Generation method of power prediction model, power prediction method and device thereof
By generating a power prediction model that includes an environmental information subordination function and RBF neural network, and the use of genetic algorithm optimization processing, the challenge of photovoltaic power generation prediction is solved, and accurate predictions for the power of photovoltaic systems are achieved.
Patent Information
- Application Number
- CN202411426434.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-12
- Publication Date
- 2025-01-21
AI Technical Summary
The prediction of photovoltaic power generation is affected by a variety of factors, including weather conditions, temperature and solar radiation. The uncertainty and nonlinear characteristics of these factors make photovoltaic power generation prediction a challenge.
A power prediction model generation method is adopted, which includes obtaining power generation information corresponding to historical dates, creating a power prediction model including environmental information membership functions and RBF neural networks, and optimizing the model through genetic algorithms to improve prediction accuracy.
This method can effectively predict the power of the photovoltaic system, improve the accuracy of forecasting, and is suitable for power grid management and energy scheduling.
Smart Images
Figure CN119338018A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of photovoltaic technology, and in particular to a method for generating a power prediction model, a power prediction method and a device thereof. Background Art
[0002] With the rapid development of renewable energy, especially the widespread application of solar photovoltaic power generation, accurate prediction of photovoltaic power generation is of great significance for grid management and energy dispatch. Photovoltaic power generation is affected by many factors, including weather conditions, temperature and solar radiation. The uncertainty and nonlinear characteristics of these factors make photovoltaic power generation prediction a challenge.
[0003] Therefore, how to predict the power of photovoltaic systems has become an urgent problem to be solved. Summary of the invention
[0004] The object of the present invention is to provide a method for generating a power prediction model, a power prediction method and a device thereof.
[0005] In order to achieve one of the above-mentioned purposes of the invention, an embodiment of the present invention provides a method for generating a power prediction model for a photovoltaic system, comprising the following steps: obtaining Num1 historical dates Date1, Date2, ..., Date Num1 Num1 pieces of power generation information correspond to each other, and each piece of power generation information includes at least Num2 pieces of environmental information Env1, Env2, ..., Env corresponding to the historical date. Num2 and photovoltaic output power P s ; Wherein, Num1 and Num2 are both natural numbers, and Num1≥2, Num2≥2; Create a power prediction model, the power prediction model includes: environmental information Env i The corresponding membership function Fun i and RBF neural network, optimizing the power prediction model based on all historical dates and the power generation information corresponding to each historical date, wherein i is a natural number, 1≤i≤Num2.
[0006] As a further improvement of an embodiment of the present invention, the membership functions Fun1, Fun2, ..., Fun Num2 are all Gaussian membership functions; in any membership function Fun i In each case, Num3 fuzzy sets Group are set. i,1 、Group i,2 、…、Group i,Num3 , where Num3 is a natural number and Num3 ≥ 2; in any membership function Fun i In the environment information Env iBelongs to the fuzzy set Group i，j The degree value And c i，1 ≤c i，2 ≤…≤c i,Num3 , σ i，1 ≤σ i，2 ≤…≤σ i，Num3 , x is the environment information Env i The value of, wherein j is a natural number, 1≤j≤Num3; the power prediction model is used to: obtain the input Num2 environmental information Env′1, Env′2, …, Env′ Num2 , any environmental information Env′ i Belongs to the fuzzy set Group i，j The degree value x' is the environment information Env i The value of; then, all degree values μ i，j (x′) is input into the RBF neural network, and the RBF neural network outputs the photovoltaic output power of the photovoltaic system; the “optimizing the power prediction model” specifically includes: in any membership function Fun i c i，j and σ i，j , and the RBF neural network weights w i，j Perform optimization processing.
[0007] As a further improvement of an embodiment of the present invention, Num2=4, the environmental information Env1 is the solar radiation intensity I r , the environmental information Env2 is the temperature T a , environmental information Env3 is humidity H, environmental information Env4 is wind speed W s .
[0008] As a further improvement of one embodiment of the present invention, the "optimizing the power prediction model based on all historical dates and the power generation information corresponding to each historical date" specifically includes: optimizing the power prediction model using a genetic algorithm based on all historical dates and the power generation information corresponding to each historical date.
[0009] As a further improvement of an embodiment of the present invention, the output of the RBF neural network Among them, φ is the radial basis function, F(Groupi,j) is the fuzzy set Group i，j The mean of .
[0010] The embodiment of the present invention also provides a device for generating a power prediction model for a photovoltaic system, comprising the following modules: an information acquisition module for acquiring Num1 historical dates Date1, Date2, ..., DateNum1 Num1 pieces of power generation information correspond to each other, and each piece of power generation information includes at least Num2 pieces of environmental information Env1, Env2, ..., Env corresponding to the historical date. Num2 and photovoltaic output power P s ; Wherein, Num1 and Num2 are both natural numbers, and Num1≥2, Num2≥2; Model creation module, used to create a power prediction model, the power prediction model includes: environmental information Env i The corresponding membership function Fun i and RBF neural network, optimizing the power prediction model based on all historical dates and the power generation information corresponding to each historical date, wherein i is a natural number, 1≤i≤Num2.
[0011] As a further improvement of an embodiment of the present invention, the membership functions Fun1, Fun2, ..., Fun Num2 are all Gaussian membership functions; in any membership function Fun i In each case, Num3 fuzzy sets Group are set. i,1 、Group i,2 、…、Group i,Num3 , where Num3 is a natural number and Num3 ≥ 2; in any membership function Fun i In the environment information Env i Belongs to the fuzzy set Group i，j The degree value And c i，1 ≤c i，2 ≤…≤c i,Num3 , σ i，1 ≤σ i，2 ≤…≤σ i，Num3 , x is the environment information Env i The value of, wherein j is a natural number, 1≤j≤Num3; the power prediction model is used to: obtain the input Num2 environmental information Env′1, Env′2, …, Env′ Num2 , any environmental information Env′ i Belongs to the fuzzy set Group i，j The degree value x' is the environment information Env i The value of; then, all degree values μ i，j (x′) is input into the RBF neural network, and the RBF neural network outputs the photovoltaic output power of the photovoltaic system; the model creation module is also used to: in any membership function Fun i c i，j and σ i，j , and the RBF neural network weights wi，j Perform optimization processing.
[0012] As a further improvement of an embodiment of the present invention, Num2=4, the environmental information Env1 is the solar radiation intensity I r , the environmental information Env2 is the temperature T a , environmental information Env3 is humidity H, environmental information Env4 is wind speed W s .
[0013] An embodiment of the present invention also provides a power prediction method for a photovoltaic system, comprising the following steps: executing the above-mentioned generation method to obtain a power prediction model; obtaining Num1 pieces of to-be-processed power generation information corresponding one to one to Num1 to-be-processed historical dates of the photovoltaic system, inputting the Num1 pieces of to-be-processed power generation information into the power prediction model to obtain the power value of the photovoltaic system.
[0014] An embodiment of the present invention also provides a power prediction device for a photovoltaic system, comprising the following modules: a model generation module, used to execute the above-mentioned generation method to obtain a power prediction model; a processing module, used to obtain Num1 pieces of to-be-processed power generation information corresponding to Num1 to-be-processed historical dates of the photovoltaic system, input the Num1 pieces of to-be-processed power generation information into the power prediction model, and obtain the power value of the photovoltaic system.
[0015] Compared with the prior art, the technical effect of the present invention is that: the embodiment of the present invention provides a method for generating a power prediction model, a power prediction method and a device thereof, the generation method comprising: obtaining Num1 historical dates Date1, Date2, ..., Date Num1 Num1 pieces of power generation information correspond to each other, and each piece of power generation information includes at least Num2 pieces of environmental information Env1, Env2, ..., Env corresponding to the historical date. Num2 and photovoltaic output power P s ; Create a power prediction model, the power prediction model includes: environmental information Env i The corresponding membership function Fun i The power prediction model is optimized based on all historical dates and the power generation information corresponding to each historical date using an RBF neural network. The power prediction model can predict the power of the photovoltaic system. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] Figure 1 It is a flowchart of a method for generating a power prediction model in an embodiment of the present invention. DETAILED DESCRIPTION
[0017] The present invention will be described in detail below in conjunction with the various embodiments shown in the accompanying drawings. However, these embodiments do not limit the present invention, and any structural, methodological, or functional changes made by a person skilled in the art based on these embodiments are all within the scope of protection of the present invention.
[0018] As used herein, terms indicating spatial relative positions such as "upper," "above," "lower," and "below" are used for the purpose of convenience to describe the relationship of one unit or feature relative to another unit or feature as shown in the accompanying drawings. Terms of spatial relative position may be intended to include different orientations of the device in use or operation in addition to the orientation shown in the drawings. For example, if the device in the figure is turned over, units described as being "below" or "beneath" other units or features will be located "above" the other units or features. Thus, the exemplary term "below" can encompass both above and below orientations. The device may be oriented in other ways (rotated 90 degrees or in other orientations), and the spatially relative descriptors used herein interpreted accordingly.
[0019] Embodiment 1 of the present invention provides a method for generating a power prediction model for a photovoltaic system, such as Figure 1 As shown, the following steps are included:
[0020] Step 101: Get Num1 historical dates Date1, Date2, ..., Date Num1 Num1 pieces of power generation information correspond to each other, and each piece of power generation information includes at least Num2 pieces of environmental information Env1, Env2, ..., Env corresponding to the historical date. Num2 and photovoltaic output power P s ; Where Num1 and Num2 are both natural numbers, and Num1≥2, Num2≥2; Here, the photovoltaic output power P s The unit can be kW.
[0021] Step 1 02: Create a power prediction model, the power prediction model includes: environment information Env i The corresponding membership function Fun i and RBF neural network, optimizing the power prediction model based on all historical dates and the power generation information corresponding to each historical date, wherein i is a natural number, 1≤i≤Num2.
[0022] RBF (Radical Basis Function) is a traditional technology for multi-dimensional space interpolation. RBF neural network belongs to the type of forward neural network, which is a three-layer forward network. The basic idea of RBF neural network is: use RBF as the "basis" of hidden units to form the hidden layer space, transform the low-dimensional pattern input data into the high-dimensional space, so that the linear inseparable problem in the low-dimensional space can be linearly separable in the commercial dimensional space. The activation function of the RBF network takes the distance between the input disk and the weight disk as the independent variable, and usually uses a Gaussian function.
[0023] In this embodiment, the membership functions Fun1, Fun2, ..., Fun Num2 are all Gaussian membership functions; in any membership function Fun i In each case, Num3 fuzzy sets Group are set. i,1 、Group i,2 、…、Group i,Num3 , where Num3 is a natural number and Num3 ≥ 2; in any membership function Fun i In the environment information Env i Belongs to the fuzzy set Group i，j The degree value And c i，1 ≤c i，2 ≤…≤c i,Num3 , σ i，1 ≤σ i，2 ≤…≤σ i，Num3 , x is the environment information Env i The value of, wherein j is a natural number, 1≤j≤Num3; the power prediction model is used to: obtain the input Num2 environmental information Env′1, Env′2, …, Env′ Num2 , any environmental information Env′ i Belongs to the fuzzy set Group i，j The degree value x' is the environment information Env i The value of; then, all degree values μ i，j (x′) is input into the RBF neural network, and the RBF neural network outputs the photovoltaic output power of the photovoltaic system; the “optimizing the power prediction model” specifically includes: in any membership function Fun i c i，j and σ i，j , and the RBF neural network weights w i，j Perform optimization processing.
[0024] exist In, c i，jis the center of the Gaussian function, that is, the central value of the jth fuzzy set; σ j is the standard deviation, which determines the width of the Gaussian curve and affects the sensitivity of the membership function.
[0025] Here, for any environment information Env i , fuzzy sets can be divided according to their numerical values. It can be understood that fuzzy set Group i,1 The center is smaller than the fuzzy set Group i,2 The center of the fuzzy set Group i,2 The center is smaller than the fuzzy set Group i,3 The center of, ..., fuzzy set Group i,Num3-1 The center is smaller than the fuzzy set Group i,Num3 center.
[0026] In this embodiment, Num2=4, and the environmental information Env1 is the solar radiation intensity I r , the environmental information Env2 is the temperature T a , environmental information Env3 is humidity H, environmental information Env4 is wind speed W s .
[0027] Here, optionally, for the solar radiation intensity I r , which corresponds to 3 fuzzy sets Group 1,1 、Group 1,2 、Group 1,3 , fuzzy set Group 1,1 The corresponding c 1,1 =200 and the corresponding σ 1,1 =50, fuzzy set Group 1,2 The corresponding c 1,2 =500 and the corresponding σ 1,2 =100, fuzzy set Group 1,3 The corresponding c 1,3 =500 and the corresponding σ 1，3 =100.
[0028] For temperature T a , which corresponds to 3 fuzzy sets Group 2,1 、Group 2,2 、Group 2,3 , fuzzy set Group 2,1 The corresponding c 2,1 =10 and the corresponding σ 2,1 =5, fuzzy set Group 2,2 The corresponding c 2,2 =20 and the corresponding σ 2,2 =5, fuzzy set Group 2.3The corresponding c 2,3 =30 and the corresponding σ 2，3 =5.
[0029] Here, the solar radiation intensity I r The unit can be W / m, temperature T a The unit can be ℃.
[0030] In this embodiment, the "optimizing the power prediction model based on all historical dates and the power generation information corresponding to each historical date" specifically includes: optimizing the power prediction model using a genetic algorithm based on all historical dates and the power generation information corresponding to each historical date.
[0031] When executing the genetic algorithm, first, generate several individuals, each of which contains Num2*Num3 triples (c i，j , σ i,j , w), 1≤i≤Num2, 1≤j≤Num3;
[0032] Perform the following steps: For any triplet, Num2 pieces of environmental information in the power generation information corresponding to any historical date Date1 are input into the power prediction model, and the power prediction model outputs the photovoltaic output power The photovoltaic output power in this power generation information The mean square error (MSE) can be used as the fitness function to evaluate the performance of each individual Thus, the MSE corresponding to each individual is obtained; the individuals with 1 / MSE < the preset threshold are discarded, and the MSEs of the remaining individuals are added to obtain the total fitness score Sum, and the selection probability of each individual is calculated, for example, the selection probability = MSE / Sum of the individual; then, a random number r (ranging from 0 to 1) is generated, and the first individual whose cumulative selection probability is greater than or equal to r is selected on the roulette wheel, so as to select several individuals, and then the selected individuals are crossover and mutation are performed to generate a new generation of populations to introduce genetic diversity. Here, crossover and mutation can be specifically as follows: for a triple (c, σ, w), the value of c, σ or w can be increased or decreased by one value. For two triplets (c, σ, w), c in one triple can be replaced by c in another triple, σ in one triple can be replaced by σ in another triple, and w in one triple can be replaced by w in another triple.
[0033] Select Example
[0034] Suppose we have 5 individuals, and their fitness (MSE) is shown in the following table:
[0035]
[0036]
[0037] Total fitness score = 5.0 + 3.33 + 2.0 + 10.0 + 2.5 = 22.83, the selection probability of individual A = 5.0 / 22.83≈0.22, the selection probability of individual B = 3.33 / 22.83≈0.15, the selection probability of individual C = 2.0 / 22.83≈0.09, the selection probability of individual D = 10.0 / 22.83≈0.44, and the selection probability of individual E = 2.5 / 22.83≈0.11.
[0038] Cumulative selection probability: Cumulative selection probability of individual A = 0.22, cumulative selection probability of individual B = 0.22 + 0.15 = 0.37, cumulative selection probability of individual C = 0.37 + 0.09 = 0.46, cumulative selection probability of individual D = 0.46 + 0.44 = 0.90, cumulative selection probability of individual E = 0.90 + 0.11 = 1.0,
[0039] Random selection: Generate a random number r. For example, if r=0.4, then select the individual whose cumulative selection probability is just greater than 0.4, that is, individual B.
[0040] Selection mechanism summary, in the selection step of the genetic algorithm, the selection probability of each individual is calculated according to its fitness, and then the individuals with higher fitness are selected to participate in the reproduction of the next generation through a random selection mechanism (such as roulette selection). Selecting individuals with higher fitness helps to improve the overall performance of the next generation and gradually optimize the photovoltaic power prediction model.
[0041] In this embodiment, the output of the RBF neural network Where φ is the radial basis function, F(Group i,j ) is the fuzzy set Group i，j The mean of .
[0042] Optionally, the radial basis functions include: Gaussian function, polyquadratic function, inverse quadratic function, inverse polyquadratic function, polyharmonic spline, thin plate spline and the like.
[0043] Embodiment 2 of the present invention provides a device for generating a power prediction model for a photovoltaic system, comprising the following modules: an information acquisition module for acquiring Num1 historical dates Date1, Date2, ..., Date Num1 Num1 pieces of power generation information correspond to each other, and each piece of power generation information includes at least Num2 pieces of environmental information Env1, Env2, ..., Env corresponding to the historical date. Num2 and photovoltaic output power P s; Wherein, Num1 and Num2 are both natural numbers, and Num1≥2, Num2≥2; Model creation module, used to create a power prediction model, the power prediction model includes: environmental information Env i The corresponding membership function Fun i and RBF neural network, optimizing the power prediction model based on all historical dates and the power generation information corresponding to each historical date, wherein i is a natural number, 1≤i≤Num2.
[0044] In this embodiment, the membership functions Fun1, Fun2, ..., Fun Num2 are all Gaussian membership functions; in any membership function Fun i In each case, Num3 fuzzy sets Group are set. i,1 、Group i,2 、…、Group i,Num3 , where Num3 is a natural number and Num3 ≥ 2; in any membership function Fun i In the environment information Env i Belongs to the fuzzy set Group i，j The degree value And c i，1 ≤c i，2 ≤…≤c i,Num3 , σ i，1 ≤σ i，2 ≤…≤σ i，Num3 , x is the environment information Env i The value of, wherein j is a natural number, 1≤j≤Num3; the power prediction model is used to: obtain the input Num2 environmental information Env′1, Env′2, …, Env′ Num2 , any environmental information Env′ i Belongs to the fuzzy set Group i，j The degree value x' is the environment information Env i The value of; then, all degree values μ i，j (x′) is input into the RBF neural network, and the RBF neural network outputs the photovoltaic output power of the photovoltaic system; the model creation module is also used to: in any membership function Fun i c i，j and σ i，j , and the RBF neural network weights w i，j Perform optimization processing.
[0045] In this embodiment, Num2=4, and the environmental information Env1 is the solar radiation intensity I r , the environmental information Env2 is the temperature T a, environmental information Env3 is humidity H, environmental information Env4 is wind speed W s .
[0046] Embodiment 3 of the present invention provides a power prediction method for a photovoltaic system, comprising the following steps: executing the generation method in embodiment 1 to obtain a power prediction model; obtaining Num1 pieces of to-be-processed power generation information corresponding to Num1 to-be-processed historical dates of the photovoltaic system, inputting the Num1 pieces of to-be-processed power generation information into the power prediction model to obtain the power value of the photovoltaic system.
[0047] Embodiment 4 of the present invention provides a power prediction device for a photovoltaic system, comprising the following modules: a model generation module, used for the generation method in embodiment 1 to obtain a power prediction model; a processing module, used for obtaining Num1 pieces of to-be-processed power generation information corresponding to Num1 to-be-processed historical dates of the photovoltaic system, inputting the Num1 pieces of to-be-processed power generation information into the power prediction model to obtain the power value of the photovoltaic system.
[0048] It should be understood that although this specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each implementation mode may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.
[0049] The series of detailed descriptions listed above are only specific descriptions of feasible implementation methods of the present invention. They are not intended to limit the scope of protection of the present invention. Any equivalent implementation methods or changes that do not deviate from the technical spirit of the present invention should be included in the scope of protection of the present invention.
Claims
1. A method for generating a power prediction model for a photovoltaic system, characterized in that: The following steps are involved: Get Num1 historical dates Date1, Date2, ..., Date Num1 Num1 pieces of power generation information correspond to each other, and each piece of power generation information includes at least Num2 pieces of environmental information Env1, Env2, ..., Env corresponding to the historical date. Num2 and photovoltaic output power P s ; Wherein, Num1 and Num2 are both natural numbers, and Num1≥2, Num2≥2; Create a power prediction model, the power prediction model includes: environment information Env o The corresponding membership function Fun i and RBF neural network, optimizing the power prediction model based on all historical dates and the power generation information corresponding to each historical date, wherein i is a natural number, 1≤i≤Num2.
2. The generation method according to claim 1, characterized in that: Membership functions Fun1, Fun2, ..., Fun Num2 All are Gaussian membership functions; In any membership function Fun i In each case, Num3 fuzzy sets Group are set. i,1 、Group i,2 , ..., Group i,Num3 , where Num3 is a natural number and Num3 ≥ 2; In any membership function Fun i In the environment information Env i Belongs to the fuzzy set Group i,j The degree value And c i,1 ≤c i,2 ≤...≤c i,Num3 , σ i,1 ≤σ i,2 ≤...≤σ i,Num3 , x is the environment information Env i The value of, where j is a natural number, 1≤j≤Num3; The power prediction model is used to obtain the input Num2 pieces of environmental information Env′1, Env′2, ..., Env′ Num2 , any environmental information Env′ i Belongs to the fuzzy set Group i,j The degree value For environment information Env i The value of; then, all degree values μ i,j (x′) is input to the RBF neural network, and the RBF neural network outputs the photovoltaic output power of the photovoltaic system; The “optimizing the power prediction model” specifically includes: i c i,j and σ i,j , and the RBF neural network weights w i,j Perform optimization processing.
3. The generation method according to claim 1, characterized in that: Num2=4, environmental information Env1 is the solar radiation intensity I r , the environmental information Env2 is the temperature T a , environmental information Env3 is humidity H, environmental information Env4 is wind speed W s .
4. The generation method according to claim 2, characterized in that: The “optimizing the power prediction model based on all historical dates and the power generation information corresponding to each historical date” specifically includes: Based on all historical dates and the power generation information corresponding to each historical date, the power prediction model is optimized using a genetic algorithm.
5. The generation method according to claim 2, characterized in that: The output of the RBF neural network Where φ is the radial basis function, F(Group i,j ) is the fuzzy set Group i,j The mean of .
6. A device for generating a power prediction model for a photovoltaic system, characterized in that: Includes the following modules: Information acquisition module, used to obtain Num1 historical dates Date1, Date2, ..., Date Num1 Num1 pieces of power generation information correspond to each other, and each piece of power generation information includes at least Num2 pieces of environmental information Env1, Env2, ..., Env corresponding to the historical date. Num2 and photovoltaic output power P s ; Wherein, Num1 and Num2 are both natural numbers, and Num1≥2, Num2≥2; A model creation module is used to create a power prediction model, wherein the power prediction model includes: environment information Env i The corresponding membership function Fun i and RBF neural network, optimizing the power prediction model based on all historical dates and the power generation information corresponding to each historical date, wherein i is a natural number, 1≤i≤Num2.
7. The generating device according to claim 6, characterized in that Membership functions Fun1, Fun2, ..., Fun Num2 All are Gaussian membership functions; In any membership function Fun i In each case, Num3 fuzzy sets Group are set. i,1 、Group i,2 , ..., Group i,Num3 , where Num3 is a natural number and Num3 ≥ 2; In any membership function Fun i In the environment information Env i Belongs to the fuzzy set Group i,j The degree value And c i,1 ≤c i,2 ≤...≤c i,Num3 , σ i,1 ≤σ i,2 ≤...≤σ i,Num3 , x is the environment information Env i The value of, where j is a natural number, 1≤j≤Num3; The power prediction model is used to obtain the input Num2 pieces of environmental information Env′1, Env′2, ..., Env′ Num2 , any environmental information Env′ i Belongs to the fuzzy set Group i,j The degree value For environment information Env i The value of; then, all degree values μ i,j (x′) is input to the RBF neural network, and the RBF neural network outputs the photovoltaic output power of the photovoltaic system; The model creation module is also used to: in any membership function Fun o c o,j and σ o,k , and the RBF neural network weights w i,j Perform optimization processing.
8. The generating device according to claim 6, characterized in that Num2=4, environmental information Env1 is the solar radiation intensity I r , the environmental information Env2 is the temperature T a , environmental information Env3 is humidity H, environmental information Env4 is wind speed W s .
9. A method for predicting power of a photovoltaic system, characterized in that: The following steps are involved: Execute the generation method described in any one of claims 1 to 5 to obtain a power prediction model; Obtain Num1 pieces of to-be-processed power generation information corresponding to Num1 to-be-processed historical dates of the photovoltaic system, input the Num1 pieces of to-be-processed power generation information into the power prediction model, and obtain the power value of the photovoltaic system.
10. A power prediction device for a photovoltaic system, characterized in that: Includes the following modules: A model generation module, used to execute the generation method according to any one of claims 1 to 5 to obtain a power prediction model; The processing module is used to obtain Num1 to-be-processed power generation information corresponding to Num1 to-be-processed historical dates of the photovoltaic system, input the Num1 to-be-processed power generation information into the power prediction model, and obtain the power value of the photovoltaic system.