Multi-model-based tracking wind and light output prediction method

The method employs multiple models trained on corrected meteorological data to enhance the accuracy of wind and solar power output predictions, addressing single-model inaccuracies and improving grid integration.

CN119089374APending Publication Date: 2024-12-06HUADIAN ELECTRIC POWER SCI INST CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410891660.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-04
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

Existing wind and solar power output prediction methods rely on a single model, which are prone to inaccuracies due to model limitations, affecting the accuracy of unified regulation predictions.

Method used

A method utilizing multiple models trained on corrected meteorological data using Gaussian kernel density estimation, divided into training, validation, and test sets, with weighted summation of predictions to enhance accuracy.

Benefits of technology

Improves the accuracy of wind and solar power output predictions, facilitating better grid integration and consumption of renewable energy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119089374A_ABST
    Figure CN119089374A_ABST
Patent Text Reader

Abstract

The invention relates to a multi-model-based unified wind and light output prediction method, and the method comprises the steps: obtaining meteorological data and unified wind and light output data of a region in a historical time period; correcting the meteorological data; taking the corrected meteorological data as a sample main body, taking the unified wind and light output data as a sample label, generating a sample, and generating a data set according to the sample; dividing the data set into a training set, a verification set and a test set; training a plurality of different preset models according to the training set and the verification set; performing tracking wind-light output prediction on each sample in the test set by using a plurality of tracking wind-light output prediction models obtained by training; and according to the tracking wind and light output prediction results corresponding to the plurality of tracking wind and light output prediction models and the sample label of each sample in the test set, calculating tracking wind power output prediction weights and tracking photovoltaic output prediction weights corresponding to the plurality of tracking wind and light output prediction models so as to accurately predict tracking wind and light output based on the prediction weights.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of wind and solar power output prediction for unified regulation, and in particular to a method for predicting wind and solar power output for unified regulation based on multiple models. Background Art

[0002] In order to promote the grid connection and consumption of wind power and photovoltaic power generation, research on wind power and photovoltaic output forecasting methods and accurate forecasting of wind and photovoltaic output have become an important means to achieve high-quality development of electricity and construction of new power systems.

[0003] The common method for predicting the wind and solar power output of unified regulation mainly inputs meteorological data and wind and solar power output data of unified regulation into the machine learning model, trains the model parameters to obtain the prediction model, and then inputs the meteorological forecast data into the prediction model to obtain the prediction result of the wind and solar power output of unified regulation. However, the wind and solar power output prediction method of unified regulation generally relies on a single model for prediction, which is easily affected by the shortcomings of the model itself and has poor prediction accuracy. Therefore, how to accurately predict the wind and solar power output of unified regulation has become a technical problem that needs to be solved urgently. Summary of the invention

[0004] The embodiment of the present application provides a method for predicting the unified wind and solar power output based on multiple models, so as to at least solve the problem of how to accurately predict the unified wind and solar power output in the related art.

[0005] In a first aspect, an embodiment of the present application provides a model training method for wind and solar power output prediction based on multiple models, the method comprising:

[0006] Obtain regional meteorological data and wind and solar output data in historical time periods;

[0007] The obtained meteorological data are corrected using Gaussian kernel density estimation algorithm;

[0008] The corrected meteorological data is used as the sample body, and the unified wind and solar power output data is used as the sample label to generate samples, and then the unified wind and solar power output prediction data set is generated;

[0009] The unified wind and solar power output prediction data set is divided into a training set, a validation set, and a test set;

[0010] Multiple different preset models are trained according to the training set and the validation set to obtain multiple unified wind and solar output prediction models;

[0011] Using multiple unified wind-solar output prediction models to predict the unified wind-solar output of each sample in the test set, and obtaining unified wind-solar output prediction results corresponding to the multiple unified wind-solar output prediction models;

[0012] According to the unified wind and solar power output prediction results corresponding to multiple unified wind and solar power output prediction models and the sample label of each sample in the test set, the unified wind power output prediction weights and the unified photovoltaic output prediction weights corresponding to the multiple unified wind and solar power output prediction models are calculated.

[0013] In some implementations of the first aspect, using a Gaussian kernel density estimation algorithm to correct the acquired meteorological data includes:

[0014] Normalize the acquired meteorological data;

[0015] The Gaussian kernel density estimation algorithm is used to estimate the probability density function of the normalized data points;

[0016] Determine outliers by comparing the estimated probability density function with the actual data points;

[0017] According to the estimated probability density function, the detected outliers are corrected.

[0018] In some possible implementations of the first aspect, dividing the unified wind and solar power output prediction data set into a training set, a validation set, and a test set includes:

[0019] Dividing the historical time period into a first historical time period, a second historical time period, and a third historical time period according to a preset division ratio;

[0020] The samples corresponding to the first historical time period in the unified wind and solar power output forecast data set are divided into a training set;

[0021] The samples corresponding to the second historical time period in the unified wind and solar power output forecast data set are divided into the validation set;

[0022] The samples corresponding to the third historical time period in the unified wind and solar power output forecast dataset are divided into the test set.

[0023] In some implementable embodiments of the first aspect, according to the unified wind and solar output prediction results corresponding to the multiple unified wind and solar output prediction models and the sample labels of the samples in the test set, calculating the unified wind output prediction weights and the unified photovoltaic output prediction weights corresponding to the multiple unified wind and solar output prediction models includes:

[0024] According to the unified wind and solar power output prediction results corresponding to the multiple unified wind and solar power output prediction models and the sample labels of each sample in the test set, the unified wind power output prediction variance and the unified photovoltaic output prediction variance corresponding to the multiple unified wind and solar power output prediction models are calculated;

[0025] Calculate the unified wind power output prediction weights corresponding to the multiple unified wind power output prediction models according to the unified wind power output prediction variances corresponding to the multiple unified wind power output prediction models;

[0026] According to the unified regulation photovoltaic output prediction variances corresponding to the multiple unified regulation wind and solar power output prediction models, the unified regulation photovoltaic output prediction weights corresponding to the multiple unified regulation wind and solar power output prediction models are calculated.

[0027] In some implementable embodiments of the first aspect, according to the unified wind and solar output prediction results corresponding to the multiple unified wind and solar output prediction models and the sample labels of the samples in the test set, calculating the unified wind output prediction variance and the unified photovoltaic output prediction variance corresponding to the multiple unified wind and solar output prediction models includes:

[0028] For any unified wind and solar output prediction model, the unified wind output prediction variance and unified photovoltaic output prediction variance corresponding to the unified wind and solar output prediction model are calculated in the following way:

[0029] Cluster the samples in the test set to obtain multiple sample clusters;

[0030] For any sample cluster, according to the unified wind-solar output prediction results corresponding to the unified wind-solar output prediction model at the sample cluster and the sample labels of each sample in the sample cluster, the unified wind-solar output prediction model at each sample in the sample cluster is calculated to predict the unified wind output error and the unified photovoltaic output prediction error; according to the unified wind output prediction error of the unified wind-solar output prediction model at each sample in the sample cluster, the average unified wind output prediction error of the unified wind-solar output prediction model at the sample cluster is calculated to predict the unified wind output error; according to the unified photovoltaic output prediction error of the unified wind-solar output prediction model at each sample in the sample cluster, the average unified photovoltaic output prediction error of the unified wind-solar output prediction model at the sample cluster is calculated to predict the unified photovoltaic output error;

[0031] According to the unified wind power output prediction error of the unified wind power output prediction model at each sample in each sample cluster and the average unified wind power output prediction error of the unified wind power output prediction model at each sample cluster, the unified wind power output prediction variance corresponding to the unified wind power output prediction model is calculated;

[0032] According to the unified regulation photovoltaic output prediction error of the unified regulation wind-solar output prediction model at each sample in each sample cluster and the average unified regulation photovoltaic output prediction error of the unified regulation wind-solar output prediction model at each sample cluster, the unified regulation photovoltaic output prediction variance corresponding to the unified regulation wind-solar output prediction model is calculated.

[0033] In some implementable embodiments of the first aspect, calculating the unified wind power output prediction weights corresponding to the multiple unified wind power output prediction models according to the unified wind power output prediction variances corresponding to the multiple unified wind power output prediction models includes:

[0034] Adding the inverses of the unified wind power output prediction variances corresponding to the multiple unified wind and solar power output prediction models;

[0035] For any unified wind and solar power output prediction model, the reciprocal of the unified wind power output prediction variance corresponding to the unified wind and solar power output prediction model is divided by the addition result to obtain the unified wind power output prediction weight corresponding to the unified wind and solar power output prediction model.

[0036] In some implementable embodiments of the first aspect, calculating the unified photovoltaic output prediction weights corresponding to the multiple unified wind and solar output prediction models according to the unified photovoltaic output prediction variances corresponding to the multiple unified wind and solar output prediction models includes:

[0037] Add the inverses of the unified photovoltaic output prediction variances corresponding to the multiple unified wind and solar output prediction models;

[0038] For any unified wind and solar power output prediction model, the inverse of the unified photovoltaic output prediction variance corresponding to the unified wind and solar power output prediction model is divided by the addition result to obtain the unified photovoltaic output prediction weight corresponding to the unified wind and solar power output prediction model.

[0039] In a second aspect, an embodiment of the present application provides a method for predicting wind and solar power output based on multiple models, the method comprising:

[0040] Obtain regional weather forecast data;

[0041] The meteorological forecast data are processed respectively using a plurality of unified wind and solar power output prediction models to complete the unified wind and solar power output prediction, and the unified wind and solar power output prediction results corresponding to the plurality of unified wind and solar power output prediction models are obtained;

[0042] According to the unified wind power output prediction weights corresponding to the multiple unified wind and solar power output prediction models, weighted summation is performed on the unified wind power output prediction results in the unified wind and solar power output prediction results corresponding to the multiple unified wind and solar power output prediction models to obtain a comprehensive unified wind power output prediction result;

[0043] According to the unified photovoltaic output prediction weights corresponding to the multiple unified wind and solar output prediction models, weighted summation is performed on the unified photovoltaic output prediction results in the unified wind and solar output prediction results corresponding to the multiple unified wind and solar output prediction models to obtain a comprehensive unified photovoltaic output prediction result;

[0044] Among them, multiple unified wind and solar power output prediction models and their corresponding unified wind power output prediction weights and unified photovoltaic power output prediction weights are obtained based on the multi-model training method as described above.

[0045] In a third aspect, an embodiment of the present application provides a model training device for wind and solar output prediction based on multiple models, the device comprising:

[0046] The acquisition module is used to obtain the meteorological data of the region in the historical period and the wind and solar power output data;

[0047] A correction module is used to correct the acquired meteorological data using a Gaussian kernel density estimation algorithm;

[0048] A generation module is used to generate samples using the corrected meteorological data as the sample body and the unified wind and solar output data as the sample label, and thereby generate a unified wind and solar output prediction data set;

[0049] A partitioning module is used to divide the unified wind and solar power output prediction data set into a training set, a validation set, and a test set;

[0050] A training module is used to train multiple different preset models according to a training set and a validation set to obtain multiple unified wind and solar output prediction models;

[0051] A prediction module is used to use multiple unified wind and solar power output prediction models to perform unified wind and solar power output prediction on each sample in the test set, and obtain unified wind and solar power output prediction results corresponding to the multiple unified wind and solar power output prediction models;

[0052] The calculation module is used to calculate the unified wind output prediction weights and unified photovoltaic output prediction weights corresponding to multiple unified wind and solar output prediction models based on the unified wind and solar output prediction results corresponding to multiple unified wind and solar output prediction models and the sample labels of each sample in the test set.

[0053] In a fourth aspect, an embodiment of the present application provides a wind and solar output prediction device based on multiple models, the device comprising:

[0054] An acquisition module is used to obtain regional weather forecast data;

[0055] A prediction module is used to process meteorological forecast data using multiple unified wind and solar power output prediction models to complete unified wind and solar power output prediction and obtain unified wind and solar power output prediction results corresponding to the multiple unified wind and solar power output prediction models;

[0056] A first summing module is used to perform weighted summation of unified wind power output prediction results in the unified wind power output prediction results corresponding to the multiple unified wind power output prediction models according to the unified wind power output prediction weights corresponding to the multiple unified wind power output prediction models, so as to obtain a comprehensive unified wind power output prediction result;

[0057] The second summing module is used to perform weighted summation of the unified photovoltaic output prediction results in the unified wind and solar output prediction results corresponding to the multiple unified wind and solar output prediction models according to the unified photovoltaic output prediction weights corresponding to the multiple unified wind and solar output prediction models, so as to obtain a comprehensive unified photovoltaic output prediction result;

[0058] Among them, multiple unified wind and solar power output prediction models and their corresponding unified wind power output prediction weights and unified photovoltaic power output prediction weights are obtained based on the multi-model training method as described above.

[0059] In a fifth aspect, an embodiment of the present application provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method as described above.

[0060] In a sixth aspect, an embodiment of the present application provides a non-transitory computer-readable storage medium storing computer instructions, where the computer instructions are used to enable a computer to execute the method as described above.

[0061] In an embodiment of the present application, meteorological data and unified wind and solar power output data of a region within a historical time period can be obtained; the obtained meteorological data can be corrected using a Gaussian kernel density estimation algorithm; samples are generated using the corrected meteorological data as the sample body and the unified wind and solar power output data as sample labels, and a data set is generated based on this; the data set is divided into a training set, a validation set, and a test set; multiple different preset models are trained based on the training set and the validation set; the unified wind and solar power output prediction models obtained through training are used to predict the unified wind and solar power output of each sample in the test set; the unified wind and solar power output prediction weights and unified photovoltaic output prediction weights corresponding to the multiple unified wind and solar power output prediction models are calculated based on the unified wind and solar power output prediction results corresponding to the multiple unified wind and solar power output prediction models and the sample labels of each sample in the test set, and then the unified wind and solar power output is accurately predicted based on this, thereby promoting the grid connection and consumption of wind power and photovoltaic power generation. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0063] Figure 1 It is a flow chart of a model training method for wind and solar power output prediction based on multi-model coordination provided in an embodiment of the present application;

[0064] Figure 2 It is a flow chart of a method for predicting wind and solar power output based on multiple models provided in an embodiment of the present application;

[0065] Figure 3 It is a structural diagram of a model training device for wind and solar power output prediction based on multiple models provided in an embodiment of the present application;

[0066] Figure 4 It is a structural diagram of a wind and solar output prediction device based on multiple models provided according to an embodiment of the present application;

[0067] Figure 5 is a structural diagram of an exemplary electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0068] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0069] In addition, the term "and / or" in this article is only a description of the association relationship between the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0070] In response to the problems arising in the background technology, the embodiments of the present invention provide a method for predicting the wind and solar power output based on multiple models. Specifically, the meteorological data and the wind and solar power output data of the region in the historical time period are obtained; the obtained meteorological data are corrected by using the Gaussian kernel density estimation algorithm; the corrected meteorological data are used as the sample body, and the wind and solar power output data of the unified regulation are used as the sample label to generate samples, and a data set is generated thereby; the data set is divided into a training set, a validation set, and a test set; a plurality of different preset models are trained according to the training set and the validation set; the wind and solar power output prediction models obtained through training are used to predict the wind and solar power output of each sample in the test set respectively; the wind and solar power output prediction weights and the photovoltaic power output prediction weights corresponding to the wind and solar power output prediction models are calculated according to the wind and solar power output prediction results corresponding to the wind and solar power output prediction models and the sample labels of each sample in the test set, and then the wind and solar power output is accurately predicted based on this, so as to promote the grid connection and consumption of wind power and photovoltaic power generation.

[0071] In the following, in conjunction with the accompanying drawings, a method for predicting wind and solar power output based on multiple models provided by an embodiment of the present invention is described in detail through specific embodiments.

[0072] Figure 1 is a flow chart of a model training method for wind and solar output prediction based on multi-model coordination provided in an embodiment of the present application, such as Figure 1 As shown, the multi-model training method 100 may include the following steps:

[0073] S110, obtaining regional meteorological data and coordinated wind and solar power output data in a historical period.

[0074] For example, the meteorological data and the wind and solar power output data of the region in the past month can be sampled to obtain the meteorological data and the wind and solar power output data of the region in the past month, wherein the sampling interval can be 15 minutes.

[0075] Exemplarily, in order to train a more general unified wind and solar power output prediction model, the region here can be multiple different regions; historical meteorological data can include: air pressure data, temperature data, humidity data, radiation data, wind data, etc.; unified wind and solar power output data includes: unified wind power output data and unified photovoltaic output data.

[0076] S120, using a Gaussian kernel density estimation algorithm to correct the acquired meteorological data.

[0077] In some embodiments, the acquired meteorological data can be normalized, and then a Gaussian kernel density estimation algorithm is used to estimate the probability density function of the normalized data points. Outliers are determined by comparing the estimated probability density function with the actual data points, and then the detected outliers are effectively corrected based on the estimated probability density function.

[0078] S130, using the corrected meteorological data as the sample body and the coordinated wind and solar power output data as the sample label, generating samples, and thereby generating a coordinated wind and solar power output prediction data set.

[0079] It is understandable that in order to achieve unified wind and solar power output forecasts of different time scales, the time scale of the samples in the unified wind and solar power output forecast data set here can be adjusted as needed. For example, for short-term unified wind and solar power output forecasts, the time scale of the samples is short-term; for medium-term unified wind and solar power output forecasts, the time scale of the samples is medium-term; for long-term unified wind and solar power output forecasts, the time scale of the samples is long-term.

[0080] S140, dividing the unified wind and solar power output prediction data set into a training set, a validation set, and a test set.

[0081] In some embodiments, the historical time period can be divided into a first historical time period, a second historical time period, and a third historical time period according to a preset division ratio, and the samples in the unified wind and solar output prediction data set corresponding to the first historical time period are divided into a training set, the samples in the unified wind and solar output prediction data set corresponding to the second historical time period are divided into a verification set, and the samples in the unified wind and solar output prediction data set corresponding to the third historical time period are divided into a test set.

[0082] For example, referring to the example in S110, the past month (30 days) can be divided into the first 20 days, the middle 5 days, and the last 5 days according to 4:1:1, and then the samples belonging to the first 20 days in the unified wind and solar output prediction dataset are divided into the training set, the samples belonging to the middle 5 days in the unified wind and solar output prediction dataset are divided into the verification set, and the samples belonging to the last 5 days in the unified wind and solar output prediction dataset are divided into the test set.

[0083] S150, training a plurality of different preset models according to the training set and the validation set to obtain a plurality of unified wind and solar power output prediction models.

[0084] Exemplarily, the preset models can be BP neural network, LSTM neural network, bidirectional LSTM neural network, GRU neural network, bidirectional GRU neural network, CNN neural network, extreme learning machine (ELM), etc., which are optimized by genetic algorithms with outstanding performance.

[0085] Taking the optimization of BP neural network by genetic algorithm as an example, the process of training the BP neural network optimized by genetic algorithm according to the training set and the verification set can be specifically as follows:

[0086] (1) Initialize the genetic algorithm to optimize the parameters of the BP neural network.

[0087] (2) Obtain the parameter population and calculate the fitness.

[0088] (3) The genetic algorithm is used to select, crossover and mutate parameters to obtain the optimal parameters (weights, biases, etc.).

[0089] (4) Substitute the obtained optimal parameters into the genetic algorithm optimized BP neural network, and then input the training set and the verification set into the genetic algorithm optimized BP neural network to train it. Among them, the evaluation index of the model can be set according to different power scenarios, which will not be elaborated here.

[0090] S160, using multiple unified wind and solar power output prediction models to perform unified wind and solar power output prediction on each sample in the test set, and obtaining unified wind and solar power output prediction results corresponding to the multiple unified wind and solar power output prediction models.

[0091] Specifically, for any unified wind-solar output prediction model, the test set is input into the unified wind-solar output prediction model, and the unified wind-solar output prediction model performs unified wind-solar output prediction on each sample in the test set to obtain a unified wind-solar output prediction result for each sample, wherein the unified wind-solar output prediction result includes: a unified wind power output prediction result and a unified photovoltaic output prediction result, which is then used as the unified wind-solar output prediction result corresponding to the unified wind-solar output prediction model.

[0092] S170, calculating the unified wind output prediction weights and the unified photovoltaic output prediction weights corresponding to the multiple unified wind output prediction models according to the unified wind output prediction results corresponding to the multiple unified wind output prediction models and the sample labels of the samples in the test set.

[0093] In some embodiments, the unified wind power output prediction variance and the unified photovoltaic output prediction variance corresponding to the multiple unified wind power output prediction models can be calculated based on the unified wind power output prediction results corresponding to the multiple unified wind power output prediction models and the sample labels of each sample in the test set.

[0094] According to the unified wind power output prediction variances corresponding to the multiple unified wind and solar power output prediction models, the unified wind power output prediction weights corresponding to the multiple unified wind and solar power output prediction models are calculated.

[0095] According to the unified regulation photovoltaic output prediction variances corresponding to the multiple unified regulation wind and solar power output prediction models, the unified regulation photovoltaic output prediction weights corresponding to the multiple unified regulation wind and solar power output prediction models are calculated.

[0096] For example, for any unified wind and solar output prediction model, the unified wind output prediction variance and unified photovoltaic output prediction variance corresponding to the unified wind and solar output prediction model can be calculated in the following manner:

[0097] Cluster the samples in the test set to obtain multiple sample clusters.

[0098] For any sample cluster, according to the unified wind and solar output prediction results corresponding to the unified wind and solar output prediction model at the sample cluster and the sample labels of each sample in the sample cluster, the unified wind output prediction error (i.e. the error between the unified wind output prediction result of the sample and the unified wind output data in the corresponding sample label) and the unified photovoltaic output prediction error (i.e. the error between the unified photovoltaic output prediction result of the sample and the unified photovoltaic output data in the corresponding sample label) of the unified wind and solar output prediction model at each sample in the sample cluster are calculated; according to the unified wind output prediction error of the unified wind and solar output prediction model at each sample in the sample cluster, the average unified wind output prediction error of the unified wind and solar output prediction model at the sample cluster is calculated; according to the unified photovoltaic output prediction error of the unified wind and solar output prediction model at each sample in the sample cluster, the average unified photovoltaic output prediction error of the unified wind and solar output prediction model at the sample cluster is calculated.

[0099] According to the unified wind power output prediction error of the unified wind-solar power output prediction model at each sample in each sample cluster and the average unified wind power output prediction error of the unified wind-solar power output prediction model at each sample cluster, the unified wind power output prediction variance corresponding to the unified wind-solar power output prediction model is calculated.

[0100] According to the unified regulation photovoltaic output prediction error of the unified regulation wind-solar output prediction model at each sample in each sample cluster and the average unified regulation photovoltaic output prediction error of the unified regulation wind-solar output prediction model at each sample cluster, the unified regulation photovoltaic output prediction variance corresponding to the unified regulation wind-solar output prediction model is calculated.

[0101] Optionally, the calculation formula used to calculate the unified wind / photovoltaic output prediction variance corresponding to the unified wind / photovoltaic output prediction model can be as follows:

[0102]

[0103] in, represents the prediction variance of the unified wind / photovoltaic output corresponding to the unified wind / photovoltaic output prediction model i; m represents the number of sample clusters; S n Represents the number of samples in sample cluster n; represents the unified wind / PV output prediction error of the unified wind / PV output prediction model i at the jth sample in sample cluster n; It represents the average unified wind / PV output prediction error of the unified wind / PV output prediction model i at sample cluster n.

[0104] For example, for any unified wind and solar power output prediction model, the unified wind power output prediction weight and the unified photovoltaic power output prediction weight corresponding to the unified wind and solar power output prediction model can be calculated in the following manner:

[0105] The reciprocals of the unified wind power output prediction variances corresponding to multiple unified wind and solar power output prediction models are added together. For any unified wind and solar power output prediction model, the reciprocal of the unified wind power output prediction variance corresponding to the unified wind and solar power output prediction model is divided by the addition result to obtain the unified wind power output prediction weight corresponding to the unified wind and solar power output prediction model.

[0106] The reciprocals of the unified photovoltaic output prediction variances corresponding to multiple unified wind and solar output prediction models are added together. For any unified wind and solar output prediction model, the reciprocal of the unified photovoltaic output prediction variance corresponding to the unified wind and solar output prediction model is divided by the addition result to obtain the unified photovoltaic output prediction weight corresponding to the unified wind and solar output prediction model.

[0107] Optionally, the calculation formula used to calculate the unified wind power / photovoltaic output prediction weights corresponding to multiple unified wind and solar power output prediction models can be as follows:

[0108]

[0109] Among them, ω i represents the unified wind / photovoltaic output prediction weight corresponding to the unified wind / photovoltaic output prediction model i, represents the unified wind / photovoltaic output prediction variance corresponding to the unified wind / photovoltaic output prediction model i, and x represents the number of unified wind / photovoltaic output prediction models.

[0110] In an embodiment of the present invention, meteorological data and unified wind and solar power output data of a region in a historical time period are obtained; the obtained meteorological data is corrected by using a Gaussian kernel density estimation algorithm; samples are generated using the corrected meteorological data as the sample body and the unified wind and solar power output data as sample labels, and a data set is generated thereby; the data set is divided into a training set, a validation set, and a test set; a plurality of different preset models are trained according to the training set and the validation set; the unified wind and solar power output prediction models obtained through training are used to perform unified wind and solar power output prediction on each sample in the test set; the unified wind and solar power output prediction weights and the unified photovoltaic output prediction weights corresponding to the plurality of unified wind and solar power output prediction models are calculated according to the unified wind and solar power output prediction results corresponding to the plurality of unified wind and solar power output prediction models and the sample labels of each sample in the test set, thereby effectively completing multi-model training.

[0111] Based on the multi-model training method 100 provided in the embodiment of the present invention, the embodiment of the present invention also provides a method for predicting wind and solar output based on multi-models, such as Figure 2 As shown, the wind and solar power output prediction method 200 may include the following steps:

[0112] S210, obtaining regional weather forecast data.

[0113] S220, using a plurality of unified wind and solar power output prediction models to process the meteorological forecast data respectively to complete the unified wind and solar power output prediction, and obtain unified wind and solar power output prediction results corresponding to the plurality of unified wind and solar power output prediction models.

[0114] Specifically, for any unified wind and solar power output prediction model, the meteorological forecast data is input into the unified wind and solar power output prediction model, and the unified wind and solar power output prediction model processes the data to complete the unified wind and solar power output prediction and obtain the unified wind and solar power output prediction result, which is then used as the unified wind and solar power output prediction result corresponding to the unified wind and solar power output prediction model.

[0115] S230, according to the unified wind power output prediction weights corresponding to the multiple unified wind and solar power output prediction models, weighted summation is performed on the unified wind power output prediction results in the unified wind and solar power output prediction results corresponding to the multiple unified wind and solar power output prediction models to obtain a comprehensive unified wind power output prediction result.

[0116] S240, according to the unified photovoltaic output prediction weights corresponding to the multiple unified wind and solar output prediction models, weighted summation of the unified photovoltaic output prediction results in the unified wind and solar output prediction results corresponding to the multiple unified wind and solar output prediction models is performed to obtain a comprehensive unified photovoltaic output prediction result.

[0117] Among them, multiple unified wind and solar power output prediction models and their corresponding unified wind power output prediction weights and unified photovoltaic power output prediction weights are obtained based on the multi-model training method as described above.

[0118] In an embodiment of the present invention, based on multiple unified wind and solar power output prediction models and their corresponding unified wind power output prediction weights and unified photovoltaic power output prediction weights, the unified wind and solar power output can be accurately predicted to promote the grid connection and consumption of wind power and photovoltaic power generation.

[0119] It is worth noting that the calculation formula used for the above weighted summation can be shown as follows:

[0120]

[0121] in, represents the forecast result of comprehensive wind power / photovoltaic output, ω i represents the unified wind power / photovoltaic output prediction weight corresponding to the unified wind / photovoltaic output prediction model i, It represents the unified wind power / photovoltaic output prediction result corresponding to the unified wind power output prediction model i, and x represents the number of unified wind power output prediction models.

[0122] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described order of actions, because according to the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0123] The above is an introduction to a method embodiment. The following is a further explanation of the solution of the present invention through an apparatus embodiment.

[0124] Figure 3 is a structural diagram of a model training device for wind and solar output prediction based on multiple models provided in an embodiment of the present application, such as Figure 3 As shown, the multi-model training device 300 may include:

[0125] The acquisition module 310 is used to obtain the meteorological data and the wind and solar power output data of the region in the historical time period.

[0126] The correction module 320 is used to correct the acquired meteorological data using a Gaussian kernel density estimation algorithm.

[0127] The generation module 330 is used to generate samples using the corrected meteorological data as the sample body and the coordinated wind and solar power output data as sample labels, and thereby generate a coordinated wind and solar power output prediction data set.

[0128] The division module 340 is used to divide the unified wind and solar power output prediction data set into a training set, a validation set, and a test set.

[0129] The training module 350 is used to train multiple different preset models according to the training set and the verification set to obtain multiple unified wind and solar power output prediction models.

[0130] The prediction module 360 ​​is used to use multiple unified wind and solar power output prediction models to perform unified wind and solar power output prediction on each sample in the test set, and obtain unified wind and solar power output prediction results corresponding to the multiple unified wind and solar power output prediction models.

[0131] The calculation module 370 is used to calculate the unified wind output prediction weights and unified photovoltaic output prediction weights corresponding to multiple unified wind and solar output prediction models according to the unified wind and solar output prediction results corresponding to the multiple unified wind and solar output prediction models and the sample labels of each sample in the test set.

[0132] Understandably, Figure 3 Each module / unit in the multi-model training device 300 shown has an implementation Figure 1 The functions of each step in the multi-model training method 100 shown can achieve its corresponding technical effects, and for the sake of brevity, they will not be repeated here.

[0133] Figure 4 is a structural diagram of a wind and solar output prediction device based on multiple models provided in an embodiment of the present application, such as Figure 4 As shown, the wind and solar power output prediction device 400 may include:

[0134] The acquisition module 410 is used to acquire regional weather forecast data.

[0135] The prediction module 420 is used to process the meteorological forecast data using multiple unified wind and solar power output prediction models to complete the unified wind and solar power output prediction and obtain the unified wind and solar power output prediction results corresponding to the multiple unified wind and solar power output prediction models.

[0136] The first summing module 430 is used to perform weighted summation of the unified wind power output prediction results in the unified wind power output prediction results corresponding to the multiple unified wind power output prediction models according to the unified wind power output prediction weights corresponding to the multiple unified wind power output prediction models, so as to obtain a comprehensive unified wind power output prediction result.

[0137] The second summing module 440 is used to perform weighted summation of the unified photovoltaic output prediction results in the unified wind and solar output prediction results corresponding to the multiple unified wind and solar output prediction models according to the unified photovoltaic output prediction weights corresponding to the multiple unified wind and solar output prediction models, so as to obtain a comprehensive unified photovoltaic output prediction result.

[0138] Among them, multiple unified wind and solar power output prediction models and their corresponding unified wind power output prediction weights and unified photovoltaic power output prediction weights are obtained based on the above multi-model training method.

[0139] Understandably, Figure 4 Each module / unit in the wind and solar power output prediction device 400 shown in the figure has the function of realizing Figure 2 The functions of each step in the coordinated wind-solar output prediction method 200 shown can achieve its corresponding technical effects, and for the sake of brevity, they will not be repeated here.

[0140] Figure 5 5 is a block diagram of an exemplary electronic device according to an embodiment of the present application. Electronic device 500 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic device 500 may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0141] like Figure 5 As shown, the electronic device 500 may include a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 can also be stored. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0142] Multiple components in the electronic device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the electronic device 500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0143] The computing unit 501 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 501 performs the various methods and processes described above, such as method 100 or 200. For example, in some embodiments, the method 100 or 200 may be implemented as a computer program product, including a computer program, which is tangibly contained in a computer-readable medium, such as a storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 500 via the ROM 502 and / or the communication unit 509. When the computer program is loaded into the RAM 503 and executed by the computing unit 501, one or more steps of the method 100 or 200 described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured to execute the method 100 or 200 in any other appropriate manner (eg, by means of firmware).

[0144] The various embodiments described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include: being implemented in one or more computer programs, which may be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general programmable processor, which may receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0145] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0146] In the context of the present invention, a computer-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. A computer-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. A more specific example of a computer-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0147] It should be noted that the present invention also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute method 100 or 200 and achieve the corresponding technical effects achieved by the embodiments of the present invention executing the methods thereof. For the sake of concise description, they will not be repeated here.

[0148] In addition, the present invention also provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the method 100 or 200 is implemented.

[0149] To provide interaction with a user, the above-described embodiments may be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0150] The embodiments described above can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0151] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0152] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution disclosed in the present invention can be achieved, and this document does not limit this.

[0153] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A model training method for wind and solar power output prediction based on multiple models, It is characterized in that The method comprises: Obtain regional meteorological data and wind and solar output data in historical time periods; The obtained meteorological data are corrected using Gaussian kernel density estimation algorithm; The corrected meteorological data is used as the sample body, and the unified wind and solar power output data is used as the sample label to generate samples, and then the unified wind and solar power output prediction data set is generated; Dividing the unified wind and solar power output prediction data set into a training set, a validation set, and a test set; Training a plurality of different preset models according to the training set and the verification set to obtain a plurality of unified wind and solar output prediction models; Using the multiple unified wind-solar output prediction models respectively to perform unified wind-solar output prediction on each sample in the test set, and obtaining unified wind-solar output prediction results corresponding to the multiple unified wind-solar output prediction models; According to the unified wind and solar output prediction results corresponding to the multiple unified wind and solar output prediction models and the sample label of each sample in the test set, the unified wind output prediction weights and the unified photovoltaic output prediction weights corresponding to the multiple unified wind and solar output prediction models are calculated.

2. The method according to claim 1, It is characterized in that The method of using the Gaussian kernel density estimation algorithm to correct the acquired meteorological data includes: Normalize the acquired meteorological data; The Gaussian kernel density estimation algorithm is used to estimate the probability density function of the normalized data points; Determine outliers by comparing the estimated probability density function with the actual data points; According to the estimated probability density function, the detected outliers are corrected.

3. The method according to claim 1, It is characterized in that The step of dividing the wind and solar power output prediction data set into a training set, a validation set, and a test set comprises: Dividing the historical time period into a first historical time period, a second historical time period, and a third historical time period according to a preset division ratio; Dividing the samples in the unified wind and solar power output prediction data set corresponding to the first historical time period into a training set; Dividing the samples in the unified wind and solar power output prediction data set corresponding to the second historical time period into a validation set; The samples in the unified wind and solar power output prediction data set corresponding to the third historical time period are divided into a test set.

4. The method according to claim 1, It is characterized in that The calculating, according to the unified wind and solar power output prediction results corresponding to the multiple unified wind and solar power output prediction models and the sample labels of the samples in the test set, the unified wind power output prediction weights and the unified photovoltaic power output prediction weights corresponding to the multiple unified wind and solar power output prediction models comprises: According to the unified wind and solar power output prediction results corresponding to the multiple unified wind and solar power output prediction models and the sample labels of the samples in the test set, calculating the unified wind power output prediction variance and the unified photovoltaic output prediction variance corresponding to the multiple unified wind and solar power output prediction models; Calculating unified wind power output prediction weights corresponding to the multiple unified wind power output prediction models according to unified wind power output prediction variances corresponding to the multiple unified wind power output prediction models; According to the unified regulation photovoltaic output prediction variances corresponding to the multiple unified regulation wind and solar power output prediction models, the unified regulation photovoltaic output prediction weights corresponding to the multiple unified regulation wind and solar power output prediction models are calculated.

5. The method according to claim 4, It is characterized in that The calculating, according to the unified wind and solar power output prediction results corresponding to the multiple unified wind and solar power output prediction models and the sample labels of each sample in the test set, the unified wind power output prediction variance and the unified photovoltaic power output prediction variance corresponding to the multiple unified wind and solar power output prediction models comprises: For any unified wind and solar power output prediction model, the unified wind power output prediction variance and unified photovoltaic power output prediction variance corresponding to the unified wind and solar power output prediction model are calculated in the following manner: Clustering the samples in the test set to obtain multiple sample clusters; For any sample cluster, according to the unified wind and solar output prediction results corresponding to the unified wind and solar output prediction model at the sample cluster and the sample labels of each sample in the sample cluster, calculate the unified wind output prediction error and the unified photovoltaic output prediction error of the unified wind and solar output prediction model at each sample in the sample cluster; according to the unified wind output prediction error of the unified wind and solar output prediction model at each sample in the sample cluster, calculate the average unified wind output prediction error of the unified wind and solar output prediction model at the sample cluster; according to the unified photovoltaic output prediction error of the unified wind and solar output prediction model at each sample in the sample cluster, calculate the average unified photovoltaic output prediction error of the unified wind and solar output prediction model at the sample cluster; Calculate the unified wind output prediction variance corresponding to the unified wind output prediction model according to the unified wind output prediction error of the unified wind and solar output prediction model at each sample in each sample cluster and the average unified wind output prediction error of the unified wind and solar output prediction model at each sample cluster; According to the unified regulation photovoltaic output prediction error of the unified regulation wind and solar output prediction model at each sample in each sample cluster and the average unified regulation photovoltaic output prediction error of the unified regulation wind and solar output prediction model at each sample cluster, the unified regulation photovoltaic output prediction variance corresponding to the unified regulation wind and solar output prediction model is calculated.

6. The method according to claim 4, It is characterized in that The calculating the unified wind power output prediction weights corresponding to the plurality of unified wind power output prediction models according to the unified wind power output prediction variances corresponding to the plurality of unified wind power output prediction models comprises: Adding the inverses of the unified wind power output prediction variances corresponding to the multiple unified wind and solar power output prediction models; For any unified wind and solar power output prediction model, the reciprocal of the unified wind power output prediction variance corresponding to the unified wind and solar power output prediction model is divided by the addition result to obtain the unified wind power output prediction weight corresponding to the unified wind and solar power output prediction model.

7. The method according to claim 4, It is characterized in that The calculating, according to the unified photovoltaic output prediction variances corresponding to the plurality of unified wind and solar output prediction models, the unified photovoltaic output prediction weights corresponding to the plurality of unified wind and solar output prediction models comprises: Adding the inverses of the unified photovoltaic output prediction variances corresponding to the multiple unified wind and solar output prediction models; For any unified wind and solar power output prediction model, the reciprocal of the unified photovoltaic output prediction variance corresponding to the unified wind and solar power output prediction model is divided by the addition result to obtain the unified photovoltaic output prediction weight corresponding to the unified wind and solar power output prediction model.

8. A method for predicting wind and solar power output based on multiple models. It is characterized in that The method comprises: Obtain regional weather forecast data; The meteorological forecast data are processed respectively using a plurality of unified wind and solar power output prediction models to complete the unified wind and solar power output prediction, and the unified wind and solar power output prediction results corresponding to the plurality of unified wind and solar power output prediction models are obtained; According to the unified wind power output prediction weights corresponding to the multiple unified wind and solar power output prediction models, weighted summation is performed on the unified wind power output prediction results in the unified wind and solar power output prediction results corresponding to the multiple unified wind and solar power output prediction models to obtain a comprehensive unified wind power output prediction result; According to the unified photovoltaic output prediction weights corresponding to the multiple unified wind and solar output prediction models, weighted summation of the unified photovoltaic output prediction results in the unified wind and solar output prediction results corresponding to the multiple unified wind and solar output prediction models is performed to obtain a comprehensive unified photovoltaic output prediction result; Among them, the multiple unified wind and solar power output prediction models and their corresponding unified wind power output prediction weights and unified photovoltaic output prediction weights are obtained based on the multi-model training method described in any one of claims 1-7.

9. A model training device for wind and solar power output prediction based on multiple models, It is characterized in that The device comprises: The acquisition module is used to obtain the meteorological data of the region in the historical period and the wind and solar power output data; A correction module is used to correct the acquired meteorological data using a Gaussian kernel density estimation algorithm; A generation module is used to generate samples using the corrected meteorological data as the sample body and the unified wind and solar output data as the sample label, and thereby generate a unified wind and solar output prediction data set; A partitioning module, used to divide the unified wind and solar power output prediction data set into a training set, a validation set, and a test set; A training module, used for training a plurality of different preset models according to the training set and the verification set to obtain a plurality of unified wind and solar output prediction models; A prediction module, used to respectively use the multiple unified wind and solar power output prediction models to perform unified wind and solar power output prediction on each sample in the test set, and obtain unified wind and solar power output prediction results corresponding to the multiple unified wind and solar power output prediction models; A calculation module is used to calculate the unified wind output prediction weights and unified photovoltaic output prediction weights corresponding to the multiple unified wind and solar output prediction models according to the unified wind and solar output prediction results corresponding to the multiple unified wind and solar output prediction models and the sample labels of each sample in the test set.

10. A wind and solar power output prediction device based on multiple models, It is characterized in that The device comprises: An acquisition module is used to obtain regional weather forecast data; A prediction module, used to process the meteorological forecast data using a plurality of unified wind and solar power output prediction models respectively, so as to complete the unified wind and solar power output prediction, and obtain the unified wind and solar power output prediction results corresponding to the plurality of unified wind and solar power output prediction models; A first summing module is used to perform weighted summation of the unified wind power output prediction results in the unified wind power output prediction results corresponding to the multiple unified wind power output prediction models according to the unified wind power output prediction weights corresponding to the multiple unified wind power output prediction models, so as to obtain a comprehensive unified wind power output prediction result; A second summing module is used to perform weighted summation of the unified photovoltaic output prediction results in the unified wind and solar output prediction results corresponding to the multiple unified wind and solar output prediction models according to the unified photovoltaic output prediction weights corresponding to the multiple unified wind and solar output prediction models, so as to obtain a comprehensive unified photovoltaic output prediction result; Among them, the multiple unified wind and solar power output prediction models and their corresponding unified wind power output prediction weights and unified photovoltaic output prediction weights are obtained based on the multi-model training method described in any one of claims 1-7.