CN114298377A - Photovoltaic power generation prediction method based on improved extreme learning machine - Google Patents
Photovoltaic power generation prediction method based on improved extreme learning machine Download PDFInfo
- Publication number
- CN114298377A CN114298377A CN202111459373.1A CN202111459373A CN114298377A CN 114298377 A CN114298377 A CN 114298377A CN 202111459373 A CN202111459373 A CN 202111459373A CN 114298377 A CN114298377 A CN 114298377A
- Authority
- CN
- China
- Prior art keywords
- power generation
- photovoltaic power
- elm
- learning machine
- extreme learning
- 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
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Photovoltaic Devices (AREA)
Abstract
The invention relates to a photovoltaic power generation prediction method based on an improved extreme learning machine, which comprises the following steps: 1) acquiring photovoltaic power generation historical data, preprocessing the photovoltaic power generation historical data, and acquiring main factors influencing photovoltaic output through correlation analysis; 2) constructing an ELM prediction model and training; 3) optimizing ELM prediction model parameters by adopting a genetic algorithm; 4) and (4) applying the ELM prediction model after optimization training to predict the photovoltaic power generation. Compared with the prior art, the method has the advantages of improving the prediction precision, reducing the redundancy, improving the network training efficiency, avoiding the greater influence of the initialized weight and threshold on the training result, and the like.
Description
Technical Field
The invention relates to the field of photovoltaic power generation prediction, in particular to a photovoltaic power generation prediction method based on an improved Extreme Learning Machine (ELM).
Background
With the implementation of the national 'push policy of the whole county of distributed photovoltaic power stations', the distributed photovoltaic installation machine shows explosive growth. By 2020, the photovoltaic accumulated grid-connected machine loading amount of China reaches 253GW, and the total photovoltaic power generation installed capacity is expected to exceed 1000GW by 2030 years in combination with the current 'double-carbon' target. However, the photovoltaic power generation capability is greatly dependent on weather conditions, and has intermittency and volatility, for example, solar irradiance is a key factor influencing the power output of photovoltaic power generation, and factors such as temperature and humidity also influence the performance of the photovoltaic power generation capability. Due to the characteristics, large-scale access of distributed photovoltaic will bring huge challenges to the aspects of power system scheduling management, consumption level, economic operation and the like.
The accuracy of photovoltaic output prediction is improved, the photovoltaic output characteristics can be obtained in advance, the operation mode and the response measures of the power grid are reasonably arranged, and the safety, the reliability and the economy of the power grid are improved. However, in recent years of research, although improvement of the photovoltaic prediction model improves the photovoltaic power prediction accuracy, a part of parameters are randomly allocated in a traditional neural network algorithm including the ELM, which may cause a certain degree of error and fluctuation to the prediction result, resulting in inaccurate prediction.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provide a photovoltaic power generation prediction method based on an improved extreme learning machine, and the accuracy of a photovoltaic power generation prediction model is improved in a mode of combining an optimized PCC and GA algorithm with an ELM model.
The purpose of the invention can be realized by the following technical scheme:
a photovoltaic power generation prediction method based on an improved extreme learning machine comprises the following steps:
1) acquiring photovoltaic power generation historical data, preprocessing the photovoltaic power generation historical data, and acquiring main factors influencing photovoltaic output through correlation analysis;
2) constructing an ELM prediction model and training;
3) optimizing ELM prediction model parameters by adopting a genetic algorithm;
4) and (4) applying the ELM prediction model after optimization training to predict the photovoltaic power generation.
In the step 1), the photovoltaic power generation historical data is normalized, and the following steps are performed:
wherein p isjIs the original photovoltaic power generation power data of the j (th), pj，minIs the minimum value of j columns of data, pj，maxIs the maximum value of the j columns of data,
the normalized photovoltaic power generation power data is obtained.
In the step 1), correlation analysis is performed on the influence factors influencing the photovoltaic output and the actual photovoltaic power generation power, a PCC correlation coefficient between each influence factor and the actual photovoltaic power generation power is calculated, and main factors are selected according to the PCC correlation coefficient.
The first k influencing factors with larger absolute value of PCC correlation coefficients are selected as main factors.
The influence factors influencing the photovoltaic output comprise irradiance, temperature, wind speed, wind direction, humidity and pressure intensity, and the selected main factors specifically comprise irradiance, temperature and humidity.
The step 2) specifically comprises the following steps:
21) construction of training sample [ x ] for prediction of photovoltaic power generationi，pi]Wherein x isi＝[xi1，xi2，...，xiN]T∈RNFor the ith one influencing the photovoltaic contributionData vector of essential factor, N is number of training samples, piE, taking the R as a sample output value, namely the photovoltaic power generation power;
22) an ELM prediction model composed of an input layer, a hidden layer and an output layer is constructed, and the expression is as follows:
wherein, betaiIs the connection weight of the hidden layer and the output layer, g (-) is the activation function of the hidden layer neuron, omegaiAs a connection weight between input layer and hidden layer neurons, biIs the threshold value of hidden layer neuron, h is the number of hidden layer neuron;
23) determining the number h of hidden layer neurons and randomly generating a connection weight ωiAnd a threshold value bi；
24) And selecting a Sigmoid function as an activation function of the hidden layer neuron, and training the model.
In step 23), determining the number h of hidden layer neurons based on an empirical formula includes:
wherein m is the number of neurons in an input layer, n is the number of neurons in an output layer, and a is a constant between 1 and 10.
The step 3) specifically comprises the following steps:
31) initializing, and setting parameters of a genetic algorithm, including a population number K, a maximum iteration number, a cross probability and a variation probability;
32) the sum of the absolute values of the errors predicted by the photovoltaic power training data is used as an individual fitness value;
33) binary coding is carried out on the connection weight value and the threshold value randomly generated by the ELM prediction model, a coding chain is constructed, and an initial population is generated;
34) training the population, calculating the fitness of each population, and performing selection, crossing and variation operations on the population to obtain a new population;
35) and judging whether the maximum iteration times is reached, if so, outputting the optimized connection weight value and the threshold value, otherwise, returning to the step 34).
In the step 33), the coding length l of the genetic algorithm individual is the sum of the weight number and the threshold number.
The method also comprises the step of carrying out performance evaluation on the ELM prediction model, specifically comprising the following steps:
and analyzing the prediction result of the ELM prediction model by adopting error general evaluation indexes, namely mean square error, a decision coefficient, a mean absolute error and a mean estimated deviation ratio.
Compared with the prior art, the invention has the following advantages:
firstly, screening input variables of a photovoltaic prediction model by a Pearson correlation coefficient method, and selecting meteorological factors with strong correlation with photovoltaic power generation as the input variables, so that the accuracy of photovoltaic prediction is greatly improved, and the redundancy of the model is reduced;
compared with the existing neural network model, the ELM model has no complex mathematical operation generated by excessive iteration in the negative feedback process, greatly improves the network training efficiency, and meets the requirement of a hidden layer for nonlinear operation;
and thirdly, the weight and the threshold of each layer of the ELM model are randomly generated, and the parameters randomly generated by the prediction model are optimized by adopting a genetic algorithm, so that the initialized weight and threshold are prevented from having great influence on the training result.
Drawings
FIG. 1 is a flow chart of the PCC-GA-ELM model.
FIG. 2 is a diagram of an ELM neural network.
Fig. 3 is a comparison graph of typical daily photovoltaic power generation prediction results.
Fig. 4 is a comparison graph of prediction errors of typical daily photovoltaic power generation.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments.
Examples
As shown in FIG. 1, the invention provides a process of a PCC-GA-ELM model, and the method for predicting photovoltaic power generation of an improved Extreme Learning Machine (ELM), which comprises the following steps:
step 1: the processing of the photovoltaic power generation historical data mainly comprises the following main steps:
step 1-1: the historical data normalization processing has the formula as follows:
wherein x isiThe original photovoltaic output data of the ith column; x is the number ofi，minIs the minimum value of i columns of data; x is the number ofi，maxIs the maximum value of i columns of data;
is normalized data.
Step 1-2: carrying out correlation analysis on influence factors influencing photovoltaic output and actual generated power, and calculating two random variables Xi，YiThe correlation between the two variables, i.e., the PCC correlation coefficient between the two variables, can be calculated by equation (2):
wherein:
are respectively variable XiAnd YiAverage value of (1), correlation coefficient ρX，YThe larger the absolute value, the stronger the correlation.
Step 1-3: determining a random variable: factor variable X influencing photovoltaic power generation prediction in historical data of certain photovoltaic power stationiThe method comprises the following steps: irradiance, temperature, wind speed, wind direction, humidity, and pressure; output variable YiActual output for photovoltaic;
step 1-4: determining random variablesThen, returning to the step 1-1 to carry out correlation analysis, and selecting rhoX，YThe influencing factors with larger absolute values are selected as input characteristic variables of the prediction model, and it can be seen from table 1 that irradiance, temperature and humidity are main meteorological factors influencing photovoltaic output.
TABLE 1 analysis of major influencing factors
Step 2, the ELM prediction model structure is shown in FIG. 2, and is composed of an input layer, a hidden layer and an output layer, and specifically comprises the following steps:
step 2-1: the training sample for predicting the generating capacity of the photovoltaic power station is xi，pi]，xi＝[xi1，xi2，...，xiN]T∈RNAs a main factor influencing the photovoltaic contribution, piAnd e, taking the E R as a trained learning sample output value, namely the photovoltaic power generation power, wherein the mathematical expression of the ELM model is as follows:
in formula (3), g (-) is the activation function of hidden layer neuron; omegaiThe connection weight between the input layer and the hidden layer neuron; biThreshold values for hidden layer neurons.
Step 2-2: determining the number of neurons in the hidden layer based on an empirical formula, and randomly generating omegaiAnd bi；
H in the formula (4) is the number of hidden layer neurons; m is the number of neurons in the input layer; n is the number of neurons in the output layer; a is a constant between 1 and 10.
Step 2-3: selecting an infinite and differentiable hidden layer neuron activation function, selecting a Sigmoid function as the activation function of the prediction model disclosed by the invention as shown in a formula 5, and calculating a hidden layer output matrix H;
if the number of neurons in the hidden layer is equal to the number of training samples, for any omegaiAnd biThe ELM neural network can approximate the training samples with zero error, i.e.
According to formula (6):
it can also be written in matrix form T ═ H β, so the hidden layer output matrix H is expressed as:
step 2-4: calculating the connection weight beta of the hidden layer and the output layer can be obtained by solving the least square solution of the following equation set:
min|||Hβ-T′||#(8)
can be expressed as:
β＝H+T′#(9)
h in formula (9)+Moore-Penrose generalized inverse of H, and T' is the transpose of the network output.
In step 2-2, h is 12, m is 4, and n is 1.
And step 3: the specific process of GA optimization ELM model comprises the following specific steps:
step 3-1: initializing, setting parameters of a genetic algorithm: the population number K, the iteration number, the cross probability, the variation probability and the like, the sum of absolute values of errors predicted by the photovoltaic power training data is used as an individual fitness value, the smaller the individual fitness is, the optimal individual is, and the fitness function is shown as a formula 10.
Wherein: o isiTo predict the power value, TiFor the actual power value, N is the total number of samples.
Step 3-2: binary coding is carried out on the connection weight and the threshold value randomly generated by the ELM, so that a coding chain is constructed, and an initial population is generated; setting the number of nodes of an input layer in the ELM model as m; the number of hidden layer nodes is h; if the number of nodes of the output layer is n, the weight from the input layer to the hidden layer is a matrix of m x h, the threshold value of the hidden layer is h, the weight from the hidden layer to the output layer is h x n, and the threshold value of the output layer is n, then the coding length of the genetic algorithm individual is as follows:
l＝m*h+h*n+h+n#(11)
wherein m + h n is the number of weights; h + n is the number of thresholds.
Step 3-3: starting to train the population, calculating the fitness of each population, and setting the population F (y), wherein y belongs to K, and K is the population number, wherein K is (y)1，y2，...，yk) For arbitrary yiThen there is yi＝(x1，x2，...，xk) I.e., each chromosome contains k genes, then:
step 3-4: according to the calculated fitness, selecting, crossing and mutating the population to obtain a new population, and adding 1 to the iteration times;
step 3-5: judging whether the maximum iteration times is reached, if so, continuing the next operation, otherwise, returning to the step 3-3 for circulation again;
step 3-6: and decoding the parameters to obtain the optimized weight and threshold, and training by using the extreme learning machine to obtain the prediction result of the optimized extreme learning machine.
In step 3-1, the set population size is 50, the crossover probability is 0.7, the mutation probability is 0.01, the number of iterations is 200, and the target error is 0.0001.
In order to evaluate the performance of the photovoltaic output prediction model provided by the invention, the invention adopts an error general evaluation index, namely Mean Square Error (MSE) and a decision coefficient (R)2) The prediction results are analyzed for Mean Absolute Error (MAE) and mean estimated deviation ratio (ADR).
In the above formula, PiThe actual value of the photovoltaic power generation power is obtained;
the photovoltaic power generation power predicted value is obtained;
is PiAverage value of (d); n is the number of samples, and 15min is one sample point, so N is 96.
According to the method, the established model is researched through an MATLAB software simulation platform, and in order to more clearly obtain the accuracy of the method, another two prediction models are selected for simulation analysis, namely an extreme learning machine model (ELM), a Person correlation coefficient method and an ELM combined model (PCC-ELM). Through training and testing of different prediction models, a photovoltaic power generation power prediction result of a typical day is obtained, and fig. 3 is a comparison graph of photovoltaic power generation prediction results of typical days under different prediction models, so that the prediction result of the PCC-GA-ELM model is closer to an actual value.
In order to visualize the prediction errors of different models, the photovoltaic output errors of the three models on a typical day are compared, and the result is shown in fig. 4, so that it can be clearly seen that the prediction errors of the PCC-GA-ELM model are respectively reduced by 68.7% and 4.7% on average compared with those of the ELM and the PCC-ELM. In addition, error indexes between predicted values and actual values obtained by different methods are counted based on simulation results, and error statistical results of three prediction models for predicting a typical day are shown in table 2.
TABLE 2 statistical table of error indexes of three prediction models
As can be seen from table 2: the prediction result based on the PCC-GA-ELM model is closer to the actual value. Wherein, Mean Square Error (MSE) is used for reflecting the difference degree between the photovoltaic power prediction quantity and the actual value, the MSE of the PCC-ELM model is reduced by 31 percent compared with the ELM model, and the MSE based on the PCC-GA-ELM model is reduced to 0.345; determining the coefficient (R)2) Is a statistical index for clarifying the reliability of the change of the dependent variable, and R is pre-measured by three methods2The values were all higher, while PCC-GA-ELM predicted the resulting R2The value is closer to 1; the Mean Absolute Error (MAE) is used for describing the error condition of a predicted value and a real value, and compared with a single ELM, the PCC-ELM and the PCC-GA-ELM model both reduce the value of the MAE to be below 1, but the mean absolute error of the PCC-GA-ELM is lower; the estimated deviation ratio (ADR) is more reflectiveAnd the ADR of the PCC-GA-ELM model is reduced by 5.5 percent compared with the PCC-ELM and is reduced by 12.8 percent compared with the ELM under the condition of error fluctuation of the predicted value and the true value.
The method analyzes and screens the factors influencing the photovoltaic output power through a Pearson correlation coefficient method, establishes an improved ELM photovoltaic prediction model, improves and optimizes the ELM model through a genetic algorithm, and iterates the optimized weight and threshold value step by step, thereby effectively avoiding the influence of the input redundancy of the model and the parameters randomly generated by the ELM on the precision of the prediction result. The PCC-GA-ELM model provided by the invention can accurately predict the photovoltaic power generation power, has better generalization performance and provides a better short-term photovoltaic power prediction method.
The above description is only of the preferred embodiments of the present invention, and it should be noted that: it will be apparent to those skilled in the art that various modifications and adaptations can be made without departing from the principles of the invention and these are intended to be within the scope of the invention.
Claims (10)
1. A photovoltaic power generation prediction method based on an improved extreme learning machine is characterized by comprising the following steps:
1) acquiring photovoltaic power generation historical data, preprocessing the photovoltaic power generation historical data, and acquiring main factors influencing photovoltaic output through correlation analysis;
2) constructing an ELM prediction model and training;
3) optimizing ELM prediction model parameters by adopting a genetic algorithm;
4) and (4) applying the ELM prediction model after optimization training to predict the photovoltaic power generation.
2. The improved extreme learning machine-based photovoltaic power generation prediction method as claimed in claim 1, wherein in the step 1), the photovoltaic power generation historical data is normalized, and the method comprises the following steps:
wherein p isjIs the original photovoltaic power generation power data of the j (th), pj，minIs the minimum value of j columns of data, pj，maxIs the maximum value of the j columns of data,
the normalized photovoltaic power generation power data is obtained.
3. The improved extreme learning machine-based photovoltaic power generation prediction method according to claim 1, wherein in the step 1), correlation analysis is performed on influence factors influencing photovoltaic output and actual photovoltaic power generation power, a PCC correlation coefficient between each influence factor and the actual photovoltaic power generation power is calculated, and main factors are selected according to the PCC correlation coefficient.
4. The improved extreme learning machine-based photovoltaic power generation prediction method according to claim 3, wherein the first k influencing factors with larger absolute values of PCC correlation coefficients are selected as main factors.
5. The improved extreme learning machine-based photovoltaic power generation prediction method as claimed in claim 4, wherein the influencing factors influencing the photovoltaic output comprise irradiance, temperature, wind speed, wind direction, humidity and pressure, and the selected main factors specifically comprise irradiance, temperature and humidity.
6. The improved extreme learning machine-based photovoltaic power generation prediction method as claimed in claim 1, wherein the step 2) specifically comprises the following steps:
21) construction of training sample [ x ] for prediction of photovoltaic power generationi，pi]Wherein x isi＝[xi1，xi2，...，xiN]T∈RNIs the data vector of the ith main factor influencing the photovoltaic output, N is the number of training samples, piE.g. R isThe sample output value is the photovoltaic power generation power;
22) an ELM prediction model composed of an input layer, a hidden layer and an output layer is constructed, and the expression is as follows:
wherein, betaiIs the connection weight of the hidden layer and the output layer, g (-) is the activation function of the hidden layer neuron, omegaiAs a connection weight between input layer and hidden layer neurons, biIs the threshold value of hidden layer neuron, h is the number of hidden layer neuron;
23) determining the number h of hidden layer neurons and randomly generating a connection weight ωiAnd a threshold value bi；
24) And selecting a Sigmoid function as an activation function of the hidden layer neuron, and training the model.
7. The improved extreme learning machine-based photovoltaic power generation prediction method as claimed in claim 6, wherein in the step 23), if the number h of hidden layer neurons is determined based on the empirical formula, the following steps are performed:
wherein m is the number of neurons in an input layer, n is the number of neurons in an output layer, and a is a constant between 1 and 10.
8. The improved extreme learning machine-based photovoltaic power generation prediction method as claimed in claim 6, wherein the step 3) specifically comprises the following steps:
31) initializing, and setting parameters of a genetic algorithm, including a population number K, a maximum iteration number, a cross probability and a variation probability;
32) the sum of the absolute values of the errors predicted by the photovoltaic power training data is used as an individual fitness value;
33) binary coding is carried out on the connection weight value and the threshold value randomly generated by the ELM prediction model, a coding chain is constructed, and an initial population is generated;
34) training the population, calculating the fitness of each population, and performing selection, crossing and variation operations on the population to obtain a new population;
35) and judging whether the maximum iteration times is reached, if so, outputting the optimized connection weight value and the threshold value, otherwise, returning to the step 34).
9. The improved extreme learning machine-based photovoltaic power generation prediction method according to claim 8, wherein in the step 33), the coding length l of the genetic algorithm is the sum of the number of the weight values and the number of the threshold values.
10. The improved extreme learning machine-based photovoltaic power generation prediction method according to claim 1, further comprising performing performance evaluation on the ELM prediction model, specifically:
and analyzing the prediction result of the ELM prediction model by adopting error general evaluation indexes, namely mean square error, a decision coefficient, a mean absolute error and a mean estimated deviation ratio.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111459373.1A CN114298377A (en) | 2021-12-02 | 2021-12-02 | Photovoltaic power generation prediction method based on improved extreme learning machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111459373.1A CN114298377A (en) | 2021-12-02 | 2021-12-02 | Photovoltaic power generation prediction method based on improved extreme learning machine |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114298377A true CN114298377A (en) | 2022-04-08 |
Family
ID=80964867
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111459373.1A Pending CN114298377A (en) | 2021-12-02 | 2021-12-02 | Photovoltaic power generation prediction method based on improved extreme learning machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114298377A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114970952A (en) * | 2022-04-14 | 2022-08-30 | 楚能新能源股份有限公司 | Photovoltaic output short-term prediction method and system considering environmental factors |
| 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 |
| CN116432991A (en) * | 2023-06-14 | 2023-07-14 | 国网浙江省电力有限公司嘉兴供电公司 | Park multi-energy supply and demand matching degree quantitative evaluation method considering space-time characteristics |
-
2021
- 2021-12-02 CN CN202111459373.1A patent/CN114298377A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114970952A (en) * | 2022-04-14 | 2022-08-30 | 楚能新能源股份有限公司 | Photovoltaic output short-term prediction method and system considering environmental factors |
| 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 |
| CN116432991A (en) * | 2023-06-14 | 2023-07-14 | 国网浙江省电力有限公司嘉兴供电公司 | Park multi-energy supply and demand matching degree quantitative evaluation method considering space-time characteristics |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108876054B (en) | Short-term power load prediction method based on improved genetic algorithm optimization extreme learning machine | |
| CN112348271A (en) | Short-term photovoltaic power prediction method based on VMD-IPSO-GRU | |
| CN114298377A (en) | Photovoltaic power generation prediction method based on improved extreme learning machine | |
| CN110751318B (en) | Ultra-short-term power load prediction method based on IPSO-LSTM | |
| CN108921339B (en) | Quantile regression-based photovoltaic power interval prediction method for genetic support vector machine | |
| CN106251001A (en) | A kind of based on the photovoltaic power Forecasting Methodology improving fuzzy clustering algorithm | |
| CN110674999A (en) | Cell load prediction method based on improved clustering and long-short term memory deep learning | |
| CN105631558A (en) | BP neural network photovoltaic power generation system power prediction method based on similar day | |
| CN113537582B (en) | Photovoltaic power ultra-short-term prediction method based on short-wave radiation correction | |
| CN109242136A (en) | A kind of micro-capacitance sensor wind power Chaos-Genetic-BP neural network prediction technique | |
| CN111898825A (en) | Photovoltaic power generation power short-term prediction method and device | |
| CN104820877A (en) | Photovoltaic system generation power prediction method based on cloud adaptive PSO-SNN | |
| CN116644970A (en) | Photovoltaic power prediction method based on VMD decomposition and lamination deep learning | |
| CN116341613A (en) | Ultra-short-term photovoltaic power prediction method based on Informar encoder and LSTM | |
| CN116826710A (en) | Peak clipping strategy recommendation method and device based on load prediction and storage medium | |
| CN113762591B (en) | Short-term electric quantity prediction method and system based on GRU and multi-core SVM countermeasure learning | |
| CN112836876B (en) | Power distribution network line load prediction method based on deep learning | |
| CN116523148B (en) | Distribution network distribution transformer overload early warning method, device and equipment | |
| CN113033898A (en) | Electrical load prediction method and system based on K-means clustering and BI-LSTM neural network | |
| CN117335425A (en) | Tidal current calculation method based on GA-BP neural network | |
| CN115392387B (en) | Low-voltage distributed photovoltaic power generation output prediction method | |
| CN111310974A (en) | Short-term water demand prediction method based on GA-ELM | |
| CN116681154A (en) | Photovoltaic power calculation method based on EMD-AO-DELM | |
| CN115796327A (en) | Wind power interval prediction method based on VMD (vertical vector decomposition) and IWOA-F-GRU (empirical mode decomposition) -based models | |
| Lv et al. | Research on short-term load forecasting approach for smart grid |
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 |