Calibration method and system for climate prediction boundary field
Through the method of intelligent calibration climate forecasting border field, climate data and machine learning models are used to solve the problem of climate prediction errors caused by inaccurate boundaries in boundary field settings, and the operating efficiency of the new energy system and the scheduling efficiency of the power system are improved.
Patent Information
- Application Number
- CN202411209719.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2025-01-24
AI Technical Summary
Inaccurate setting of boundary fields can lead to climate prediction errors, affecting the evaluation and management of wind and solar resources.
By obtaining climate data, prepro processing and generating training data sets. Based on training data set training prediction models, the boundary field of the predictive model is calibrated by real -time climate conditions.
It significantly improves the accuracy of climate prediction, reduces the errors of wind and solar resources prediction, and optimize the operating efficiency of the new energy system and the scheduling efficiency of the power system.
Smart Images
Figure CN119357657A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of climate prediction, and more specifically, to a calibration method and system for a climate prediction boundary field. Background Art
[0002] As the global demand for renewable energy continues to grow, the development and utilization of wind and solar energy resources has become increasingly important. However, the generation of wind and solar energy is highly dependent on meteorological conditions, so accurate climate forecasting plays a key role in the assessment, planning and management of wind and solar energy resources. Climate forecasting provides an estimate of meteorological conditions for a period of time in the future, including wind speed, wind direction, solar radiation intensity, temperature, etc. These meteorological factors directly affect the power generation capacity and efficiency of wind farms and photovoltaic power plants.
[0003] In the field of wind energy, accurate climate forecasts can help determine the best location and layout of wind farms, optimize the scheduling of wind turbines, and reduce the volatility of power generation caused by changes in wind speed. In the field of solar energy, climate forecasts can be used to assess the solar energy potential in different regions, optimize the design and layout of photovoltaic systems, and develop reasonable energy management strategies to improve the utilization of solar energy.
[0004] Climate model prediction is to simulate the earth system (such as atmosphere, ocean, land and ice sheet) through numerical models to predict future climate conditions. Climate models require initial conditions and boundary conditions to operate. Boundary fields refer to external conditions or constraints that climate models must follow during simulation. These conditions include but are not limited to atmospheric temperature, humidity, wind speed, sea surface temperature, solar radiation, surface temperature, etc.
[0005] The setting of boundary fields is usually based on historical data, reanalysis data or output of other models to provide necessary external inputs for climate models. These inputs ensure that the models can reasonably reflect the physical processes in the real world, making the climate prediction results more credible.
[0006] The setting of boundary fields has a crucial impact on the accuracy of climate prediction. If the boundary fields are not set accurately, it may cause significant deviations between the meteorological elements (such as wind speed, temperature, precipitation, etc.) output by the model and the actual situation, thus affecting the reliability and effectiveness of the prediction.
[0007] In the forecast of new energy wind and solar power, the error in boundary field setting may lead to deviations in the forecast of key meteorological elements such as wind speed and solar radiation, which in turn affects the evaluation and management of wind and solar power generation. For example, in wind energy forecast, if the wind speed setting in the boundary field is inaccurate, it may lead to over- or underestimation of wind power generation potential, thus affecting the planning and economic benefits of wind farms; in solar energy forecast, inaccurate solar radiation boundary field may lead to incorrect assessment of solar power generation potential, affecting the design and layout of photovoltaic power plants.
[0008] Therefore, accurately setting the boundary field and performing effective calibration can significantly improve the accuracy of climate forecasts, thereby providing a more reliable basis for resource assessment and power generation scheduling of new energy (wind and solar energy).
[0009] In summary, accurate climate prediction is crucial to the development and utilization of new energy wind and solar resources, and the boundary field setting in the climate model directly affects the accuracy of the prediction. Therefore, developing a system that can intelligently calibrate the boundary field is of great practical significance for improving the accuracy of wind and solar energy predictions and reducing prediction errors. This will not only help optimize the utilization of new energy, but also promote the larger-scale application of renewable energy and contribute to addressing global climate change. Summary of the invention
[0010] In view of the above problems, the present invention proposes a calibration method for a climate prediction boundary field, comprising:
[0011] Acquiring climate data, preprocessing the climate data, and generating a training data set;
[0012] Based on the training data set, a prediction model is obtained by training;
[0013] Calibrate the climate prediction boundary fields of the forecast model based on real-time climate condition data.
[0014] Optional, climate data, including: historical climate data, real-time observation data, and climate prediction model output data;
[0015] The historical climate data include: wind speed, wind direction, solar radiation intensity, temperature and humidity data;
[0016] The real-time observation data includes: real-time wind speed, wind direction and solar radiation intensity data;
[0017] The output data of the weather prediction model include: output data of existing climate models or numerical weather prediction models.
[0018] Optional preprocessing includes: denoising, normalization and missing value filling.
[0019] Optionally, based on the training data set, training a prediction model includes:
[0020] Extracting feature vectors from the training data set;
[0021] The preset model is used to learn the feature vector using a preset machine learning algorithm, and an optimization algorithm is used to adjust the parameters of the preset model to generate a prediction model.
[0022] Optional, pre-set machine learning algorithms, including:
[0023] Neural network algorithm, random forest algorithm, or support vector machine algorithm.
[0024] Optional, optimization algorithms, including:
[0025] Gradient descent algorithm.
[0026] Optionally, calibrate the climate prediction boundary fields of the forecast model based on real-time climate condition data, including:
[0027] Using real-time climate conditions as input features of a prediction model, inputting the input features into the prediction model, and predicting optimal boundary field configuration parameters;
[0028] The optimal boundary field configuration parameters are output as the calibration pattern.
[0029] Optionally, the method further includes:
[0030] Calculate the error between the predicted value of the prediction model and the actual observed value;
[0031] Determine the source of the error, and based on the source of the error, adjust the parameters of the prediction model until the error requirements are met.
[0032] In another aspect, the present invention further provides a climate prediction boundary field calibration system, comprising:
[0033] A data processing unit, used to obtain climate data, pre-process the climate data, and generate a training data set;
[0034] A training unit, used for training a prediction model based on the training data set;
[0035] The calibration unit is used to calibrate the climate prediction boundary field of the prediction model according to the real-time climate condition data.
[0036] Optional, climate data, including: historical climate data, real-time observation data, and climate prediction model output data;
[0037] The historical climate data include: wind speed, wind direction, solar radiation intensity, temperature and humidity data;
[0038] The real-time observation data includes: real-time wind speed, wind direction and solar radiation intensity data;
[0039] The output data of the weather prediction model include: output data of existing climate models or numerical weather prediction models.
[0040] Optional preprocessing includes: denoising, normalization and missing value filling.
[0041] Optionally, based on the training data set, training a prediction model includes:
[0042] Extracting feature vectors from the training data set;
[0043] The preset model is used to learn the feature vector using a preset machine learning algorithm, and an optimization algorithm is used to adjust the parameters of the preset model to generate a prediction model.
[0044] Optional, pre-set machine learning algorithms, including:
[0045] Neural network algorithm, random forest algorithm, or support vector machine algorithm.
[0046] Optional, optimization algorithms, including:
[0047] Gradient descent algorithm.
[0048] Optionally, calibrate the climate prediction boundary fields of the forecast model based on real-time climate condition data, including:
[0049] Using real-time climate conditions as input features of a prediction model, inputting the input features into the prediction model, and predicting optimal boundary field configuration parameters;
[0050] The optimal boundary field configuration parameters are output as the calibration pattern.
[0051] Optionally, the calibration unit is also used to:
[0052] Calculate the error between the predicted value of the prediction model and the actual observed value;
[0053] Determine the source of the error, and based on the source of the error, adjust the parameters of the prediction model until the error requirements are met.
[0054] In yet another aspect, the present invention further provides a computing device, comprising: one or more processors;
[0055] a processor for executing one or more programs;
[0056] When the one or more programs are executed by the one or more processors, the above-described method is implemented.
[0057] In yet another aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed, the method described above is implemented.
[0058] Compared with the prior art, the present invention has the following beneficial effects:
[0059] The present invention provides a method for calibrating a climate prediction boundary field, comprising: obtaining climate data, preprocessing the climate data, and generating a training data set; training a prediction model based on the training data set; and calibrating the climate prediction boundary field of the prediction model according to real-time climate condition data. The application of the present invention can improve the operating efficiency of the new energy system, and effectively improve the dispatching efficiency and economic benefits of the power system. BRIEF DESCRIPTION OF THE DRAWINGS
[0060] Figure 1 is a flow chart of the method of the present invention;
[0061] Figure 2 A flowchart of an embodiment of the method of the present invention;
[0062] Figure 3 It is a structural diagram of the system of the present invention. DETAILED DESCRIPTION
[0063] Now, exemplary embodiments of the present invention are described with reference to the accompanying drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to disclose the present invention in detail and completely and to fully convey the scope of the present invention to those skilled in the art. The terms used in the exemplary embodiments shown in the accompanying drawings are not intended to limit the present invention. In the accompanying drawings, the same units / elements are marked with the same reference numerals.
[0064] Unless otherwise specified, the terms (including technical terms) used herein have the commonly understood meanings to those skilled in the art. In addition, it is understood that the terms defined in commonly used dictionaries should be understood to have the same meanings as those in the context of the relevant fields, and should not be understood as idealized or overly formal meanings.
[0065] Embodiment 1:
[0066] The present invention proposes a calibration method for climate prediction boundary fields, such as Figure 1 As shown, including:
[0067] Step 1: Acquire climate data, pre-process the climate data, and generate a training data set;
[0068] Step 2: Based on the training data set, a prediction model is obtained by training;
[0069] Step 3: Calibrate the climate prediction boundary field of the prediction model based on real-time climate condition data.
[0070] Among them, climate data includes: historical climate data, real-time observation data and climate prediction model output data;
[0071] The historical climate data include: wind speed, wind direction, solar radiation intensity, temperature and humidity data;
[0072] The real-time observation data includes: real-time wind speed, wind direction and solar radiation intensity data;
[0073] The output data of the weather prediction model include: output data of existing climate models or numerical weather prediction models.
[0074] Among them, preprocessing includes: denoising, normalization and missing value filling.
[0075] Wherein, based on the training data set, the prediction model is trained, including:
[0076] Extracting feature vectors from the training data set;
[0077] The preset model is used to learn the feature vector using a preset machine learning algorithm, and an optimization algorithm is used to adjust the parameters of the preset model to generate a prediction model.
[0078] Among them, the preset machine learning algorithms include:
[0079] Neural network algorithm, random forest algorithm, or support vector machine algorithm.
[0080] Among them, the optimization algorithm includes:
[0081] Gradient descent algorithm.
[0082] Among them, according to the real-time climate condition data, the climate prediction boundary field of the prediction model is calibrated, including:
[0083] Using real-time climate conditions as input features of a prediction model, inputting the input features into the prediction model, and predicting optimal boundary field configuration parameters;
[0084] The optimal boundary field configuration parameters are output as the calibration pattern.
[0085] The method further includes:
[0086] Calculate the error between the predicted value of the prediction model and the actual observed value;
[0087] Determine the source of the error, and based on the source of the error, adjust the parameters of the prediction model until the error requirements are met.
[0088] The following is further described in conjunction with the embodiments of the present invention:
[0089] Example steps Figure 2 As shown, including:
[0090] (1) Data collection and processing:
[0091] Data sources include historical climate data, real-time observation data, and climate prediction model output data. The specific steps are as follows:
[0092] 1) Historical climate data: Collect climate data from the past few years or even decades, such as wind speed, wind direction, solar radiation intensity, temperature, humidity, etc. These data can come from weather stations, satellite observations, and reanalysis datasets.
[0093] 2) Real-time observation data: Get current meteorological observation data, such as real-time wind speed, wind direction, solar radiation intensity, etc. These data can come from weather stations, anemometers, radiometers, radars or satellites.
[0094] 3) Climate prediction model output data: Output from existing climate models or numerical weather prediction models, which serve as reference data to help the model identify and correct possible prediction errors.
[0095] These data are preprocessed (including denoising, normalization, missing value filling, etc.) to form a training data set for machine learning model training.
[0096] (2) Model training:
[0097] Model training is the core of intelligent boundary field calibration. By training a machine learning model, it can predict and adjust the boundary field configuration under different conditions. The specific steps include:
[0098] 1) Feature extraction:
[0099] Extract feature vectors from the collected multi-source data. For wind energy prediction, extract features such as the annual average, seasonal changes, and diurnal changes of wind speed; for solar energy prediction, extract features such as solar radiation intensity and sunshine duration.
[0100] 2) Model selection:
[0101] Choose appropriate machine learning algorithms such as neural networks, random forests, support vector machines, etc.
[0102] 3) Training process:
[0103] The training data is fed into the model and the model parameters are adjusted through optimization algorithms such as gradient descent so that the model can accurately predict the boundary field settings.
[0104] (3) Adaptive calibration:
[0105] Adaptive calibration uses the model to analyze current climate conditions in real time and dynamically adjust boundary field input parameters to optimize wind and solar resource forecast results. The main steps include:
[0106] 1) Real-time input:
[0107] Real-time acquisition of current meteorological conditions data as input features of the model, such as real-time wind speed, wind direction, solar radiation intensity, etc.
[0108] 2) Predict boundary field configuration:
[0109] Use the trained machine learning model to predict the optimal boundary field configuration parameters p based on real-time input opt :
[0110] p opt =f(x real-time )
[0111] Among them, f is the trained model, x real-time is the current real-time input.
[0112] 3) Adjust the boundary field:
[0113] Apply the predicted boundary field configuration parameters to a wind or solar forecast model to generate adjusted model output.
[0114] (4) Error assessment and feedback:
[0115] Error assessment and feedback are important steps in optimizing the model. By comparing the model prediction results with the actual observation data, the calibration ability of the model is continuously optimized. The specific steps are as follows:
[0116] 1) Error calculation:
[0117] Calculate the model prediction value y pred The actual observed value y obs The error E between:
[0118] E=y pred -y obs
[0119] 2) Error analysis:
[0120] The source of analysis error may be due to model bias, quality of input data or inaccurate boundary condition setting.
[0121] 3) Model feedback:
[0122] Adjust model parameters through error feedback. For example, in a neural network, the weight matrix W and bias b can be adjusted through the back-propagation algorithm to reduce the error:
[0123]
[0124] Here, α is the learning rate.
[0125] 4) Iterative optimization:
[0126] Continuously carry out error assessment and feedback, and through multiple iterative optimizations, the predictive ability of the model is continuously improved.
[0127] The present invention integrates multi-source meteorological data and machine learning algorithms to intelligently calibrate and optimize the boundary field input parameters in the climate model, and adjusts them in real time to adapt to different regions and meteorological conditions. Through this innovative method, the prediction accuracy of new energy wind and solar resources has been significantly improved, enabling power dispatch agencies to more accurately estimate the power generation capacity of wind and solar energy, thereby formulating more accurate and timely dispatch strategies.
[0128] This intelligent calibration mechanism solves the problem of wind and solar energy resource prediction errors caused by inaccurate boundary field settings in the current field. It can assist power dispatching agencies in reducing the negative impact of fluctuations in wind speed and solar radiation intensity on the power grid, and ensure the stable operation and power supply security of the power system. At the same time, the calibration system can also optimize the utilization of new energy resources, improve the economic benefits of wind farms and photovoltaic power stations, and reduce power generation losses caused by inaccurate predictions. In short, while improving the operating efficiency of the new energy system, the present invention effectively improves the dispatching efficiency and economic benefits of the power system, providing key technical support for the large-scale application of renewable energy.
[0129] Embodiment 2:
[0130] The present invention also proposes a climate prediction boundary field calibration system 200, such as Figure 3 As shown, including:
[0131] The data processing unit 201 is used to obtain climate data, pre-process the climate data, and generate a training data set;
[0132] A training unit 202, configured to train a prediction model based on the training data set;
[0133] The calibration unit 203 is used to calibrate the climate prediction boundary field of the prediction model according to the real-time climate condition data.
[0134] Among them, climate data includes: historical climate data, real-time observation data and climate prediction model output data;
[0135] The historical climate data include: wind speed, wind direction, solar radiation intensity, temperature and humidity data;
[0136] The real-time observation data includes: real-time wind speed, wind direction and solar radiation intensity data;
[0137] The output data of the weather prediction model include: output data of existing climate models or numerical weather prediction models.
[0138] Among them, preprocessing includes: denoising, normalization and missing value filling.
[0139] Wherein, based on the training data set, the prediction model is trained, including:
[0140] Extracting feature vectors from the training data set;
[0141] The preset model is used to learn the feature vector using a preset machine learning algorithm, and an optimization algorithm is used to adjust the parameters of the preset model to generate a prediction model.
[0142] Among them, the preset machine learning algorithms include:
[0143] Neural network algorithm, random forest algorithm, or support vector machine algorithm.
[0144] Among them, the optimization algorithm includes:
[0145] Gradient descent algorithm.
[0146] Among them, according to the real-time climate condition data, the climate prediction boundary field of the prediction model is calibrated, including:
[0147] Using real-time climate conditions as input features of a prediction model, inputting the input features into the prediction model, and predicting optimal boundary field configuration parameters;
[0148] The optimal boundary field configuration parameters are output as the calibration pattern.
[0149] The calibration unit 203 is further used for:
[0150] Calculate the error between the predicted value of the prediction model and the actual observed value;
[0151] Determine the source of the error, and based on the source of the error, adjust the parameters of the prediction model until the error requirements are met.
[0152] The application of the present invention can improve the operating efficiency of the new energy system, and at the same time effectively improve the dispatching efficiency and economic benefits of the power system.
[0153] Embodiment 3:
[0154] Based on the same inventive concept, the present invention also provides a computer device, which includes a processor and a memory, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, and is specifically suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding functions, so as to implement the steps of the method in the above embodiment.
[0155] Embodiment 4:
[0156] Based on the same inventive concept, the present invention also provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device for storing programs and data. It is understandable that the computer-readable storage medium here can include both a built-in storage medium in a computer device and an extended storage medium supported by the computer device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by a processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the steps of the method in the above embodiment.
[0157] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The schemes in the embodiments of the present invention may be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.
[0158] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0159] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0160] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.
[0161] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.
[0162] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.
Claims
1. A method for calibrating a climate prediction boundary field, characterized in that: include: Acquiring climate data, preprocessing the climate data, and generating a training data set; Based on the training data set, a prediction model is obtained by training; Calibrate the climate prediction boundary fields of the forecast model based on real-time climate condition data.
2. The calibration method according to claim 1, characterized in that: The climate data include: historical climate data, real-time observation data and climate prediction model output data; The historical climate data include: wind speed, wind direction, solar radiation intensity, temperature and humidity data; The real-time observation data includes: real-time wind speed, wind direction and solar radiation intensity data; The output data of the weather prediction model include: output data of existing climate models or numerical weather prediction models.
3. The calibration method according to claim 1, characterized in that: The preprocessing includes: denoising processing, normalization processing and missing value filling processing.
4. The calibration method according to claim 1, characterized in that: The step of training a prediction model based on the training data set includes: Extracting feature vectors from the training data set; The preset model is used to learn the feature vector using a preset machine learning algorithm, and an optimization algorithm is used to adjust the parameters of the preset model to generate a prediction model.
5. The calibration method according to claim 4, characterized in that: The preset machine learning algorithm includes: Neural network algorithm, random forest algorithm, or support vector machine algorithm.
6. The calibration method according to claim 4, characterized in that: The optimization algorithm comprises: Gradient descent algorithm.
7. The calibration method according to claim 1, characterized in that: The method of calibrating the climate prediction boundary field of the prediction model according to the real-time climate condition data includes: Using real-time climate conditions as input features of a prediction model, inputting the input features into the prediction model, and predicting optimal boundary field configuration parameters; The optimal boundary field configuration parameters are output as the calibration pattern.
8. The calibration method according to claim 1, characterized in that: The method further comprises: Calculate the error between the predicted value of the prediction model and the actual observed value; Determine the source of the error, and based on the source of the error, adjust the parameters of the prediction model until the error requirements are met.
9. A climate prediction boundary field calibration system, characterized in that: include: A data processing unit, used to obtain climate data, pre-process the climate data, and generate a training data set; A training unit, used for training a prediction model based on the training data set; The calibration unit is used to calibrate the climate prediction boundary field of the prediction model according to the real-time climate condition data.
10. The calibration system according to claim 9, characterized in that The climate data include: historical climate data, real-time observation data and climate prediction model output data; The historical climate data include: wind speed, wind direction, solar radiation intensity, temperature and humidity data; The real-time observation data includes: real-time wind speed, wind direction and solar radiation intensity data; The output data of the weather prediction model include: output data of existing climate models or numerical weather prediction models.
11. The calibration system according to claim 9, characterized in that The preprocessing includes: denoising processing, normalization processing and missing value filling processing.
12. The calibration system according to claim 9, characterized in that The step of training a prediction model based on the training data set includes: Extracting feature vectors from the training data set; The preset model is used to learn the feature vector using a preset machine learning algorithm, and an optimization algorithm is used to adjust the parameters of the preset model to generate a prediction model.
13. The calibration system according to claim 12, characterized in that The preset machine learning algorithm includes: Neural network algorithm, random forest algorithm, or support vector machine algorithm.
14. The calibration system according to claim 12, characterized in that: The optimization algorithm comprises: Gradient descent algorithm.
15. The calibration system according to claim 9, characterized in that The method of calibrating the climate prediction boundary field of the prediction model according to the real-time climate condition data includes: Using real-time climate conditions as input features of a prediction model, inputting the input features into the prediction model, and predicting optimal boundary field configuration parameters; The optimal boundary field configuration parameters are output as the calibration pattern.
16. The calibration system according to claim 9, characterized in that The calibration unit is also used for: Calculate the error between the predicted value of the prediction model and the actual observed value; Determine the source of the error, and based on the source of the error, adjust the parameters of the prediction model until the error requirements are met.
17. A computer device, characterized in that: include: one or more processors; a processor for executing one or more programs; When the one or more programs are executed by the one or more processors, the method according to any one of claims 1 to 7 is implemented.
18. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed, the method according to any one of claims 1 to 7 is implemented.