Photovoltaic power prediction model construction method and photovoltaic power prediction method
By constructing a photovoltaic power prediction model, learning the correspondence between the historical irradiance and power data of flat single-axis photovoltaic modules, the problem of inaccurate power prediction of photovoltaic power plants in flat single-axis photovoltaic modules is solved, and the surface radiation of photovoltaic panels is realized is realized Adaptability prediction of degree changes.
Patent Information
- Application Number
- CN202411072747.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-06
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-08-06
AI Technical Summary
In the prior art, the power prediction method of a photovoltaic power station equipped with flat single-axis photovoltaic modules is inaccurate, and the influence of the change in the surface irradiance of the photovoltaic panel on the photovoltaic power caused by the installation of flat single-axis photovoltaic brackets cannot be effectively utilized. .
Construct a photovoltaic power prediction model, calculate the historical surface irradiance data of flat single-axis photovoltaic modules by obtaining historical measured irradiance and power data, and divide it into point value sequences as input features and historical power point value The sequence is used as the target value to build the training data set, and the pre-constructed photovoltaic power prediction model is used for training, learning the correspondence between irradiance and power, and photovoltaic panels in the form of flat single-axis photovoltaic brackets are used for prediction.
It improves the accuracy of photovoltaic power prediction, can adapt to the surface irradiance changes of photovoltaic panels in the form of flat single-axis photovoltaic brackets, accurately predicts photovoltaic power, and overcomes the problem of inaccurate prediction results in traditional methods.
Smart Images

Figure CN118940041A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of new energy technology, and in particular to a photovoltaic power prediction model training method and a photovoltaic power prediction method. Background Art
[0002] At present, photovoltaic panels in photovoltaic power stations are usually installed in the form of fixed photovoltaic brackets. When photovoltaic power is predicted based on this installation form, the output characteristics of the photovoltaic panels often show a linear function relationship with the total horizontal radiation. The total horizontal radiation and this linear function relationship are used in the prediction to obtain the photovoltaic power generation power corresponding to the total horizontal radiation.
[0003] In the prior art, photovoltaic panels in photovoltaic power stations are also installed in the form of flat single-axis photovoltaic brackets; the flat single-axis photovoltaic bracket is horizontal in the north-south direction, and the array plane rotates east-west with the rotation axis according to the position of the sun; compared with the fixed photovoltaic bracket installation, this installation method makes full use of direct sunlight in the morning and evening, and has better photovoltaic power generation efficiency; however, under this installation method, although the flat single-axis output curve has a strong positive correlation with the irradiance, the output characteristics and the total radiation are no longer a linear function relationship. Therefore, for photovoltaic power stations equipped with flat single-axis single-sided photovoltaic modules, when the traditional fixed photovoltaic bracket installation form prediction method is still used for power prediction, the power prediction result is inaccurate. Summary of the invention
[0004] In view of this, the present invention provides a photovoltaic power prediction model training method and a photovoltaic power prediction method to solve the problem of inaccurate power prediction for a photovoltaic power station equipped with flat single-axis single-sided photovoltaic modules.
[0005] In a first aspect, the present invention provides a method for constructing a photovoltaic power prediction model, which is applied to a target photovoltaic power station equipped with flat single-axis single-sided photovoltaic modules; the photovoltaic power prediction model includes: obtaining historical measured irradiance data and historical power data of the target photovoltaic power station within a preset time period; calculating the historical surface irradiance data of each of the flat single-axis single-sided photovoltaic modules based on each of the historical measured irradiance data; dividing the historical surface irradiance data and the historical power data according to preset time intervals to obtain a historical surface irradiance point value sequence and a historical power point value sequence; using the historical surface irradiance point value sequence as an input feature and the historical power point value sequence as a target value to construct a first training data set; inputting the first training data set into a pre-constructed photovoltaic power prediction model for model training to obtain the photovoltaic power prediction model; in the process of model training, learning the correspondence between each of the historical surface irradiance point value sequences and the historical power point value sequences.
[0006] As an exemplary embodiment, the method of using the historical surface irradiance point value sequence as input features and the historical power point value sequence as target values to construct a first training data set includes: obtaining historical difference features of historical surface irradiance data and historical power data at each moment on a time scale of days, and constructing a time-series related historical difference feature sequence; using the historical difference feature sequence and the historical surface irradiance point value sequence as input features and the historical power point value sequence as target values to construct a first training data set.
[0007] As an exemplary embodiment, the method of obtaining historical difference characteristics of historical surface irradiance data and historical power data at each moment on a day time scale and constructing a time-series-related historical difference characteristic sequence includes: calculating the historical change trend of the historical surface irradiance point value sequence relative to the historical power point value sequence at each moment on a day time scale; dividing the historical change trend into a first sub-historical change trend in which the historical surface irradiance point value sequence is positively correlated with the historical power point value sequence and a second sub-historical change trend in which the historical surface irradiance point value sequence is not positively correlated with the historical power point value sequence; and constructing a first historical difference characteristic sequence and a second historical difference characteristic sequence that are time-series-related based on the first sub-historical change trend and the second sub-historical change trend, respectively.
[0008] As an exemplary embodiment, the method for constructing the photovoltaic power prediction model also includes: respectively obtaining a first historical change period corresponding to the first historical difference feature sequence and a second historical change period corresponding to the second historical difference feature sequence; dividing the first training data set based on the first historical change period and the second historical change period to obtain a second training data set; wherein the second training data set includes a second training data subset consisting of a plurality of historical surface irradiance point value sequences and historical power point value sequences with different historical change trends divided according to the first historical change period and the second historical change period; inputting each of the second training data subsets into a pre-constructed photovoltaic power prediction model for model training to obtain a first photovoltaic power prediction sub-model corresponding to each of the historical change periods; in the process of model training, learning the correspondence between each of the historical surface irradiance point value sequences and the historical power point value sequences in the second training data subset corresponding to each of the historical change periods based on each of the first photovoltaic power prediction sub-models.
[0009] As an exemplary embodiment, the method for constructing the photovoltaic power prediction model also includes: obtaining the historical temperature point value sequence and / or the historical wind speed point value sequence corresponding to each of the second training data subsets; adding the historical temperature point value sequence and / or the historical wind speed point value sequence to the corresponding second training data subset to obtain a third training data subset; inputting each of the third training data subsets into a pre-constructed photovoltaic power prediction model for model training to obtain a second photovoltaic power prediction sub-model corresponding to each of the historical change time periods; during the model training process, based on each of the second photovoltaic power prediction sub-models, learning the correspondence between each of the historical temperature point value sequence and / or the historical wind speed point value sequence, the historical surface irradiance point value sequence and the historical power point value sequence in the third training data subset corresponding to each of the historical change time periods.
[0010] In a second aspect, the present invention provides a photovoltaic power prediction method, which is applied to a target photovoltaic power station equipped with flat single-axis single-sided photovoltaic modules; the photovoltaic power prediction method includes: obtaining predicted irradiance data of the target photovoltaic power station; calculating the predicted surface irradiance data of each of the flat single-axis single-sided photovoltaic modules based on the predicted irradiance data; dividing the predicted surface irradiance data according to a preset time interval to obtain a predicted surface irradiance point value sequence; inputting the predicted surface irradiance point value sequence into a pre-trained photovoltaic power prediction model for power prediction to obtain a power prediction result; wherein the pre-trained photovoltaic power prediction model is trained by using a historical surface irradiance point value sequence as an input feature and a historical power point value sequence as a target value to construct a first training data set, and the historical surface irradiance point value sequence and the historical power point value sequence are obtained by obtaining the historical measured irradiance data and historical power data of the target photovoltaic power station within a preset time period, calculating the historical surface irradiance data of each of the flat single-axis single-sided photovoltaic modules based on each of the historical measured irradiance data, and dividing them according to a preset time interval.
[0011] As an exemplary embodiment, the photovoltaic power prediction method also includes: obtaining a time-series-related historical difference feature sequence constructed by constructing historical surface irradiance data and historical power data of the target photovoltaic power station at each moment on a daily time scale; selecting the historical difference feature sequence corresponding to the historical surface irradiance data whose similarity with the predicted surface irradiance point value sequence is greater than a preset degree as a predicted difference feature sequence and inputting it into a pre-trained photovoltaic power prediction model for power prediction to obtain a photovoltaic power prediction result.
[0012] As an exemplary embodiment, the method of obtaining the historical surface irradiance data and the historical power data of the target photovoltaic power station at each moment on a day time scale to construct a time-related historical difference feature sequence includes: calculating the historical change trend of the historical surface irradiance point value sequence at each moment on a day time scale relative to the historical power point value sequence; dividing the historical change trend into a first sub-historical change trend in which the historical surface irradiance point value sequence is positively correlated with the historical power point value sequence and a second sub-historical change trend in which the historical surface irradiance point value sequence is not positively correlated with the historical power point value sequence; and constructing a first historical difference feature sequence and a second historical difference feature sequence that are time-related based on the first sub-historical change trend and the second sub-historical change trend, respectively.
[0013] As an exemplary embodiment, the photovoltaic power prediction method also includes: obtaining the predicted change period corresponding to each of the predicted surface irradiance point value sequences with days as the time scale; based on the predicted change period, inputting the predicted surface irradiance point value corresponding to the predicted change period into the first photovoltaic power prediction sub-model corresponding to the historical change period matching the predicted change period to perform power prediction, and obtain a photovoltaic power prediction result; wherein, the first photovoltaic power prediction sub-model is trained by a second training data subset consisting of a historical surface irradiance point value sequence and a historical power point value sequence according to the historical change trend corresponding to the historical change period, and during the training process, the correspondence between each of the historical surface irradiance point value sequences and the historical power point value sequence in the second training data subset corresponding to each of the historical change periods is learned.
[0014] As an exemplary embodiment, the photovoltaic power prediction method also includes: obtaining a predicted temperature point value sequence and / or a predicted wind speed point value sequence corresponding to the predicted surface irradiance point value corresponding to each of the predicted change time periods; respectively inputting the predicted surface irradiance point value and the predicted temperature point value sequence and / or the predicted wind speed point value sequence into a pre-trained second photovoltaic power prediction sub-model, wherein the second photovoltaic power prediction sub-model is a third training data subset consisting of a predicted temperature point value sequence and / or a predicted wind data point value sequence according to a predicted change trend under the predicted change time period, a predicted surface irradiance point value sequence, and the predicted surface irradiance point value sequence, and the correspondence between the predicted temperature point value sequence and / or the predicted wind data point value sequence, the predicted surface irradiance point value sequence, and the predicted power point value sequence is obtained by learning during the training process.
[0015] The present invention provides a method for constructing a photovoltaic power prediction model, which is applied to a target photovoltaic power station equipped with flat single-axis single-sided photovoltaic modules; the photovoltaic power prediction model comprises: obtaining historical measured irradiance data and historical power data of the target photovoltaic power station within a preset time period; calculating the historical surface irradiance data of each of the flat single-axis single-sided photovoltaic modules based on each of the historical measured irradiance data; dividing the historical surface irradiance data and the historical power data according to preset time intervals to obtain a historical surface irradiance point value sequence and a historical power point value sequence; using the historical surface irradiance point value sequence as an input feature and the historical power point value sequence as a target value to construct a first training data set; inputting the first training data set into a pre-constructed photovoltaic power prediction model for model training to obtain the photovoltaic power prediction model; in the process of model training, learning the corresponding relationship between each of the historical surface irradiance point value sequences and the historical power point value sequences.
[0016] As an exemplary embodiment, the method of using the historical surface irradiance point value sequence as input features and the historical power point value sequence as target values to construct a first training data set includes: obtaining historical difference features of historical surface irradiance data and historical power data at each moment on a time scale of days, and constructing a time-series related historical difference feature sequence; using the historical difference feature sequence and the historical surface irradiance point value sequence as input features and the historical power point value sequence as target values to construct a first training data set.
[0017] As an exemplary embodiment, the method of obtaining historical difference characteristics of historical surface irradiance data and historical power data at each moment on a day time scale and constructing a time-series-related historical difference characteristic sequence includes: calculating the historical change trend of the historical surface irradiance point value sequence relative to the historical power point value sequence at each moment on a day time scale; dividing the historical change trend into a first sub-historical change trend in which the historical surface irradiance point value sequence is positively correlated with the historical power point value sequence and a second sub-historical change trend in which the historical surface irradiance point value sequence is not positively correlated with the historical power point value sequence; and constructing a first historical difference characteristic sequence and a second historical difference characteristic sequence that are time-series-related based on the first sub-historical change trend and the second sub-historical change trend, respectively.
[0018] As an exemplary embodiment, the method for constructing the photovoltaic power prediction model also includes: respectively obtaining a first historical change period corresponding to the first historical difference feature sequence and a second historical change period corresponding to the second historical difference feature sequence; dividing the first training data set based on the first historical change period and the second historical change period to obtain a second training data set; wherein the second training data set includes a second training data subset consisting of a plurality of historical surface irradiance point value sequences and historical power point value sequences with different historical change trends divided according to the first historical change period and the second historical change period; inputting each of the second training data subsets into a pre-constructed photovoltaic power prediction model for model training to obtain a first photovoltaic power prediction sub-model corresponding to each of the historical change periods; in the process of model training, learning the correspondence between each of the historical surface irradiance point value sequences and the historical power point value sequences in the second training data subset corresponding to each of the historical change periods based on each of the first photovoltaic power prediction sub-models.
[0019] As an exemplary embodiment, the method for constructing the photovoltaic power prediction model also includes: obtaining the historical temperature point value sequence and / or the historical wind speed point value sequence corresponding to each of the second training data subsets; adding the historical temperature point value sequence and / or the historical wind speed point value sequence to the corresponding second training data subset to obtain a third training data subset; inputting each of the third training data subsets into a pre-constructed photovoltaic power prediction model for model training to obtain a second photovoltaic power prediction sub-model corresponding to each of the historical change time periods; during the model training process, based on each of the second photovoltaic power prediction sub-models, learning the correspondence between each of the historical temperature point value sequence and / or the historical wind speed point value sequence, the historical surface irradiance point value sequence and the historical power point value sequence in the third training data subset corresponding to each of the historical change time periods.
[0020] The present invention provides a method for constructing a photovoltaic power prediction model, which is applied to a target photovoltaic power station equipped with flat single-axis single-sided photovoltaic modules; the photovoltaic power prediction model comprises: obtaining historical measured irradiance data and historical power data of the target photovoltaic power station within a preset time period; calculating the historical surface irradiance data of each of the flat single-axis single-sided photovoltaic modules based on each of the historical measured irradiance data; dividing the historical surface irradiance data and the historical power data according to preset time intervals to obtain a historical surface irradiance point value sequence and a historical power point value sequence; using the historical surface irradiance point value sequence as an input feature and the historical power point value sequence as a target value to construct a first training data set; inputting the first training data set into a pre-constructed photovoltaic power prediction model for model training to obtain the photovoltaic power prediction model; in the model During the training process, the correspondence between each of the historical surface irradiance point value sequences and the historical power point value sequence is learned; in the method for constructing a photovoltaic power prediction model of the present application, the first training data set is constructed using the historical surface irradiance point value sequence as an input feature, and the historical power point value sequence is constructed as a target value, and the historical surface irradiance point value sequence is based on each of the historical measured irradiance data, taking into account the change in irradiance caused by the plane of the photovoltaic panel array installed on the flat single-axis photovoltaic bracket with the rotation axis rotating east-west with the position of the sun. Therefore, the photovoltaic power prediction model of this embodiment can learn the impact of changes in the surface irradiance of the photovoltaic panel caused by the installation form of the flat single-axis photovoltaic bracket on the photovoltaic power prediction during the training process, and can adaptively predict photovoltaic power for photovoltaic panels installed in the form of flat single-axis photovoltaic brackets. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0022] Figure 1 It is a graph of irradiance and power of a photovoltaic panel installed with a flat single-axis photovoltaic bracket in the related art;
[0023] Figure 2 is a flow chart of a method for constructing a photovoltaic power prediction model according to an embodiment of the present invention;
[0024] Figure 3 is a flow chart of a photovoltaic power prediction method according to an embodiment of the present invention. DETAILED DESCRIPTION
[0025] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution 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 those skilled in the art without creative work are within the scope of protection of the present invention.
[0026] In the related art, photovoltaic panels in photovoltaic power stations are usually installed in the form of fixed photovoltaic brackets. When photovoltaic power prediction is performed based on this installation form, the output characteristics of the photovoltaic panels often show a linear function relationship with the total horizontal radiation. The total horizontal radiation and this linear function relationship are used in the prediction to obtain the photovoltaic power generation power corresponding to the total horizontal radiation.
[0027] However, the photovoltaic panels in the photovoltaic power station are still installed in the form of flat single-axis photovoltaic brackets; the flat single-axis photovoltaic bracket is horizontal in the north-south direction, and the array plane rotates east-west with the rotation axis according to the position of the sun; compared with the fixed photovoltaic bracket installation, this installation method fully utilizes the direct sunlight in the morning and evening, and has better photovoltaic power generation efficiency; however, under this installation method, although the flat single-axis output curve has a strong positive correlation with the irradiance, the output characteristics are no longer a linear function relationship with the total radiation. Specifically, Figure 1 It is a curve diagram of the output characteristics of the fixed photovoltaic panels and the photovoltaic panels installed by the flat single-axis photovoltaic bracket in the related art, such as Figure 1 As shown, the first output characteristic curve 11 is a curve of the output characteristics of the photovoltaic panel installed with a flat single-axis photovoltaic bracket, and the second output characteristic curve 12 is a curve of the output characteristics of the fixedly installed photovoltaic panel. The trend of the fixed photovoltaic basically changes with the trend of the solar radiation intensity, but the irradiance received by the flat single-axis photovoltaic changes with the rotation of the bracket and the azimuth, altitude and incident angle. The angle changes continuously with time, resulting in the final output characteristic being the highest output in the morning and evening, and the output showing a "depression" at noon. Taking the day as the time scale, it can be distinguished into the first output characteristic events 13 and 15 in which the solar irradiance and power change in a positive correlation, and the second output characteristic event 14 in which the solar irradiance and power change in a non-positive correlation; therefore, for a photovoltaic power station equipped with a flat single-axis single-sided photovoltaic module, when the power prediction method of the traditional fixed photovoltaic bracket installation is still used for power prediction, the power prediction result is inaccurate.
[0028] To solve the above problems, in the first aspect, according to an embodiment of the present invention, a method for constructing a photovoltaic power prediction model is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0029] In this embodiment, a method for constructing a photovoltaic power prediction model is provided. Figure 2 is a flow chart of a method for constructing a photovoltaic power prediction model according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps:
[0030] Step S101, obtaining historical measured irradiance data and historical power data of a target photovoltaic power station within a preset time period.
[0031] Exemplarily, the historical measured irradiance data can be measured by a sensor that detects solar irradiance, for example, by setting a solar total radiation sensor in the target photovoltaic power station and obtaining the data by reading the historical data in the solar total radiation sensor; the historical power data can be obtained by reading the historical data of the target photovoltaic power station.
[0032] In the present invention, the power generation power expected to be predicted by the target photovoltaic power station may be an ultra-short-term power generation power; wherein, the ultra-short-term power generation power prediction period is usually 24 hours, and within each day, the time resolution of the predicted power data is 15 minutes; therefore, in the present invention, the preset time length is a time length greater than 24 hours.
[0033] Step S102, calculating the historical surface irradiance data of each flat single-axis single-sided photovoltaic module based on the historical measured irradiance data.
[0034] As mentioned above, the irradiance received by the photovoltaic panels installed on the horizontal single-axis photovoltaic bracket changes with the rotation of the bracket and the azimuth, altitude and incident angle over time, resulting in the final output characteristics being the highest output in the morning and evening, and the output showing a "sag" at noon; therefore, after obtaining the historical measured irradiance data, the irradiance of the horizontal single-axis monofacial photovoltaic module (Monofacial PV with Horizontal Single-axis Tracking, MHST) can be calculated according to the latitude and longitude, altitude, incident angle and other factors of the target photovoltaic power station according to the following formula; specifically, the MHST can be calculated using the formula (1):
[0035]
[0036] In formula (1), ω is the inclination angle of the flat uniaxial photovoltaic module to the horizontal plane following the change of the solar hour angle ω, ρ is the ground reflection, is the local latitude, α is the solar altitude angle, i.e. the angle between the sun’s rays and the horizon, and δ is the solar declination angle; where Q DNI Horizontal direct radiation irradiance, Q DHI is the horizontal diffuse radiation irradiance, Q GRI is the ground reflected radiation irradiance, Q MHST The calculated historical surface irradiance data of each flat uniaxial monofacial photovoltaic module.
[0037] After obtaining the historical measured irradiance data, the Q under each historical measured irradiance data can be calculated by formula (1). MHST .
[0038] The above method of this embodiment calculates the historical surface irradiance data of each flat single-axis single-sided photovoltaic module based on the historical measured irradiance data. It can take into account the change in irradiance caused by the plane of the photovoltaic panel array installed on the flat single-axis photovoltaic bracket rotating east-west with the position of the sun, and then consider the impact of the change in irradiance on the output power of the photovoltaic panel through the model during subsequent model training.
[0039] Step S103, dividing the historical surface irradiance data and the historical power data according to a preset time interval to obtain a historical surface irradiance point value sequence and a historical power point value sequence;
[0040] Exemplarily, in order to make the time scale of the input data match the time scale of the predicted power data during model training, the preset time interval may be 24 hours; in this embodiment, the historical surface irradiance data and the historical power data are divided with a period of 24 hours and a time resolution of 15 minutes to obtain a historical surface irradiance point value sequence and a historical power point value sequence.
[0041] Step S104: construct a first training data set by taking the historical surface irradiance point value sequence as an input feature and the historical power point value sequence as a target value.
[0042] Step S105, inputting the first training data set into a pre-built photovoltaic power prediction model for model training to obtain a photovoltaic power prediction model; during the model training process, learning the correspondence between each historical surface irradiance point value sequence and the historical power point value sequence.
[0043] In this embodiment, the photovoltaic power prediction model may be a convolutional neural network (CNN) model; wherein the input feature of the CNN model is a surface irradiance point value sequence, and the output feature is a power point value sequence.
[0044] The method for constructing a photovoltaic power prediction model of the present application is that the first training data set is constructed by using a historical surface irradiance point value sequence as an input feature and a historical power point value sequence as a target value, and the historical surface irradiance point value sequence is calculated based on various historical measured irradiance data taking into account the change in irradiance caused by the plane of the photovoltaic panel array installed on a flat single-axis photovoltaic bracket with the rotation axis rotating east-west with the position of the sun. Therefore, the photovoltaic power prediction model of this embodiment can learn the impact of changes in the surface irradiance of the photovoltaic panel caused by the installation form of the flat single-axis photovoltaic bracket on the photovoltaic power prediction during the training process, and can adaptively predict photovoltaic power for photovoltaic panels installed in the form of flat single-axis photovoltaic brackets.
[0045] As an exemplary embodiment, taking the historical surface irradiance point value sequence as input features and the historical power point value sequence as target values to construct a first training data set includes: obtaining historical difference features of historical surface irradiance data and historical power data at each moment on a time scale of days, and constructing a time-series related historical difference feature sequence; taking the historical difference feature sequence and the historical surface irradiance point value sequence as input features and the historical power point value sequence as target values to construct a first training data set.
[0046] The irradiance received by the flat single-axis photovoltaic system changes continuously with the rotation of the bracket and the azimuth, altitude and incident angle over time, resulting in the final output characteristic being the highest output in the morning and evening and a concave output at noon; in order to enable the model to consider the difference between the above-mentioned output characteristics and the prior art in which irradiance and power are in a linear function relationship during model training, in this embodiment, a historical difference feature sequence is constructed based on the historical difference characteristics of the historical surface irradiance data and the historical power data at each moment on a time scale of day; wherein the historical difference feature sequence is used to represent the output characteristics suitable for the flat single-axis photovoltaic system, specifically, the historical difference feature may include the historical change trend of the historical surface irradiance point value sequence at each moment relative to the historical power point value sequence, and the historical change trend may be the change slope, change difference, etc. of the historical surface irradiance point value sequence relative to the historical power point value sequence.
[0047] Furthermore, a time-series related historical difference feature sequence is constructed; the historical difference feature sequence and the historical surface irradiance point value sequence are used as input features, and the historical power point value sequence is used as the target value to construct a first training data set, and the model is trained based on the first training data set so that the model can consider the correspondence between the historical surface irradiance point value sequence and the power sequence corresponding to the historical difference feature sequence reflected by the difference in output characteristics brought about by the installation form of parallel single-axis photovoltaics.
[0048] Based on this, as an exemplary embodiment, historical difference characteristics of historical surface irradiance data and historical power data at each moment on a day time scale are obtained, and a time-series-related historical difference characteristic sequence is constructed, including: calculating the historical change trend of the historical surface irradiance point value sequence at each moment on a day time scale relative to the historical power point value sequence; dividing the historical change trend into a first sub-historical change trend in which the historical surface irradiance point value sequence is positively correlated with the historical power point value sequence and a second sub-historical change trend in which the historical surface irradiance point value sequence is not positively correlated with the historical power point value sequence; and constructing a first historical difference characteristic sequence and a second historical difference characteristic sequence that are time-series-related based on the first sub-historical change trend and the second sub-historical change trend, respectively.
[0049] In this embodiment, the historical change trend is divided into a first sub-historical change trend in which the historical surface irradiance point value sequence is positively correlated with the historical power point value sequence and a second sub-historical change trend in which the historical surface irradiance point value sequence is not positively correlated with the historical power point value sequence, and then the first sub-historical change trend and the second sub-historical change trend are used to construct a first historical difference feature sequence and a second historical difference feature sequence that are time-series related; according to the above-mentioned construction method of the difference feature sequence, the first historical difference feature sequence can be used to describe the state in which the irradiance and power are positively correlated in the installation method of the flat single-axis photovoltaic, and the second historical difference feature can be used to describe the state in which the irradiance and power are not positively correlated in the installation method of the flat single-axis photovoltaic.
[0050] Furthermore, a first training data set is constructed by taking a historical difference feature sequence and a historical surface irradiance point value sequence as input features, and taking a historical power point value sequence as a target value; when the first training data set is input into a pre-constructed photovoltaic power prediction model for model training, the model can determine the output state to which the currently input historical irradiance point value sequence belongs through the above-mentioned first historical difference feature sequence and second historical difference feature sequence, and then can adaptively learn the positive correlation or non-positive correlation between the historical power point value sequence and the historical irradiance point value sequence, so as to obtain a power prediction model of photovoltaic output characteristics suitable for the installation method of flat single-axis photovoltaics.
[0051] In order to enable the model to learn the correspondence between irradiance and power in different time periods, as an exemplary embodiment, the method for constructing a photovoltaic power prediction model also includes: respectively obtaining a first historical change period corresponding to a first historical difference feature sequence and a second historical change period corresponding to a second historical difference feature; dividing the first training data set based on the first historical change period and the second historical change period to obtain a second training data set; wherein the second training data set includes a second training data subset consisting of a plurality of historical surface irradiance point value sequences and historical power point value sequences with different historical change trends divided according to the first historical change period and the second historical change period; inputting each second training data subset into a pre-constructed photovoltaic power prediction model for model training to obtain a first photovoltaic power prediction sub-model corresponding to each historical change period; in the process of model training, learning the correspondence between each historical surface irradiance point value sequence and historical power point value sequence in the second training data subset corresponding to each historical change period based on each first photovoltaic power prediction sub-model.
[0052] In this embodiment, by obtaining the first historical change period corresponding to the first historical difference feature sequence and the second historical change period corresponding to the second historical difference feature, the first training data set is further divided based on the first historical change period and the second historical change period, respectively, to obtain the second training data set. The first historical change period can be used to describe the time corresponding to the positive correlation state between irradiance and power in the installation method of flat single-axis photovoltaics, and the second historical difference feature can be used to describe the time corresponding to the non-positive correlation state between irradiance and power in the installation method of flat single-axis photovoltaics.
[0053] Furthermore, the first training data set is divided based on the first historical change period and the second historical change period to obtain a second training data set, and each second training data subset is input into a pre-constructed photovoltaic power prediction model for model training to obtain a first photovoltaic power prediction sub-model corresponding to each historical change period. Each first photovoltaic power prediction sub-model can determine the correspondence between the output state and time of the currently input historical irradiance point value sequence through the above-mentioned first historical change period and second historical change period, and then can adaptively learn the time period to which the positive correlation between the historical power point value sequence and the historical irradiance point value sequence belongs or the time period to which the non-positive correlation belongs, so as to obtain a power prediction model of photovoltaic output characteristics suitable for the installation method of flat single-axis photovoltaics.
[0054] As an exemplary embodiment, the method for constructing a photovoltaic power prediction model also includes: obtaining a historical temperature point value sequence and / or a historical wind speed point value sequence corresponding to each second training data subset; adding the historical temperature point value sequence and / or the historical wind speed point value sequence to the corresponding second training data subset to obtain a third training data subset; inputting each third training data subset into a pre-constructed photovoltaic power prediction model for model training to obtain a second photovoltaic power prediction sub-model corresponding to each historical change period; during the model training process, based on each second photovoltaic power prediction sub-model, learning the correspondence between each historical temperature point value sequence and / or historical wind speed point value sequence, historical surface irradiance point value sequence and historical power point value sequence in the third training data subset corresponding to each historical change period.
[0055] Based on the theoretical power generation curve of photovoltaic panels, the actual power generation of photovoltaic panels is affected by temperature; under the same irradiance, the power generation of photovoltaic panels shows a trend of first increasing and then decreasing with the increase of temperature; as a possible implementation method, in order to consider the influence of temperature on the power generation of photovoltaic panels, the historical temperature point value sequence corresponding to each second training data subset is obtained; the historical temperature point value sequence is added to the corresponding second training data subset to obtain a third training data subset; each third training data subset is input into a pre-built photovoltaic power prediction model for model training to obtain a second photovoltaic power prediction sub-model corresponding to each historical change period; in the process of model training, the correspondence between each historical temperature point value sequence, historical surface irradiance point value sequence and historical power point value sequence in the third training data subset corresponding to each historical change period is learned based on each second photovoltaic power prediction sub-model.
[0056] Moreover, due to the installation form of the flat single-axis photovoltaic bracket, the north-south direction of the flat single-axis photovoltaic bracket is horizontal, and the array plane rotates east-west with the rotation axis according to the position of the sun. The rotation angle of the array plane that rotates east-west may be affected by the wind force in the east-west direction; as a possible implementation method, in order to consider the influence of wind speed on the power generation of photovoltaic panels, the historical wind speed point value sequence corresponding to each second training data subset is obtained; the historical wind speed point value sequence is added to the corresponding second training data subset to obtain a third training data subset; each third training data subset is respectively input into a pre-constructed photovoltaic power prediction model for model training to obtain a second photovoltaic power prediction sub-model corresponding to each historical change period; in the process of model training, the correspondence between each historical wind speed point value sequence, historical surface irradiance point value sequence and historical power point value sequence in the third training data subset corresponding to each historical change period is learned based on each second photovoltaic power prediction sub-model.
[0057] Exemplarily, the historical wind speed point value sequence corresponding to each second training data subset may be the wind speed value that will actually affect the east-west rotation of the array plane with the rotation axis following the position of the sun; in the present embodiment, the historical wind speed point value sequence may be the component wind speed in the east-west direction of the historical wind speed obtained by the historical wind speed, so as to take into account that the rotation angle of the array plane rotating east-west may be affected by the wind force in the east-west direction.
[0058] As another exemplary embodiment, in order to simultaneously consider the impact of temperature and wind speed on the power generation of photovoltaic panels, the method for constructing a photovoltaic power prediction model also includes: obtaining the historical temperature point value sequence and the historical wind speed point value sequence corresponding to each second training data subset; adding the historical temperature point value sequence and the historical wind speed point value sequence to the corresponding second training data subset to obtain a third training data subset; inputting each third training data subset into a pre-constructed photovoltaic power prediction model for model training to obtain a second photovoltaic power prediction sub-model corresponding to each historical change period; in the process of model training, based on each second photovoltaic power prediction sub-model, learning the correspondence between each historical temperature point value sequence and historical wind speed point value sequence, historical surface irradiance point value sequence and historical power point value sequence in the third training data subset corresponding to each historical change period.
[0059] In a second aspect, the present invention provides a photovoltaic power prediction method, which is applied to a target photovoltaic power station equipped with flat single-axis single-sided photovoltaic modules; Figure 3 A photovoltaic power prediction method according to an embodiment of the present invention includes:
[0060] Step S201, obtaining predicted irradiance data of a target photovoltaic power station.
[0061] In this embodiment, the predicted irradiance data of the target photovoltaic power station can be obtained through weather forecast; wherein the predicted irradiance data at least includes horizontal direct radiation irradiance, horizontal scattered radiation irradiance and ground reflected radiation irradiance.
[0062] Step S202, calculating predicted surface irradiance data of each flat uniaxial monofacial photovoltaic module based on the predicted irradiance data.
[0063] After the predicted irradiance data is obtained, the predicted surface irradiance data of each flat uniaxial single-sided photovoltaic module can be calculated using formula (1) based on the predicted irradiance data.
[0064] Step S203 , dividing the predicted surface irradiance data according to preset time intervals to obtain a predicted surface irradiance point value sequence.
[0065] In this embodiment, the predicted surface irradiance data needs to be converted according to the data input format of the power prediction model; wherein the preset time interval is 15 minutes, so as to convert the predicted surface irradiance data into a sequence of predicted surface irradiance point values at intervals of 15 minutes.
[0066] Step S204, input the predicted surface irradiance point value sequence into a pre-trained photovoltaic power prediction model for power prediction to obtain a power prediction result; wherein the pre-trained photovoltaic power prediction model is trained by using the historical surface irradiance point value sequence as input features and the historical power point value sequence as the target value to construct a first training data set, and the historical surface irradiance point value sequence and the historical power point value sequence are obtained by obtaining the historical measured irradiance data and historical power data of the target photovoltaic power station within a preset time period, calculating the historical surface irradiance data of each flat single-axis single-sided photovoltaic module based on each historical measured irradiance data, and dividing the data according to preset time intervals.
[0067] In this embodiment, the pre-trained photovoltaic power prediction model is obtained by training the first training data set by taking the historical surface irradiance point value sequence as input features and the historical power point value sequence as the target value, and the historical surface irradiance point value sequence and the historical power point value sequence are obtained by obtaining the historical measured irradiance data and historical power data of the target photovoltaic power station within a preset time period, calculating the historical surface irradiance data of each flat single-axis single-sided photovoltaic module based on each historical measured irradiance data, and dividing the data according to a preset time interval; the pre-trained photovoltaic power prediction model is obtained by taking the historical surface irradiance point value sequence as input features and the historical power point value sequence as the target value, and the historical surface irradiance point value sequence is based on each historical measured irradiance data The change in irradiance caused by the plane of the photovoltaic panel array installed with a flat single-axis photovoltaic bracket rotating east-west with the rotation axis following the position of the sun is calculated. Therefore, the photovoltaic power prediction model of this embodiment can learn the influence of the change in surface irradiance of the photovoltaic panel caused by the installation form of the photovoltaic panel flat single-axis photovoltaic bracket on the photovoltaic power prediction during the training process, and can adaptively predict the photovoltaic power for the photovoltaic panels installed in the form of the flat single-axis photovoltaic bracket. The photovoltaic power prediction model of this embodiment can consider the influence of the change in surface irradiance of the photovoltaic panel caused by the installation form of the photovoltaic panel flat single-axis photovoltaic bracket on the photovoltaic power prediction during the prediction process, and can adaptively predict the photovoltaic power for the photovoltaic panels installed in the form of the flat single-axis photovoltaic bracket.
[0068] As an exemplary embodiment, the photovoltaic power prediction method also includes: obtaining a time-series-related historical difference feature sequence constructed by constructing historical surface irradiance data and historical power data of the target photovoltaic power station at each moment on a daily time scale; selecting a historical difference feature sequence corresponding to historical surface irradiance data whose similarity with the predicted surface irradiance point value sequence is greater than a preset degree as a predicted difference feature sequence and inputting it into a pre-trained photovoltaic power prediction model for power prediction to obtain a photovoltaic power prediction result.
[0069] In this embodiment, during the model training process, in order to enable the model to take into account the difference between the above-mentioned output characteristics and the prior art in which irradiance and power are in a linear function relationship, a historical difference feature sequence is constructed according to the historical difference characteristics of the historical surface irradiance data and the historical power data at each moment on a day time scale; wherein, the historical difference feature sequence is used to indicate that the above-mentioned output characteristics are expressed in a numerical form in which the output is highest in the morning and evening and the output is concave at noon. Specifically, the historical difference feature may include the historical change trend of the historical surface irradiance point value sequence at each moment relative to the historical power point value sequence, and the historical change trend may be the historical surface irradiance. The slope of change, the difference of change, etc. of the degree point value sequence relative to the historical power point value sequence when it changes; based on this, in this embodiment, the historical difference feature sequence corresponding to the historical surface irradiance data whose similarity with the predicted surface irradiance point value sequence is greater than a preset degree is selected as the predicted difference feature sequence, so that in the prediction process, the actual output characteristic conditions corresponding to the current predicted surface irradiance point value sequence can be considered, and the impact of the change in the surface irradiance of the photovoltaic panel caused by the installation form of the flat single-axis photovoltaic bracket of the photovoltaic panel on the photovoltaic power prediction can be considered, and the photovoltaic power prediction can be adaptively performed for the photovoltaic panels installed in the form of a flat single-axis photovoltaic bracket.
[0070] As an exemplary embodiment, a time-series-related historical difference feature sequence is constructed by obtaining historical surface irradiance data and historical power data of a target photovoltaic power station at each moment on a day time scale, including: calculating the historical change trend of the historical surface irradiance point value sequence at each moment on a day time scale relative to the historical power point value sequence; dividing the historical change trend into a first sub-historical change trend in which the historical surface irradiance point value sequence is positively correlated with the historical power point value sequence and a second sub-historical change trend in which the historical surface irradiance point value sequence is not positively correlated with the historical power point value sequence; and constructing a first historical difference feature sequence and a second historical difference feature sequence that are time-series-related based on the first sub-historical change trend and the second sub-historical change trend, respectively.
[0071] In this embodiment, the historical change trend is divided into a first sub-historical change trend in which the historical surface irradiance point value sequence is positively correlated with the historical power point value sequence and a second sub-historical change trend in which the historical surface irradiance point value sequence is not positively correlated with the historical power point value sequence, and then the first sub-historical change trend and the second sub-historical change trend are used to construct a first historical difference feature sequence and a second historical difference feature sequence that are time-series related; according to the above-mentioned construction method of the difference feature sequence, the first historical difference feature sequence can be used to describe the feature label of the state in which the irradiance and power are positively correlated in the installation method of the flat single-axis photovoltaic, and the second historical difference feature can be used to describe the feature label of the state in which the irradiance and power are not positively correlated in the installation method of the flat single-axis photovoltaic.
[0072] Furthermore, a first training data set is constructed by taking historical difference feature sequences and historical surface irradiance point value sequences as input features, and taking historical power point value sequences as target values; when selecting historical difference feature sequences corresponding to historical surface irradiance data whose similarity with the predicted surface irradiance point value sequence is greater than a preset degree as a predicted difference feature sequence, and inputting the predicted difference feature sequence into a pre-trained photovoltaic power prediction model for power prediction, the model can determine the output state to which the currently input predicted irradiance point value sequence belongs through the above-mentioned predicted difference features, and then can adaptively perform power prediction according to the output state.
[0073] As an exemplary embodiment, the photovoltaic power prediction method also includes: obtaining a predicted change period corresponding to each predicted surface irradiance point value sequence based on a time scale of days; based on the predicted change period, inputting the predicted surface irradiance point value corresponding to the predicted change period into a first photovoltaic power prediction sub-model corresponding to a historical change period matching the predicted change period to perform power prediction, thereby obtaining a photovoltaic power prediction result; wherein the first photovoltaic power prediction sub-model is trained by a second training data subset consisting of a historical surface irradiance point value sequence and a historical power point value sequence according to a historical change trend corresponding to the historical change period, and during the training process, learning the correspondence between each historical surface irradiance point value sequence and each historical power point value sequence in the second training data subset corresponding to each historical change period.
[0074] In this embodiment, the first photovoltaic power prediction sub-model is trained by a second training data subset consisting of a historical surface irradiance point value sequence and a historical power point value sequence according to the historical change trend corresponding to the historical change period. During the training process, the correspondence between the historical surface irradiance point value sequence and the historical power point value sequence in the second training data subset corresponding to each historical change period is learned. The model is obtained by obtaining the first historical change period corresponding to the first historical difference feature sequence and the second historical change period corresponding to the second historical difference feature. Further, the first training data set is divided based on the first historical change period and the second historical change period, respectively, to obtain the second training data set. The first historical change period can be used to describe the time label of the positive correlation state between irradiance and power in the installation method of flat single-axis photovoltaics, and the second historical difference feature can be used to describe the time label of the non-positive correlation state between irradiance and power in the installation method of flat single-axis photovoltaics.
[0075] As an exemplary embodiment, the photovoltaic power prediction method also includes: obtaining a predicted temperature point value sequence and / or a predicted wind speed point value sequence corresponding to the predicted surface irradiance point value corresponding to each predicted change period; respectively inputting the predicted surface irradiance point value and the predicted temperature point value sequence and / or the predicted wind speed point value sequence into a pre-trained second photovoltaic power prediction sub-model, wherein the second photovoltaic power prediction sub-model is a third training data subset consisting of a predicted temperature point value sequence and / or a predicted wind data point value sequence according to the predicted change trend under the predicted change period, a predicted surface irradiance point value sequence and a predicted power point value sequence, and the correspondence between the predicted temperature point value sequence and / or the predicted wind data point value sequence, the predicted surface irradiance point value sequence and the predicted power point value sequence is obtained by learning during the training process.
[0076] Based on the theoretical power generation curve of photovoltaic panels, the actual power generation of photovoltaic panels is affected by temperature; under the same irradiance, the power generation of photovoltaic panels shows a trend of first increasing and then decreasing with the increase of temperature; as a possible implementation method, in order to consider the influence of temperature on the power generation of photovoltaic panels, the predicted temperature point value sequence corresponding to the predicted surface irradiance point value corresponding to each predicted change period is obtained; the predicted temperature point value sequence and the predicted surface irradiance point value sequence are input into the pre-trained second photovoltaic power prediction sub-model for power prediction to obtain the power prediction result.
[0077] Moreover, due to the installation form of the flat single-axis photovoltaic bracket, the north-south direction of the flat single-axis photovoltaic bracket is horizontal, and the array plane rotates east-west with the rotation axis according to the position of the sun, the rotation angle of the array plane that rotates east-west may be affected by the wind force in the east-west direction; as a possible implementation method, in order to consider the influence of wind speed on the power generation of photovoltaic panels, the predicted wind speed point value sequence corresponding to the predicted surface irradiance point value corresponding to each predicted change period is obtained; the predicted wind speed point value sequence and the predicted irradiance point value sequence are input into the pre-constructed second photovoltaic power prediction sub-model for power prediction to obtain the power prediction result.
[0078] Exemplarily, the predicted wind speed point value sequence may be the wind speed values that will actually affect the east-west rotation of the array plane with the rotation axis following the position of the sun; in the present embodiment, the predicted wind speed point value sequence may be the component wind speed in the east-west direction of the historical wind speed obtained by using the historical wind speed, so as to take into account that the rotation angle of the array plane rotating east-west may be affected by the wind force in the east-west direction.
[0079] As another exemplary embodiment, in order to simultaneously consider the impact of temperature and wind speed on the power generation of photovoltaic panels, the method for constructing a photovoltaic power prediction model also includes: obtaining a predicted temperature point value sequence and a predicted wind speed point value sequence corresponding to the predicted surface irradiance point value corresponding to each predicted change period; inputting the predicted surface irradiance point value, the predicted temperature point value sequence and the predicted wind speed point value sequence into a pre-trained second photovoltaic power prediction sub-model, respectively, wherein the second photovoltaic power prediction sub-model is a third training data subset consisting of a predicted temperature point value sequence and a predicted wind data point value sequence according to the predicted change trend under the predicted change period, a predicted surface irradiance point value sequence and a predicted power point value sequence, and the correspondence between the predicted temperature point value sequence and the predicted wind data point value sequence, the predicted surface irradiance point value sequence and the predicted power point value sequence is learned during the training process.
[0080] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.
[0081] If the integrated units in the above embodiments are implemented in the form of software functional units and sold or used as independent products, they can be stored in the above computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for enabling one or more computer devices (which may be personal computers, servers or network devices, etc.) to execute all or part of the steps of the method in the above embodiments.
[0082] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.
[0083] In the several embodiments provided in the present application, it should be understood that the disclosed client can be implemented in other ways. Among them, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.
[0084] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution provided in this embodiment.
[0085] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.
[0086] The above are only preferred implementations of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.
[0087] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.
[0088] The above are only preferred implementations of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.
Claims
1. A method for constructing a photovoltaic power prediction model, characterized in that: Applicable to a target photovoltaic power station equipped with flat single-axis single-sided photovoltaic modules; the photovoltaic power prediction model includes: Obtain the historical measured irradiance data and historical power data of the target photovoltaic power station within a preset time period; Calculate the historical surface irradiance data of each of the flat single-axis single-sided photovoltaic modules based on each of the historical measured irradiance data; Dividing the historical surface irradiance data and the historical power data according to preset time intervals to obtain a historical surface irradiance point value sequence and a historical power point value sequence; Using the historical surface irradiance point value sequence as input features and the historical power point value sequence as target values to construct a first training data set; The first training data set is input into a pre-built photovoltaic power prediction model for model training to obtain the photovoltaic power prediction model; during the model training process, the correspondence between each of the historical surface irradiance point value sequences and the historical power point value sequences is learned.
2. The method for constructing a photovoltaic power prediction model according to claim 1, characterized in that: The step of using the historical surface irradiance point value sequence as an input feature and the historical power point value sequence as a target value to construct a first training data set comprises: Obtain the historical difference characteristics of the historical surface irradiance data and historical power data at each moment on a daily time scale, and construct a time-series-related historical difference characteristic sequence; The historical difference feature sequence and the historical surface irradiance point value sequence are used as input features, and the historical power point value sequence is used as a target value to construct a first training data set.
3. The method for constructing a photovoltaic power prediction model according to claim 2, characterized in that: The method of obtaining the historical difference characteristics of the historical surface irradiance data and the historical power data at each moment on a day time scale and constructing a time-series-related historical difference characteristic sequence includes: Calculate the historical variation trend of the historical surface irradiance point value sequence at each moment with a day as the time scale relative to the historical power point value sequence; Dividing the historical change trend into a first sub-historical change trend in which the historical surface irradiance point value sequence is positively correlated with the historical power point value sequence and a second sub-historical change trend in which the historical surface irradiance point value sequence is not positively correlated with the historical power point value sequence; A first historical difference feature sequence and a second historical difference feature sequence that are time-series related are constructed based on the first sub-historical change trend and the second sub-historical change trend, respectively.
4. The method for constructing a photovoltaic power prediction model according to claim 3, characterized in that: The method for constructing the photovoltaic power prediction model also includes: Respectively obtaining a first historical change period corresponding to the first historical difference feature sequence and a second historical change period corresponding to the second historical difference feature sequence; The first training data set is divided based on the first historical change period and the second historical change period to obtain a second training data set; wherein the second training data set includes a second training data subset consisting of a plurality of historical surface irradiance point value sequences and historical power point value sequences with different historical change trends divided according to the first historical change period and the second historical change period; Each of the second training data subsets is respectively input into a pre-built photovoltaic power prediction model for model training to obtain a first photovoltaic power prediction sub-model corresponding to each of the historical change time periods; during the model training process, the correspondence between each of the historical surface irradiance point value sequences and the historical power point value sequences in the second training data subset corresponding to each of the historical change time periods is learned based on each of the first photovoltaic power prediction sub-models.
5. The method for constructing a photovoltaic power prediction model according to claim 4, characterized in that: The method for constructing the photovoltaic power prediction model also includes: Obtaining a historical temperature point value sequence and / or a historical wind speed point value sequence corresponding to each of the second training data subsets; Adding the historical temperature point value sequence and / or the historical wind speed point value sequence to the corresponding second training data subset to obtain a third training data subset; Each of the third training data subsets is respectively input into a pre-constructed photovoltaic power prediction model for model training to obtain a second photovoltaic power prediction sub-model corresponding to each of the historical change time periods; during the model training process, the correspondence between each of the historical temperature point value sequences and / or historical wind speed point value sequences, the historical surface irradiance point value sequence and the historical power point value sequence in the third training data subset corresponding to each of the historical change time periods is learned based on each of the second photovoltaic power prediction sub-models.
6. A photovoltaic power prediction method, characterized in that: Applied to a target photovoltaic power station equipped with flat single-axis single-sided photovoltaic modules; the photovoltaic power prediction method includes: Obtain the predicted irradiance data of the target photovoltaic power station; Calculate the predicted surface irradiance data of each of the flat uniaxial single-sided photovoltaic modules based on the predicted irradiance data; Dividing the predicted surface irradiance data according to preset time intervals to obtain a predicted surface irradiance point value sequence; The predicted surface irradiance point value sequence is input into a pre-trained photovoltaic power prediction model for power prediction to obtain a power prediction result; wherein the pre-trained photovoltaic power prediction model is trained by using the historical surface irradiance point value sequence as input features and the historical power point value sequence as the target value to construct a first training data set, and the historical surface irradiance point value sequence and the historical power point value sequence are obtained by obtaining the historical measured irradiance data and historical power data of the target photovoltaic power station within a preset time period, calculating the historical surface irradiance data of each of the flat single-axis single-sided photovoltaic modules based on each of the historical measured irradiance data, and dividing the data according to preset time intervals.
7. The photovoltaic power prediction method according to claim 6, characterized in that: The photovoltaic power prediction method further comprises: Obtain the time-series-related historical difference feature sequence constructed by constructing the historical surface irradiance data and historical power data of the target photovoltaic power station at each moment on a daily time scale; The historical difference feature sequence corresponding to the historical surface irradiance data having a similarity with the predicted surface irradiance point value sequence greater than a preset degree is selected as the predicted difference feature sequence and input into a pre-trained photovoltaic power prediction model for power prediction to obtain a photovoltaic power prediction result.
8. The photovoltaic power prediction method according to claim 7, characterized in that: The method of obtaining the time-series-related historical difference feature sequence of the historical surface irradiance data and the historical power data of the target photovoltaic power station at each moment on a day time scale comprises: Calculate the historical variation trend of the historical surface irradiance point value sequence at each moment with a day as the time scale relative to the historical power point value sequence; Dividing the historical change trend into a first sub-historical change trend in which the historical surface irradiance point value sequence is positively correlated with the historical power point value sequence and a second sub-historical change trend in which the historical surface irradiance point value sequence is not positively correlated with the historical power point value sequence; A first historical difference feature sequence and a second historical difference feature sequence that are time-series related are constructed based on the first sub-historical change trend and the second sub-historical change trend, respectively.
9. The photovoltaic power prediction method according to claim 6, characterized in that: The photovoltaic power prediction method further comprises: Obtaining the predicted change period corresponding to each of the predicted surface irradiance point value sequences with days as the time scale; Based on the predicted change period, the predicted surface irradiance point value corresponding to the predicted change period is input into the first photovoltaic power prediction sub-model corresponding to the historical change period matching the predicted change period to perform power prediction, and obtain a photovoltaic power prediction result; wherein, the first photovoltaic power prediction sub-model is trained by a second training data subset consisting of a historical surface irradiance point value sequence and a historical power point value sequence according to the historical change trend corresponding to the historical change period, and during the training process, the correspondence between each of the historical surface irradiance point value sequences and the historical power point value sequence in the second training data subset corresponding to each of the historical change periods is learned.
10. The photovoltaic power prediction method according to claim 9, characterized in that: The photovoltaic power prediction method further comprises: Obtain a predicted temperature point value sequence and / or a predicted wind speed point value sequence corresponding to the predicted surface irradiance point value corresponding to each of the predicted change time periods; The predicted surface irradiance point value and the predicted temperature point value sequence and / or predicted wind speed point value sequence are respectively input into a pre-trained second photovoltaic power prediction sub-model, wherein the second photovoltaic power prediction sub-model is a third training data subset consisting of a predicted temperature point value sequence and / or predicted wind data point value sequence according to a predicted change trend under a predicted change period, a predicted surface irradiance point value sequence and the predicted surface irradiance point value sequence, and the correspondence between the predicted temperature point value sequence and / or predicted wind data point value sequence, the predicted surface irradiance point value sequence and the predicted power point value sequence is learned during the training process.
Citation Information
Patent Citations
Photovoltaic power prediction method, system and equipment based on segmented modeling and medium
CN117763258A
Training method of generating capacity prediction model and generating capacity prediction method
CN117763344A
Power calculation model training method and device, power calculation method and device, equipment and medium
CN118228577A
Construction method and prediction method of short-term power prediction model based on unmanned aerial vehicle
CN118296507A
Cited By
Photovoltaic power generation efficiency adjusting method, system, program product, equipment and medium
CN120601842A
