Photovoltaic power station power generation prediction method, device, equipment and storage medium
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHENZHEN XINWANGDA SMART ENERGY CO LTD
- Filing Date
- 2026-03-16
- Publication Date
- 2026-06-19
Smart Images
Figure CN122241079A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of photovoltaic technology, and in particular to a method, apparatus, equipment and storage medium for predicting the power generation of a photovoltaic power plant. Background Technology
[0002] Photovoltaic power generation has become an important component of the power system due to its cleanliness. However, the intermittency and volatility of photovoltaic power generation (significantly affected by meteorological factors such as solar intensity, cloud cover, and temperature changes) make it difficult to predict its output power stably, posing a serious challenge to real-time grid dispatch and the balance of power supply and demand.
[0003] In real-world scenarios, forecasting power generation for the next hour (i.e., short-term forecasting) is a crucial element in ensuring the safe operation of the power grid. For example, in power grid dispatching scenarios, dispatch centers need to dynamically adjust power generation plans based on power generation forecasts to balance the volatility of photovoltaic power generation; in power plant operation and maintenance scenarios, maintenance personnel need to optimize equipment operation status based on forecast results to reduce power generation losses caused by sudden weather changes.
[0004] However, the accuracy of existing power generation prediction methods is low and cannot meet the accuracy requirements of real-world scenarios. Summary of the Invention
[0005] This application provides a method, apparatus, equipment, and storage medium for predicting the power generation of a photovoltaic power plant, in order to improve the accuracy of predicting the power generation of a photovoltaic power plant.
[0006] In a first aspect, embodiments of this application provide a method for predicting the power generation of a photovoltaic power plant, the method comprising:
[0007] Acquire historical monitoring data of the photovoltaic power station within a preset time period. The historical monitoring data includes historical power generation information, historical equipment status information, and historical meteorological data.
[0008] Based on historical monitoring data within a preset time period, the hyperparameters of the initial machine learning model are iteratively optimized using a Bayesian optimization framework, where the Bayesian optimization framework is used to indicate the search range of the hyperparameters.
[0009] The process continues until the initial machine learning model meets the preset training conditions, resulting in a trained machine learning model. The trained machine learning model is then used to predict the power generation of the photovoltaic power station in the next hour.
[0010] Optionally, the step of iteratively optimizing the hyperparameters of the initial machine learning model using a Bayesian optimization framework based on historical monitoring data within a preset time period includes: using a Bayesian optimization framework to determine the value range of the hyperparameters in the initial machine learning model, wherein the hyperparameters include one or more of learning rate, tree depth, sampling ratio, and regularization parameter; dividing the historical monitoring data within the preset time period into a training set and a test set; training the initial machine learning model using the training set and the test set, and iteratively optimizing the hyperparameters of the initial machine learning model within the value range of the hyperparameters with the goal of minimizing the root mean square error of the predicted values.
[0011] Optionally, the step of iteratively optimizing the hyperparameters of the initial machine learning model within the range of the hyperparameters with the goal of minimizing the root mean square error of the predicted values includes: using a tree structure estimation algorithm to dynamically adjust the values of the hyperparameters within the range of the hyperparameters based on historical experimental results, with the goal of minimizing the root mean square error of the predicted values, to iteratively optimize the hyperparameters of the initial machine learning model; wherein, the tree structure estimation algorithm is used to predict the values of the hyperparameters based on historical experimental results.
[0012] Optionally, dividing the historical monitoring data within a preset time period into a training set and a test set includes: standardizing the historical monitoring data within the preset time period, the standardization process including timestamp alignment and categorical variable encoding; identifying and deleting redundant monitoring data in the historical monitoring data to obtain processed historical monitoring data, wherein the redundant monitoring data is monitoring data with a relevance greater than a preset threshold; and dividing the processed historical monitoring data into a training set and a test set according to chronological order.
[0013] Optionally, until the initial machine learning model meets the preset training conditions to obtain a trained machine learning model, the following steps are taken: until the number of iterations of the initial machine learning model reaches a preset number of iterations to obtain a trained machine learning model; or, the change in the root mean square error of the predicted values of the initial machine learning model is determined until the number of consecutive times the change is less than a preset value reaches a preset number to obtain a trained machine learning model.
[0014] Optionally, predicting the power generation of the photovoltaic power station in the next hour using a trained machine learning model includes: acquiring the current equipment status information of the photovoltaic power station and the weather forecast data for the next hour; inputting the current equipment status information of the photovoltaic power station and the weather forecast data for the next hour into the trained machine learning model, and outputting the power generation of the photovoltaic power station in the next hour.
[0015] Secondly, embodiments of this application provide a device for predicting the power generation of a photovoltaic power plant, the speed control device comprising:
[0016] The acquisition module is used to acquire historical monitoring data of the photovoltaic power station within a preset time period. The historical monitoring data includes historical power generation information, historical equipment status information, and historical meteorological data.
[0017] The iteration module is used to iteratively optimize the hyperparameters of the initial machine learning model based on historical monitoring data within a preset time period using a Bayesian optimization framework, where the Bayesian optimization framework is used to indicate the search range of the hyperparameters.
[0018] The prediction module is used to obtain a trained machine learning model until the initial machine learning model meets the preset training conditions, and then use the trained machine learning model to predict the power generation of the photovoltaic power station in the next hour.
[0019] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;
[0020] The memory stores instructions that the computer executes;
[0021] The processor executes computer execution instructions stored in memory, causing the processor to perform various possible implementations as described above.
[0022] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement various possible implementations as described in any of the above aspects.
[0023] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements various possible implementations as described in any of the above aspects.
[0024] The photovoltaic power generation prediction method, apparatus, equipment, and storage medium provided in this application embodiment include: acquiring historical monitoring data of the photovoltaic power station within a preset time period, the historical monitoring data including historical power generation information, historical equipment status information, and historical meteorological data; iteratively optimizing the hyperparameters of an initial machine learning model using a Bayesian optimization framework based on the historical monitoring data within the preset time period, wherein the Bayesian optimization framework is used to indicate the search range of the hyperparameters; until the initial machine learning model meets preset training conditions, a trained machine learning model is obtained, and the power generation of the photovoltaic power station in the next hour is predicted using the trained machine learning model. In this application embodiment, because the hyperparameter combination of the machine learning model is iteratively optimized using a Bayesian optimization framework, the problems of low efficiency, reliance on human experience, and difficulty in finding the global optimum of traditional grid search or random search methods are solved. Furthermore, the Bayesian optimization framework can dynamically adjust the parameter search strategy based on historical experimental results, which can reduce the number of iterations of the training model. That is, this method significantly improves the efficiency and stability of hyperparameter tuning by intelligently guiding parameter space exploration, thus improving the prediction accuracy of the model under complex meteorological conditions. Attached Figure Description
[0025] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.
[0026] Figure 1 A flowchart illustrating a method for predicting the power generation of a photovoltaic power plant, as provided in this application embodiment;
[0027] Figure 2 A schematic diagram illustrating a method for predicting the power generation of a photovoltaic power plant, provided in an embodiment of this application;
[0028] Figure 3 A structural diagram of a photovoltaic power plant power generation prediction device provided in this application embodiment;
[0029] Figure 4 A schematic diagram of the structure of the electronic device provided in this application.
[0030] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation
[0031] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.
[0032] Photovoltaic power generation has become an important component of the power system due to its cleanliness. However, the intermittency and volatility of photovoltaic power generation (significantly affected by meteorological factors such as solar intensity, cloud cover, and temperature changes) make it difficult to predict its output power stably, posing a serious challenge to real-time grid dispatch and the balance of power supply and demand.
[0033] In real-world scenarios, forecasting power generation for the next hour (i.e., short-term forecasting) is a crucial element in ensuring the safe operation of the power grid. For example, in power grid dispatching scenarios, dispatch centers need to dynamically adjust power generation plans based on power generation forecasts to balance the volatility of photovoltaic power generation; in power plant operation and maintenance scenarios, maintenance personnel need to optimize equipment operation based on forecast results to reduce power generation losses caused by sudden weather changes. However, existing power generation forecasting methods have low accuracy and cannot meet the accuracy requirements of real-world scenarios.
[0034] Therefore, there is an urgent need for an efficient and intelligent prediction method to meet the requirements of high accuracy, low error, and rapid deployment in real-world scenarios.
[0035] To address the aforementioned technical problems, this application provides a method for predicting the power generation of a photovoltaic power plant. The method includes: acquiring historical monitoring data of the photovoltaic power plant within a preset time period, including historical power generation information, historical equipment status information, and historical meteorological data; iteratively optimizing the hyperparameters of an initial machine learning model using a Bayesian optimization framework based on the historical monitoring data within the preset time period, wherein the Bayesian optimization framework is used to indicate the search range of the hyperparameters; and obtaining a trained machine learning model until the initial machine learning model meets preset training conditions, and then predicting the power generation of the photovoltaic power plant for the next hour using the trained machine learning model.
[0036] In this embodiment, a Bayesian optimization framework is used to iteratively optimize the hyperparameter combination of a machine learning model, solving the problems of low efficiency, reliance on human experience, and difficulty in finding the global optimum in traditional grid search or random search methods. The Bayesian optimization framework dynamically adjusts the parameter search strategy based on historical experimental results, for example, prioritizing the exploration of regions that may achieve better results in the learning rate search (such as a combination of low learning rate and high tree depth), thereby reducing the number of invalid trials. This technique is based on the mathematical principles of the Sequence Model Optimization (TPE) algorithm, and through intelligent guidance of parameter space exploration, it significantly improves the efficiency and stability of hyperparameter tuning, and enhances the prediction accuracy of the model under complex weather conditions.
[0037] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.
[0038] Figure 1 This is a flowchart illustrating a method for predicting the power generation of a photovoltaic power plant, as provided in an embodiment of this application. Figure 1 As shown, the method for predicting the power generation of this photovoltaic power station includes:
[0039] S101. Obtain historical monitoring data of the photovoltaic power station within a preset time period. The historical monitoring data includes historical power generation information, historical equipment status information, and historical meteorological data.
[0040] In this embodiment, the historical power generation information includes power generation over multiple time periods within a preset duration. For example, the preset duration can be one day, and the time period can be 15 minutes. The historical power generation information includes the power generation every 15 minutes within a day. In this embodiment, the value of the preset duration is not specifically limited and can be set and modified as needed.
[0041] The historical equipment status information includes the status information of each photovoltaic power generation device in the photovoltaic power plant. Optionally, the status information of the photovoltaic power generation device includes any information related to power generation. For example, the status information of the photovoltaic power generation device includes: the voltage, current, power, and tilt angle of the photovoltaic panel.
[0042] Historical meteorological data includes one or more of the following: irradiance information, temperature information, wind speed information, and weather type. Optionally, irradiance information includes total irradiance and hourly cumulative irradiance.
[0043] Optionally, historical monitoring data includes multi-source heterogeneous data. For example, such as... Figure 2As shown, the historical monitoring data includes: historical grid-connected power generation time-series data provided by the photovoltaic power station monitoring system, weather forecast data for the next hour (such as irradiance, temperature, wind speed, weather conditions, etc.), and power station equipment status data.
[0044] S102. Based on historical monitoring data within a preset time period, the hyperparameters of the initial machine learning model are iteratively optimized using a Bayesian optimization framework, wherein the Bayesian optimization framework is used to indicate the search range of the hyperparameters.
[0045] In this embodiment, a parameter search space specifically designed for photovoltaic prediction models can be constructed. Optionally, based on historical monitoring data within a preset time period, the hyperparameters of the initial machine learning model are iteratively optimized using a Bayesian optimization framework, including the following steps (1) to (3):
[0046] (1) Using the Bayesian optimization framework, determine the range of values for hyperparameters in the initial machine learning model. Hyperparameters include one or more of the following: learning rate, tree depth, sampling ratio, and regularization parameter.
[0047] Optionally, the learning rate has a non-linear effect on the model, with smaller values better suited for photovoltaic prediction. Optionally, the learning rate is searched on a logarithmic scale. The logarithmic scale more evenly covers the small value range, avoiding the situation where most trials fall into the invalid large value range during uniform sampling. For example, the learning rate is set to a range of 0.00001 to 1.0.
[0048] Optionally, the tree depth can be set to a range of 2 to 10. A tree depth of 2 to 10 is sufficient to fit nonlinear relationships, while a tree depth greater than 10 is prone to overfitting.
[0049] Optionally, the sampling ratio can range from 0.1 to 1.0. The sampling ratio can include a subsample and a colsample. By controlling the range of the sampling ratio, overfitting of the time series data can be avoided.
[0050] Optionally, the regularization parameter can be set to search on a logarithmic scale.
[0051] (2) Divide the historical monitoring data within the preset time period into training set and test set.
[0052] Optionally, the training set and the test set can be divided according to a preset ratio. For example, the ratio of historical monitoring data in the training set and the test set could be 0.8:0.2.
[0053] In some embodiments, such as Figure 2 As shown, historical monitoring data within a preset time period can be divided into training and testing sets according to chronological order.
[0054] For example, with a preset duration of 10 days, the historical monitoring data of the first 8 days can be divided into a training set, and the historical monitoring data of the last 2 days can be divided into a test set.
[0055] In this embodiment of the disclosure, data leakage can be avoided by ensuring that the data is divided into training and testing sets in strict chronological order.
[0056] In other embodiments, features that contribute highly to power generation prediction can be retained. Accordingly, historical monitoring data within a preset time period is divided into a training set and a test set, including: standardizing the historical monitoring data within the preset time period, the standardization process including timestamp alignment and categorical variable encoding; identifying and deleting redundant monitoring data from the historical monitoring data to obtain processed historical monitoring data, where redundant monitoring data are monitoring data with a correlation greater than a preset threshold; and dividing the processed historical monitoring data into a training set and a test set according to chronological order.
[0057] In this embodiment of the disclosure, the value of the preset threshold is not specifically limited. The timestamp alignment includes standardizing the time field into timestamps and sorting them chronologically. The categorical variable encoding can be numerical encoding of categorical features (such as weather or wind direction). Optionally, the standardization process may further include checking for and handling missing and outlier values.
[0058] (3) The initial machine learning model is trained using the training set and the test set, and the hyperparameters of the initial machine learning model are iteratively optimized within the range of hyperparameter values with the goal of minimizing the root mean square error of the predicted values.
[0059] In some embodiments, with the goal of minimizing the root mean square error of the predicted values, the hyperparameters of the initial machine learning model are iteratively optimized within the range of hyperparameter values. This includes: with the goal of minimizing the root mean square error of the predicted values, a tree structure estimation algorithm is used to dynamically adjust the values of the hyperparameters within the range of hyperparameter values based on historical experimental results, so as to iteratively optimize the hyperparameters of the initial machine learning model; wherein, the tree structure estimation algorithm is used to predict the values of the hyperparameters based on historical experimental results.
[0060] For example, such as Figure 2 As shown, with the goal of minimizing the root mean square error (RMSE) of the predicted values, a tree-structured estimation algorithm is used to iteratively explore the parameter space. In each trial, the algorithm intelligently suggests the next set of parameters that are more likely to achieve better results based on the results of historical trials, thereby finding the globally or near-globally optimal combination of hyperparameters with fewer trials.
[0061] S103. Continue until the initial machine learning model meets the preset training conditions to obtain a trained machine learning model, and use the trained machine learning model to predict the power generation of the photovoltaic power station in the next hour.
[0062] In the embodiments of this disclosure, the machine learning model can be an extreme gradient boosting tree model, or a gradient boosting decision tree model, a category feature gradient boosting model, or other tree-based machine learning models.
[0063] In some embodiments, obtaining a trained machine learning model until the initial machine learning model meets preset training conditions includes: obtaining a trained machine learning model until the number of iterations of the initial machine learning model reaches a preset number of iterations; or, determining the change in the root mean square error of the predicted values of the initial machine learning model, and obtaining a trained machine learning model until the number of consecutive times the change is less than a preset value reaches a preset number. In this embodiment, the preset number of iterations and the numerical value of the preset number are not specifically limited. Optionally, the preset number of iterations can be 50, 60, or 100. The preset number can be 2, 3, or 5, etc.
[0064] For example, such as Figure 2 As shown, the change in the root mean square error of the predictions of the initial machine learning model can be determined through the validation set.
[0065] In some embodiments, predicting the power generation of a photovoltaic power station in the next hour using a trained machine learning model includes: acquiring the current equipment status information of the photovoltaic power station and the weather forecast data for the next hour; inputting the current equipment status information of the photovoltaic power station and the weather forecast data for the next hour into the trained machine learning model, and outputting the power generation of the photovoltaic power station in the next hour.
[0066] It should be noted that the embodiments in this application are illustrated with the goal of minimizing the root mean square error of the predicted value. The root mean square error can also be replaced by mean absolute percentage error, mean square error, mean absolute error, or quantile loss, etc.
[0067] This application provides a method for predicting the power generation of a photovoltaic power station: acquiring historical monitoring data of the photovoltaic power station within a preset time period, including historical power generation information, historical equipment status information, and historical meteorological data; based on the historical monitoring data within the preset time period, iteratively optimizing the hyperparameters of an initial machine learning model using a Bayesian optimization framework, wherein the Bayesian optimization framework is used to indicate the search range of the hyperparameters; until the initial machine learning model meets preset training conditions, a trained machine learning model is obtained, and the power generation of the photovoltaic power station in the next hour is predicted using the trained machine learning model. In this application embodiment, because the hyperparameter combination of the machine learning model is iteratively optimized using a Bayesian optimization framework, the problems of low efficiency, reliance on human experience, and difficulty in finding the global optimum of traditional grid search or random search methods are solved. Furthermore, the Bayesian optimization framework can dynamically adjust the parameter search strategy based on historical experimental results, which can reduce the number of iterations of the training model. That is, this method significantly improves the efficiency and stability of hyperparameter tuning by intelligently guiding the exploration of the parameter space, thus improving the prediction accuracy of the model under complex meteorological conditions.
[0068] Figure 3 This is a schematic diagram of the structure of a photovoltaic power plant power generation prediction device provided in an embodiment of this application. Figure 3 As shown, the photovoltaic power plant power generation prediction device includes:
[0069] The acquisition module 301 is used to acquire historical monitoring data of the photovoltaic power station within a preset time period. The historical monitoring data includes historical power generation information, historical equipment status information, and historical meteorological data.
[0070] The iteration module 302 is used to iteratively optimize the hyperparameters of the initial machine learning model based on historical monitoring data within a preset time period using a Bayesian optimization framework, wherein the Bayesian optimization framework is used to indicate the search range of the hyperparameters.
[0071] The prediction module 303 is used to obtain a trained machine learning model until the initial machine learning model meets the preset training conditions, and to predict the power generation of the photovoltaic power station in the next hour through the trained machine learning model.
[0072] In one possible implementation, the iteration module 302 uses a Bayesian optimization framework to iteratively optimize the hyperparameters of the initial machine learning model based on historical monitoring data within a preset time period. This includes: using a Bayesian optimization framework to determine the value range of the hyperparameters in the initial machine learning model, whereby the hyperparameters include one or more of the following: learning rate, tree depth, sampling ratio, and regularization parameter; dividing the historical monitoring data within the preset time period into a training set and a test set; training the initial machine learning model using the training set and the test set, and iteratively optimizing the hyperparameters of the initial machine learning model within the value range of the hyperparameters with the goal of minimizing the root mean square error of the predicted values.
[0073] In one possible implementation, the iterative module 302 aims to minimize the root mean square error of the predicted values and iteratively optimizes the hyperparameters of the initial machine learning model within the range of hyperparameter values. This includes: using a tree structure estimation algorithm to dynamically adjust the hyperparameter values within the range of hyperparameter values based on historical experimental results, with the goal of minimizing the root mean square error of the predicted values, to iteratively optimize the hyperparameters of the initial machine learning model; wherein, the tree structure estimation algorithm is used to predict the hyperparameter values based on historical experimental results.
[0074] In one possible implementation, the iterative module 302 divides the historical monitoring data within a preset time period into a training set and a test set, including: standardizing the historical monitoring data within the preset time period, the standardization process including timestamp alignment and categorical variable encoding; identifying and deleting redundant monitoring data in the historical monitoring data to obtain processed historical monitoring data, the redundant monitoring data being monitoring data with a relevance greater than a preset threshold; and dividing the processed historical monitoring data into a training set and a test set according to the chronological order.
[0075] In one possible implementation, the prediction module 303 obtains a trained machine learning model until the initial machine learning model meets the preset training conditions, including: obtaining a trained machine learning model until the number of iterations of the initial machine learning model reaches the preset number of iterations; or, determining the change in the root mean square error of the predicted values of the initial machine learning model until the number of consecutive times the change is less than the preset value reaches the preset number of iterations, thereby obtaining a trained machine learning model.
[0076] In one possible implementation, the prediction module 303 predicts the power generation of the photovoltaic power station in the next hour using a trained machine learning model, including: acquiring the current equipment status information of the photovoltaic power station and the weather forecast data for the next hour; inputting the current equipment status information of the photovoltaic power station and the weather forecast data for the next hour into the trained machine learning model, and outputting the power generation of the photovoltaic power station in the next hour.
[0077] The photovoltaic power generation prediction device provided in this application solves the problems of low efficiency, reliance on human experience, and difficulty in finding the global optimum of traditional grid search or random search methods by iteratively optimizing the hyperparameter combination of the machine learning model through a Bayesian optimization framework. Furthermore, the Bayesian optimization framework can dynamically adjust the parameter search strategy based on historical experimental results, which can reduce the number of iterations of the training model. In other words, this method significantly improves the efficiency and stability of hyperparameter tuning by intelligently guiding the exploration of the parameter space, thereby improving the prediction accuracy of the model under complex weather conditions.
[0078] The photovoltaic power generation prediction device provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.
[0079] Figure 4 A schematic diagram of the structure of the electronic device provided in this application. Figure 4 As shown, the electronic device 40 provided in this embodiment includes at least one processor 401 and a memory 402. Optionally, the device 40 further includes a communication component 403. The processor 401, memory 402, and communication component 403 are connected via a bus 404.
[0080] In a specific implementation, at least one processor 401 executes computer execution instructions stored in memory 402, causing at least one processor 401 to perform the above-described method.
[0081] The specific implementation process of processor 401 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.
[0082] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.
[0083] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.
[0084] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.
[0085] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.
[0086] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.
[0087] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.
[0088] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.
[0089] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or electrical connection shown or discussed may be indirect coupling or electrical connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.
[0090] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.
[0091] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.
[0092] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0093] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.
[0094] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.
Claims
1. A method for predicting the power generation of a photovoltaic power plant, characterized in that, The method includes: Acquire historical monitoring data of the photovoltaic power station within a preset time period. The historical monitoring data includes historical power generation information, historical equipment status information, and historical meteorological data. Based on historical monitoring data within a preset time period, the hyperparameters of the initial machine learning model are iteratively optimized using a Bayesian optimization framework, where the Bayesian optimization framework is used to indicate the search range of the hyperparameters. The process continues until the initial machine learning model meets the preset training conditions, resulting in a trained machine learning model. The trained machine learning model is then used to predict the power generation of the photovoltaic power station in the next hour.
2. The method of claim 1, wherein, The step of iteratively optimizing the hyperparameters of the initial machine learning model using a Bayesian optimization framework based on historical monitoring data within a preset time period includes: A Bayesian optimization framework is used to determine the range of values for hyperparameters in the initial machine learning model. These hyperparameters include one or more of the following: learning rate, tree depth, sampling ratio, and regularization parameter. The historical monitoring data within a preset time period is divided into a training set and a test set; The initial machine learning model is trained using training and test sets, and the hyperparameters of the initial machine learning model are iteratively optimized within the range of the hyperparameter values with the goal of minimizing the root mean square error of the predicted values.
3. The method of claim 2, wherein, The step of iteratively optimizing the hyperparameters of the initial machine learning model within the range of the hyperparameter values, with the objective of minimizing the root mean square error of the predicted values, includes: With the goal of minimizing the root mean square error of the predicted values, a tree structure estimation algorithm is used to dynamically adjust the values of the hyperparameters within the range of the predicted values based on historical experimental results, so as to iteratively optimize the hyperparameters of the initial machine learning model; wherein, the tree structure estimation algorithm is used to predict the values of the hyperparameters based on historical experimental results.
4. The method according to claim 2, characterized in that, The step of dividing historical monitoring data within a preset time period into a training set and a test set includes: The historical monitoring data within a preset time period is standardized, and the standardization process includes timestamp alignment and categorical variable encoding. Redundant monitoring data in the historical monitoring data is identified and deleted to obtain processed historical monitoring data. The redundant monitoring data is monitoring data with a correlation greater than a preset threshold. The processed historical monitoring data is divided into a training set and a test set according to the chronological order.
5. The method according to claim 1, characterized in that, The process continues until the initial machine learning model meets the preset training conditions, resulting in a trained machine learning model, including: The training process continues until the initial machine learning model has undergone a preset number of iterations, resulting in a trained machine learning model; or, the change in the root mean square error of the predicted values of the initial machine learning model is determined until the number of consecutive times the change is less than a preset value reaches a preset number, resulting in a trained machine learning model.
6. The method according to claim 1, characterized in that, The prediction of the photovoltaic power station's power generation for the next hour using a trained machine learning model includes: Obtain the current equipment status information of the photovoltaic power station and the weather forecast data for the next hour; The current equipment status information of the photovoltaic power station and the weather forecast data for the next hour are input into the trained machine learning model, and the power generation of the photovoltaic power station for the next hour is output.
7. A device for predicting the power generation of a photovoltaic power station, characterized in that, The device includes: The acquisition module is used to acquire historical monitoring data of the photovoltaic power station within a preset time period. The historical monitoring data includes historical power generation information, historical equipment status information, and historical meteorological data. The iteration module is used to iteratively optimize the hyperparameters of the initial machine learning model based on historical monitoring data within a preset time period using a Bayesian optimization framework, where the Bayesian optimization framework is used to indicate the search range of the hyperparameters. The prediction module is used to obtain a trained machine learning model until the initial machine learning model meets the preset training conditions, and then use the trained machine learning model to predict the power generation of the photovoltaic power station in the next hour.
8. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-6.
9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-6.
10. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method described in any one of claims 1-6.