Photovoltaic ultra-short-term power prediction method and equipment based on adaptive meteorological classification
Through the adaptive meteorological classification method, the meteorological category and irradiance prediction model is constructed, which solves the accuracy of ultra-short-term power prediction of photovoltaic power generation, realizes high-precision prediction under different weather conditions, improves the grid operation stability and photovoltaic power stations profit.
Patent Information
- Application Number
- CN202210053799.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-01-18
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2042-01-18
AI Technical Summary
The existing technology is difficult to accurately predict the ultra-short-term power of photovoltaic power generation, resulting in unstable grid operation and loss of profits of photovoltaic power plants, mainly due to the diversity of solar irradiance prediction modes and the violent fluctuations in meteorological factors.
Using the ultra-short-term power prediction method of photovoltaic based on adaptive meteorological classification, by obtaining meteorological factor vectors, constructing meteorological category classification model and irradiance prediction model, training solar irradiance-photovoltaic power generation power conversion model, and using the forecast data of the National Meteorological Administration Make predictions.
It improves the accuracy of ultra-short-term power prediction of photovoltaic power generation, can make predictions separately under different weather conditions, and solves the problem of high-precision prediction under complex and variable meteorological conditions.
Smart Images

Figure CN114492181B_ABST
Abstract
Description
Technical Field
[0001] The embodiments of the present invention relate to the field of solar power generation technology, and in particular to a photovoltaic ultra-short-term power prediction method and device based on adaptive meteorological classification. Background Art
[0002] Solar power generation is intermittent, and the power generation at different times fluctuates greatly. Directly feeding the power generation into the power grid will have a great impact on the smooth operation of the power grid. In order to ensure the smooth operation of the power grid, clear assessment requirements are put forward for ultra-short-term power prediction: the accuracy of daily ultra-short-term power prediction shall not be lower than the corresponding threshold of 90%. If the corresponding prediction accuracy is lower than the threshold, the power grid will assess the power of the photovoltaic power station and cause profit losses. The main meteorological factor affecting photovoltaic power generation is solar irradiance. According to the technical characteristics of photovoltaic panels, solar irradiance and photovoltaic power generation are linearly related. Therefore, the key to improving the accuracy of photovoltaic ultra-short-term power prediction lies in improving the accuracy of ultra-short-term prediction of solar irradiance. Solar irradiance is strongly correlated with other meteorological factors such as atmospheric humidity and atmospheric temperature. Various meteorological influencing factors fluctuate violently under different weather conditions, resulting in the diversity of solar irradiance prediction modes. At present, there is no method to accurately predict photovoltaic ultra-short-term power. Therefore, developing a photovoltaic ultra-short-term power prediction method and equipment based on adaptive meteorological classification can effectively overcome the defects in the above-mentioned related technologies, which has become a technical problem that needs to be solved urgently in the industry. Summary of the invention
[0003] In view of the above problems existing in the prior art, an embodiment of the present invention provides a photovoltaic ultra-short-term power prediction method and device based on adaptive meteorological classification.
[0004] In the first aspect, an embodiment of the present invention provides a photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification, comprising: obtaining a number of meteorological factors at any historical moment, and encapsulating the obtained meteorological factors into a meteorological factor vector; using an adaptive iterative algorithm at the historical moment to construct a meteorological category classification model and several irradiance prediction models, the input of the meteorological category classification model is the meteorological factor vector, and the output is a meteorological category value represented by an integer, each meteorological category value corresponds to an irradiance prediction model, and the total number of meteorological category values output by the meteorological category classification model is the same as the total number of irradiance prediction models; the meteorological factor vector at a certain moment is input into the meteorological category classification model, and the output is the meteorological category value at that moment. Category value, find the corresponding irradiance prediction model, input the meteorological factor vector at that moment into the corresponding irradiance prediction model, and output the predicted value of solar irradiance at that moment; according to the solar irradiance observation value and photovoltaic power generation observation value at the historical moment, train the solar irradiance-photovoltaic power generation conversion model, which is a linear regression model; for any moment in the future 16 moments, obtain the meteorological factor vector forecast value at that moment from the National Meteorological Administration, substitute it into the meteorological category classification model to obtain the meteorological category value at that moment, substitute it into the corresponding irradiance prediction model to obtain the solar irradiance prediction value, substitute the solar irradiance prediction value into the solar irradiance-photovoltaic power generation conversion model, and obtain the ultra-short-term prediction value of photovoltaic power generation.
[0005] Based on the content of the above method embodiment, the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided in the embodiment of the present invention, at multiple historical moments, the latitude and longitude information is input on the website of the Numerical Meteorological Forecast Center of the National Meteorological Administration to obtain the observed values of m meteorological factors in a certain area, and the observed values of the m meteorological factors at time t constitute a meteorological factor vector:
[0006]
[0007] The observed irradiance at time t is x t ,s 1t ,s 2t ,…,s mt is the observed value of m meteorological factors at time t. t=1,2,…,N, where N is the total number of historical moments.
[0008] Based on the content of the above method embodiment, the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided in the embodiment of the present invention, in the meteorological factor vector of N historical moments The meteorological categories are initialized and k irradiance prediction models are trained. Then, the meteorological categories and irradiance prediction models are updated according to the principle of minimum prediction deviation, including: setting a very small positive number ε, using the clustering algorithm to It is classified into k meteorological categories, which are represented by integer labels 1, 2, ..., k. Each meteorological category corresponds to Several vectors in; train the irradiance prediction model on several meteorological factor vectors corresponding to meteorological categories 1, 2, ..., k, and obtain k irradiance prediction models g1, g2, ..., g k The root mean square error of all meteorological factor vectors generated during the irradiance prediction model training process is recorded as r1, which is called the first root mean square error. From the irradiance prediction model g1,g2,…,g k Search for the model with the smallest prediction deviation and The meteorological category is reset to the subscript of the model with the smallest prediction deviation, so The k meteorological categories to which they belong are updated, and the irradiance prediction models are trained on the several meteorological factor vectors corresponding to the updated meteorological categories 1, 2, ..., k, so that the k irradiance prediction models g1, g2, ..., g k The root mean square error of all meteorological factor vectors generated during the training process of the updated irradiance prediction model is recorded as r2, which is called the second root mean square error.
[0009] Based on the content of the above method embodiment, the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided in the embodiment of the present invention sets a very small positive number ε, and calculates the absolute value of the difference between the second root mean square error and the first root mean square error, and determines whether the absolute value of the difference is less than or equal to the positive number ε to decide whether to update The meteorological category to which it belongs and whether to update the k irradiance prediction models include: if the absolute value of the difference is greater than the positive number ε, rename the second root mean square error to the first root mean square error, and for any meteorological factor vector From the irradiance prediction model g1,g2,…,g k Search for the model with the smallest prediction deviation and The meteorological category is reset to the subscript of the model with the smallest prediction deviation, so The k meteorological categories to which they belong are updated, and the irradiance prediction models are trained on the several meteorological factor vectors corresponding to the updated meteorological categories 1, 2, ..., k, so that the k irradiance prediction models g1, g2, ..., g kis also updated. The root mean square error generated by all meteorological factor vectors during the training process of the updated irradiance prediction model is recorded as r2, named the second root mean square error, so that the second root mean square error is also updated. The absolute value of the difference between the second root mean square error and the first root mean square error is recalculated. If the absolute value of the difference between the second root mean square error and the first root mean square error is greater than the positive number ε, the above process is re-executed; if the absolute value of the difference is less than or equal to the positive number ε, then The meteorological category and the k irradiance prediction models remain unchanged and will not be updated.
[0010] Based on the content of the above method embodiment, the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided in the embodiment of the present invention is based on the meteorological factor vector The meteorological classification model h is trained with the corresponding meteorological category observations. The input of h is the meteorological factor vector, and the output is an integer in 1,…,k, representing the meteorological category. The irradiance prediction models corresponding to the k weather categories 1,…,k are g1, g2,…, g k For the meteorological factor vector at any time t The predicted value of the meteorological category at time t is According to the corresponding relationship between meteorological categories and irradiance prediction models, the meteorological category values The corresponding irradiance prediction model is So the predicted irradiance at time t is
[0011] Based on the content of the above method embodiment, the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided in the embodiment of the present invention constructs a solar irradiance-photovoltaic power conversion model, which is a linear model, including: training a linear model
[0012]
[0013] The coefficients a and b of the model are unknown. The values of a and b are given by the following formula
[0014]
[0015] Among them, p t and x t They are the observed values of photovoltaic power generation and solar irradiance at time t, t = 1, 2, …, N, and N is the total number of historical moments.
[0016] On the basis of the contents of the above method embodiments, the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided in the embodiments of the present invention applies the solar irradiance prediction model and the solar irradiance-photovoltaic power conversion model to perform ultra-short-term photovoltaic power prediction for each of the future moments N+1, N+2, ..., N+16, including: obtaining the meteorological factor vector forecast values for the future moments N+1, N+2, ..., N+16 from the National Meteorological Administration according to the longitude and latitude Substitute the 16 meteorological factor vector forecast values into the irradiance prediction formula to obtain the irradiance forecast values at these 16 moments Substituting the irradiance prediction values at these 16 moments into the solar irradiance-photovoltaic power conversion model, we can obtain the ultra-short-term prediction values of photovoltaic power generation at these moments.
[0017]
[0018] In the second aspect, an embodiment of the present invention provides a photovoltaic ultra-short-term power prediction device based on adaptive meteorological classification, including: a first main module, used to obtain a number of meteorological factors and solar irradiance at multiple historical moments, and encapsulate the obtained meteorological factors into meteorological factor vectors; a second main module, used to use an adaptive iterative algorithm to construct a meteorological category classification model and a number of irradiance prediction models; a third main module, training a linear regression model based on solar irradiance sample data and photovoltaic power generation sample data to obtain a solar irradiance-photovoltaic power generation power conversion model; a fourth main module, obtaining the meteorological factor vector forecast value at a future moment from the National Meteorological Administration, substituting it into the meteorological category classification model to obtain the meteorological category value at that moment, substituting it into the corresponding irradiance prediction model to obtain the solar irradiance prediction value, and substituting the solar irradiance prediction value into the solar irradiance-photovoltaic power generation power conversion model to obtain the ultra-short-term prediction value of photovoltaic power generation power.
[0019] In a third aspect, an embodiment of the present invention provides an electronic device, including:
[0020] at least one processor; and
[0021] at least one memory in communication with the processor, wherein:
[0022] The memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided by any one of the various implementations of the first aspect.
[0023] In a fourth aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions enable a computer to execute a photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided by any one of the various implementation methods of the first aspect.
[0024] The photovoltaic ultra-short-term power prediction method and device based on adaptive meteorological classification provided by the embodiment of the present invention encapsulates several acquired meteorological factors into vectors, adopts an adaptive iterative algorithm to find the meteorological category corresponding to each vector, obtains a meteorological category classification model and an irradiance prediction model, further obtains a solar irradiance-photovoltaic power conversion model, and obtains a photovoltaic ultra-short-term power prediction based on this. The photovoltaic power generation power can be predicted separately under each type of weather condition, solving the problem of high-precision prediction of photovoltaic power generation power under complex and changeable meteorological conditions. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0026] Figure 1 A flow chart of a photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided by an embodiment of the present invention;
[0027] Figure 2 A schematic diagram of the structure of a photovoltaic ultra-short-term power prediction device based on adaptive meteorological classification provided by an embodiment of the present invention;
[0028] Figure 3 A schematic diagram of the physical structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0029] 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 in 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. In addition, the technical features in the various embodiments or single embodiments provided by the present invention can be arbitrarily combined with each other to form a feasible technical solution. This combination is not subject to the constraints of the sequence of steps and / or the structural composition mode, but must be based on the ability of ordinary technicians in this field to achieve. When the combination of technical solutions is contradictory or cannot be achieved, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.
[0030] The embodiment of the present invention provides a photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification, see Figure 1 The method includes: obtaining a number of meteorological factors at any historical moment, and encapsulating the obtained meteorological factors into a meteorological factor vector; using an adaptive iterative algorithm to construct a meteorological category classification model and a number of irradiance prediction models at the historical moment, the input of the meteorological category classification model is the meteorological factor vector, and the output is a meteorological category value represented by an integer, each meteorological category value corresponds to an irradiance prediction model, and the total number of meteorological category values output by the meteorological category classification model is the same as the total number of irradiance prediction models; the meteorological factor vector at a certain moment is input to the meteorological category classification model, the output is the meteorological category value at that moment, and the corresponding irradiance prediction model is found, and the The meteorological factor vector at that moment is input into the corresponding irradiance prediction model, and the output is the predicted value of solar irradiance at that moment; according to the solar irradiance observation values and photovoltaic power generation observation values at historical moments, the solar irradiance-photovoltaic power generation conversion model is trained, and the model is a linear regression model; for any moment in the future 16 moments, the meteorological factor vector forecast value at that moment is obtained from the National Meteorological Administration, substituted into the meteorological category classification model to obtain the meteorological category value at that moment, substituted into the corresponding irradiance prediction model to obtain the solar irradiance prediction value, and the solar irradiance prediction value is substituted into the solar irradiance-photovoltaic power generation conversion model to obtain the ultra-short-term prediction value of photovoltaic power generation.
[0031] Specifically, for a real-valued time series {p t |t=1,2,…,N}, if p t is the instantaneous power generation of the photovoltaic power station at any time t (unit: megawatt), then the sequence is called the photovoltaic power generation time series. For a real-valued time series {x t |t=1,2,…,N}, if x tis the instantaneous observed value of solar irradiance at any time t (unit: watt / square meter), then the sequence is called solar irradiance time series. Standing at time N, the photovoltaic power generation power of 16 future moments N+1, N+2, ..., N+16 is predicted, which is called photovoltaic power ultra-short-term prediction, and the solar irradiance of these 16 future moments is predicted, which is called solar irradiance ultra-short-term prediction.
[0032] Based on the content of the above method embodiment, as an optional embodiment, the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided in the embodiment of the present invention, specifically, for any historical moment t, assuming that there are m meteorological factors (such as atmospheric temperature, humidity, wind speed, etc.) and the solar irradiance x at the same moment t Strong correlation, input the latitude and longitude information on the website of the National Meteorological Administration Numerical Meteorological Forecast Center to obtain the observed values of the m meteorological factors in a certain area. The observed values of the m meteorological factors at time t constitute the meteorological factor vector:
[0033]
[0034] The observed irradiance at time t is x t ,s 1t ,s 2t ,…,s mt is the observed value of m meteorological factors at time t. t=1,2,…,N, where N is the total number of historical moments.
[0035] Based on the content of the above method embodiment, as an optional embodiment, the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided in the embodiment of the present invention, specifically, according to the strength of solar irradiance, it is assumed that weather conditions can be divided into k categories. In the meteorological factor vectors at the N historical moments The meteorological categories are initialized and k irradiance prediction models are trained. Then, the meteorological categories and irradiance prediction models are updated according to the principle of minimum prediction deviation, including: setting a very small positive number ε, using the clustering algorithm to It is classified into k meteorological categories, which are represented by integer labels 1, 2, ..., k. Each meteorological category corresponds to Several vectors in; train the irradiance prediction model on several meteorological factor vectors corresponding to meteorological categories 1, 2, ..., k, and obtain k irradiance prediction models g1, g2, ..., g k The root mean square error of all meteorological factor vectors generated during the irradiance prediction model training process is recorded as r1, which is called the first root mean square error. From the irradiance prediction model g1,g2,…,g k Search for the model with the smallest prediction deviation and The meteorological category is reset to the subscript of the model with the smallest prediction deviation, so The k meteorological categories to which they belong are updated, and the irradiance prediction models are trained on the several meteorological factor vectors corresponding to the updated meteorological categories 1, 2, ..., k, so that the k irradiance prediction models g1, g2, ..., g k The root mean square error of all meteorological factor vectors generated during the training process of the updated irradiance prediction model is recorded as r2, which is called the second root mean square error.
[0036] Based on the content of the above method embodiment, as an optional embodiment, the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided in the embodiment of the present invention sets a very small positive number ε, and calculates the absolute value of the difference between the second root mean square error and the first root mean square error, and determines whether the absolute value of the difference is less than or equal to the positive number ε to decide whether to update The meteorological category to which it belongs and whether to update the k irradiance prediction models include: if the absolute value of the difference is greater than the positive number ε, rename the second root mean square error to the first root mean square error, and for any meteorological factor vector From the irradiance prediction model g1,g2,…,g k Search for the model with the smallest prediction deviation and The meteorological category is reset to the subscript of the model with the smallest prediction deviation, so The k meteorological categories to which they belong are updated, and the irradiance prediction models are trained on the several meteorological factor vectors corresponding to the updated meteorological categories 1, 2, ..., k, so that the k irradiance prediction models g1, g2, ..., g k is also updated. The root mean square error generated by all meteorological factor vectors during the training process of the updated irradiance prediction model is recorded as r2, named the second root mean square error, so that the second root mean square error is also updated. The absolute value of the difference between the second root mean square error and the first root mean square error is recalculated. If the absolute value of the difference between the second root mean square error and the first root mean square error is greater than the positive number ε, the above process is re-executed; if the absolute value of the difference is less than or equal to the positive number ε, then The meteorological category and the k irradiance prediction models remain unchanged and will not be updated.
[0037] Based on the content of the above method embodiment, as an optional embodiment, the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided in the embodiment of the present invention is based on the meteorological factor vector The meteorological classification model h is trained with the corresponding meteorological category observations. The input of h is the meteorological factor vector, and the output is an integer in 1,…,k, representing the meteorological category. The irradiance prediction models corresponding to the k weather categories 1,…,k are g1, g2,…, g k For the meteorological factor vector at any time t The predicted value of the meteorological category at time t is According to the corresponding relationship between meteorological categories and irradiance prediction models, the meteorological category values The corresponding irradiance prediction model is So the predicted irradiance at time t is
[0038] Based on the content of the above method embodiment, as an optional embodiment, the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided in the embodiment of the present invention constructs a solar irradiance-photovoltaic power conversion model, which is a linear model. Specifically, according to the physical characteristics and technical indicators of the photovoltaic panel, the photovoltaic power generation power is linearly related to the solar irradiance. In the solar irradiance sample data {x t |t=1,2,…,N} and photovoltaic power generation power sample data {p t |t=1,2,…,N} to train a linear regression model and obtain the following solar irradiance-photovoltaic power conversion model
[0039]
[0040] The coefficients a and b of the model are unknown. The values of a and b are given by the following formula
[0041]
[0042] Among them, p t and x t They are the observed values of photovoltaic power generation and solar irradiance at time t, t = 1, 2, …, N, and N is the total number of historical moments.
[0043] Based on the content of the above method embodiment, as an optional embodiment, the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided in the embodiment of the present invention applies the solar irradiance prediction model and the solar irradiance-photovoltaic power conversion model to perform ultra-short-term photovoltaic power prediction for each of the future moments N+1, N+2, ..., N+16. Specifically, the meteorological factor vector forecast values for the future moments N+1, N+2, ..., N+16 are obtained from the National Meteorological Administration according to the longitude and latitude. Substitute the 16 meteorological factor vector forecast values into the ground irradiance prediction model to obtain the irradiance forecast values at these 16 moments Substituting the irradiance prediction values at these 16 moments into the solar irradiance-photovoltaic power conversion model, we can obtain the ultra-short-term prediction values of photovoltaic power generation at these moments.
[0044]
[0045]
[0046]
[0047]
[0048] The photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification provided by the embodiment of the present invention encapsulates several acquired meteorological factors into vectors, adopts an adaptive iterative algorithm to find the meteorological category corresponding to each vector, obtains a meteorological category classification model and an irradiance prediction model, further obtains a solar irradiance-photovoltaic power conversion model, and obtains a photovoltaic ultra-short-term power prediction based on this. The photovoltaic power generation power can be predicted separately under each type of weather condition, solving the problem of high-precision prediction of photovoltaic power generation power under complex and changeable meteorological conditions.
[0049] The implementation basis of each embodiment of the present invention is to implement programmed processing through a device with a processor function. Therefore, in engineering practice, the technical solutions and functions of each embodiment of the present invention can be encapsulated into various modules. Based on this reality, on the basis of the above embodiments, the embodiment of the present invention provides a photovoltaic ultra-short-term power prediction device based on adaptive meteorological classification, which is used to execute the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification in the above method embodiment. Figure 2 The device includes: a first main module, which is used to obtain a number of meteorological factors and solar irradiance at multiple historical moments, and encapsulate the obtained meteorological factors into meteorological factor vectors; a second main module, which is used to construct a meteorological category classification model and a number of irradiance prediction models using an adaptive iterative algorithm; a third main module, which trains a linear regression model according to solar irradiance sample data and photovoltaic power generation sample data to obtain a solar irradiance-photovoltaic power generation power conversion model; a fourth main module, which obtains the meteorological factor vector forecast value at a future moment from the National Meteorological Administration, substitutes it into the meteorological category classification model to obtain the meteorological category value at the moment, substitutes it into the corresponding irradiance prediction model to obtain the solar irradiance prediction value, and substitutes the solar irradiance prediction value into the solar irradiance-photovoltaic power generation power conversion model to obtain the ultra-short-term prediction value of photovoltaic power generation.
[0050] The photovoltaic ultra-short-term power prediction device based on adaptive meteorological classification provided by the embodiment of the present invention adopts Figure 2By encapsulating several meteorological factors into vectors, an adaptive iterative algorithm is used to find the meteorological category corresponding to each vector, and a meteorological category classification model and an irradiance prediction model are obtained. A solar irradiance-photovoltaic power conversion model is further obtained, and based on this, an ultra-short-term photovoltaic power forecast is obtained. The photovoltaic power generation can be predicted separately under each type of weather condition, solving the problem of high-precision prediction of photovoltaic power under complex and changeable meteorological conditions.
[0051] It should be noted that the device in the device embodiment provided by the present invention can be used to implement the method in the above method embodiment as well as the method in other method embodiments provided by the present invention. The only difference is that the corresponding functional modules are set, and the principle is basically the same as the principle of the above device embodiment provided by the present invention. As long as the technical personnel in the field refer to the specific technical solutions in other method embodiments on the basis of the above device embodiment, obtain the corresponding technical means and the technical solutions composed of these technical means by combining technical features, the device in the above device embodiment can be improved on the premise of ensuring the practicality of the technical solution, thereby obtaining the corresponding device class embodiment, which is used to implement the methods in other method class embodiments. For example:
[0052] Based on the content of the above device embodiment, as an optional embodiment, the photovoltaic ultra-short-term power prediction device based on adaptive meteorological classification provided in the embodiment of the present invention further includes: a first submodule, which is used to input longitude and latitude information on the website of the Numerical Meteorological Forecast Center of the National Meteorological Administration to obtain the observed values of m meteorological factors in a certain area, and the observed values of the m meteorological factors at time t constitute a meteorological factor vector:
[0053]
[0054] The observed irradiance at time t is x t ,s 1t ,s 2t ,…,s mt is the observed value of m meteorological factors at time t. t=1,2,…,N, where N is the total number of historical moments.
[0055] Based on the content of the above device embodiment, as an optional embodiment, the photovoltaic ultra-short-term power prediction device based on adaptive meteorological classification provided in the embodiment of the present invention further includes: a second submodule, which is used to construct the one meteorological category classification model and several irradiance prediction models by using an adaptive iterative algorithm, and performs three steps in sequence, including:
[0056] Step 1. Initialize the meteorological classification results and the irradiance prediction model. Use the clustering algorithm to It is classified into k meteorological categories, which are represented by integer labels 1, 2, ..., k. Each meteorological category corresponds to Several vectors in; train the irradiance prediction model on several meteorological factor vectors corresponding to meteorological categories 1, 2, ..., k, and obtain k irradiance prediction models g1, g2, ..., g k The root mean square error of all meteorological factor vectors generated during the irradiance prediction model training process is recorded as r1, which is called the first root mean square error. From the irradiance prediction model g1,g2,…,g k Search for the model with the smallest prediction deviation and The meteorological category is reset to the subscript of the model with the smallest prediction deviation, so The k meteorological categories to which they belong are updated, and the irradiance prediction models are trained on the several meteorological factor vectors corresponding to the updated meteorological categories 1, 2, ..., k, so that the k irradiance prediction models g1, g2, ..., g k The root mean square error of all meteorological factor vectors generated during the training process of the updated irradiance prediction model is recorded as r2, which is called the second root mean square error.
[0057] Step 2. Update the meteorological classification results and the irradiance prediction model. Set a very small positive number ε, calculate the absolute value of the difference between the second root mean square error and the first root mean square error, and determine whether the absolute value of the difference is less than or equal to the positive number ε to decide whether to update The meteorological category to which it belongs and whether to update the k irradiance prediction models include: if the absolute value of the difference is greater than the positive number ε, rename the second root mean square error to the first root mean square error, and for any meteorological factor vector From the irradiance prediction model g1,g2,…,g k Search for the model with the smallest prediction deviation and The meteorological category is reset to the subscript of the model with the smallest prediction deviation, so The k meteorological categories to which they belong are updated, and the irradiance prediction models are trained on the several meteorological factor vectors corresponding to the updated meteorological categories 1, 2, ..., k, so that the k irradiance prediction models g1, g2, ..., g k is also updated. The root mean square error generated by all meteorological factor vectors during the training process of the updated irradiance prediction model is recorded as r2, named the second root mean square error, so that the second root mean square error is also updated. The absolute value of the difference between the second root mean square error and the first root mean square error is recalculated. If the absolute value of the difference between the second root mean square error and the first root mean square error is greater than the positive number ε, the above process is re-executed; if the absolute value of the difference is less than or equal to the positive number ε, then The meteorological category and the k irradiance prediction models remain unchanged and will not be updated.
[0058] Step 3. Output the meteorological classification model and irradiance prediction model. Based on the meteorological factor vector The meteorological classification model h is trained with the corresponding meteorological category observations. The input of h is the meteorological factor vector, and the output is an integer in 1,…,k, representing the meteorological category. The irradiance prediction models corresponding to the k weather categories 1,…,k are g1, g2,…, g k For the meteorological factor vector at any time t The predicted value of the meteorological category at time t is According to the corresponding relationship between meteorological categories and irradiance prediction models, the meteorological category values The corresponding irradiance prediction model is So the predicted irradiance at time t is
[0059] Based on the content of the above device embodiment, as an optional embodiment, the photovoltaic ultra-short-term power prediction device based on adaptive meteorological classification provided in the embodiment of the present invention also includes: a third submodule for constructing the solar irradiance-photovoltaic power conversion model, including: training linear model
[0060]
[0061] The coefficients a and b of the model are unknown. The values of a and b are given by the following formula
[0062]
[0063] Among them, p t and x t They are the observed values of photovoltaic power generation and solar irradiance at time t, t = 1, 2, …, N, and N is the total number of historical moments.
[0064] Based on the contents of the above-mentioned device embodiment, as an optional embodiment, the photovoltaic ultra-short-term power prediction device based on adaptive meteorological classification provided in the embodiment of the present invention also includes: a fourth submodule, applying the solar irradiance prediction model and the solar irradiance-photovoltaic power conversion model to perform ultra-short-term photovoltaic power prediction for each of the future moments N+1, N+2, ..., N+16, including: obtaining the meteorological factor vector forecast values for the future moments N+1, N+2, ..., N+16 from the National Meteorological Administration according to the longitude and latitude Substitute the 16 meteorological factor vector forecast values into the irradiance prediction formula to obtain the irradiance forecast values at these 16 moments Substituting the irradiance prediction values at these 16 moments into the solar irradiance-photovoltaic power conversion model, we can obtain the ultra-short-term prediction values of photovoltaic power generation at these moments.
[0065]
[0066] The method of the embodiment of the present invention is implemented by relying on electronic devices, so it is necessary to introduce the relevant electronic devices. Based on this purpose, the embodiment of the present invention provides an electronic device, such as Figure 3 As shown, the electronic device includes: at least one processor, a communication interface, at least one memory, and a communication bus, wherein at least one processor, the communication interface, and at least one memory communicate with each other through the communication bus. At least one processor can call the logic instructions in at least one memory to execute all or part of the steps of the method provided by the aforementioned various method embodiments.
[0067] In addition, the logic instructions in the at least one memory mentioned above can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each method embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.
[0068] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.
[0069] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0070] The flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the systems, methods and computer program products according to multiple embodiments of the present invention. Based on this understanding, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and sometimes in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or the flowchart, and the combination of the boxes in the block diagram and / or the flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.
[0071] In this patent, the terms "include", "comprises" or any other variation thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of more restrictions, the elements defined by the sentence "includes..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.
[0072] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification, characterized in that: include: At any historical moment, several meteorological factors are obtained and encapsulated into meteorological factor vectors; at the historical moment, an adaptive iterative algorithm is used to construct a meteorological category classification model and several irradiance prediction models. The input of the meteorological category classification model is the meteorological factor vector, and the output is the meteorological category value represented by an integer. Each meteorological category value corresponds to an irradiance prediction model, and the total number of meteorological category values output by the meteorological category classification model is the same as the total number of irradiance prediction models; the meteorological factor vector at a certain moment is input into the meteorological category classification model, and the meteorological category value at that moment is output, and the corresponding irradiance prediction model is found, and the meteorological category value at that moment is calculated. The meteorological factor vector is input to the corresponding irradiance prediction model, and the output is the predicted solar irradiance value at that moment; according to the solar irradiance observation value and photovoltaic power generation observation value at the historical moment, the solar irradiance-photovoltaic power generation conversion model is trained, and the model is a linear regression model; for any moment in the future 16 moments, the meteorological factor vector forecast value of that moment is obtained from the National Meteorological Administration, substituted into the meteorological category classification model to obtain the meteorological category value at that moment, substituted into the corresponding irradiance prediction model to obtain the solar irradiance prediction value, and the solar irradiance prediction value is substituted into the solar irradiance-photovoltaic power generation conversion model to obtain the ultra-short-term prediction value of photovoltaic power generation; At any historical time t, input the latitude and longitude information on the website of the National Meteorological Administration Numerical Meteorological Forecast Center to obtain the observed values of m meteorological factors in a certain area. The observed values of m meteorological factors at time t constitute the meteorological factor vector: The observed irradiance at time t is x t ,s 1t ,s 2t ,…,s mt is the observed value of m meteorological factors at time t, t = 1, 2, ..., N, N is the total number of historical moments; The meteorological factor vector for N historical moments Initialize the meteorological category and train k irradiance prediction models, then update the meteorological category and irradiance prediction model according to the principle of minimum prediction deviation, including: setting a very small positive number ε, using the clustering algorithm to It is classified into k meteorological categories, which are represented by integer labels 1, 2, ..., k. Each meteorological category corresponds to Several vectors in; train the irradiance prediction model on several meteorological factor vectors corresponding to meteorological categories 1, 2, ..., k, and obtain k irradiance prediction models g1, g2, ..., g k The root mean square error of all meteorological factor vectors generated during the irradiance prediction model training process is recorded as r1, which is called the first root mean square error. From the irradiance prediction model g1,g2,…,g k Search for the model with the smallest prediction deviation and The meteorological category is reset to the subscript of the model with the smallest prediction deviation, so The k meteorological categories to which they belong are updated, and the irradiance prediction models are trained on the several meteorological factor vectors corresponding to the updated meteorological categories 1, 2, ..., k, so that the k irradiance prediction models g1, g2, ..., g k It is also updated, and the root mean square error of all meteorological factor vectors generated in the training process of the updated irradiance prediction model is recorded as r2, which is called the second root mean square error; Set a very small positive number ε to determine whether the absolute value of the difference between the second RMS error and the first RMS error is less than or equal to the positive number ε to decide whether to update The meteorological category to which it belongs and whether to update k irradiance prediction models include: if the absolute value of the difference between the second root mean square error and the first root mean square error is greater than the positive number ε, the second root mean square error is renamed as the first root mean square error, and for any meteorological factor vector From the irradiance prediction model g1,g2,…,g k Search for the model with the smallest prediction deviation and The meteorological category is reset to the subscript of the model with the smallest prediction deviation, so The k meteorological categories to which they belong are updated, and the irradiance prediction models are trained on the several meteorological factor vectors corresponding to the updated meteorological categories 1, 2, ..., k, so that the k irradiance prediction models g1, g2, ..., g k is also updated; the root mean square error generated by all meteorological factor vectors in the training process of the updated irradiance prediction model is recorded as r2, named the second root mean square error, so that the second root mean square error is also updated, and the absolute value of the difference between the second root mean square error and the first root mean square error is recalculated. If the absolute value of the difference between the second root mean square error and the first root mean square error is greater than the positive number ε, the above process is re-executed; if the absolute value of the difference between the second root mean square error and the first root mean square error is less than or equal to the positive number ε, then The meteorological category and the k irradiance prediction models remain unchanged and will not be updated.
2. The photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification according to claim 1 is characterized in that: Meteorological factor vector The corresponding meteorological categories are all known. The meteorological classification model h is trained based on all meteorological factor vectors and the corresponding meteorological category values. The input of h is the meteorological factor vector, and the output is an integer in 1,…,k, representing the meteorological category. The irradiance prediction models corresponding to the k weather categories 1,…,k are g1, g2,…, g k , for the meteorological factor vector at any time t The predicted value of the meteorological category at time t is According to the corresponding relationship between meteorological categories and irradiance prediction models, the meteorological category values The corresponding irradiance prediction model is So the predicted irradiance at time t is 3. The photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification according to claim 2 is characterized in that: Construct a solar irradiance-photovoltaic power conversion model, which is a linear model, including: training linear model The coefficients a and b of the model are unknown. The values of a and b are given by the following formula Among them, p t and x t They are the observed values of photovoltaic power generation and solar irradiance at time t, t = 1, 2, …, N, and N is the total number of historical moments.
4. The photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification according to claim 2 or claim 3 is characterized in that: For each of the future moments N+1, N+2, ..., N+16, first calculate the predicted solar irradiance value at that moment, substitute the predicted solar irradiance value into the solar irradiance-photovoltaic power conversion model, and obtain the predicted photovoltaic power generation value at that moment, including: querying the meteorological factor vector forecast value of the future moments N+1, N+2, ..., N+16 from the National Meteorological Administration according to the longitude and latitude Substitute the 16 meteorological factor vector forecast values into the irradiance prediction formula to obtain the irradiance forecast values at these 16 moments Substituting the irradiance prediction values at these 16 moments into the solar irradiance-photovoltaic power conversion model, we can obtain the ultra-short-term prediction values of photovoltaic power generation at these moments.
5. A photovoltaic ultra-short-term power prediction device based on adaptive meteorological classification, using the photovoltaic ultra-short-term power prediction method based on adaptive meteorological classification according to any one of claims 1 to 4, characterized in that: include: The first main module is used to obtain a number of meteorological factors and solar irradiance at multiple historical moments, and encapsulate the obtained meteorological factors into a meteorological factor vector; The second main module is used to construct a meteorological category classification model and several irradiance prediction models using an adaptive iterative algorithm; the third main module trains a linear regression model based on solar irradiance sample data and photovoltaic power generation sample data to obtain a solar irradiance-photovoltaic power generation conversion model; the fourth main module obtains the meteorological factor vector forecast value at a future moment from the National Meteorological Administration, substitutes it into the meteorological category classification model to obtain the meteorological category value at that moment, substitutes it into the corresponding irradiance prediction model to obtain the solar irradiance prediction value, and substitutes the solar irradiance prediction value into the solar irradiance-photovoltaic power generation conversion model to obtain the ultra-short-term prediction value of photovoltaic power generation.
6. An electronic device, characterized in that: include: At least one processor, at least one memory and a communication interface; wherein, The processor, memory and communication interface communicate with each other; The memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the method according to any one of claims 1 to 4.
7. A non-transitory computer-readable storage medium, characterized in that: The non-transitory computer-readable storage medium stores computer instructions, which cause the computer to execute the method of any one of claims 1 to 4.
