CN112669169A - Short-term photovoltaic power prediction device and method - Google Patents
Short-term photovoltaic power prediction device and method Download PDFInfo
- Publication number
- CN112669169A CN112669169A CN202011479990.3A CN202011479990A CN112669169A CN 112669169 A CN112669169 A CN 112669169A CN 202011479990 A CN202011479990 A CN 202011479990A CN 112669169 A CN112669169 A CN 112669169A
- Authority
- CN
- China
- Prior art keywords
- data
- module
- real
- value
- population
- 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
- 238000000034 method Methods 0.000 title claims abstract description 48
- 230000005540 biological transmission Effects 0.000 claims abstract description 12
- 238000012806 monitoring device Methods 0.000 claims abstract description 12
- 238000013528 artificial neural network Methods 0.000 claims description 27
- 230000000739 chaotic effect Effects 0.000 claims description 20
- 238000013507 mapping Methods 0.000 claims description 18
- 238000012545 processing Methods 0.000 claims description 13
- 238000005286 illumination Methods 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 5
- 230000008569 process Effects 0.000 claims description 5
- 230000006872 improvement Effects 0.000 claims description 4
- 238000012544 monitoring process Methods 0.000 claims description 4
- 241000288673 Chiroptera Species 0.000 claims description 3
- 230000006978 adaptation Effects 0.000 claims description 3
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000012937 correction Methods 0.000 claims description 3
- 238000012804 iterative process Methods 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 3
- 238000005295 random walk Methods 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 238000012986 modification Methods 0.000 claims description 2
- 230000004048 modification Effects 0.000 claims description 2
- QENJZWZWAWWESF-UHFFFAOYSA-N tri-methylbenzoic acid Natural products CC1=CC(C)=C(C(O)=O)C=C1C QENJZWZWAWWESF-UHFFFAOYSA-N 0.000 claims 1
- 238000013480 data collection Methods 0.000 abstract description 4
- 238000010248 power generation Methods 0.000 description 16
- 230000008859 change Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 238000012795 verification Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 239000003245 coal Substances 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 239000003344 environmental pollutant Substances 0.000 description 1
- 238000007429 general method Methods 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 239000013618 particulate matter Substances 0.000 description 1
- 231100000719 pollutant Toxicity 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
- 238000010792 warming Methods 0.000 description 1
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A short-term photovoltaic power prediction device and a method belong to the technical field of power systems, and comprise a real-time meteorological data collection device, a historical database module, a power prediction module, a data transmission module, a man-machine interface and a power station monitoring device, wherein the real-time meteorological data collection device is connected with the historical database module and the power prediction module, and the power station monitoring device is connected with the data transmission module through the man-machine interface; the real-time meteorological data collecting device is used for comparing and combining the collected real-time meteorological data and meteorological data received from the meteorological station, and the like.
Description
Technical Field
The invention belongs to the technical field of power systems, and particularly relates to a short-term photovoltaic power prediction device and method.
Background
With the continuous development of society and economy, energy is reflected in the aspect of life, the demand of people for the energy is continuously increased, and the social development is restricted by the energy problem. In China, the energy mainly depends on traditional fossil energy, but the fossil energy, such as coal, generates a large amount of pollutants in the using process, pollutes air to cause greenhouse effect and causes global warming. Compared with the traditional energy sources and the renewable energy sources, the renewable energy sources have the advantages of zero energy consumption, zero particulate matter emission and zero generated pollution gas. China actively encourages large-scale development and utilization of photovoltaic power generation systems, and the photovoltaic power generation industry develops rapidly in recent years.
After 2009, the proportion of photovoltaic power stations with installed capacity at megawatt level is getting larger, and by combining weather reasons and new energy power generation characteristics, the photovoltaic power station characteristics have higher volatility and certain influence on the safety and stability of the power system. On one hand, in recent years, the phenomenon that grid-connected photovoltaic power stations are separated from a power system for many times causes huge energy loss and threatens the safety of the power system; on the other hand, the uncertainty of the output power of the photovoltaic power station brings great difficulty to grid connection of the power system. In order to ensure the safe and reliable operation of the power system, the accurate prediction of the output power of the photovoltaic power station is very important.
At present, the output power of a photovoltaic power generation system can be predicted by establishing a prediction model through the methods such as the multivariate linear regression method, the Artificial Neural Network (ANN) method, the Support Vector Machine (SVM) method, the gray theory method and the like, the photovoltaic power generation power prediction can improve the utilization efficiency of solar photovoltaic power generation resources, but the prediction methods are only based on the predictions obtained on the basis of historical same-period and same-day types, the historical same-period refers to the same historical time period such as one month, the same-day type refers to the same weather condition such as sunny days, the prediction methods have the defects that the predicted days are only classified into a certain day type, such as sunny days, the weather change and the power generation of photovoltaic power stations in the same historical period and the same-day type are considered to have similarity, but the real-time change information processing of weather elements in the predicted days is insufficient, and the local searching capability is not strong, the prediction accuracy is not accurate enough.
Disclosure of Invention
Based on the technical problems, the invention provides a short-term photovoltaic power prediction device and method based on MTMBA-BP, the method has strong local search capability, good global optimization effect can be obtained, and the prediction result is more accurate than that of a common method.
The invention adopts the following technical scheme:
a short-term photovoltaic power prediction device comprises a real-time meteorological data collection device, a historical database module, a power prediction module, a data transmission module, a man-machine interface and a power station monitoring device, wherein the real-time meteorological data receiving device is connected with the historical database module and the power prediction module; the real-time meteorological data collecting device is used for comparing and combining the collected real-time meteorological data and meteorological data received from a meteorological station, and performing dimensionality reduction screening processing on the received meteorological data with high dimensionality to obtain predicted meteorological data; the historical database module contains data including real-time and historical power, temperature and illumination intensity of the photovoltaic power station, the resolution of each data is 60 minutes, the historical database module can receive the power data of the photovoltaic power station transmitted by the power station monitoring device and the real-time meteorological data transmitted by the real-time meteorological collection device, and the historical data and the real-time data are used for comparing and analyzing errors; the power prediction module comprises a weight threshold processing module and a neural network module, wherein the weight threshold processing module updates data individuals to obtain a new position solution after batxing the weight threshold of the data by using a population position and a speed formula, the new solution corresponds to a neural network structure through the neural network module, and the weight threshold is updated again by using a neural network method to obtain a new solution.
Furthermore, the power station monitoring device monitors the power prediction module in real time, sends real-time meteorological data and prediction data of the photovoltaic power station to the historical database module, and monitors and manages the transmitted data in real time through the human-computer interface, wherein the monitored data comprise temperature, illumination intensity and real-time power of the photovoltaic power station, and the monitored data resolution is 60 minutes.
A prediction method of a short-term photovoltaic power prediction device is characterized in that a weight threshold processing module adopts a TMBA (transient matrix based architecture) improvement method based on chaotic disturbance, and the method comprises the following steps:
(1) initializing a method, randomly generating n bat individuals in a search range to form an initial population, giving initial values to the position, speed, frequency, pulse frequency and pulse sound intensity of each bat, wherein a parameter beststep represents the continuous and invariable times of the optimal individuals and is initially set to be 0;
(2) evaluating the population, calculating the adaptive value of each bat in the population, and finding out the current global optimal value;
(3) updating the population, and introducing speed weight in the iterative process
And chaotic mapping to update the frequency f of the individual respectivelyiVelocity viIn spatial position xi；
fi＝fmin+(fmax-fmin)CMi (1)
Wherein CMiAs a general term for chaotic mapping, CMi∈[0，1]For a confused number, X*Is the current local optimal solution in the population;
(4) generating a random number rand, and if rand is larger than r, obtaining a local new solution of the optimal bat individual by a formula (5) through a random walk rule;
(5) evaluating the population, namely evaluating each individual in the current population, if a certain individual is better than the global optimum, updating the bat with the global optimum, juxtaposing beststep with 0, and otherwise, updating beststep by adding 1;
(6) judging whether the variation condition, beststep reaches the maximum value maxstep or whether the variation of the global optimum value of two successive iterations is small (< eta), if yes, executing (7), otherwise, executing (8);
(7) performing variation operation, namely performing Gaussian variation on the optimal bat individuals in the current population, and performing t-distribution variation on other bats; evaluating new population fitness, if a certain bat adaptation value is better than the global optimum, replacing, setting beststep as 0, otherwise, adding 1 to beststep for updating;
(8) and (4) judging termination conditions: whether the iteration times reach the maximum value Maxgen or whether the optimal solution reaches the preset precision, if not, updating beststep by adding 1, and turning to (3) to continue executing the next generation bat optimization process, otherwise, turning to (9) to execute;
(9) the method is terminated, and the optimal solution is output;
the neural network module is a BP neural network learning module and comprises the following steps:
(1) initializing, namely firstly setting the weight and the threshold value in the module as an arbitrary value and updating;
(2) random learning input vector (X)k,Yk) Into a neural network;
(3) data is transmitted to an input layer, transmitted to a hidden layer and finally transmitted to an output layer, and the number of output nodes is the vector number obtained by the prediction model;
(4) obtaining corresponding data of hidden layer by combining the following formulas
(5) Obtaining corresponding data of output point by combining the following formulas
(6) According to the related data output by the output layer, the formula (8) is used for corresponding calculation, so that the corresponding numerical value and the correction error of each node can be obtained
(7) Calculating the error of the corresponding hidden layer node value according to the formula (9),
(8) the chaos mapping is introduced to enhance the randomness and accelerate the searching step to adjust the connecting value V between the hidden layer and the output layer and the threshold value gamma of the output layer,
in the formula:
CMias a general term for chaotic mapping, CMi∈[0，1]，
(9) Further modification is made to the weights and thresholds according to chaotic mapping,
(10) selecting an input vector, and repeatedly returning to the step (3) until all data are completely trained;
(11) checking all calculated errors to see if the criteria are met; if yes, directly turning to the step (13);
(12) adjusting the training times of the neural network model according to the requirement, and directly turning to the step (2) if the adjusted training times are less than or equal to a set value;
(13) and finishing the learning of the neural network model.
Further, the data transmission module transmits measurement data, alarm events and file data to the monitoring system, and various types of data are modeled according to the DL/T860 standard.
The invention has the advantages and effects that:
the invention relates to a short-term photovoltaic power prediction device based on a chaos disturbance improved self-adaptive t-distribution bat method and a chaos mapping improved back propagation method (MTMBA-BP), which comprises a real-time meteorological data collection device, a historical database module, a power prediction module, a data transmission module, a man-machine interface and a power station monitoring device. The power prediction module combines a TMBA method based on chaos disturbance improvement and a chaos mapping improved BP neural network learning method. And utilizing the improved population position and speed formula to bat the weight threshold value of the data, updating the data individual to obtain a new position solution, corresponding the new position solution to a neural network structure, and utilizing a neural network method to update the weight threshold value again to obtain a new solution. The whole device has stronger local searching capability, the searching step is executed at a higher speed, a good global optimization effect can be obtained, and the prediction result is more accurate than that of the general method.
Drawings
FIG. 1 is a schematic diagram of a short-term photovoltaic power prediction device based on the MTMBA-BP method;
FIG. 2 is a flow chart of a chaotic perturbation improved TMBA method;
FIG. 3 is a flow chart of a MTMBA-BP neural network learning method;
FIG. 4 is a verification comparison of TMBA-BP predicted power generation amount based on chaos disturbance improvement;
fig. 5 is a verification comparison of BP predicted power generation amount.
Detailed Description
As shown in the figure, the system comprises a real-time meteorological data collecting and receiving device, a historical database module, a power prediction module, a data transmission module, a man-machine interface and a power station monitoring device. The meteorological data receiving device is connected with the historical power generation server and the power prediction module; the power station monitoring module is connected with the data transmission module through a man-machine interface, and meanwhile, the data transmission module, the power prediction module and the historical power generation server are monitored in real time and data management is carried out.
The real-time meteorological collecting device compares and combines the collected real-time meteorological data and the meteorological data received from the meteorological station, and performs dimensionality reduction screening processing on the received meteorological data with high dimensionality to obtain more precise and accurate predicted meteorological data.
The historical database module, the data that contain in the module include photovoltaic power station real-time and historical power, temperature, illumination intensity, and the resolution ratio of each data is 60 minutes, and can also receive the real-time meteorological data that photovoltaic power station power data and real-time meteorological collection device that the power station monitoring devices conveys convey conveys to utilize historical data and real-time data to carry out the comparison and be convenient for analysis error.
The power prediction module comprises a weight threshold processing module and a neural network module. And the weight threshold processing module updates the data individuals to obtain a new position solution after batxing the weight threshold of the data by using the improved population position and speed formula method. And the new solution is corresponding to a neural network structure through a neural network module, and the weight threshold value is updated again by using a neural network method to obtain the new solution.
The weight threshold processing module adopts a chaotic disturbance improved TMBA method, and is characterized by comprising the following steps:
(1) the method is initialized. And randomly generating n bat individuals in the search range to form an initial population. Initial values are given to the position, speed, frequency, pulse frequency and pulse sound intensity of each bat. The parameter beststep represents the continuous and unchangeable times of the optimal individual, and the initial value is 0;
(2) the population was evaluated. Calculating the adaptive value of each bat in the population, and finding out the current global optimal value;
(3) and updating the population. In an iterative process, velocity weights are introduced
And chaotic mapping to update the frequency f of the individual respectivelyiVelocity viIn spatial position xi；
fi＝fmin+(fmax-fmin)CMi (1)
Wherein CMiAs a general term for chaotic mapping, CMi∈[0，1]For a confused number, X*Is the current locally optimal solution (bit) in the populationSet). In the experimental process, a corresponding frequency change interval can be set according to the requirement of a problem.
(4) A random number rand is generated. If rand is larger than r, obtaining a local new solution of the optimal bat individual by the formula (5) through a random walk rule;
(5) evaluating each individual in the current population, if a certain individual is better than the global optimum, updating the bat to be the global optimum, setting the beststep to be 0, and if not, updating the beststep by adding 1;
(6) and (5) judging variation conditions. Whether beststep reaches the maximum value maxstep or whether the change of the global optimum value is small (eta) after two successive iterations is performed (7) if yes, and otherwise, performing (8);
(7) and (5) performing mutation operation. Firstly, carrying out Gaussian variation on the optimal bat individuals in the current population, and carrying out t-distribution variation on other bats; evaluating new population fitness, if a certain bat adaptation value is better than the global optimum, replacing, setting beststep as 0, otherwise, adding 1 to beststep for updating;
(8) and (4) judging termination conditions: whether the iteration times reach the maximum value Maxgen or whether the optimal solution reaches the preset precision, if not, updating beststep by adding 1, and turning to (3) to continue executing the next generation bat optimization process, otherwise, turning to (9) to execute;
(9) and (5) ending the method and outputting the optimal solution.
The BP neural network learning method module adopts a chaotic mapping to improve a BP method, and is characterized by comprising the following steps:
(1) and (5) initializing. Firstly, the weight value and the threshold value in the module are set as any value for updating.
(2) Random learning input vector (X)k,Yk) Into a neural network.
(3) And (3) data is transmitted to the input layer, the hidden layer and the output layer, and the quantity of the output nodes is the vector quantity obtained by the prediction model.
(4) Obtaining corresponding data of hidden layer by combining the following formulas
(5) Obtaining corresponding data of output point by combining the following formulas
(6) According to the related data output by the output layer, the formula (8) is used for corresponding calculation, so that the corresponding numerical value and the correction error of each node can be obtained
(7) The error of the corresponding hidden layer node value is calculated according to the formula (9).
(8) In order to obtain the global optimal weight and threshold, a chaotic mapping enhanced randomness accelerated search step is introduced to adjust the connection weight V between the hidden layer and the output layer and the threshold gamma of the output layer.
In the formula:
CMias a general term for chaotic mapping, CMi∈[0，1]。
(9) And further correcting the weight value and the threshold value according to the chaotic mapping.
(10) And (4) selecting an input vector, and repeatedly returning to the step (3) until all data are completely trained.
(11) Checking all calculated errors to see if the criteria are met; if yes, go to step (13).
(12) And (4) adjusting the training times of the neural network model according to the requirement, and directly turning to the step (2) if the adjusted training times are less than or equal to a set value.
(13) And finishing the learning of the neural network model.
And the data transmission module transmits measurement data, alarm events and file data to the monitoring system, and various data are modeled according to the DL/T860 standard.
The power station monitoring device monitors the power prediction module in real time, sends real-time meteorological data and prediction data of the photovoltaic power station to the historical database module, monitors and manages transmitted data in real time through the human-computer interface, and is convenient for finding fault data in time. The monitored data comprises temperature, illumination intensity, real-time power of a photovoltaic power station and the like, and the resolution ratio of the monitored data is 60 minutes.
Example 1
And comparing the method provided by the invention with the traditional BP network model for verification analysis by using the electricity quantity data of a certain photovoltaic station 2019 and 6 months acquired by the electricity energy acquisition system. The daily sampling frequency is 5min per point, and after data processing, 8: 00-16: 009 time data values per day are selected, as shown in Table 1.
The simulation results of simulation prediction of the 21-month-10-21 solar photovoltaic power generation in 2019 by using the trained chaotic disturbance improved TMBA-BP model and the trained BP model are shown in fig. 4-5, and the results show that the absolute error between the predicted power generation amount and the expected power generation amount based on the chaotic disturbance improved TMBA-BP model is within 20%, and the absolute error between the predicted power generation amount and the expected power generation amount of the BP model exceeds 20% in a plurality of time periods. The fact that the TMBA-BP neural network is improved based on chaotic disturbance is shown to have higher prediction precision.
Claims (4)
1. A short-term photovoltaic power prediction apparatus, characterized by: the real-time meteorological data receiving device is connected with the historical database module and the power prediction module, the power station monitoring device is connected with the data transmission module through the human-computer interface, and meanwhile, the data transmission module, the power prediction module and the historical database module are monitored in real time and managed; the real-time meteorological data collecting device is used for comparing and combining the collected real-time meteorological data and meteorological data received from a meteorological station, and performing dimensionality reduction screening processing on the received meteorological data with high dimensionality to obtain predicted meteorological data; the historical database module contains data including real-time and historical power, temperature and illumination intensity of the photovoltaic power station, the resolution of each data is 60 minutes, the historical database module can receive the power data of the photovoltaic power station transmitted by the power station monitoring device and the real-time meteorological data transmitted by the real-time meteorological collection device, and the historical data and the real-time data are used for comparing and analyzing errors; the power prediction module comprises a weight threshold processing module and a neural network module, wherein the weight threshold processing module updates data individuals to obtain a new position solution after batxing the weight threshold of the data by using a population position and a speed formula, the new solution corresponds to a neural network structure through the neural network module, and the weight threshold is updated again by using a neural network method to obtain a new solution.
2. The short-term photovoltaic power prediction apparatus as claimed in claim 1, wherein: the power station monitoring device monitors the power prediction module in real time, sends real-time meteorological data and prediction data of the photovoltaic power station to the historical database module, and monitors and manages transmitted data in real time through the human-computer interface, the monitored data comprise temperature, illumination intensity and real-time power of the photovoltaic power station, and the monitored data resolution is 60 minutes.
3. The method of claim 1, wherein the method comprises: the weight threshold processing module adopts a chaos disturbance-based TMBA improvement method, and comprises the following steps:
(1) initializing a method, randomly generating n bat individuals in a search range to form an initial population, giving initial values to the position, speed, frequency, pulse frequency and pulse sound intensity of each bat, wherein a parameter beststep represents the continuous and invariable times of the optimal individuals and is initially set to be 0;
(2) evaluating the population, calculating the adaptive value of each bat in the population, and finding out the current global optimal value;
(3) updating the population, and introducing a speed weight omega in an iterative processi tAnd chaotic mapping to update the frequency f of the individual respectivelyiVelocity viIn spatial position xi；
fi＝fmin+(fmax-fmin)CMi (1)
Wherein CMiAs a general term for chaotic mapping, CMi∈[0，1]For a confused number, X*Is the current local optimal solution in the population;
(4) generating a random number rand, and if rand is larger than r, obtaining a local new solution of the optimal bat individual by a formula (5) through a random walk rule;
(5) evaluating the population, namely evaluating each individual in the current population, if a certain individual is better than the global optimum, updating the bat with the global optimum, juxtaposing beststep with 0, and otherwise, updating beststep by adding 1;
(6) judging whether the variation condition, beststep reaches the maximum value maxstep or whether the variation of the global optimum value of two successive iterations is small (< eta), if yes, executing (7), otherwise, executing (8);
(7) performing variation operation, namely performing Gaussian variation on the optimal bat individuals in the current population, and performing t-distribution variation on other bats; evaluating new population fitness, if a certain bat adaptation value is better than the global optimum, replacing, setting beststep as 0, otherwise, adding 1 to beststep for updating;
(8) and (4) judging termination conditions: whether the iteration times reach the maximum value Maxgen or whether the optimal solution reaches the preset precision, if not, updating beststep by adding 1, and turning to (3) to continue executing the next generation bat optimization process, otherwise, turning to (9) to execute;
(9) the method is terminated, and the optimal solution is output;
the neural network module is a BP neural network learning module and comprises the following steps:
(1) initializing, namely firstly setting the weight and the threshold value in the module as an arbitrary value and updating;
(2) random learning input vector (X)k,Yk) ToAmong the neural networks;
(3) data is transmitted to an input layer, transmitted to a hidden layer and finally transmitted to an output layer, and the number of output nodes is the vector number obtained by the prediction model;
(4) obtaining corresponding data of hidden layer by combining the following formulas
(5) Obtaining corresponding data of output point by combining the following formulas
(6) According to the related data output by the output layer, the formula (8) is used for corresponding calculation, so that the corresponding numerical value and the correction error of each node can be obtained
(7) Calculating the error of the corresponding hidden layer node value according to the formula (9),
(8) the chaos mapping is introduced to enhance the randomness and accelerate the searching step to adjust the connecting value V between the hidden layer and the output layer and the threshold value gamma of the output layer,
in the formula:
CMias a general term for chaotic mapping, CMi∈[0，1]，
(9) Further modification is made to the weights and thresholds according to chaotic mapping,
(10) selecting an input vector, and repeatedly returning to the step (3) until all data are completely trained;
(11) checking all calculated errors to see if the criteria are met; if yes, directly turning to the step (13);
(12) adjusting the training times of the neural network model according to the requirement, and directly turning to the step (2) if the adjusted training times are less than or equal to a set value;
(13) and finishing the learning of the neural network model.
4. The method of claim 3, wherein the method comprises: the data transmission module transmits measurement data, alarm events and file data to the monitoring system, and various data are modeled according to the DL/T860 standard.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011479990.3A CN112669169B (en) | 2020-12-15 | 2020-12-15 | Short-term photovoltaic power prediction device and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011479990.3A CN112669169B (en) | 2020-12-15 | 2020-12-15 | Short-term photovoltaic power prediction device and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112669169A true CN112669169A (en) | 2021-04-16 |
| CN112669169B CN112669169B (en) | 2024-04-30 |
Family
ID=75404902
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011479990.3A Active CN112669169B (en) | 2020-12-15 | 2020-12-15 | Short-term photovoltaic power prediction device and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112669169B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018072351A1 (en) * | 2016-10-20 | 2018-04-26 | 北京工业大学 | Method for optimizing support vector machine on basis of particle swarm optimization algorithm |
| CN108765951A (en) * | 2018-06-11 | 2018-11-06 | 广东工业大学 | Method for identifying traffic status of express way based on bat algorithm support vector machines |
| CN109344990A (en) * | 2018-08-02 | 2019-02-15 | 中国电力科学研究院有限公司 | A kind of short-term load forecasting method and system based on DFS and SVM feature selecting |
| CN109426889A (en) * | 2017-09-01 | 2019-03-05 | 南京理工大学 | Short-term load forecasting method based on KPCA in conjunction with improvement neural network |
| CN110852902A (en) * | 2019-11-14 | 2020-02-28 | 合肥工业大学 | Photovoltaic power generation power prediction method based on BAS-BP |
| WO2020063690A1 (en) * | 2018-09-25 | 2020-04-02 | 新智数字科技有限公司 | Electrical power system prediction method and apparatus |
| CN111260116A (en) * | 2020-01-10 | 2020-06-09 | 河南理工大学 | Time-interval refined short-term load prediction method based on BOA-SVR and fuzzy clustering |
| CN112052913A (en) * | 2020-09-27 | 2020-12-08 | 国网江苏省电力有限公司南京供电分公司 | Distributed photovoltaic power station power data virtual acquisition method |
-
2020
- 2020-12-15 CN CN202011479990.3A patent/CN112669169B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018072351A1 (en) * | 2016-10-20 | 2018-04-26 | 北京工业大学 | Method for optimizing support vector machine on basis of particle swarm optimization algorithm |
| CN109426889A (en) * | 2017-09-01 | 2019-03-05 | 南京理工大学 | Short-term load forecasting method based on KPCA in conjunction with improvement neural network |
| CN108765951A (en) * | 2018-06-11 | 2018-11-06 | 广东工业大学 | Method for identifying traffic status of express way based on bat algorithm support vector machines |
| CN109344990A (en) * | 2018-08-02 | 2019-02-15 | 中国电力科学研究院有限公司 | A kind of short-term load forecasting method and system based on DFS and SVM feature selecting |
| WO2020063690A1 (en) * | 2018-09-25 | 2020-04-02 | 新智数字科技有限公司 | Electrical power system prediction method and apparatus |
| CN110852902A (en) * | 2019-11-14 | 2020-02-28 | 合肥工业大学 | Photovoltaic power generation power prediction method based on BAS-BP |
| CN111260116A (en) * | 2020-01-10 | 2020-06-09 | 河南理工大学 | Time-interval refined short-term load prediction method based on BOA-SVR and fuzzy clustering |
| CN112052913A (en) * | 2020-09-27 | 2020-12-08 | 国网江苏省电力有限公司南京供电分公司 | Distributed photovoltaic power station power data virtual acquisition method |
Non-Patent Citations (3)
| Title |
|---|
| 吴云，雷建文，鲍丽山，李春哲: "基于改进灰色关联分析与蝙蝠优化神经网络的短期负荷预测", 电力系统自动化, vol. 42, no. 20, pages 1 - 8 * |
| 许德刚，赵萍: "蝙蝠算法研究及应用综述", 计算机工程与应用, vol. 55, no. 15 * |
| 许童羽，马艺铭, 曹英丽, 唐瑞，陈俊杰: "基于主成分分析和遗传优化BP 神经网络的光伏输出功率短期预测", 电力系统保护与控制, vol. 44, no. 22, pages 1 - 6 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112669169B (en) | 2024-04-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111626506B (en) | Regional photovoltaic power probability prediction method based on federal learning and cooperative regulation and control system thereof | |
| Yadav et al. | Short-term PV power forecasting using empirical mode decomposition in integration with back-propagation neural network | |
| CN110334870B (en) | Photovoltaic power station short-term power prediction method based on gated cyclic unit network | |
| CN110263915A (en) | A kind of wind power forecasting method based on deepness belief network | |
| CN109242136A (en) | A kind of micro-capacitance sensor wind power Chaos-Genetic-BP neural network prediction technique | |
| Wang et al. | Hybrid deep learning model for short-term wind speed forecasting based on time series decomposition and gated recurrent unit | |
| CN109255728A (en) | The photovoltaic generation power neural network prediction method of chaos phase space optimal reconfiguration | |
| Luo et al. | Short-term photovoltaic generation forecasting based on similar day selection and extreme learning machine | |
| Chen et al. | Research on wind power prediction method based on convolutional neural network and genetic algorithm | |
| Wang et al. | Ultra-short-term wind power forecasting based on deep belief network | |
| CN103886223B (en) | Power forecasting method and system | |
| Wu et al. | Combined IXGBoost-KELM short-term photovoltaic power prediction model based on multidimensional similar day clustering and dual decomposition | |
| Sunny et al. | Design and simulation of maximum power point tracking of photovoltaic system using ANN | |
| Zahraoui et al. | ANN-LSTM Based Tool For Photovoltaic Power Forecasting. | |
| Wang et al. | A novel structure adaptive grey seasonal model with data reorganization and its application in solar photovoltaic power generation prediction | |
| CN112669169B (en) | Short-term photovoltaic power prediction device and method | |
| CN116128211A (en) | Wind-light-water combined short-term optimization scheduling method based on wind-light uncertainty prediction scene | |
| Kehe et al. | Research of wind power prediction model based on RBF neural network | |
| Wang et al. | Wind power output prediction: a comparative study of extreme learning machine | |
| Ding et al. | Photovoltaic array power prediction model based on EEMD and PSO-KELM | |
| Sulaiman et al. | Optimization of an Artificial Neural Network using Firefly Algorithm for modeling AC power from a photovoltaic system | |
| Zim et al. | Short-term Weather Forecasting for Wind Energy Generation using A Deep Learning Technique | |
| Jabari et al. | Long-term solar irradiance forecasting using feed-forward back-propagation neural network | |
| CN109165792B (en) | Photovoltaic short-term output power prediction method based on SOA-WNN | |
| Zhou et al. | Short-term power forecast of wind power generation based on genetic algorithm optimized neural network |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20211228 Address after: 123000 No. 53 Jiefang Avenue, Haizhou District, Liaoning, Fuxin Applicant after: STATE GRID FUXIN ELECTRIC POWER SUPPLY Co. Applicant after: STATE GRID CORPORATION OF CHINA Address before: 123200 No.53 Jiefang street, Haizhou District, Fuxin City, Liaoning Province Applicant before: STATE GRID FUXIN ELECTRIC POWER SUPPLY Co. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |