Photovoltaic power plant power prediction methods, devices, equipment, storage media and products

Abstract

The present invention relates to the field of power prediction technology and discloses a photovoltaic power station power prediction method, apparatus, device, storage medium, and product. The method comprises: integrating meteorological data from at least two meteorological data sources to obtain integrated multi-source meteorological data; predicting the predicted daily solar irradiance of the photovoltaic power station based on the multi-source meteorological data; and utilizing a deep learning algorithm to predict the predicted daily output power data of the photovoltaic power station based on the predicted daily solar irradiance and real-time plant environmental parameters. The present invention fully considers multiple influencing factors, updates power prediction results in real time, promptly reflects the latest meteorological changes, and provides accurate and reliable power generation predictions for photovoltaic power stations.

Description

Technical Field

[0001] This invention relates to the field of power prediction technology, specifically to a method, apparatus, equipment, storage medium, and product for predicting the power of a photovoltaic power plant. Background Technology

[0002] With the increasing global demand for renewable energy, solar energy, as a clean and sustainable energy source, has received widespread attention. In recent years, advancements in solar technology and cost reductions have led to its rapid global adoption. Especially in regions with abundant sunshine, solar power generation has become one of the main energy production methods. However, influenced by factors such as climate and geography, the efficiency and stability of solar power generation still face numerous challenges. Addressing these challenges is crucial for improving the efficiency of solar resource utilization and reducing energy costs.

[0003] High-altitude regions typically possess abundant solar energy resources, especially in summer with long hours of sunshine and high solar irradiance, providing favorable natural conditions for solar power generation. However, the complex and variable meteorological conditions in these regions, such as drastic temperature variations and frequent wind speed changes, place higher demands on the stability and power prediction of solar power generation equipment. Furthermore, the thin air at high altitudes also presents challenges to the accuracy of solar irradiance prediction. Therefore, developing efficient solar energy prediction technologies tailored to the unique climatic conditions of high-altitude regions is particularly important.

[0004] Current photovoltaic (PV) power forecasting technologies mostly rely on a single meteorological data source, and the forecasting models are primarily based on simple physical or statistical methods. This limits the accuracy and practicality of the forecasts to some extent. Especially under complex climatic conditions, a single data source cannot fully capture meteorological changes, leading to significant forecast errors. Furthermore, current PV power forecasting only considers meteorological factors and neglects environmental parameters of the power plant, such as the temperature of the photovoltaic panels, further amplifying the forecast error. Summary of the Invention

[0005] In view of this, the present invention provides a method, apparatus, equipment, storage medium and product for predicting the power of a photovoltaic power plant, in order to solve the technical problem of low prediction accuracy of existing photovoltaic power plant power prediction methods.

[0006] In a first aspect, the present invention provides a method for predicting the power output of a photovoltaic power plant, comprising: integrating meteorological data from at least two meteorological data sources to obtain integrated multi-source meteorological data; predicting the solar irradiance of the photovoltaic power plant on a predicted day based on the multi-source meteorological data; and using a deep learning algorithm to predict the predicted daily output power data of the photovoltaic power plant based on the predicted daily solar irradiance and real-time power plant environmental parameters.

[0007] This invention discloses a photovoltaic power plant power prediction method. By integrating meteorological data from at least two meteorological data sources, it reduces prediction errors caused by a single data source, improves the comprehensiveness and accuracy of meteorological information, predicts the solar irradiance of the photovoltaic power plant on the predicted day based on the integrated multi-source meteorological data, and predicts the output power data of the photovoltaic power plant on the predicted day by combining the predicted solar irradiance with real-time power plant environmental parameters. This method can fully consider various influencing factors, update the power prediction results in real time, and promptly reflect the latest meteorological changes. It provides accurate and reliable power generation predictions for photovoltaic power plants, improves the efficiency of solar energy resource utilization, optimizes power plant operation and management, and significantly improves economic benefits and power supply stability.

[0008] Optionally, predicting the predicted daily solar irradiance of the photovoltaic power station based on the multi-source meteorological data includes: inputting the multi-source meteorological data into a pre-constructed solar irradiance prediction model, wherein the solar irradiance prediction model is a prediction model combining a physical model and a machine learning model; and predicting the predicted daily solar irradiance of the photovoltaic power station through the solar irradiance prediction model.

[0009] By combining physical and machine learning models, solar irradiance prediction models can be developed to predict solar irradiance based on various meteorological data, making the prediction results more accurate.

[0010] Optionally, the deep learning algorithm is a Long Short-Term Memory (LSTM) algorithm. Correspondingly, using the deep learning algorithm to predict the predicted daily output power data of the photovoltaic power station based on the predicted daily solar irradiance and real-time power station environmental parameters includes: training the LTM network based on historical solar irradiance, historical power output data, and historical power station environmental parameters; inputting the predicted daily solar irradiance and the real-time power station environmental parameters into the trained LTM network; and predicting the predicted daily output power data of the photovoltaic power station through the LTM algorithm.

[0011] By utilizing the Long Short-Term Memory (LSTM) algorithm to predict power, it is possible to comprehensively consider historical data and current meteorological conditions, providing highly accurate prediction results.

[0012] Optionally, after obtaining the integrated multi-source meteorological data, the method further includes cleaning, aligning, and denoising the multi-source meteorological data.

[0013] By cleaning, aligning, and denoising the multi-source meteorological data, redundancy and noise in the data are removed, thereby improving data quality.

[0014] Optionally, meteorological data from at least two meteorological data sources are integrated to obtain integrated multi-source meteorological data, including: determining the optimal weight combination for the meteorological data from each meteorological data source; and integrating the meteorological data from each meteorological data source through the optimal weight combination to obtain integrated multi-source meteorological data.

[0015] By integrating data from various meteorological data sources through determining the optimal weight combination, the advantages and disadvantages of different data sources can be balanced, thereby improving the overall accuracy of forecasts.

[0016] Optionally, determining the optimal weight combination for meteorological data from various meteorological data sources includes: determining an objective function, which measures the power prediction accuracy of different weight combinations; initializing a prior distribution of the weight combinations; extracting a set of weight combinations from the prior distribution and integrating the meteorological data from various meteorological data sources using the extracted weight combinations, and obtaining a power prediction result based on the integrated multi-source meteorological data; calculating the power prediction accuracy under the current weight combination according to the objective function; calculating the posterior distribution using Bayes' theorem based on the power prediction accuracy under the current weight combination and the current prior distribution; using the posterior distribution as the new prior distribution and extracting a new set of weight combinations from the new prior distribution to iteratively update the prior distribution; and after satisfying the iteration conditions, selecting the weight combination with the highest prediction accuracy during the iteration process as the optimal weight combination.

[0017] Combining Bayesian optimization methods to determine the optimal weight combination can improve the optimization efficiency of the optimal weight combination.

[0018] Secondly, the present invention provides a photovoltaic power plant power prediction device, comprising: a multi-source data integration module for integrating meteorological data from at least two meteorological data sources to obtain integrated multi-source meteorological data; an irradiance prediction module for predicting the solar irradiance of the photovoltaic power plant on the predicted day based on the multi-source meteorological data; and a power prediction module for predicting the output power data of the photovoltaic power plant on the predicted day based on the solar irradiance on the predicted day and real-time power plant environmental parameters using a deep learning algorithm.

[0019] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the photovoltaic power plant power prediction method of the first aspect or any corresponding embodiment described above.

[0020] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the photovoltaic power plant power prediction method of the first aspect or any corresponding embodiment described above.

[0021] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute a photovoltaic power plant power prediction method as described in the first aspect or any corresponding embodiment thereof. Attached Figure Description

[0022] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 This is a flowchart illustrating the photovoltaic power plant power prediction method according to an embodiment of the present invention;

[0024] Figure 2 This is a structural block diagram of the photovoltaic power plant power prediction device according to an embodiment of the present invention;

[0025] Figure 3 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] In related technologies, photovoltaic power plant power prediction methods have the following problems:

[0028] 1. Limitations of a single data source:

[0029] Most existing photovoltaic (PV) power plants rely on a single meteorological data source, such as data from the European Centre for Medium-Range Weather Forecasts (ECMWF). This single-data-source approach has significant limitations because each meteorological data source has its unique advantages and disadvantages, making it difficult for a single data source to comprehensively and accurately reflect complex meteorological changes. Especially in high-altitude areas, where meteorological conditions are more variable, the accuracy of predictions from a single data source is greatly reduced, affecting the accuracy of solar irradiance predictions and consequently impacting the accuracy of PV power plant power predictions.

[0030] 2. Lack of data fusion and processing methods:

[0031] Current solar irradiance prediction systems suffer from shortcomings in data fusion and processing. Although some systems attempt to use multi-source data, data preprocessing, alignment, and fusion techniques are still immature, leading to data redundancy and severe noise problems, which affect the accuracy and reliability of prediction models. Existing technologies lack effective multimodal data fusion methods, making it difficult to fully utilize information from different meteorological data sources, resulting in inconsistent and unstable prediction results.

[0032] 3. The power prediction model lacks accuracy and real-time performance:

[0033] Most existing photovoltaic (PV) power prediction models are based on simple physical or statistical models, failing to adequately consider various influencing factors such as temperature, wind speed, cloud cover, and snow cover, as well as environmental parameters of the power plant such as PV panel temperature, resulting in low prediction accuracy. Furthermore, these models have low update frequency and poor real-time performance, making them ill-suited for rapidly changing weather conditions, especially in high-altitude areas. In such cases, the prediction results cannot reflect the latest weather changes in a timely manner, impacting the operation and management efficiency of PV power plants.

[0034] Therefore, there is an urgent need to develop a new generation of multi-source data fusion and high-precision prediction models to improve the efficiency and reliability of solar power generation. The photovoltaic power plant power prediction method of this invention aims to improve the accuracy and real-time performance of solar power generation prediction in high-altitude areas. By integrating multi-source meteorological data, employing advanced data processing and fusion technologies, and constructing a high-precision solar irradiance and power prediction model, it comprehensively optimizes the efficiency of solar resource utilization, improves the operation and management level of photovoltaic power plants, and achieves higher economic benefits and power supply stability.

[0035] According to an embodiment of the present invention, a method for predicting the power of a photovoltaic power plant is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0036] This embodiment provides a photovoltaic power plant power prediction method, which can be used on mobile terminals such as computers and mobile phones. Figure 1 As shown, the process includes the following steps:

[0037] Step S101: Integrate meteorological data from at least two meteorological data sources to obtain integrated multi-source meteorological data.

[0038] Specifically, the integrated meteorological data comes from multiple different meteorological data sources, such as the European Centre for Medium-Range Weather Forecasts (ECMWF), the National Weather Service, and the U.S. Center for Environmental Prediction (GFS). By integrating meteorological data from different sources through multi-source data fusion technology, the advantages of each data source are combined to overcome the limitations of a single data source and provide comprehensive and accurate meteorological information. This enables a more comprehensive capture of the complex and variable meteorological conditions in high-altitude areas, laying the foundation for subsequent solar irradiance and power prediction.

[0039] Step S102: Predict the predicted daily solar irradiance of the photovoltaic power station based on multi-source meteorological data.

[0040] Specifically, multi-source meteorological data includes at least two meteorological factors that affect solar irradiance, such as temperature, cloud cover, wind speed, and humidity.

[0041] The pre-trained solar irradiance prediction model accurately predicts solar irradiance based on various meteorological factors such as temperature, cloud cover, wind speed, and humidity.

[0042] Step S103: Using a deep learning algorithm, predict the predicted daily output power data of the photovoltaic power station based on the predicted daily solar irradiance and real-time power station environmental parameters.

[0043] Specifically, real-time power plant environmental parameters are real-time operating parameters that affect the power generation of the power plant, such as the temperature of the photovoltaic panels.

[0044] Since photovoltaic power plants generate electricity using solar energy, solar irradiance and power generation are closely related. Power generation can be predicted relatively accurately based on solar irradiance. In addition, the influence of real-time power plant environmental parameters on power generation is also considered, making the power prediction results more accurate.

[0045] This invention discloses a photovoltaic power plant power prediction method that integrates meteorological data from at least two meteorological data sources to reduce prediction errors caused by a single data source, improves the comprehensiveness and accuracy of meteorological information, predicts the predicted daily solar irradiance of the photovoltaic power plant based on the integrated multi-source meteorological data, and predicts the predicted daily output power data of the photovoltaic power plant by combining the predicted daily solar irradiance with real-time power plant environmental parameters. This method can fully consider various influencing factors, update the power prediction results in real time, and promptly reflect the latest meteorological changes, providing accurate and reliable power generation predictions for photovoltaic power plants, improving the efficiency of solar energy resource utilization, optimizing power plant operation and management, and significantly improving economic benefits and power supply stability.

[0046] In some embodiments, step S102, predicting the predicted daily solar irradiance of the photovoltaic power station based on multi-source meteorological data, includes:

[0047] Step S1021: Input multi-source meteorological data into a pre-built solar irradiance prediction model, wherein the solar irradiance prediction model is a prediction model that combines a physical model and a machine learning model.

[0048] Step S1022: The predicted daily solar irradiance of the photovoltaic power station is predicted using the solar irradiance prediction model.

[0049] Specifically, this invention combines physical and machine learning models to construct a high-precision solar irradiance prediction model. This model can consider various factors affecting solar irradiance, such as temperature, cloud cover, wind speed, and humidity. By combining real-time updated data and Bayesian methods for iterative optimization, the solar irradiance prediction model can quickly respond to changes in meteorological conditions, providing accurate solar irradiance prediction results and a reliable basis for solar power generation.

[0050] The embodiments of the present invention, through a solar irradiance prediction model that combines physical models and machine learning models, can predict solar irradiance based on a variety of meteorological data, making the prediction results more accurate.

[0051] In some embodiments, the deep learning algorithm is a Long Short-Term Memory (LSTM) algorithm.

[0052] Correspondingly, step S102 involves using a deep learning algorithm to predict the predicted daily output power data of the photovoltaic power station based on the predicted daily solar irradiance and real-time power station environmental parameters, including:

[0053] Step S1021: Train the Long Short-Term Memory Network based on historical solar irradiance, historical power output data, and historical power plant environmental parameters.

[0054] Step S1022: Input the predicted daily solar irradiance and real-time power plant environmental parameters into the trained long short-term memory network.

[0055] Step S1023: The predicted daily output power data of the photovoltaic power station is obtained by using the Long Short-Term Memory algorithm.

[0056] Specifically, to achieve high-precision real-time power prediction and optimization, this embodiment of the invention employs an advanced deep learning algorithm—Long Short-Term Memory (LSTM). LSTM is a special type of recurrent neural network that can effectively process and predict time-series data, and is particularly suitable for capturing the complex time dependencies and nonlinear characteristics in solar power generation.

[0057] LSTM, through its unique memory cells and gating mechanism, is able to remember long-term time dependencies and ignore irrelevant information. This allows LSTM to provide highly accurate predictions by comprehensively considering historical and current meteorological data when forecasting changes in solar irradiance and power.

[0058] It should be understood that historical solar irradiance, historical power output data, and historical power plant environmental parameters are solar irradiance, power, and environmental parameter data prior to the prediction date, obtained through the power plant's historical records. Real-time power plant environmental parameters are obtained through real-time monitoring of the power plant's operating parameters.

[0059] In one example, historical solar irradiance and predicted daily solar irradiance are used to form irradiance time series data, historical power plant environmental parameters and real-time power plant environmental parameters are used to form environmental parameter time series data, and historical power output data is used as power time series data. These time series data are learned using an LSTM network with a long short-term memory algorithm, and predicted daily output power data is output.

[0060] In the application of this invention, an LSTM network is trained using historical solar irradiance, historical power output data, and historical power plant environmental parameters to learn patterns and relationships within this data. This enables real-time prediction of the output power of solar power plants in practical applications. This real-time power prediction and optimization algorithm, combining LSTM deep learning and reinforcement learning, allows for accurate prediction of changes in photovoltaic power generation under complex and variable weather conditions.

[0061] In some embodiments, step S101, after obtaining the integrated multi-source meteorological data, further includes cleaning, aligning, and denoising the multi-source meteorological data.

[0062] Specifically, after data integration, advanced data preprocessing and noise filtering techniques are employed to clean, align, and denoise the multi-source meteorological data. Machine learning algorithms and statistical methods are used to remove redundancy and noise from the data, improving data quality. This step ensures the accuracy and reliability of the data input to the prediction model, thereby enhancing the precision of the prediction results.

[0063] In some embodiments, step S101 involves integrating meteorological data from at least two meteorological data sources to obtain integrated multi-source meteorological data, including:

[0064] Step S1011: Determine the optimal weight combination for meteorological data from each meteorological data source.

[0065] Specifically, the Bayesian optimization method is used to gradually approximate the optimal weight combination using a probabilistic model, thereby achieving the optimal weight allocation for the data source and improving prediction accuracy.

[0066] The specific process of the Bayesian optimization method includes the following steps:

[0067] 1. Define the objective function: First, define an objective function (root mean square error, correlation coefficient, etc.) to measure the prediction accuracy under different weight combinations.

[0068] 2. Initialize the prior distribution: Initialize the prior distribution of the weight combination by selecting a uniform distribution as the prior distribution, which represents the initial uncertainty of the weight combination.

[0069] 3. Sampling: A set of weight combinations is randomly selected from the prior distribution, and then these weight combinations are used to fuse multi-source meteorological data to obtain power prediction results.

[0070] 4. Evaluate the objective function: Use historical data for cross-validation, calculate the prediction accuracy under the current weight combination based on the objective function, and ensure the reliability of the evaluation results.

[0071] 5. Update the posterior distribution: Based on the prediction accuracy of the current weight combination and the prior distribution, calculate the posterior distribution using Bayes' theorem.

[0072] 6. Select a new weight combination: Based on the posterior distribution, select the weight combination with the highest prediction accuracy as the starting point for the next iteration.

[0073] 7. Repeat steps 3-6 until the stopping condition is met: You can set the number of iterations (e.g., 10 times), prediction accuracy threshold, or computational resource limit as the stopping condition.

[0074] 8. Output the optimal weight combination: After the stopping condition is met, select the weight combination with the highest prediction accuracy as the final result.

[0075] Step S1012: The meteorological data from each meteorological data source is integrated by combining the optimal weights to obtain the integrated multi-source meteorological data.

[0076] Specifically, the optimal weight combination determines the weight of each meteorological data point, and weights are assigned to the meteorological data from each meteorological data source based on the optimal weight combination. For example, taking temperature data as an example, there are two meteorological data sources, 35℃ and 30℃, and the weights obtained according to the optimal weight combination are 0.6 and 0.4 respectively. Therefore, the integrated multi-source meteorological data of temperature is 0.6×35℃+0.4×30℃.

[0077] This invention combines Bayesian optimization methods to determine the optimal weight combination, which can improve the optimization efficiency of the optimal weight combination, balance the advantages and disadvantages of different data sources, and improve the overall prediction accuracy.

[0078] This invention discloses a photovoltaic power plant power prediction method. By integrating multi-source meteorological data and employing advanced data processing and fusion technologies, a high-precision solar irradiance prediction model is constructed. This model is then combined with a long short-term memory algorithm to predict the predicted daily solar irradiance and consequently, the predicted daily output power of the photovoltaic power plant. This significantly improves the accuracy and real-time performance of solar power generation prediction. Firstly, multi-source data fusion leverages the advantages of various meteorological data sources, reducing prediction errors from single data sources and enhancing the comprehensiveness and accuracy of meteorological information. Secondly, advanced data processing and fusion technologies effectively filter data noise, optimize data quality, and enhance the stability of the prediction model. Finally, the high-precision solar irradiance prediction model considers multiple influencing factors, updates the predicted daily solar irradiance in real time, and promptly reflects the latest meteorological changes. This provides an accurate and reliable basis for photovoltaic power plant power generation prediction. Furthermore, by combining real-time power plant environmental parameters with learning and prediction, accurate power generation power is obtained, optimizing power plant operation and management, and significantly improving economic efficiency and power supply stability.

[0079] This invention also provides a photovoltaic power plant power prediction device, such as... Figure 2 As shown, it includes:

[0080] The multi-source data integration module 201 is used to integrate meteorological data from at least two meteorological data sources to obtain integrated multi-source meteorological data.

[0081] Irradiance prediction module 202 is used to predict the predicted daily solar irradiance of a photovoltaic power station based on multi-source meteorological data;

[0082] The power prediction module 203 is used to predict the predicted daily output power data of the photovoltaic power station based on the predicted daily solar irradiance and real-time power station environmental parameters using a deep learning algorithm.

[0083] The photovoltaic power plant power prediction device of this invention reduces the prediction error caused by a single data source by integrating meteorological data from at least two meteorological data sources, improves the comprehensiveness and accuracy of meteorological information, predicts the solar irradiance of the photovoltaic power plant on the predicted day based on the integrated multi-source meteorological data, and predicts the output power data of the photovoltaic power plant on the predicted day by using the predicted solar irradiance and real-time power plant environmental parameters. It can fully consider various influencing factors, update the power prediction results in real time, and reflect the latest meteorological changes in a timely manner, providing accurate and reliable power generation predictions for photovoltaic power plants, improving the efficiency of solar energy resource utilization, optimizing power plant operation and management, and significantly improving economic benefits and power supply stability.

[0084] Furthermore, the irradiance prediction module 202 includes:

[0085] The meteorological data input module is used to input multi-source meteorological data into a pre-built solar irradiance prediction model, which is a prediction model that combines a physical model and a machine learning model.

[0086] The model prediction module is used to predict the daily solar irradiance of a photovoltaic power station using a solar irradiance prediction model.

[0087] Furthermore, the deep learning algorithm is a long short-term memory algorithm; correspondingly, the power prediction module 203 includes:

[0088] The network training module is used to train the long short-term memory network based on historical solar irradiance, historical power output data, and historical power plant environmental parameters.

[0089] The parameter input module is used to input the predicted daily solar irradiance and real-time power plant environmental parameters into the trained long short-term memory network;

[0090] The power output module is used to predict the daily output power data of the photovoltaic power station through a long short-term memory algorithm.

[0091] Furthermore, the photovoltaic power plant power prediction device also includes:

[0092] The data processing module is used to clean, align, and denoise multi-source meteorological data.

[0093] Furthermore, the multi-source data integration module includes:

[0094] The weight optimization module is used to determine the optimal weight combination for meteorological data from various meteorological data sources.

[0095] The weighting integration module is used to integrate meteorological data from various meteorological data sources through the optimal weight combination to obtain integrated multi-source meteorological data.

[0096] Furthermore, the weight optimization module includes:

[0097] The function determination module is used to determine the objective function, which is used to measure the power prediction accuracy of different weight combinations;

[0098] The prior distribution initialization module is used to initialize the prior distribution of the weight combination;

[0099] The prediction result acquisition module is used to extract a set of weight combinations from the prior distribution and integrate the meteorological data from various meteorological data sources using the extracted weight combinations, and obtain the power prediction result based on the integrated multi-source meteorological data.

[0100] The prediction accuracy calculation module is used to calculate the power prediction accuracy under the current weight combination based on the objective function.

[0101] The posterior distribution calculation module is used to calculate the posterior distribution using Bayes' theorem based on the power prediction accuracy under the current weight combination and the current prior distribution.

[0102] The iterative module is used to take the posterior distribution as the new prior distribution and draw a new set of weight combinations from the new prior distribution to iteratively update the prior distribution.

[0103] The weight output module is used to select the weight combination with the highest prediction accuracy during the iteration process as the optimal weight combination after the iteration conditions are met.

[0104] This invention also provides a schematic diagram of the structure of a computer device, such as... Figure 3 As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 3 Take a processor 10 as an example.

[0105] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0106] The memory 20 stores instructions executable by at least one processor 10 to cause at least one processor 10 to perform the method shown in the above embodiments.

[0107] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0108] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0109] The computer device also includes an input device 30 and an output device 40. The processor 10, memory 20, input device 30, and output device 40 can be connected via a bus or other means. Figure 3 The bus connection is taken as an example.

[0110] Input device 30 can receive input numerical or character information, and generate key signal inputs related to user settings and function control of the computer device, such as a touchscreen, keypad, mouse, trackpad, touchpad, joystick, one or more mouse buttons, trackball, joystick, etc. Output device 40 may include display devices, auxiliary lighting devices (e.g., LEDs), and haptic feedback devices (e.g., vibration motors). The aforementioned display devices include, but are not limited to, liquid crystal displays, light-emitting diodes, displays, and plasma displays. In some alternative embodiments, the display device may be a touchscreen.

[0111] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0112] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0113] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and all such modifications and variations fall within the scope of protection.

Claims

1. A method for predicting the power output of a photovoltaic power plant, characterized in that, include: Integrating meteorological data from at least two meteorological data sources to obtain integrated multi-source meteorological data includes: determining the optimal weight combination for meteorological data from each meteorological data source; and integrating the meteorological data from each meteorological data source using the optimal weight combination to obtain integrated multi-source meteorological data. The predicted daily solar irradiance of the photovoltaic power station is predicted based on the multi-source meteorological data. Using deep learning algorithms, the predicted daily output power data of the photovoltaic power station is predicted based on the predicted daily solar irradiance and real-time power station environmental parameters; This includes determining the optimal weight combination for meteorological data from various meteorological data sources, including: Determine the objective function, which is used to measure the power prediction accuracy of different weight combinations; Initialize the prior distribution of the weight combination; A set of weight combinations is extracted from the prior distribution, and the extracted weight combinations are used to integrate the meteorological data from various meteorological data sources. Based on the integrated multi-source meteorological data, the power prediction result is obtained. Calculate the power prediction accuracy under the current weight combination based on the objective function; Based on the power prediction accuracy under the current weight combination and the current prior distribution, the posterior distribution is calculated using Bayes' theorem. The posterior distribution is used as the new prior distribution, and a new set of weight combinations is drawn from the new prior distribution to iteratively update the prior distribution. After the iteration conditions are met, the weight combination with the highest prediction accuracy during the iteration process is selected as the optimal weight combination.

2. The photovoltaic power plant power prediction method according to claim 1, characterized in that, The predicted daily solar irradiance of the photovoltaic power station is predicted based on the multi-source meteorological data, including: The multi-source meteorological data is input into a pre-constructed solar irradiance prediction model, wherein the solar irradiance prediction model is a prediction model that combines a physical model and a machine learning model. The predicted daily solar irradiance of the photovoltaic power station is obtained by using the solar irradiance prediction model.

3. The photovoltaic power plant power prediction method according to claim 1, characterized in that, The deep learning algorithm is a long short-term memory algorithm; Correspondingly, using deep learning algorithms, the predicted daily output power data of the photovoltaic power station is predicted based on the predicted daily solar irradiance and real-time power station environmental parameters, including: The long short-term memory network was trained based on historical solar irradiance, historical power output data, and historical power plant environmental parameters. The predicted daily solar irradiance and the real-time power plant environmental parameters are input into a trained long short-term memory network. The predicted daily output power data of the photovoltaic power station is obtained by using the Long Short-Term Memory algorithm.

4. The photovoltaic power plant power prediction method according to claim 1, characterized in that, After obtaining the integrated multi-source meteorological data, it also includes: The multi-source meteorological data is cleaned, aligned, and denoised.

5. A photovoltaic power plant power prediction device, characterized in that, include: The multi-source data integration module is used to integrate meteorological data from at least two meteorological data sources to obtain integrated multi-source meteorological data. The multi-source data integration module includes: a weight optimization module, used to determine the optimal weight combination for meteorological data from each meteorological data source; and a weight integration module, used to integrate meteorological data from each meteorological data source through the optimal weight combination to obtain integrated multi-source meteorological data. The irradiance prediction module is used to predict the predicted daily solar irradiance of the photovoltaic power station based on the multi-source meteorological data. The power prediction module is used to predict the predicted daily output power data of the photovoltaic power station based on the predicted daily solar irradiance and real-time power station environmental parameters using a deep learning algorithm. The weight optimization module includes: The function determination module is used to determine the objective function, which is used to measure the power prediction accuracy of different weight combinations; The prior distribution initialization module is used to initialize the prior distribution of the weight combination; The prediction result acquisition module is used to extract a set of weight combinations from the prior distribution and integrate the meteorological data from various meteorological data sources using the extracted weight combinations, and obtain the power prediction result based on the integrated multi-source meteorological data. The prediction accuracy calculation module is used to calculate the power prediction accuracy under the current weight combination based on the objective function. The posterior distribution calculation module is used to calculate the posterior distribution using Bayes' theorem based on the power prediction accuracy under the current weight combination and the current prior distribution. The iterative module is used to take the posterior distribution as the new prior distribution and draw a new set of weight combinations from the new prior distribution to iteratively update the prior distribution. The weight output module is used to select the weight combination with the highest prediction accuracy during the iteration process as the optimal weight combination after the iteration conditions are met.

6. A computer device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the photovoltaic power plant power prediction method according to any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the photovoltaic power plant power prediction method according to any one of claims 1 to 4.

8. A computer program product, characterized in that, It includes computer instructions for causing a computer to execute the photovoltaic power plant power prediction method as described in any one of claims 1 to 4.

Images