大学 | Output power predicting method, device, equipment and the medium of photovoltaic generating system |
| US20180375334A1 (en) * | 2017-06-26 | 2018-12-27 | Tsinghua University | Method and apparatus for photovoltaic power forecast based on numerical weather prediction |
-
2019
- 2019-02-28 CN CN201910151971.9A patent/CN109902874A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105005825A (en) * | 2015-07-03 | 2015-10-28 | 深圳供电局有限公司 | Method and system for predicting photovoltaic power based on dynamic neural network |
| CN105426956A (en) * | 2015-11-06 | 2016-03-23 | 国家电网公司 | Ultra-short-period photovoltaic prediction method |
| US20180375334A1 (en) * | 2017-06-26 | 2018-12-27 | Tsinghua University | Method and apparatus for photovoltaic power forecast based on numerical weather prediction |
| CN108205717A (en) * | 2017-12-30 | 2018-06-26 | 国网江苏省电力公司无锡供电公司 | A kind of photovoltaic generation power Multiple Time Scales Forecasting Methodology |
| CN108280551A (en) * | 2018-02-02 | 2018-07-13 | 华北电力大学 | A kind of photovoltaic power generation power prediction method using shot and long term memory network |
| CN108734331A (en) * | 2018-03-23 | 2018-11-02 | 武汉理工大学 | Short-term photovoltaic power generation power prediction method based on LSTM and system |
| CN108985521A (en) * | 2018-08-02 | 2018-12-11 | 广东工业大学 | Output power predicting method, device, equipment and the medium of photovoltaic generating system |
Non-Patent Citations (6)
| Title |
|---|
| ANDRÉ GENSLER;JANOSCH HENZE;BERNHARD SICK;NILS RAABE: "Deep Learning for solar power forecasting — An approach using AutoEncoder and LSTM Neural Networks", 《2016 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN, AND CYBERNETICS (SMC)》, 12 October 2016 (2016-10-12), pages 2858 - 2865, XP055559556, DOI: 10.1109/SMC.2016.7844673 * |
| 史佳琪等: "基于深度学习的超短期光伏精细化预测模型研究", 《电力建设》 * |
| 史佳琪等: "基于深度学习的超短期光伏精细化预测模型研究", 《电力建设》, no. 06, 1 June 2017 (2017-06-01), pages 32 - 39 * |
| 朱永强等: "最小二乘支持向量机在光伏功率预测中的应用", 《电网技术》 * |
| 朱永强等: "最小二乘支持向量机在光伏功率预测中的应用", 《电网技术》, no. 07, 5 July 2011 (2011-07-05), pages 60 - 65 * |
| 郝志峰主编,郝志峰主编；李杨，刘小兰，廖芹副主编: "《数据科学与数学建模》", 华中科技大学出版社, pages: 270 - 151 * |
Cited By (37)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110377447A (en) * | 2019-07-17 | 2019-10-25 | 腾讯科技（深圳）有限公司 | A kind of abnormal deviation data examination method, device and server |
| CN110543988A (en) * | 2019-08-28 | 2019-12-06 | 上海电力大学 | Photovoltaic short-term output prediction system and method based on XGboost algorithm |
| CN110738363A (en) * | 2019-09-30 | 2020-01-31 | 华中科技大学 | photovoltaic power generation power prediction model and construction method and application thereof |
| CN110738363B (en) * | 2019-09-30 | 2022-04-01 | 华中科技大学 | Photovoltaic power generation power prediction method |
| CN110766212A (en) * | 2019-10-15 | 2020-02-07 | 哈尔滨工程大学 | Ultra-short-term photovoltaic power prediction method for historical data missing electric field |
| CN110766212B (en) * | 2019-10-15 | 2023-01-03 | 哈尔滨工程大学 | Ultra-short-term photovoltaic power prediction method for historical data missing electric field |
| CN110991689A (en) * | 2019-10-17 | 2020-04-10 | 国网河南省电力公司鹤壁供电公司 | Distributed photovoltaic power generation system short-term prediction method based on LSTM-Morlet model |
| CN110991689B (en) * | 2019-10-17 | 2022-10-04 | 国网河南省电力公司鹤壁供电公司 | Distributed photovoltaic power generation system short-term prediction method based on LSTM-Morlet model |
| CN112734073A (en) * | 2019-10-28 | 2021-04-30 | 国网河北省电力有限公司 | Photovoltaic power generation short-term prediction method based on long and short-term memory network |
| CN111092454A (en) * | 2019-11-04 | 2020-05-01 | 重庆大学 | Unit combination rapid calculation method based on characteristic scheduling points |
| CN111092454B (en) * | 2019-11-04 | 2021-06-22 | 重庆大学 | Unit combination rapid calculation method based on characteristic scheduling points |
| CN111461297A (en) * | 2020-03-25 | 2020-07-28 | 河海大学 | Solar irradiation quantity optimization prediction algorithm based on MPC and E L M neural network |
| CN111461297B (en) * | 2020-03-25 | 2022-04-29 | 河海大学 | Solar irradiation quantity optimization prediction algorithm based on MPC and ELM neural network |
| CN111461890B (en) * | 2020-04-21 | 2024-02-02 | 上海积成能源科技有限公司 | Distributed power generation transaction system and method based on multiple prediction and blockchain technologies |
| CN111461890A (en) * | 2020-04-21 | 2020-07-28 | 上海积成能源科技有限公司 | Distributed power generation transaction system and method based on multiple prediction and block chain technologies |
| CN111626506A (en) * | 2020-05-27 | 2020-09-04 | 华北电力大学 | Regional photovoltaic power probability prediction method based on federal learning and cooperative regulation and control system thereof |
| CN112561189A (en) * | 2020-12-23 | 2021-03-26 | 宁夏中科嘉业新能源研究院（有限公司） | Method suitable for predicting power generation capacity of photovoltaic power station |
| CN112686443A (en) * | 2020-12-30 | 2021-04-20 | 南京工程学院 | Photovoltaic power generation prediction method based on artificial intelligence |
| CN112749904A (en) * | 2021-01-14 | 2021-05-04 | 国网湖南省电力有限公司 | Power distribution network fault risk early warning method and system based on deep learning |
| CN112749904B (en) * | 2021-01-14 | 2023-06-27 | 国网湖南省电力有限公司 | Power distribution network fault risk early warning method and system based on deep learning |
| CN112862630B (en) * | 2021-03-08 | 2024-03-22 | 海南省电力学校(海南省电力技工学校) | Photovoltaic power prediction method, terminal and medium based on weather type index interval |
| CN112862630A (en) * | 2021-03-08 | 2021-05-28 | 海南省电力学校(海南省电力技工学校) | Weather type index interval-based photovoltaic power prediction method, terminal and medium |
| CN112949210A (en) * | 2021-03-31 | 2021-06-11 | 上海电机学院 | Micro-grid renewable energy source prediction method based on UT (user input) transformation and width learning |
| CN112989706A (en) * | 2021-04-01 | 2021-06-18 | 东北大学 | Tunnel lamp illumination attenuation prediction method |
| CN112989706B (en) * | 2021-04-01 | 2023-09-29 | 东北大学 | Tunnel lamp illumination attenuation prediction method |
| CN113392972A (en) * | 2021-06-23 | 2021-09-14 | 广东工业大学 | Photovoltaic short-term power prediction model training method, prediction method and device |
| CN113763184A (en) * | 2021-08-26 | 2021-12-07 | 甘肃同兴智能科技发展有限责任公司 | Carbon asset assessment method |
| CN113962441A (en) * | 2021-09-26 | 2022-01-21 | 华能澜沧江水电股份有限公司 | Short-term irradiance prediction method and prediction system based on historical data analysis |
| CN113887809A (en) * | 2021-10-11 | 2022-01-04 | 国网新疆电力有限公司巴州供电公司 | Power distribution network supply and demand balance method, system, medium and computing equipment under double-carbon target |
| CN115034484A (en) * | 2022-06-17 | 2022-09-09 | 国网江苏省电力有限公司经济技术研究院 | New energy output prediction method based on deep learning |
| CN115310657A (en) * | 2022-06-22 | 2022-11-08 | 江苏润和软件股份有限公司 | Online distributed photovoltaic power generation prediction method based on lstm neural network |
| CN116245259A (en) * | 2023-05-11 | 2023-06-09 | 华能山东发电有限公司众泰电厂 | Photovoltaic power generation prediction method and device based on depth feature selection and electronic equipment |
| CN116245259B (en) * | 2023-05-11 | 2023-10-31 | 华能山东泰丰新能源有限公司 | Photovoltaic power generation prediction method and device based on depth feature selection and electronic equipment |
| CN116307291A (en) * | 2023-05-22 | 2023-06-23 | 国网信息通信产业集团有限公司 | Distributed photovoltaic power generation prediction method and prediction terminal based on wavelet decomposition |
| CN116307291B (en) * | 2023-05-22 | 2023-08-29 | 国网信息通信产业集团有限公司 | Distributed photovoltaic power generation prediction method and prediction terminal based on wavelet decomposition |
| CN117713238A (en) * | 2023-12-14 | 2024-03-15 | 国网青海省电力公司清洁能源发展研究院 | Random optimization operation strategy combining photovoltaic power generation and energy storage micro-grid |
| CN117713238B (en) * | 2023-12-14 | 2024-05-07 | 国网青海省电力公司清洁能源发展研究院 | Random optimization operation strategy combining photovoltaic power generation and energy storage micro-grid |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109902874A (en) | A kind of micro-capacitance sensor photovoltaic power generation short term prediction method based on deep learning | |
| Li et al. | Using enhanced crow search algorithm optimization-extreme learning machine model to forecast short-term wind power | |
| CN103390116B (en) | Use the photovoltaic power station power generation power forecasting method of stepping mode | |
| CN109002915B (en) | Photovoltaic power station short-term power prediction method based on Kmeans-GRA-Elman model | |
| CN113496311A (en) | Photovoltaic power station generated power prediction method and system | |
| CN108985521A (en) | Output power predicting method, device, equipment and the medium of photovoltaic generating system | |
| CN108388956A (en) | Consider the photovoltaic power prediction technique of attenuation | |
| CN109165774A (en) | A kind of short-term photovoltaic power prediction technique | |
| Yang et al. | Day-ahead forecasting of photovoltaic output power with similar cloud space fusion based on incomplete historical data mining | |
| CN104899665A (en) | Wind power short-term prediction method | |
| CN114792156B (en) | Photovoltaic output power prediction method and system based on curve characteristic index clustering | |
| CN110334870B (en) | Photovoltaic power station short-term power prediction method based on gated cyclic unit network | |
| CN110111003A (en) | A kind of new energy typical scene construction method based on improvement FCM clustering algorithm | |
| CN105574615A (en) | Spatial correlation and genetic algorithm (GA) based wind power forecast method for wavelet-BP neural network | |
| CN109034464A (en) | A kind of method that short-term photovoltaic generating system power prediction and result are checked | |
| CN109002926A (en) | The photovoltaic power generation quantity prediction model and its construction method of a kind of high accuracy and application | |
| CN104463356A (en) | Photovoltaic power generation power prediction method based on multi-dimension information artificial neural network algorithm | |
| CN110070228A (en) | BP neural network wind speed prediction method for neuron branch evolution | |
| CN108596242A (en) | Power grid meteorology load forecasting method based on wavelet neural network and support vector machines | |
| CN104638672A (en) | Determining method of photovoltaic transmission power limit considering variable correlation | |
| CN110276472A (en) | A kind of offshore wind farm power ultra-short term prediction method based on LSTM deep learning network | |
| CN113962441A (en) | Short-term irradiance prediction method and prediction system based on historical data analysis | |
| Lodge et al. | Short term wind speed prediction using artificial neural networks | |
| CN110084430A (en) | A method of considering space-time characterisation design distributed photovoltaic power output prediction model | |
| Luo et al. | Short-term photovoltaic generation forecasting based on similar day selection and extreme learning machine |
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: 20190618 |
|
| RJ01 | Rejection of invention patent application after publication |