Construction Method and Device for Prediction Model of Influence of Wind Farm on Surface Temperature

Abstract

The present disclosure provides a method and device for constructing a prediction model of the influence of a wind farm on surface temperature, which relates to the technical field of data processing. The method includes: determining the spatial ranges of the wind farm area and the control area based on satellite remote sensing data, and obtaining the surface temperature data of the wind farm and the control surface temperature data by using the spatial ranges; simulating the meteorological variable data of the wind farm area by using a meteorological numerical model, and screening the operation period of the wind farm by using the wind speed variable in the meteorological variables; determining the difference between the surface temperature data of the wind farm and the control surface temperature data to obtain the influence data of the wind farm on the surface temperature; and constructing a prediction model of the influence of the wind farm on the surface temperature based on a machine learning algorithm with the meteorological variable data corresponding to the operation period of the wind farm as the input and the influence data as the output. The technical solution in the present disclosure can accurately predict the influence of the wind farm on the surface temperature through a machine learning algorithm, and improve the prediction accuracy and applicability.

Description

Background Art

[0002] Under the background of global climate change, as an important part of renewable energy, wind power generation has developed rapidly globally. The installed capacity of wind power generation continues to grow, especially in arid and semi-arid regions, where the development efforts are constantly increasing. However, the ecological environment in these regions is relatively fragile, and the operation of wind power generation facilities may have complex impacts on the local atmosphere and surface environment. Therefore, while promoting the development of the wind power industry, it is urgent to pay attention to its potential impact on the local climate to ensure the coordination and unity of energy development and ecological protection.

[0003] However, for the impact of wind power generation facilities on the local surface temperature, the existing technologies mainly rely on ground observations, remote sensing monitoring, or numerical simulation methods. Although ground observation data can provide high-precision information, due to the limited distribution of measurement points, it is difficult to comprehensively describe the temperature change characteristics of a large-scale area. Remote sensing monitoring methods can obtain large-scale spatial data, but are limited by cloud cover, sensor resolution, and imaging frequency, resulting in deficiencies in data continuity and accuracy. Numerical simulation methods calculate the impact of wind power generation facilities on the local climate through physical models. Although they can provide results with high spatio-temporal resolution, the calculation cost is high, and the simulation accuracy is greatly affected by the quality of input data, model parameter settings, and the ability to describe physical processes, leading to a high degree of uncertainty in the prediction results. Therefore, the existing technologies have insufficient accuracy and applicability in analyzing the impact of wind farms on surface temperature.

[0004] It should be noted that the information disclosed in the above background art section is only used to enhance the understanding of the background of the present disclosure, and thus may include information that does not constitute relevant technologies known to those of ordinary skill in the art. Summary of the Invention

[0005] The purpose of the embodiments of the present disclosure is to provide a method for constructing a prediction model of the impact of a wind farm on surface temperature, a prediction method of the impact of a wind farm on surface temperature, a device for constructing a prediction model of the impact of a wind farm on surface temperature, a prediction device of the impact of a wind farm on surface temperature, an electronic device, and a computer-readable storage medium, so as to accurately predict the impact of a wind farm on surface temperature through machine learning algorithms and improve the prediction accuracy and applicability.

[0006] Other features and advantages of the present disclosure will become apparent through the following detailed description, or will be partially learned through the practice of the present disclosure.

[0007] According to the first aspect of the embodiments of the present disclosure, a method for constructing a prediction model of the impact of a wind farm on surface temperature is provided. The method includes: determining the spatial ranges of the wind farm area and the control area based on satellite remote sensing data, and obtaining the surface temperature data of the wind farm and the control surface temperature data using the spatial ranges; simulating the meteorological variable data of the wind farm area using a meteorological numerical model, and screening the operation periods of the wind farm using the wind speed variable in the meteorological variables; within the operation periods of the wind farm, determining the difference between the surface temperature data of the wind farm and the control surface temperature data to obtain the impact data of the wind farm on surface temperature; and constructing a prediction model of the impact of the wind farm on surface temperature based on a machine learning algorithm with the meteorological variable data corresponding to the operation periods of the wind farm as the input and the impact data as the output.

[0008] In some exemplary embodiments of the present disclosure, based on the foregoing solution, the method for constructing a prediction model of the impact of a wind farm on surface temperature further includes: calculating a prediction residual based on the prediction result of the prediction model and the impact data; constructing a residual prediction model based on the prediction residual and the meteorological variable data; and post-processing the prediction result of the prediction model of the impact of the wind farm on surface temperature using the residual prediction model to obtain the final prediction result of the impact of the wind farm on surface temperature.

[0009] In some exemplary embodiments of the present disclosure, based on the foregoing solution, the determining the spatial ranges of the wind farm area and the control area based on satellite remote sensing data, and obtaining the surface temperature data of the wind farm and the control surface temperature data using the spatial ranges includes: determining the spatial distribution positions of the wind turbines in the wind farm area using satellite remote sensing data, and generating the spatial range of the wind farm area based on the spatial distribution of the wind turbines; generating an initial spatial range of the control area in the external space of the wind farm area, and obtaining the geographical feature data of the control area using satellite remote sensing data; screening the spatial range of the control area based on the geographical feature data to determine the spatial range of the control area with the same geographical background as the wind farm area; and respectively obtaining the surface temperature data of the wind farm and the control surface temperature data from the satellite remote sensing data based on the spatial ranges of the wind farm area and the control area.

[0010] In some exemplary embodiments of the present disclosure, based on the foregoing solution, the simulating the meteorological variable data of the wind farm area using a meteorological numerical model, and screening the operation periods of the wind farm using the wind speed variable in the meteorological variables includes: simulating and obtaining the hourly meteorological variable data of the wind farm area using a meteorological numerical model based on meteorological reanalysis data; and determining the periods when the wind speed meets the preset operating conditions of the wind turbines according to the wind speed variable in the meteorological variable data to screen out the operation periods of the wind farm.

[0011] In some exemplary embodiments of the present disclosure, based on the foregoing solution, during the operation period of the wind farm, determining the difference between the surface temperature data of the wind farm and the control surface temperature data to obtain the influence data of the wind farm on the surface temperature includes: determining the average surface temperature of the wind farm area and the control area at the same moment; calculating the difference between the average surface temperature of the wind farm area and the average surface temperature of the control area at each moment to obtain the influence data of the wind farm on the surface temperature.

[0012] In some exemplary embodiments of the present disclosure, based on the foregoing solution, using the meteorological variable data corresponding to the operation period of the wind farm as input and the influence data as output, constructing a prediction model of the influence of the wind farm on the surface temperature based on a machine learning algorithm includes: using a machine learning algorithm to construct multiple prediction models, using the meteorological variable data corresponding to the operation period of the wind farm as input and the influence data as output to train the multiple prediction models respectively; using validation set data to evaluate the prediction accuracy of the multiple prediction models, and selecting the machine learning model with the highest prediction accuracy as the prediction model of the influence of the wind farm on the surface temperature.

[0013] In some exemplary embodiments of the present disclosure, based on the foregoing solution, constructing a residual prediction model based on the prediction residual and the meteorological variable data includes: using the meteorological variable data corresponding to the operation period of the wind farm as input and the prediction residual as output, and constructing a residual prediction model using a linear regression calibration algorithm.

[0014] In some exemplary embodiments of the present disclosure, based on the foregoing solution, using the meteorological variable data corresponding to the operation period of the wind farm as input and the influence data as output, constructing a prediction model of the influence of the wind farm on the surface temperature based on a machine learning algorithm includes: constructing an input matrix based on the meteorological variable data corresponding to the operation period of the wind farm, and constructing a target vector based on the influence data of the wind farm on the surface temperature; using the bootstrap sampling method to extract a training sample set from the input matrix and the target vector, and constructing a random forest model based on the training sample set, which is composed of independently trained decision trees to form a decision tree ensemble:

[0015]

[0016] where represents the decision tree ensemble of the random forest model, represents the th decision tree;

[0017] Determining the optimal feature partitioning node based on information gain and recursively constructing a binary decision tree:

[0018]

[0019] Among them, represents the optimal partition of the th decision tree at feature , is the partitioning criterion, represents selecting the feature index that makes reach the maximum value, represents that the value range of the feature index is from 1 to n , n represents the total number of features, and the information gain calculation formula is: Among them, represents the data entropy, represents the conditional entropy after partitioning according to feature , and a decision tree structure is constructed based on the optimal partition;

[0020] Based on the trained random forest model, calculate the predicted value of the surface temperature of the wind farm, and perform weighted calculation using the prediction results of all decision trees:

[0021]

[0022] Among them, represents the final predicted value at time step , represents the th decision tree's prediction output for the input sample , and a prediction model for the impact of the wind farm on the surface temperature is obtained.

[0023] According to the second aspect of the embodiments of the present disclosure, a method for predicting the impact of a wind farm on the surface temperature is provided. The method includes: obtaining meteorological variable data corresponding to the operation period of the wind farm within the target wind farm area; inputting the meteorological variable data into the prediction model for the impact of the wind farm on the surface temperature to obtain the impact data of the wind farm on the surface temperature within the target wind farm area; wherein, the prediction model for the impact of the wind farm on the surface temperature is trained according to the construction method of the prediction model for the impact of the wind farm on the surface temperature as described in the above embodiments.

[0024] According to a third aspect of the embodiments of the present disclosure, there is provided an apparatus for constructing a prediction model of the impact of a wind farm on surface temperature. The apparatus includes: a temperature data acquisition module, configured to determine the spatial ranges of the wind farm area and the control area based on satellite remote sensing data, and acquire the surface temperature data of the wind farm and the control surface temperature data using the spatial ranges; a meteorological data acquisition module, configured to simulate the meteorological variable data of the wind farm area using a meteorological numerical model, and screen the operation periods of the wind farm using the wind speed variable in the meteorological variables; an impact data acquisition module, configured to determine the difference between the surface temperature data of the wind farm and the control surface temperature data during the operation period of the wind farm to obtain the impact data of the wind farm on surface temperature; and a prediction model construction module, configured to construct a prediction model of the impact of the wind farm on surface temperature based on a machine learning algorithm, with the meteorological variable data corresponding to the operation period of the wind farm as the input and the impact data as the output.

[0025] According to a fourth aspect of the embodiments of the present disclosure, there is provided a prediction apparatus for the impact of a wind farm on surface temperature. The apparatus includes: a target data acquisition module, configured to acquire the meteorological variable data corresponding to the operation period of a target wind farm area; and an impact prediction module, configured to input the meteorological variable data into a prediction model of the impact of the wind farm on surface temperature to obtain the impact data of the wind farm on surface temperature in the target wind farm area, where the prediction model of the impact of the wind farm on surface temperature is trained according to the construction method of the prediction model of the impact of the wind farm on surface temperature as described in the above embodiments.

[0026] According to a fifth aspect of the embodiments of the present disclosure, there is provided an electronic device, including: a processor; and a memory storing computer-readable instructions thereon, where the computer-readable instructions, when executed by the processor, implement the above construction method of the prediction model of the impact of the wind farm on surface temperature or the prediction method of the impact of the wind farm on surface temperature.

[0027] According to a sixth aspect of the embodiments of the present disclosure, there is provided a computer-readable storage medium storing a computer program thereon, where the computer program, when executed by a processor, implements the construction method of the prediction model of the impact of the wind farm on surface temperature or the prediction method of the impact of the wind farm on surface temperature as described above.

[0028] The technical solutions provided by the embodiments of the present disclosure may include the following beneficial effects:

[0029] The method for constructing a prediction model of the impact of a wind farm on surface temperature in the exemplary embodiments of the present disclosure, on the one hand, determines the spatial ranges of the wind farm area and the control area based on satellite remote sensing data, and uses these spatial ranges to obtain the surface temperature data of the wind farm and the control surface temperature data, improving the accuracy of the variation characteristics of the surface temperature of the wind farm. On the other hand, by using a meteorological numerical model to simulate the meteorological variable data of the wind farm area and combining the wind speed variable in the meteorological variables to screen the operation periods of the wind farm, the data input not only includes the comprehensive impact of the atmospheric environment during the operation of the wind farm, but also can eliminate the data in the non-operation state, improving the representativeness of the data samples, making the model training more targeted, and thus enhancing the prediction ability. On the further hand, by using the meteorological variable data corresponding to the operation periods of the wind farm as the input and the impact data of the wind farm on the surface temperature as the output, a prediction model of the impact of the wind farm on the surface temperature is constructed using a machine learning algorithm, enabling the modeling process to combine multi-source data and extract the complex non-linear relationships between the data based on the adaptive learning ability. Compared with the traditional physical model calculation method, it can effectively improve the prediction accuracy and enhance the adaptability of the model to the impact of the wind farm under different environmental conditions.

[0030] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The drawings herein are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure. Obviously, the drawings in the following description are only some embodiments of the present disclosure, and those of ordinary skill in the art can obtain other drawings without creative efforts based on these drawings.

[0032] Figure 1 Schematically shows a flowchart of a method for constructing a prediction model of the impact of a wind farm on surface temperature according to some embodiments of the present disclosure.

[0033] Figure 2 Schematically shows a comparison diagram of the impact of a wind farm on surface temperature predicted by a random forest model and the observed values according to some embodiments of the present disclosure.

[0034] Figure 3 Schematically shows a comparison diagram of the impact of a wind farm on surface temperature after post-processing and the observed values according to some embodiments of the present disclosure.

[0035] Figure 4 Schematically shows a diagram of the change with hours of the impact of a wind farm on surface temperature after post-processing and the observed values according to some embodiments of the present disclosure.

[0036] Figure 5 Schematically shows a comparison diagram of the hourly variation of the impact of a post - processed wind farm on the surface temperature and the observed values according to some embodiments of the present disclosure.

[0037] Figure 6 Schematically shows a flowchart of a method for predicting the impact of a wind farm on the surface temperature according to some embodiments of the present disclosure.

[0038] Figure 7 Schematically shows a block diagram of a device for constructing a prediction model of the impact of a wind farm on the surface temperature according to some embodiments of the present disclosure.

[0039] Figure 8 Schematically shows a block diagram of a prediction device for the impact of a wind farm on the surface temperature according to some embodiments of the present disclosure.

[0040] Figure 9 Schematically shows a structural diagram of a computer system of an electronic device according to some embodiments of the present disclosure.

[0041] Figure 10 Schematically shows a diagram of a computer - readable storage medium according to some embodiments of the present disclosure.

[0042] In the drawings, the same or corresponding reference numerals indicate the same or corresponding parts. Detailed Description of Specific Embodiments

[0043] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this specification. On the contrary, they are merely examples of devices and methods consistent with some aspects of this specification as detailed in the appended claims.

[0044] The terms used in this specification are for the purpose of describing specific embodiments only and are not intended to limit this specification. The singular forms “a”, “the” and “said” used in this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0045] It should be understood that although the terms first, second, third, etc. may be used in this specification to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other. For example, without departing from the scope of this specification, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "when" or "while" or "in response to determining".

[0046] Example embodiments will now be described more fully with reference to the accompanying drawings. However, the example embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the example embodiments to those skilled in the art.

[0047] In addition, the described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a thorough understanding of the embodiments of the present disclosure. However, those skilled in the art will realize that the technical solutions of the present disclosure may be practiced without one or more of the specific details, or may be implemented using other methods, components, devices, steps, etc. In other cases, well-known methods, devices, implementations, or operations are not shown or described in detail to avoid obscuring aspects of the present disclosure.

[0048] In addition, the drawings are only schematic illustrations and are not necessarily drawn to scale. The block diagrams shown in the drawings are only functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices.

[0049] The impact of a wind farm on the surface temperature mainly comes from the downwind turbulence caused by the rotation of the wind turbine blades in the atmospheric boundary layer. This turbulence mixes the surface layer and the upper air layer, resulting in the redistribution of elements such as heat. For example, due to the change in solar radiation, the temperature difference (temperature vertical lapse rate) between the upper and surface air layers changes drastically within a day. During the day, due to strong solar radiation, the temperature vertical lapse rate is relatively high. At this time, the wind farm will cause the cooling of the surface air. While at night, the temperature vertical lapse rate is relatively low, and the wind farm will cause the warming of the surface air. This process will also be conducted to the ground surface, resulting in the change of the ground surface temperature, which can thus be observed by satellite remote sensing means.

[0050] The related technologies mainly rely on ground observations, remote sensing monitoring, or numerical simulation methods to analyze the impact of wind power generation facilities on surface temperature. Although ground observations have high precision, their coverage is limited and it is difficult to reflect large-scale changes. Although remote sensing monitoring can obtain large-scale data, due to limitations such as clouds, resolution, and imaging frequency, the data continuity and accuracy are insufficient. Although numerical simulation methods can provide results with high spatio-temporal resolution, the calculation cost is high, and the prediction results are unstable due to the quality of input data, model parameters, and the description of physical processes, resulting in limited accuracy and adaptability.

[0051] To solve all or part of the technical problems in the above related technologies, in the exemplary embodiments of the present disclosure, a method for constructing a prediction model of the impact of a wind farm on surface temperature is first proposed. Figure 1 A schematic flow diagram of a method for constructing a prediction model of the impact of a wind farm on surface temperature according to some embodiments of the present disclosure is schematically shown. Refer to Figure 1 As shown, the method for constructing a prediction model of the impact of a wind farm on surface temperature may include the following steps:

[0052] Step S110, determining the spatial ranges of the wind farm area and the control area based on satellite remote sensing data, and obtaining the surface temperature data of the wind farm and the control surface temperature data using the spatial ranges.

[0053] Step S120, simulating the meteorological variable data of the wind farm area using a meteorological numerical model, and screening the operation periods of the wind farm using the wind speed variable in the meteorological variables.

[0054] Step S130, determining the difference between the surface temperature data of the wind farm and the control surface temperature data during the operation period of the wind farm to obtain the impact data of the wind farm on surface temperature.

[0055] Step S140, using the meteorological variable data corresponding to the operation period of the wind farm as input and the impact data as output, constructing a prediction model of the impact of the wind farm on surface temperature based on a machine learning algorithm.

[0056] This method first extracts the spatial extents of the wind farm area and the control area based on satellite remote sensing data, and determines the dataset that meets the research requirements through coordinate matching and spatial consistency screening. In the screened dataset, the surface temperature data of the wind farm and the control surface temperature data are extracted, and the data quality is controlled to eliminate abnormal data and ensure the integrity and reliability of the input data. After determining the spatial extent, the meteorological variable data of the wind farm area are further simulated using a meteorological numerical model. By inputting the initial meteorological field data and setting the simulation parameters, the meteorological variables of the wind farm area are calculated, the wind speed variable is extracted from them, and threshold conditions are set to screen the operating periods of the wind farm. By screening the periods that meet the operating requirements of the wind farm, the data input becomes more targeted, avoiding the influence of non-operating periods and improving the representativeness of the data.

[0057] During the operating period of the wind farm, the surface temperature data of the wind farm are matched with the control surface temperature data, and alignment calculations are performed in the time series. The impact data of the wind farm on the surface temperature are determined through difference operations. Outlier removal is performed on the calculated impact data to ensure the stability of the data. After completing the data calculations, a prediction model is constructed based on a machine learning algorithm using the meteorological variable data corresponding to the operating period of the wind farm as the input and the impact data of the wind farm on the surface temperature as the output. The parameter combination is optimized through model training, the model accuracy is evaluated using the validation set data, the prediction error is analyzed, and the model structure is optimized in combination with the error correction method to improve the accuracy and generalization ability of the prediction results, enabling it to be applicable to different wind farm areas and environmental conditions.

[0058] Next, the method for constructing the prediction model of the impact of the wind farm on the surface temperature in the above exemplary embodiment will be further described.

[0059] Step S110, determine the spatial extents of the wind farm area and the control area based on satellite remote sensing data, and obtain the surface temperature data of the wind farm and the control surface temperature data using the spatial extents.

[0060] Among them, the wind farm area can represent the geographical extent where the wind power generation facilities are located determined based on satellite remote sensing data, and this extent is used to extract the surface temperature data of the wind farm. The control area can represent the geographical extent used to compare with the wind farm area, which has similar geographical characteristics to the wind farm area but does not contain wind power generation facilities, and is used to extract the control surface temperature data. The surface temperature data of the wind farm can represent the surface temperature data extracted within the scope of the wind farm area, and this data is used to analyze the impact of the operation of the wind power generation facilities on the surface temperature. The control surface temperature data can represent the surface temperature data extracted within the scope of the control area, and this data is used to compare with the surface temperature data of the wind farm to determine the impact of the wind farm operation on the surface temperature.

[0061] In some embodiments, the spatial ranges of the wind farm area and the control area are determined based on satellite remote sensing data, and the surface temperature data of the wind farm and the control surface temperature data are obtained using the spatial ranges. The specific technical steps are as follows:

[0062] First, the spatial distribution positions of the wind turbines in the wind farm area are determined using satellite remote sensing data, and the spatial range of the wind farm area is generated based on the spatial distribution of the wind turbines. Exemplarily, satellite remote sensing data can be used to extract the spatial distribution position information of the wind turbines in a certain plateau area through map tools such as Google Earth. On this basis, the wind farms with the most concentrated distribution are selected, with a total of 1721 wind turbines. The wind turbines are spatially marked on the map tool, and finally a.shp file is exported for spatial analysis. Based on the spatial distribution of the wind turbines, using geographic information system tools, a buffer zone with a fixed radius is constructed centered on each wind turbine, and all buffer zone areas are merged to generate the spatial range of the wind farm area, so as to ensure that the wind farm range covers all wind turbines and accurately represents the overall influence area of the wind farm.

[0063] Then, an initial spatial range of the control area is generated in the external space of the wind farm area, and the geographic feature data of the control area are obtained using satellite remote sensing data. Exemplarily, based on the principle of spatial consistency, a strip buffer zone is set outside the wind farm area to ensure that the control area is outside the influence range of the wind farm while maintaining sufficient spatial distribution uniformity. The initial spatial range of the control area is obtained by establishing a strip buffer zone within a range of 3 - 8 km outside the wind farm area, and the geographic feature data of the control area, including terrain, altitude, land cover type, etc., are extracted in combination with satellite remote sensing data to ensure the reasonable spatial distribution of the control area and provide basic data support for subsequent screening.

[0064] Next, based on the geographic feature data, the spatial range of the control area is screened to determine the spatial range of the control area that is geographically consistent with the wind farm area. Specifically, satellite remote sensing data is used to analyze the geographic features of the wind farm area and the control area, and the areas with large differences in topography and landforms in the control area are excluded, including the parts with large altitude changes and significantly different surface cover types, to ensure a high degree of consistency in the geographic background between the control area and the wind farm area. The spatial range of the control area after screening is optimized through spatial analysis methods to enable it to accurately reflect the surface temperature characteristics under the condition of no wind turbine operation, thereby improving the comparability of subsequent analyses.

[0065] Finally, based on the spatial extents of the wind farm area and the control area, surface temperature data of the wind farm and control surface temperature data are respectively obtained from satellite remote sensing data. Exemplarily, the surface temperature retrieval data of the Fengyun-4 satellite can be selected, and hourly data with a spatial resolution of 4 km and a temporal resolution of 15 - 60 minutes are obtained based on the Fengyun satellite remote sensing data, and the data time range is set to ensure that the data covers the operation impact period of the wind farm. The downloaded data format is in the.nc format. By programming to look up the table and match the actual longitude and latitude coordinates, it is converted into the geotiff format containing spatial coordinate information to ensure the accuracy of data geographical matching. During the data processing, screening is carried out according to the data quality flag, and data with large cloud cover influence is excluded to ensure that the obtained surface temperature data of the wind farm and the control surface temperature data are both high-quality data and can accurately reflect the temperature change characteristics of the study area.

[0066] Step S120: Use a meteorological numerical model to simulate the meteorological variable data of the wind farm area, and use the wind speed variable in the meteorological variables to screen the operation period of the wind farm.

[0067] Among them, the meteorological numerical model can represent a mathematical model for calculating atmospheric physical and dynamic processes, which, based on the input initial and boundary conditions, simulates the temporal evolution of the atmospheric state through numerical calculation methods. The meteorological variable data can represent the atmospheric state parameters calculated by the meteorological numerical model, including but not limited to wind speed, temperature, humidity, air pressure, etc., and are used to describe the meteorological characteristics within the wind farm area. The wind speed variable can represent the wind speed parameter in the meteorological variable data, which is used to characterize the rate of air flow within the wind farm area and can be used to judge the operation state of the wind farm. The operation period of the wind farm can represent the time interval when the wind speed variable meets the normal operation conditions of the wind power generation facilities, and the data during this period is used for subsequent analysis of the impact of wind farm operation on the surface temperature.

[0068] In some embodiments, using a meteorological numerical model to simulate the meteorological variable data of the wind farm area and using the wind speed variable in the meteorological variables to screen the operation period of the wind farm specifically includes the following technical steps: Based on the meteorological reanalysis data, use the meteorological numerical model to simulate and obtain the hourly meteorological variable data of the wind farm area; according to the wind speed variable in the meteorological variable data, determine the period when the wind speed meets the preset operation conditions of the wind turbine to screen out the operation period of the wind farm.

[0069] Specifically, based on meteorological reanalysis data, meteorological numerical models are used to simulate and obtain hourly meteorological variable data in the wind farm area. First, meteorological reanalysis data is obtained from the target website. Exemplarily, the meteorological reanalysis data can be ERA5 (Fifth Generation ECMWF Atmospheric Reanalysis) data, which is stored in GRIB format and needs to be preprocessed before being used for numerical model calculations. The ERA5 data is converted through the Weather Research and Forecasting Preprocessing System (WPS) module of the WRF model, which specifically includes three steps. First, run geogrid.exe, set the operating area and nested area of the model, and load static geographical data to ensure that the numerical simulation range meets the research requirements. Then, run ungrib.exe to extract meteorological variables from the ERA5 data to ensure that the required atmospheric elements are available for subsequent simulations. Finally, run metgrid.exe to grid the data processed in the previous two steps to generate a standardized data set that can be read by the WRF model. After completing the WPS preprocessing, link the result file generated by metgrid.exe to the WRF directory and further configure the physical schemes adopted by the WRF model, including radiation schemes, boundary layer schemes, cumulus schemes, etc., to ensure that the model can accurately simulate the meteorological characteristics of the wind farm area. Subsequently, run real.exe to initialize the data into a format recognizable by WRF and run wrf.exe for the final numerical simulation to obtain hourly meteorological variable data in the wind farm area. Since the WRF model usually has a higher calculation accuracy for meteorological variables than its driving data, the simulation results can provide more refined meteorological variable information and high-quality data for subsequent analysis.

[0070] According to the wind speed variable in the meteorological variable data, determine the time period when the wind speed meets the preset operating conditions of the wind turbine to screen out the operating time period of the wind farm. The operating status of a wind farm can usually be judged by its real-time capacity factor, but this data is usually internal data of the operator and is difficult to obtain. Therefore, in this embodiment, the wind speed variable is used as the discrimination criterion for the operation of the wind farm. Generally speaking, the operation of a wind turbine depends on whether the wind speed is within its designed operating range, that is, when the wind speed exceeds the cut-in wind speed but does not reach the cut-out wind speed, the wind turbine is in a normal operating state. According to the general design parameters of existing wind turbines, a wind speed range of 3 m / s to 20 m / s is selected as the operation discrimination condition for the wind farm. Extract the wind speed variable from the hourly meteorological variable data and screen out the time periods that meet this wind speed condition to obtain the wind farm operation time period data set. This method uses the widely available wind speed data to replace the traditional capacity factor data, making the screening of the wind farm operation status more feasible and applicable.

[0071] Step S130, within the wind farm operation time period, determine the difference between the surface temperature data of the wind farm and the reference surface temperature data to obtain the influence data of the wind farm on the surface temperature.

[0072] Among them, the influence data of the wind farm on the surface temperature can represent the result calculated based on the difference between the surface temperature data of the wind farm and the reference surface temperature data within the wind farm operation time period. This data is used to describe the influence degree of the wind farm operation on the surface temperature and its spatio-temporal variation characteristics.

[0073] In some embodiments, within the wind farm operation time period, determining the difference between the surface temperature data of the wind farm and the reference surface temperature data to obtain the influence data of the wind farm on the surface temperature specifically includes the following technical steps: determining the average surface temperature of the wind farm area and the reference area at the same moment; calculating the difference between the average surface temperature of the wind farm area and the average surface temperature of the reference area at each moment to obtain the influence data of the wind farm on the surface temperature.

[0074] Exemplarily, when determining the average surface temperature of the wind farm area and the control area at the same moment, the average surface temperature of the wind farm area and the control area at the same moment can be determined based on the surface temperature data of the Fengyun-4 satellite to construct a temperature comparison data set. To ensure the accuracy of the calculation, it is necessary to extract the temperature at the pixel level for the wind farm area and the control area respectively. Based on the spatial range of the wind farm area, the surface temperature pixels in the Fengyun-4 satellite remote sensing data are screened, and the average surface temperature of the wind farm area is calculated at the same time step. The same method is used for the calculation process of the surface temperature of the control area. Based on the selected range of the control area, the remote sensing temperature data at the corresponding moment is extracted, and the average value of the pixels is calculated. Since the surface temperatures of the wind farm and the control area are jointly affected by the regional climate background, selecting a control area with a consistent geographical background can effectively reduce the interference of climate element changes on temperature calculation and ensure that the calculation results can accurately reflect the independent impact of wind farm operation on surface temperature.

[0075] After calculating the surface temperatures of the wind farm area and the control area, calculate the difference between the average surface temperature of the wind farm area and the average surface temperature of the control area at each moment to obtain the data on the impact of the wind farm on surface temperature. Using the time alignment method, ensure that the calculation at each moment is based on the matching data of the wind farm area and the control area to reduce the error caused by time differences. The specific calculation method is as follows:

[0076]

[0077] where represents the impact of the wind farm on the surface temperature at a certain moment, represents the average value of all surface temperature pixels in the wind farm area at a certain moment, represents the average value of all surface temperature pixels in the control area at a certain moment.

[0078] Step S140: Using the meteorological variable data corresponding to the operation period of the wind farm as the input and the impact data as the output, a prediction model of the impact of the wind farm on the surface temperature is constructed based on a machine learning algorithm. In the embodiments of the present disclosure, when constructing the prediction model of the impact of the wind farm on the surface temperature, first, variables that meet the actual engineering requirements are screened from the meteorological variable data output by the WRF model to construct an efficient and robust data input set. In the data input preparation stage, based on the WRF model output data during the operation period of the wind farm, wind speed, temperature, humidity, air pressure, and their derived variables are extracted to ensure that the input data can reflect the main meteorological characteristics of the wind farm area. Although the WRF model can provide rich meteorological variables, to ensure the engineering applicability of the model, only variables that can be actually measured during the wind farm planning and design stage are selected, and the characteristic contribution degrees of each variable are evaluated based on data analysis methods to ensure that the selected variables have a high explanatory power for the impact of the wind farm on the surface temperature. For time series data, the sliding window method is used to reconstruct the meteorological variables to capture short-term change trends, and at the same time, the data is normalized to ensure that the numerical ranges of different variables are consistent, so as to improve the convergence speed and stability of the model. In the model construction stage, the impact data of the wind farm on the surface temperature is used as the output, and machine learning algorithms with different principles are selected for modeling to ensure the adaptability of the model to data characteristics.

[0079] Exemplarily, the machine learning algorithm may include, but is not limited to, a support vector machine model, a K-nearest neighbor model, a random forest model, and a gradient boosting model. Among them, the support vector machine model is a binary classification model, and its basic idea is to find an optimal decision boundary (also called a hyperplane) that can maximize the separation of two categories and minimize the classification error of unknown data. The K-nearest neighbor model is an instance-based learning method. For a new instance with an unknown class, the K nearest known instances most similar to this instance are found in the training set, and the class or value of the new instance is predicted based on the classes or values of these nearest neighbors. The random forest model is a tree-based ensemble learning algorithm that improves accuracy and stability by constructing multiple decision trees and voting or averaging, and introduces randomness to reduce the risk of overfitting. The gradient boosting model is an optimization algorithm based on gradient boosting decision trees, which forms an ensemble model by iteratively constructing and combining multiple weak learners (usually decision trees). In this embodiment, machine learning modeling is implemented by calling the Scikit-learn function library through Python programming.

[0080] In some embodiments, a prediction model for the impact of a wind farm on surface temperature is constructed based on a machine learning algorithm, with the meteorological variable data corresponding to the operating period of the wind farm as the input and the impact data as the output. The specific technical steps are as follows: Construct multiple prediction models using a machine learning algorithm, and use the meteorological variable data corresponding to the operating period of the wind farm as the input and the impact data as the output to train the multiple prediction models separately; Use the validation set data to evaluate the prediction accuracy of the multiple prediction models, and select the machine learning model with the highest prediction accuracy as the prediction model for the impact of the wind farm on surface temperature.

[0081] Specifically, first, construct multiple prediction models using a machine learning algorithm. Use the meteorological variable data corresponding to the operating period of the wind farm as the input and the impact data of the wind farm on surface temperature as the output to train the multiple prediction models separately. In the data input stage, screen the wind speed, air temperature, air pressure, humidity, and their derived variables from the meteorological variable data output by the WRF model to ensure that the model input features are closely related to the meteorological characteristics of the wind farm area. Normalize the time series data to eliminate the scale differences between different variables and improve the stability of the training process. Subsequently, based on machine learning algorithms with different principles, including support vector machine models, K-nearest neighbor models, random forest models, and gradient boosting models, etc., construct prediction models respectively, and use the training data set to optimize the parameters and fit each model to obtain preliminary prediction results.

[0082] Next, use the validation set data to evaluate the prediction accuracy of the multiple prediction models, and select the machine learning model with the highest prediction accuracy as the final prediction model for the impact of the wind farm on surface temperature. To quantify the model prediction accuracy, compare the impact data of the wind farm on surface temperature observed by the Fengyun-4 satellite with the prediction results of each model, and use the coefficient of determination (R 2 ), root mean square error (RMSE), and mean square error (MSE) for performance evaluation. Among them, the coefficient of determination can measure the fitting degree between the predicted data and the observed data, indicating the explanatory ability of the model for the impact of the wind farm on surface temperature. The value range is 0-1, and the closer the value is to 1, the stronger the explanatory ability of the model. The mean square error can measure the error between the predicted data and the observed data, indicating the accuracy of the model. The lower the value, the smaller the prediction error of the model. The calculation of these evaluation indicators is implemented by calling the statsmodels function library through Python programming and is compared among multiple models. After the model evaluation is completed, select the model with the highest prediction accuracy based on the evaluation results as the prediction model for the impact of the wind farm on surface temperature.

[0083] In some embodiments, a random forest algorithm can also be used to construct a prediction model for the impact of a wind farm on surface temperature. The specific construction process can include the following technical steps:

[0084] First, construct an input matrix based on the meteorological variable data corresponding to the operating period of the wind farm, and construct a target vector based on the data of the impact of the wind farm on the surface temperature. Among them, the input matrix can be expressed as:

[0085]

[0086] represents the meteorological variable data matrix of the operating period of the wind farm, represents the total number of time steps, represents the number of meteorological variables, represents the time step at the -th value of the meteorological variable. The target vector can be expressed as:

[0087]

[0088] represents the impact vector of the wind farm on the surface temperature, represents the impact data of the wind farm on the surface temperature at time step .

[0089] Second, use the bootstrap sampling method to extract a training sample set from the input matrix and the target vector, and construct a random forest model based on the training sample set, which consists of independently trained decision trees to form a decision tree ensemble:

[0090]

[0091] Among them, represents the decision tree ensemble of the random forest model, represents the -th decision tree.

[0092] Third, determine the optimal feature division node based on the information gain, and recursively construct a binary decision tree:

[0093]

[0094] Among them, represents the optimal division of the -th decision tree at feature , is the division criterion, represents selecting the feature index that makes reach the maximum value, represents the range of values of the feature index from 1 to n , n represents the total number of features. The information gain calculation formula is: , where, represents the data entropy, represents the conditional entropy after partitioning by feature and constructs a decision tree structure based on the optimal partitioning.

[0095] Fourthly, calculate the predicted value of the impact of the wind farm on the surface temperature based on the trained random forest model, and perform weighted calculation using the prediction results of all decision trees:

[0096]

[0097] wherein, represents the time step of the final predicted value, represents the th decision tree's prediction output for the input sample to obtain a prediction model for the impact of the wind farm on the surface temperature.

[0098] In some embodiments, the method for constructing a prediction model for the impact of a wind farm on the surface temperature includes the following steps: calculating a prediction residual based on the prediction result of the prediction model and the impact data; constructing a residual prediction model based on the prediction residual and meteorological variable data; using the residual prediction model to post-process the prediction result of the prediction model for the impact of the wind farm on the surface temperature to obtain the final prediction result of the impact of the wind farm on the surface temperature.

[0099] Specifically, during the construction of the prediction model for the impact of a wind farm on the surface temperature, if the accuracy of the initial prediction result is not ideal, the prediction accuracy can be improved through a post-processing method. In this embodiment, a residual modeling method is used for post-processing to correct the systematic error of the original prediction result, thereby obtaining a more accurate prediction result for the impact of the wind farm on the surface temperature. The residual modeling process includes calculating the prediction residual, constructing a residual prediction model, and using the residual prediction model to post-process the original prediction result to optimize the final prediction result.

[0100] First, calculate the prediction residual based on the prediction result of the prediction model and the impact data to quantify the prediction error. For each wind farm operation period, calculate the difference between the predicted value and the data on the impact of the wind farm on the surface temperature observed by the Fengyun-4 satellite, i.e.:

[0101]

[0102] wherein, represents the prediction residual, represents the predicted value of the prediction model for the impact of the wind farm on the surface temperature, It represents the data on the impact of the observed wind farm on the surface temperature. Next, a residual prediction model is constructed based on the prediction residuals and meteorological variable data to identify the relationship between the errors and the input features. The same data processing method as the initial modeling is adopted, using the meteorological variable data as the input and the prediction residuals as the output to train the residual prediction model. This process is consistent with the original machine learning modeling steps, except that the target variable of the model is replaced by the prediction residual value from the surface temperature impact value. Multiple machine learning methods, including random forest, support vector machine, and neural network, are used to train the model, and the fitting capabilities of different models are evaluated through the cross-validation method to select the optimal residual prediction model.

[0103] Finally, the residual prediction model is used to post-process the prediction results of the prediction model for the impact of the wind farm on the surface temperature to obtain the final prediction results. The predicted residual value at each moment is obtained through the residual prediction model, and this value is used to correct the initial prediction results. The predicted value of the impact of the wind farm on the surface temperature after post-processing is calculated. The specific calculation formula is as follows:

[0104]

[0105] Where, represents the final prediction result after post-processing, represents the correction value of the residual prediction model for the prediction residuals. This correction process effectively reduces the systematic error, improves the prediction accuracy of the model, and enhances its adaptability to the impact of the wind farm under different environmental conditions.

[0106] In some embodiments, a residual prediction model is constructed based on the prediction residuals and meteorological variable data, specifically including the following steps: Using the meteorological variable data corresponding to the wind farm operation period as the input and the prediction residuals as the output, a residual prediction model is constructed using the linear regression calibration algorithm.

[0107] Specifically, first obtain the input data and output data for residual modeling to ensure that the data format is consistent with the model training requirements. Let the meteorological variable data corresponding to the wind farm operation period be the input matrix:

[0108]

[0109] Where, represents the meteorological variable data matrix corresponding to the wind farm operation period, represents the total number of the wind farm operation period, represents the number of meteorological variables, represents the time step at the value of the

[0110] th meteorological variable. Correspondingly, the prediction residuals are used as the output data to construct the target vector:

[0111]

[0112] Among them, represents the predicted residual vector of the impact of the wind farm on the surface temperature, represents the time step the predicted residual of the impact of the wind farm on the surface temperature at the downwind.

[0113] In the model construction stage, a linear regression calibration algorithm is adopted to establish a residual prediction model to optimally fit the relationship between meteorological variable data and predicted residuals. The expression of the linear regression model is:

[0114]

[0115] Among them, represents the predicted output of the residual prediction model, represents the regression coefficient vector to be optimized, represents the intercept term. The regression coefficient and the intercept can be obtained by minimizing the mean square error, minimizing the squared error of the predicted residuals:

[0116]

[0117] Among them, represents the predicted value of the residual prediction model at the time step , represents the actually calculated predicted residual at the time step . By using the gradient descent optimization algorithm or the method of analytical solution to solve the optimal and , finally, a residual prediction model is obtained to correct the initial prediction result.

[0118] In some embodiments, the model construction method for predicting the impact of a wind farm on surface temperature in the above embodiments is used for model construction. First, the average values of meteorological elements such as wind speed, temperature, humidity, and air pressure, which are actually measured in engineering, between the FY-4 surface temperature product and the WRF output are calculated using the generated.shp files of a certain plateau wind farm area and the control area. Align the calculation results of the two in time and uniformly convert them to Beijing time. Perform quality control on the FY-4 surface temperature product. After removing the data of poor-quality time steps, 5076 hours of data remain out of the 8760 hours of data throughout the year. Then, use the WRF wind speed data to screen the operating time steps of the wind farm, with the screening range being 3 - 20 m / s. After screening, 3462 hours of data remain. Use these 3462 sets of data for modeling, and divide these data into a training set and a validation set according to the ratio of 80% and 20% using the train_test_split function library in Python.

[0119] In the training set, use various machine learning models and parameter combinations, flexibly select and combine the meteorological elements output by the WRF model as the model input data, and use the impact of the wind farm on the surface temperature observed by the FY-4 satellite as the model output data. Select the optimal settings based on the prediction accuracy of the validation set. The optimal model is the random forest model, with its parameter settings being the number of decision trees n_estimators = 50, the maximum depth of the decision tree max_depth = 10, the minimum number of samples required for node splitting min_samples_split = 10, the minimum number of samples required for leaf nodes min_samples_leaf = 15. The selected WRF meteorological elements include the vertical lapse rate of air temperature, wind speed, air pressure, etc. Based on the validation set prediction of this model, the R 2 is 0.49 and the MSE is 0.12 °C. Figure 2 is the comparison schematic diagram of the impact of the wind farm on the surface temperature predicted by the random forest model ( ), and the observed value ( ). Each black scatter point represents a set of observed values and their corresponding predicted values. The diagonal dashed line represents the ideal prediction result, that is, the case where the model predicted value is exactly the same as the observed value. The black solid line represents the best linear fitting relationship between the model predicted value and the observed value, which is used to evaluate the trend error of the model.

[0120] Figure 3 is the comparison schematic diagram of the impact of the post-processed wind farm on the surface temperature and the observed value. By comparing the adjusted predicted value ( ) and the observed value ( ), the effect of post - processing is evaluated. Compared with the previous prediction model, this method uses a gradient - boosting model for residual modeling and corrects the initial prediction result using the prediction error to improve the overall prediction accuracy. The MSE is reduced to 0.08 °C, and the error is reduced compared with 0.12 °C without residual modeling, indicating that the post - processing effectively improves the prediction accuracy. It is increased to 0.67, compared with 0.49 before residual modeling, indicating that the interpretability of the model is enhanced and it can more accurately describe the impact of the wind farm on the surface temperature.

[0121] Figure 4 It is a schematic diagram showing the variation of the impact of the post - processed wind farm on the surface temperature and the observed value with time. This figure shows the impact of the wind farm on the surface temperature observed by the Fengyun - 4 satellite ( ) and the predicted value adjusted after residual modeling ( ) within 24 hours of a day. The impact of the wind farm has obvious diurnal variation characteristics, that is, it cools down during the day and warms up at night. The two curves in the figure represent the observed value and the predicted value respectively, and the overall variation trends of the two curves are relatively consistent, indicating that the model can effectively capture the intraday variation pattern of the impact of the wind farm on the surface temperature.

[0122] Figure 5 It is a schematic diagram comparing the hourly variation of the impact of the post - processed wind farm on the surface temperature with the observed value. Among them, each black dot represents the predicted value of the impact of the wind farm on the surface temperature within one hour ( ) and the corresponding observed value ( ). In this figure, MSE = 0.01 °C, , indicating that after the model is optimized by post - processing, the prediction accuracy is greatly improved.

[0123] Furthermore, in the embodiments of the present disclosure, a prediction method for the impact of a wind farm on the surface temperature is also provided. As shown in Figure 6 , this method includes the following steps:

[0124] Step S610, obtain the meteorological variable data corresponding to the operation period of the wind farm within the target wind farm area.

[0125] Step S620, input the meteorological variable data into the prediction model for the impact of the wind farm on the surface temperature to obtain the impact data of the wind farm on the surface temperature within the target wind farm area; among them, the prediction model for the impact of the wind farm on the surface temperature is trained according to the construction method of the prediction model for the impact of the wind farm on the surface temperature in the above - mentioned embodiments.

[0126] It should be noted that although the steps of the method in the present disclosure are described in a specific order in the drawings, this does not require or imply that these steps must be performed in that specific order, or that all the steps shown must be performed to achieve the desired result. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step for execution, and / or one step may be decomposed into multiple steps for execution, etc.

[0127] Next, in the embodiments of the present disclosure, there is also provided a device for constructing a prediction model of the impact of a wind farm on surface temperature. Referring to Figure 7 as shown in, the device 700 for constructing a prediction model of the impact of a wind farm on surface temperature may be composed of a temperature data acquisition module 701, a meteorological data acquisition module 702, an impact data acquisition module 703, and a prediction model construction module 704. Among them: the temperature data acquisition module may be used to determine the spatial ranges of the wind farm area and the control area based on satellite remote sensing data, and obtain the surface temperature data of the wind farm and the control surface temperature data using the spatial ranges; the meteorological data acquisition module may be used to simulate the meteorological variable data of the wind farm area using a meteorological numerical model, and screen the operation periods of the wind farm using the wind speed variable in the meteorological variables; the impact data acquisition module may be used to determine the difference between the surface temperature data of the wind farm and the control surface temperature data during the operation period of the wind farm to obtain the impact data of the wind farm on the surface temperature; the prediction model construction module may be used to construct a prediction model of the impact of the wind farm on the surface temperature based on a machine learning algorithm, with the meteorological variable data corresponding to the operation period of the wind farm as the input and the impact data as the output.

[0128] In the embodiments of the present disclosure, there is also provided a prediction device for the impact of a wind farm on surface temperature. Referring to Figure 8 as shown in, the prediction device 800 for the impact of a wind farm on surface temperature may be composed of a target data acquisition module 801 and an impact prediction module 802. Among them, the data acquisition module 801 may be used to acquire the meteorological variable data corresponding to the operation period of the wind farm in the target wind farm area; the impact prediction module 802 may be used to input the meteorological variable data into the prediction model of the impact of the wind farm on the surface temperature to obtain the impact data of the wind farm on the surface temperature in the target wind farm area; among them, the prediction model of the impact of the wind farm on the surface temperature is trained according to the method for constructing a prediction model of the impact of a wind farm on surface temperature in the above embodiments.

[0129] It should be noted that the specific details of each part in the above device for constructing a prediction model of the impact of a wind farm on surface temperature have been described in detail in the implementation manners of the method for constructing a prediction model of the impact of a wind farm on surface temperature. The undisclosed detailed content can be referred to the implementation manners of the method part, and thus will not be elaborated here.

[0130] In addition, in an exemplary embodiment of the present disclosure, an electronic device capable of implementing the above method for constructing a prediction model of the impact of a wind farm on surface temperature is also provided.

[0131] Those skilled in the art of the present disclosure can understand that various aspects of the present disclosure can be implemented as a system, a method, or a program product. Therefore, various aspects of the present disclosure can be specifically implemented in the following forms, namely: a complete hardware embodiment, a complete software embodiment (including firmware, microcode, etc.), or an embodiment combining hardware and software aspects, which can be collectively referred to as "circuitry", "module", or "system" here.

[0132] The following refers to Figure 7 to describe the electronic device 900 according to this embodiment of the present disclosure. Figure 9 The illustrated electronic device 900 is merely an example and should not impose any limitation on the functions and scope of use of the embodiments of the present disclosure.

[0133] As Figure 9 shown, the electronic device 900 is presented in the form of a general-purpose computing device. The components of the electronic device 900 may include, but are not limited to: the at least one processing unit 910 described above, the at least one storage unit 920 described above, a bus 930 connecting different system components (including the storage unit 920 and the processing unit 910), and a display unit 940.

[0134] Among them, the storage unit stores program code, and the program code can be executed by the processing unit 910, so that the processing unit 910 executes the steps according to various exemplary embodiments of the present disclosure described in the above "exemplary method" section of this specification.

[0135] The storage unit 920 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 921 and / or a cache storage unit 922, and may further include a read-only storage unit (ROM) 923.

[0136] The storage unit 920 may further include a program / utility 924 having a set (at least one) of program modules 925. Such program modules 925 include, but are not limited to: an operating system, one or more application programs, other program modules, and program data. The implementation of a network environment may be included in each or some combination of these examples.

[0137] The bus 930 may be a storage unit bus, a peripheral bus, or a local bus of any bus structure.

[0138] The electronic device 900 may also communicate with one or more external devices 970 (such as a keyboard, a pointing device, a Bluetooth device, etc.), and may also communicate with one or more devices that enable a user to interact with the electronic device 900, and / or communicate with any device that enables the electronic device 900 to communicate with one or more other computing devices (such as a router, a modem, etc.). Such communication may be carried out through the input / output (I / O) interface 950. Moreover, the electronic device 900 may also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through the network adapter 960. As shown in the figure, the network adapter 960 communicates with other modules of the electronic device 900 through the bus 930. It should be understood that although not shown in the figure, other hardware and / or software modules may be used in combination with the electronic device 900, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0139] Through the description of the above embodiments, those skilled in the art can easily understand that the exemplary embodiments described herein can be implemented by software or by a combination of software and necessary hardware. Therefore, the technical solutions according to the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the embodiments of the present disclosure.

[0140] In an exemplary embodiment of the present disclosure, there is also provided a computer-readable storage medium having stored thereon a program product capable of implementing the above method of the present specification. In some possible embodiments, various aspects of the present disclosure may also be implemented in the form of a program product, which includes program code. When the program product runs on a terminal device, the program code is used to cause the terminal device to execute the steps according to various exemplary embodiments of the present disclosure described in the above "Exemplary Method" section of the present specification.

[0141] Reference Figure 10 As shown, a program product 1000 for implementing the above method for constructing a prediction model of the impact of a wind farm on surface temperature according to an embodiment of the present disclosure is described. It may be a portable compact disc read-only memory (CD-ROM) and includes program code, and may run on a terminal device, such as a personal computer. However, the program product of the present disclosure is not limited thereto. In this document, the readable storage medium may be any tangible medium that contains or stores a program, and the program may be used by or in combination with an instruction execution system, apparatus, or device.

[0142] The program product may adopt any combination of one or more readable media. The readable media may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory, an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0143] The computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which the readable program code is carried. Such a propagated data signal may take various forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The readable signal medium may also be any readable medium other than the readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0144] The program code contained on the readable medium may be transmitted by any appropriate medium, including but not limited to wireless, wired, optical fiber cable, electromagnetic wave, etc., or any suitable combination of the above.

[0145] The program code for performing the operations of the present disclosure may be written in any combination of one or more programming languages. The programming languages include object-oriented programming languages such as Java, C++, etc., and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on the user's computing device, partially on the user's device, executed as a stand-alone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (for example, by using an Internet service provider to connect through the Internet).

[0146] It should be understood that the present disclosure is not limited to the exact structures described above and shown in the drawings, and various modifications and changes may be made without departing from its scope. The scope of the present disclosure is only limited by the appended claims.

Claims

1. A method for constructing a prediction model of the influence of a wind farm on the surface temperature, characterized in that, Including: Determine the spatial ranges of the wind farm area and the control area based on satellite remote sensing data, and obtain the surface temperature data of the wind farm and the control surface temperature data by using the spatial ranges; Simulate the meteorological variable data of the wind farm area by using a meteorological numerical model, and screen the operation periods of the wind farm by using the wind speed variable in the meteorological variables; During the operation period of the wind farm, determine the difference between the surface temperature data of the wind farm and the control surface temperature data to obtain the influence data of the wind farm on the surface temperature; Using the meteorological variable data corresponding to the operation period of the wind farm as input and the influence data as output, construct a prediction model for the influence of the wind farm on the surface temperature based on a machine learning algorithm; Calculate the prediction residual based on the prediction result of the prediction model and the influence data; Construct a residual prediction model based on the prediction residual and the meteorological variable data; Use the residual prediction model to post-process the prediction result of the prediction model for the influence of the wind farm on the surface temperature to obtain the final prediction result of the influence of the wind farm on the surface temperature.

2. The method for constructing a prediction model of the influence of a wind farm on the surface temperature according to claim 1, characterized in that The determining the spatial ranges of the wind farm area and the control area based on satellite remote sensing data, and obtaining the surface temperature data of the wind farm and the control surface temperature data by using the spatial ranges includes: Determine the spatial distribution positions of the wind turbines in the wind farm area by using satellite remote sensing data, and generate the spatial range of the wind farm area based on the spatial distribution of the wind turbines; Generate an initial spatial range of the control area in the external space of the wind farm area, and obtain the geographical feature data of the control area by using satellite remote sensing data; Based on the geographical feature data, screen the spatial range of the control area to determine the spatial range of the control area with the same geographical background as the wind farm area; Based on the spatial ranges of the wind farm area and the control area, respectively obtain the surface temperature data of the wind farm and the control surface temperature data from the satellite remote sensing data.

3. The method for constructing a prediction model of the influence of a wind farm on the surface temperature according to claim 1, wherein, The simulating the meteorological variable data of the wind farm area by using a meteorological numerical model, and screening the operation periods of the wind farm by using the wind speed variable in the meteorological variables includes: Based on meteorological reanalysis data, simulate and obtain the hourly meteorological variable data of the wind farm area by using a meteorological numerical model; According to the wind speed variable in the meteorological variable data, determine the periods when the wind speed meets the preset operating conditions of the wind turbines to screen out the operation periods of the wind farm.

4. The method for constructing a prediction model of the influence of a wind farm on surface temperature according to claim 1, wherein, The determining the difference between the surface temperature data of the wind farm and the control surface temperature data during the operation period of the wind farm to obtain the influence data of the wind farm on the surface temperature includes: Determine the average surface temperature of the wind farm area and the control area at the same moment; Calculate the difference between the average surface temperature of the wind farm area and the average surface temperature of the control area at each moment to obtain the influence data of the wind farm on the surface temperature.

5. The method for constructing a prediction model of the influence of a wind farm on surface temperature according to claim 1, wherein The using the meteorological variable data corresponding to the operation period of the wind farm as input and the influence data as output, and constructing a prediction model for the influence of the wind farm on the surface temperature based on a machine learning algorithm includes: Construct multiple prediction models using machine learning algorithms, and use the meteorological variable data corresponding to the operation period of the wind farm as input and the impact data as output to train the multiple prediction models respectively; Evaluate the prediction accuracy of the multiple prediction models using validation set data, and select the machine learning model with the highest prediction accuracy as the prediction model for the impact of the wind farm on the surface temperature.

6. The method for constructing a prediction model of the influence of a wind farm on surface temperature according to claim 1, characterized in that, The constructing a residual prediction model based on the prediction residuals and the meteorological variable data includes: Use the meteorological variable data corresponding to the operation period of the wind farm as input and the prediction residuals as output, and construct a residual prediction model using a linear regression calibration algorithm.

7. The method for constructing a prediction model of the influence of a wind farm on the surface temperature according to claim 1, wherein The constructing a prediction model for the impact of the wind farm on the surface temperature based on the meteorological variable data corresponding to the operation period of the wind farm as input and the impact data as output using a machine learning algorithm includes: Construct an input matrix based on the meteorological variable data corresponding to the operation period of the wind farm, and construct a target vector based on the data of the impact of the wind farm on the surface temperature; Use the bootstrap sampling method to extract a training sample set from the input matrix and the target vector, and construct a random forest model based on the training sample set, which consists of decision trees independently trained to form a decision tree ensemble: Among them, represents the set of decision trees of the random forest model, represents the th decision tree; Determine the optimal feature division node based on information gain and recursively construct a binary decision tree: Among them, represents the optimal partition of the th decision tree at the feature ; is the partitioning criterion, represents selecting the feature index that maximizes ; represents that the value range of the feature index is from 1 to n ; n represents the total number of features. The information gain calculation formula is: Among them, represents the data entropy, represents the conditional entropy after partitioning by the feature , and a decision tree structure is constructed based on the optimal partition; Calculate the predicted value of the impact of the wind farm on the surface temperature based on the trained random forest model, and perform weighted calculation using the prediction results of all decision trees: Among them, represents the final predicted value at the time step . represents the predicted output of the -th decision tree for the input sample , and a prediction model for the influence of the wind farm on the surface temperature is obtained.​ 8. A prediction method for the influence of a wind farm on surface temperature, characterized in that, including: Obtain the meteorological variable data corresponding to the operation period of the wind farm within the target wind farm area; Input the meteorological variable data into the prediction model for the impact of the wind farm on the surface temperature to obtain the data of the impact of the wind farm on the surface temperature within the target wind farm area; wherein, the prediction model for the impact of the wind farm on the surface temperature is trained according to the method for constructing the prediction model for the impact of the wind farm on the surface temperature described in any one of claims 1-7.

9. An apparatus for constructing a prediction model of the influence of a wind farm on surface temperature, characterized in that, including: A temperature data acquisition module for determining the spatial ranges of the wind farm area and the control area based on satellite remote sensing data, and using the spatial ranges to obtain the surface temperature data of the wind farm and the control surface temperature data; A meteorological data acquisition module for simulating the meteorological variable data of the wind farm area using a meteorological numerical model, and screening the operation period of the wind farm using the wind speed variable in the meteorological variables; An impact data acquisition module for determining the difference between the surface temperature data of the wind farm and the control surface temperature data during the operation period of the wind farm to obtain the data of the impact of the wind farm on the surface temperature; A prediction model construction module for constructing a prediction model for the impact of the wind farm on the surface temperature based on the meteorological variable data corresponding to the operation period of the wind farm as input and the impact data as output using a machine learning algorithm; Calculate the prediction residuals based on the prediction results of the prediction model and the impact data; construct a residual prediction model based on the prediction residuals and the meteorological variable data; and post-process the prediction results of the prediction model for the impact of the wind farm on the surface temperature using the residual prediction model to obtain the final prediction result of the impact of the wind farm on the surface temperature.

Images