An Adaptive Correction Method and System for New Energy Meteorological Monitoring Data

Abstract

This invention belongs to the field of new energy meteorological monitoring technology, specifically relating to an adaptive correction method and system for new energy meteorological monitoring data. The method includes: preprocessing multi-source meteorological monitoring data collected from target monitoring stations to obtain processed data; determining the path code corresponding to the processed data using a pre-established multi-dimensional decision tree model; and correcting the processed data according to the path code and the corresponding model to obtain the final corrected meteorological monitoring data. The multi-dimensional decision tree model is constructed using multi-source meteorological monitoring data from historical monitoring stations. The technical solution provided by this invention achieves automatic matching of optimal correction strategies under different scenarios, significantly improving data quality and providing a more reliable data foundation for new energy power prediction.

Description

Technical Field

[0001] This invention belongs to the field of new energy meteorological monitoring technology, specifically relating to an adaptive correction method and system for new energy meteorological monitoring data. Background Technology

[0002] With the increasing penetration of intermittent renewable energy sources such as wind and solar power into the power system, high-precision meteorological monitoring data has become an indispensable cornerstone for renewable energy power plant planning, power forecasting, and the safe and stable operation of the power grid. However, the harsh environments and complex terrains of renewable energy power plants pose a serious challenge to the quality of their meteorological monitoring data, and data redundancy and anomaly identification have become recognized technical bottlenecks in the industry.

[0003] Existing technical solutions have significant shortcomings and limitations in addressing such issues, primarily in the following three aspects: First, redundant processing technologies at the data acquisition and transmission level are lagging behind. At the data source, multiple sensors, multiple backup acquisition terminals, and high-frequency sampling strategies generate massive amounts of data, containing a large amount of spatiotemporal redundant information. Existing technologies mostly employ simple data compression algorithms or fixed-threshold deduplication strategies, lacking intelligent perception of the spatiotemporal correlation characteristics of new energy meteorological data. Second, the intelligence level of data quality control and anomaly identification technologies is insufficient. In terms of anomaly data identification, current industry practices mostly rely on rule-based "island-style" inspection methods, failing to deeply explore the inherent physical correlations between meteorological elements. Data quality control is ineffective in areas with significant micro-meteorological phenomena, such as complex mountains and valleys. Third, a correlation feature index system and adaptive processing framework for new energy applications are lacking. Existing research and technologies mostly focus on "data cleaning" itself, neglecting the fact that the ultimate goal of data quality is to serve the core application of "new energy power generation." There is a lack of a correlation feature index system that integrates original meteorological indicators with derived power generation indicators for comprehensive evaluation of data quality, leading to a disconnect between data quality assessment and final power generation prediction performance. In summary, existing technologies are characterized by "emphasizing single points and neglecting collaboration," "emphasizing statistics and neglecting physics," and "emphasizing cleaning and neglecting application," which cannot meet the urgent needs of new power systems for highly reliable new energy meteorological data. Summary of the Invention

[0004] To overcome the problems existing in the above-mentioned related technologies, the present invention provides an adaptive correction method and system for new energy meteorological monitoring data.

[0005] According to a first aspect of the present invention, an adaptive correction method for meteorological monitoring data of new energy sources is provided, comprising: The multi-source meteorological monitoring data collected from the target monitoring stations are preprocessed to obtain the processed data. The path code corresponding to the processed data is determined using a pre-established multidimensional decision tree model. Based on the model corresponding to the path coding scheduling, the processed data is corrected to obtain the final corrected meteorological monitoring data. The multidimensional decision tree model is constructed using multi-source meteorological monitoring data from historical monitoring stations.

[0006] Preferably, the preprocessing of the multi-source meteorological monitoring data collected from the target monitoring stations includes: The multi-source meteorological monitoring data from the target monitoring station are cleaned to obtain the processed data.

[0007] Preferably, determining the path code corresponding to the processed data using a pre-established multidimensional decision tree model includes: Feature extraction is performed on the processed data to obtain a first multidimensional feature vector; Using the first multidimensional feature vector as the input to the multidimensional decision tree model, the path code is output.

[0008] Preferably, the step of extracting features from the processed data to obtain a first multidimensional feature vector includes: The missing data rate statistical method is used to quantify the degree of data missing in the processed data. The distance matrix and elevation difference between the target monitoring station and other stations are determined using a GIS platform, and the spatial topology relationship of the processed data is determined based on the distance matrix and the elevation difference. Based on a pre-set meteorological physics knowledge base, determine the physical interpretability strength indicator of the processed data; The timeliness requirement level is determined based on the collection time of the processed data. The degree of data missing, the spatial topological relationship, the physical interpretability strength indicator, and the timeliness requirement level constitute the first multidimensional feature vector.

[0009] Preferably, the process of establishing the multidimensional decision tree model includes: Data cleaning was performed on the multi-source meteorological monitoring data from the historical monitoring stations to obtain cleaned data; Feature extraction is performed on the cleaned data to obtain a second multidimensional feature vector; Using the second multidimensional feature vector and the preset decision logic, the initial decision tree model is trained to obtain the multidimensional decision tree model; The preset decision logic includes: path direction and its corresponding path code.

[0010] Preferably, the step of extracting features from the cleaned data to obtain a second multidimensional feature vector includes: The missing data rate statistical method is used to quantify the degree of missing data in the cleaned data; The distance matrix and elevation difference between the historical monitoring station and other stations are determined using a GIS platform, and the spatial topology of the cleaned data is determined based on the distance matrix and elevation difference. Based on a pre-set meteorological physics knowledge base, determine the physical interpretability strength indicator of the cleaned data; The timeliness requirement level is determined based on the collection time of the cleaned data. The degree of data missing, the spatial topological relationship, the physical interpretability strength indicator, and the timeliness requirement level constitute the second multidimensional feature vector.

[0011] Preferably, the multidimensional decision tree model includes: a root node, second-level nodes, third-level nodes, and leaf nodes; The root node represents the degree of data missing, the second-level nodes represent the spatial topological relationships, the third-level nodes represent the physical interpretability strength indicators, and the leaf nodes represent the timeliness requirement level.

[0012] Preferably, the step of using a missing rate statistical method to quantify the degree of data missing in the processed data includes: Calculate the missing rate of the processed data; If the missing rate is less than a first preset threshold, the data missing degree is isolated missing; if the missing rate is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, the data missing degree is partial missing; if the missing rate is greater than the second preset threshold, the data missing degree is continuous missing.

[0013] Preferably, determining the spatial topological relationship of the processed data based on the distance matrix and the elevation difference includes: If the distance between the target monitoring station and its nearest other station is greater than a first distance threshold, then the spatial topology is an isolated station. If the number of other stations located less than or equal to the second distance threshold from the target monitoring station is greater than or equal to a preset number, then the spatial topology is a dense monitoring network. If the standard deviation of elevation calculated with the target monitoring station as the center within the grid where the target monitoring station is located is greater than the third distance threshold, then the spatial topology is a complex terrain area.

[0014] Preferably, determining the physical interpretability strength indicator of the processed data based on a preset meteorological physics knowledge base includes: Determine whether the meteorological elements in the processed data have corresponding physical laws or equations in a preset meteorological physics knowledge base. If they do, determine the list of parameters that must be input for the calculation corresponding to the physical law or equation. If they do not, mark the physical interpretability as weak. After determining the parameter list, it is determined whether each parameter in the parameter list is supported by a data source. If each parameter is supported by a data source, the physical interpretability strength is marked as strong; otherwise, the physical interpretability strength is marked as weak.

[0015] Preferably, determining the timeliness requirement level based on the collection time of the processed data includes: If the acquisition time is less than or equal to the first time threshold, the timeliness requirement level is real-time; if the acquisition time is greater than the first time threshold and less than the second time threshold, the timeliness requirement level is near real-time; if the acquisition time is greater than or equal to the second time threshold, the timeliness requirement level is offline.

[0016] Preferably, the step of correcting the processed data according to the model corresponding to the path coding scheduling to obtain the final corrected meteorological monitoring data includes: From a pre-defined model library, a correction technique configuration corresponding to the path encoding is scheduled; wherein the correction technique configuration defines at least one model or method and specifies the collaboration paradigm and execution parameters between models; The processed data is corrected using the correction technology to obtain the final corrected meteorological monitoring data.

[0017] Preferably, the formula for calculating the missing rate includes: n=N_missing / N_total In the above formula, or The missing rate, N_missing This represents the number of missing data points in the processed data. N_total This represents the total amount of data after processing.

[0018] According to a second aspect of the present invention, a new energy meteorological monitoring data adaptive correction system is provided, comprising: The processing unit is used to preprocess the multi-source meteorological monitoring data collected from the target monitoring stations to obtain the processed data; The determining unit is used to determine the path code corresponding to the processed data using a pre-established multidimensional decision tree model; The correction unit is used to correct the processed data according to the model corresponding to the path coding scheduling to obtain the final corrected meteorological monitoring data. The multidimensional decision tree model is constructed using multi-source meteorological monitoring data from historical monitoring stations.

[0019] Preferably, the processing unit includes: The first acquisition module is used to clean the multi-source meteorological monitoring data of the target monitoring station to obtain the processed data.

[0020] Preferably, the determining unit includes: The second acquisition module is used to extract features from the processed data to obtain a first multidimensional feature vector; The third acquisition module is used to output the path code by taking the first multidimensional feature vector as the input of the multidimensional decision tree model.

[0021] Preferably, the second acquisition module includes: The first quantification submodule is used to quantify the degree of data missing in the processed data using a missing rate statistical method. The first determining submodule is used to determine the distance matrix and elevation difference between the target monitoring station and other stations using a GIS platform, and to determine the spatial topology relationship of the processed data based on the distance matrix and the elevation difference. The second determining submodule is used to determine the physical interpretability strength identifier of the processed data based on a preset meteorological and physical knowledge base. The third determining submodule is used to determine the timeliness requirement level based on the collection time of the processed data; The fourth determining submodule is used to construct the first multidimensional feature vector by considering the degree of data missing, the spatial topological relationship, the physical interpretability strength identifier, and the timeliness requirement level.

[0022] Preferably, it further includes: a building unit for building the multidimensional decision tree model; the building unit includes: The fourth acquisition module is used to clean the multi-source meteorological monitoring data from the historical monitoring stations to obtain cleaned data. The fifth acquisition module is used to extract features from the cleaned data to obtain a second multidimensional feature vector; The training module is used to train the initial decision tree model using the second multidimensional feature vector and the preset decision logic to obtain the multidimensional decision tree model. The preset decision logic includes: path direction and its corresponding path code.

[0023] Preferably, the fifth acquisition module includes: The second quantification submodule is used to quantify the degree of data missing in the cleaned data using the missing rate statistical method. The fifth determination submodule is used to determine the distance matrix and elevation difference between the historical monitoring station and other stations using the GIS platform, and to determine the spatial topology relationship of the cleaned data based on the distance matrix and the elevation difference; The sixth determination submodule is used to determine the physical interpretability strength identifier of the cleaned data based on a preset meteorological and physical knowledge base; The seventh determination submodule is used to determine the timeliness requirement level based on the collection time of the cleaned data; The eighth determining submodule is used to construct the second multidimensional feature vector based on the degree of data missing, the spatial topological relationship, the physical interpretability strength identifier, and the timeliness requirement level.

[0024] Preferably, the multidimensional decision tree model includes: a root node, second-level nodes, third-level nodes, and leaf nodes; The root node represents the degree of data missing, the second-level nodes represent the spatial topological relationships, the third-level nodes represent the physical interpretability strength indicators, and the leaf nodes represent the timeliness requirement level.

[0025] Preferably, the quantization submodule is specifically used for: Calculate the missing rate of the processed data; If the missing rate is less than a first preset threshold, the data missing degree is isolated missing; if the missing rate is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, the data missing degree is partial missing; if the missing rate is greater than the second preset threshold, the data missing degree is continuous missing.

[0026] Preferably, the first determining submodule is specifically used for: If the distance between the target monitoring station and its nearest other station is greater than a first distance threshold, then the spatial topology is an isolated station. If the number of other stations located less than or equal to the second distance threshold from the target monitoring station is greater than or equal to a preset number, then the spatial topology is a dense monitoring network. If the standard deviation of elevation calculated with the target monitoring station as the center within the grid where the target monitoring station is located is greater than the third distance threshold, then the spatial topology is a complex terrain area.

[0027] Preferably, the second determining submodule is specifically used for: Determine whether the meteorological elements in the processed data have corresponding physical laws or equations in a preset meteorological physics knowledge base. If they do, determine the list of parameters that must be input for the calculation corresponding to the physical law or equation. If they do not, mark the physical interpretability as weak. After determining the parameter list, it is determined whether each parameter in the parameter list is supported by a data source. If each parameter is supported by a data source, the physical interpretability strength is marked as strong; otherwise, the physical interpretability strength is marked as weak.

[0028] Preferably, the third determining submodule is specifically used for: If the acquisition time is less than or equal to the first time threshold, the timeliness requirement level is real-time; if the acquisition time is greater than the first time threshold and less than the second time threshold, the timeliness requirement level is near real-time; if the acquisition time is greater than or equal to the second time threshold, the timeliness requirement level is offline.

[0029] Preferably, the correction unit includes: The scheduling module is used to schedule the correction technology configuration corresponding to the path code from a preset model library; wherein the correction technology configuration defines at least one model or method and specifies the collaboration paradigm and execution parameters between models; The correction module is used to correct the processed data using the correction technology to obtain the final corrected meteorological monitoring data.

[0030] Preferably, the formula for calculating the missing rate includes: n=N_missing / N_total In the above formula, or The missing rate, N_missing This represents the number of missing data points in the processed data. N_total This represents the total amount of data after processing.

[0031] According to a third aspect of the present invention, an electronic device is provided, comprising: at least one processor and a memory; the memory and the processor are connected via a bus; The memory is used to store one or more programs; The method is implemented when the one or more programs are executed by the at least one processor.

[0032] According to a fourth aspect of the present invention, a readable storage medium is provided having an executable program stored thereon, wherein when the executable program is executed, the method is implemented.

[0033] The technical solution provided by this invention has the following beneficial effects: This invention provides an adaptive correction method and system for meteorological monitoring data of new energy sources. By preprocessing multi-source meteorological monitoring data collected from target monitoring stations, the system obtains processed data, enabling intelligent identification of redundancy and accurate anomaly identification, thus providing high-quality data support for the planning and operation of new energy sources. By utilizing a pre-established multi-dimensional decision tree model, the system determines the path code corresponding to the processed data, schedules the corresponding model according to the path code, and corrects the processed data to obtain the final corrected meteorological monitoring data. This achieves automatic matching of the optimal correction strategy under different scenarios, significantly improving data quality and providing a more reliable data foundation for new energy power prediction. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 This is a flowchart of an adaptive correction method for meteorological monitoring data of new energy provided in an embodiment of the present invention; Figure 2 This is a flowchart of an adaptive correction method for meteorological monitoring data of new energy provided in an embodiment of the present invention; Figure 3 This is a structural block diagram of an adaptive correction system for meteorological monitoring data of new energy provided in an embodiment of the present invention; Figure 4 This is a structural block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the following embodiments are only some embodiments of this invention, not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0037] Example 1 This invention provides an adaptive correction method for meteorological monitoring data of new energy sources, such as... Figure 1 As shown, it includes the following steps: Step 11: Preprocess the multi-source meteorological monitoring data collected from the target monitoring stations to obtain the processed data; Step 12: Using a pre-established multidimensional decision tree model, determine the path code corresponding to the processed data; Step 13: Correct the processed data according to the model corresponding to the path coding scheduling to obtain the final corrected meteorological monitoring data; The multidimensional decision tree model is constructed using multi-source meteorological monitoring data from historical monitoring stations.

[0038] This invention can intelligently identify redundancy and accurately identify anomalies, providing high-quality data support for the planning and operation of new energy sources.

[0039] Further, step 11 includes: Step 111: Perform data cleaning on the multi-source meteorological monitoring data of the target monitoring station to obtain the processed data.

[0040] It should be noted that the "data cleaning" method involved in the embodiments of the present invention is well known to those skilled in the art, therefore, its specific implementation will not be described in detail.

[0041] Further, step 12 includes: Step 121: Extract features from the processed data to obtain the first multidimensional feature vector; Step 122: Use the first multidimensional feature vector as the input to the multidimensional decision tree model and output the path code.

[0042] It should be noted that the embodiments of the present invention do not limit the "path encoding", which can be set by those skilled in the art according to engineering needs, experimental data or expert experience.

[0043] Further, step 121 includes: Step 1211: Use the missing rate statistical method to quantify the degree of missing data in the processed data; Step 1212: Use the GIS platform to determine the distance matrix and elevation difference between the target monitoring station and other stations, and determine the spatial topology relationship of the processed data based on the distance matrix and elevation difference; Step 1213: Determine the physical interpretability strength indicator of the processed data based on the preset meteorological physics knowledge base; Step 1214: Determine the timeliness requirement level based on the collection time of the processed data; Step 1215: The degree of data missing, spatial topological relationship, physical interpretability strength indicator and timeliness requirement level constitute the first multidimensional feature vector.

[0044] It should be noted that the method of "using a GIS platform to determine the distance matrix and elevation difference between the target monitoring station and other stations" involved in the embodiments of the present invention is well known to those skilled in the art, therefore, its specific implementation will not be described in detail.

[0045] This invention provides a perceptual and adaptive data correction framework. By constructing a multidimensional decision tree model, it dynamically evaluates four core dimensions—data missingness, spatial topological relationship, physical interpretability, and timeliness requirements—to intelligently select or combine the optimal correction strategy for each anomalous data instance.

[0046] Furthermore, the method also includes: Step 10: Establishing a multidimensional decision tree model; Step 10 includes: Step 101: Clean the multi-source meteorological monitoring data from historical monitoring stations to obtain cleaned data; Step 102: Extract features from the cleaned data to obtain the second multidimensional feature vector; Step 103: Using the second multidimensional feature vector and the preset decision logic, train the initial decision tree model to obtain a multidimensional decision tree model; The preset decision-making logic includes: path direction and its corresponding path code.

[0047] Further, step 102 includes: Step 1021: Use the missing rate statistical method to quantify the degree of missing data in the cleaned data; Step 1022: Use the GIS platform to determine the distance matrix and elevation difference between historical monitoring stations and other stations, and determine the spatial topology of the cleaned data based on the distance matrix and elevation difference; Step 1023: Determine the physical interpretability strength indicator of the cleaned data based on the preset meteorological physics knowledge base; Step 1024: Determine the timeliness requirement level based on the collection time of the cleaned data; Step 1025: The degree of data missing, spatial topological relationship, physical interpretability strength indicator and timeliness requirement level constitute the second multidimensional feature vector.

[0048] Further, step 1021 includes: Calculate the missing rate of the cleaned data; If the missing rate is less than the first preset threshold, the data missing level is isolated missing; if the missing rate is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, the data missing level is partial missing; if the missing rate is greater than the second preset threshold, the data missing level is continuous missing.

[0049] Further, step 1022 includes: If the distance between a historical monitoring station and its nearest other station is greater than a first distance threshold, then the spatial topology is that of an isolated station. If the number of other stations that are less than or equal to the second distance threshold from the historical monitoring station is greater than or equal to the preset number, then the spatial topology is a dense monitoring network. If the standard deviation of elevation calculated with the historical monitoring station as the center within the grid is greater than the third distance threshold, then the spatial topology is a complex terrain area.

[0050] Further, step 1023 includes: Determine whether the meteorological elements in the cleaned data have corresponding physical laws or equations in the preset meteorological physics knowledge base. If they do, determine the list of parameters that must be input for the calculation corresponding to the physical law or equation; if they do not, mark the physical interpretability as weak. After determining the parameter list, it is determined whether each parameter in the parameter list is supported by a data source. If each parameter is supported by a data source, the physical interpretability strength is marked as strong; otherwise, the physical interpretability strength is marked as weak.

[0051] Further, step 1024 includes: If the collection time is less than or equal to the first time threshold, the timeliness requirement level is real-time; if the collection time is greater than the first time threshold but less than the second time threshold, the timeliness requirement level is near real-time; if the collection time is greater than or equal to the second time threshold, the timeliness requirement level is offline.

[0052] Furthermore, the multidimensional decision tree model includes: root node, second-level nodes, third-level nodes, and leaf nodes; The root node represents the degree of data missing, the second-level nodes represent the spatial topological relationships, the third-level nodes represent the physical interpretability strength indicators, and the leaf nodes represent the timeliness requirement level.

[0053] For example, a four-layer sequential decision structure can be constructed, with the degree of data missing as the root node, spatial topological relationships as the second layer, physical interpretability strength indicators as the third layer, and timeliness requirement levels as leaf nodes. The first multidimensional feature vector is input into the multidimensional decision tree model, initiating the judgment process: First, the degree of data missing is determined; if continuous missing data occurs, time-series models relying solely on their own historical data are excluded, and the process shifts to branches incorporating external data sources. Second, spatial topological relationships are determined; if isolated sites exist, the weight of spatial interpolation methods is reduced, and the process shifts to branches relying on physical models. Then, the physical interpretability strength indicator is determined; for elements with strong physical interpretability, the process shifts to branches using equation inversion or physical constraint models, while for elements with weak physical interpretability, the process shifts to data-driven or statistical learning branches. Finally, the timeliness requirement level is determined; for data with high real-time requirements, the process shifts to lightweight, low-latency model branches, while for data with low real-time requirements, the process shifts to models with higher accuracy and larger computational requirements.

[0054] Further, step 1211 includes: Step 1211a: Calculate the missing rate of the processed data; Specifically, the formula for calculating the missing rate includes: n=N_missing / N_total In the above formula, or The missing rate, N_missing This represents the number of missing data points in the processed data. N_total The total amount of data after processing; Step 1211b: If the missing rate is less than the first preset threshold, the data missing degree is isolated missing; if the missing rate is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, the data missing degree is partial missing; if the missing rate is greater than the second preset threshold, the data missing degree is continuous missing.

[0055] It should be noted that the embodiments of the present invention do not limit the "first preset threshold" and the "second preset threshold," and these can be set by those skilled in the art based on engineering needs, experimental data, or expert experience. For example, let the missing rate be... or, Isolated missing as or <5%, partially missing is 5%≤ or ≤30%, consecutive deletions are or >30%.

[0056] Further, step 1212 includes: If the distance between the target monitoring station and its nearest other station is greater than a first distance threshold, the spatial topology is that of an isolated station. If the number of other stations that are less than or equal to the second distance threshold from the target monitoring station is greater than or equal to the preset number, then the spatial topology is a dense monitoring network. If the standard deviation of elevation calculated with the target monitoring station as the center within the grid where the target monitoring station is located is greater than the third distance threshold, then the spatial topology is a complex terrain area.

[0057] It should be noted that the calculation method of the "elevation standard deviation" involved in the embodiments of the present invention is well known to those skilled in the art, therefore, its specific implementation will not be described in detail. For example, within the grid where the target monitoring station is located, the elevation standard deviation is calculated using the elevation difference between the target monitoring station and other stations, with the target monitoring station as the center.

[0058] The embodiments of the present invention do not limit the "first distance threshold", "second distance threshold" and "third distance threshold", which can be set by those skilled in the art according to engineering needs, experimental data or expert experience. For example, the first distance threshold may be, but is not limited to, 20 kilometers, the second distance threshold may be, but is not limited to, 5 kilometers, and the third distance threshold may be, but is not limited to, 100 meters.

[0059] Further, step 1213 includes: Determine whether the meteorological elements in the processed data have corresponding physical laws or equations in the preset meteorological physics knowledge base. If they do, determine the list of parameters that must be input for the calculation corresponding to the physical law or equation; if they do not, mark the physical interpretability as weak. After determining the parameter list, it is determined whether each parameter in the parameter list is supported by a data source. If each parameter is supported by a data source, the physical interpretability strength is marked as strong; otherwise, the physical interpretability strength is marked as weak.

[0060] Further, step 1214 includes: If the collection time is less than or equal to the first time threshold, the timeliness requirement level is real-time; if the collection time is greater than the first time threshold but less than the second time threshold, the timeliness requirement level is near real-time; if the collection time is greater than or equal to the second time threshold, the timeliness requirement level is offline.

[0061] It should be noted that the embodiments of the present invention do not limit the "first time threshold" and the "second time threshold," which can be set by those skilled in the art based on engineering needs, experimental data, or expert experience. For example, the first time threshold may be, but is not limited to, 5 minutes, and the second time threshold may be, but is not limited to, 1 hour.

[0062] Further, step 13 includes: Step 131: From the preset model library, schedule the correction technology configuration corresponding to the path encoding; wherein, the correction technology configuration defines at least one model or method, and specifies the collaboration paradigm and execution parameters between models; Step 132: Use the correction technology to correct the processed data to obtain the final corrected meteorological monitoring data.

[0063] Understandably, the path encoding uniquely maps to at least one correction model in a pre-defined model library, such as WRF-PINN, which represents: WRF downscaling simulation + physical information neural network (PINN) fusion correction.

[0064] For example, from a preset model library, a correction technique configuration corresponding to the path coding is scheduled; wherein, the correction technique configuration defines at least one model or method and specifies the collaboration paradigm and execution parameters between models; the correction technique configuration is executed to jointly correct the processed data according to the specified collaboration paradigm and execution parameters; after the joint correction is completed, the final corrected meteorological monitoring data is output.

[0065] To further illustrate the above-mentioned adaptive correction method for new energy meteorological monitoring data, this invention also provides a specific example, such as... Figure 2 As shown, it includes the following steps: Step 21: Input multi-source meteorological monitoring data from the target monitoring stations. First, perform data cleaning and standardization, including format unification, time alignment, and invalid value removal. Second, perform parallel computing on four core decision features: use the missing rate statistical method to quantify the degree of data missingness. or Based on GIS spatial analysis technology, the distance between stations and the difference in terrain elevation are calculated to determine the spatial topological relationship. The meteorological and physical knowledge base is queried to determine the physical interpretability strength of the elements, and the timeliness level label is configured according to the business scenario. Finally, the processed data and its corresponding first multidimensional feature vector are output.

[0066] Step 22: Based on the first multidimensional feature vector generated in Step 21, an automatic judgment is performed using a multidimensional decision tree model that combines rules and machine learning: first, the degree of missing data is determined; second, the spatial topology type is considered; third, physical interpretability is evaluated; and finally, timeliness requirements are matched. This is done through decision tree classification rules (e.g., if...). or If the area is greater than 30% and the terrain is complex, then trigger path P3 and output an optimal or multiple combined data correction technique path number.

[0067] Step 23: Based on the technology path number output in Step 22, schedule the corresponding algorithm model instance from the preset model library. The model library includes: physical models (such as the WRF atmospheric model), statistical models (such as spatiotemporal kriging), and AI models (such as LSTM, PINN), etc.; at the same time, allocate computing resources according to timeliness requirements; finally, output the ready signal of the initialized and configured model instance.

[0068] Step 24: Based on the model scheduled in Step 23, perform correction calculations on the processed data from Step 21 to obtain the final corrected meteorological monitoring data.

[0069] This invention provides an adaptive correction method for new energy meteorological monitoring data. Addressing the operational needs of handling data redundancy and anomalies in new energy meteorological monitoring, it constructs an adaptive data correction technology system through a multi-dimensional decision engine based on data missingness, spatial topology, physical interpretability, and timeliness requirements. This invention achieves automatic matching of optimal correction strategies under different scenarios by intelligently identifying data characteristics and dynamically scheduling physical models, statistical models, and AI algorithms. This effectively solves the pain points of traditional methods, such as rigid rules and inability to adapt to complex terrain and variable meteorological conditions. This technology significantly improves data quality, providing a more reliable data foundation for new energy power prediction.

[0070] Example 2 To further illustrate the above-mentioned adaptive correction method for new energy meteorological monitoring data, this invention also provides a specific example, such as... Figure 2 As shown, it includes the following steps: Step 31: Multi-source data preprocessing and feature extraction: The system receives heterogeneous data from multiple sources, including ground-based weather stations and satellite remote sensing. First, the raw data is cleaned and time-series aligned to eliminate format differences and obvious errors, resulting in processed data. Then, four feature metrics are computed in parallel. ① Degree of data missing: The missing rate is quantified using a missing rate statistical algorithm, and the missing rate η is defined as follows: n=N_missing / N_total In the above formula, or The missing rate, N_missing This represents the number of missing data points in the processed data. N_total This represents the total amount of data after processing.

[0071] Based on the missing rate, it is divided into isolated missing (… or <5%), partially missing (5%≤) or ≤30%) and consecutive missing ( or (>30%) in three levels.

[0072] ② Spatial topological relationships: Based on the GIS platform, the distance matrix and elevation difference between the target monitoring station and other stations are calculated, and the spatial topological relationship is determined according to the preset threshold. Isolated site: The distance between the target monitoring site and the nearest site is greater than 20 kilometers; Dense monitoring network: station spacing less than or equal to 5 kilometers and number of stations greater than or equal to 3; Complex terrain areas: Within the grid where the target monitoring station is located, the standard deviation of the elevation calculated with the target monitoring station as the center is greater than 100 meters.

[0073] ③ Physical interpretability: The system queries the meteorological physics knowledge base to determine whether the current meteorological elements in the processed data have clear physical laws or equations to support them. If there are physical laws or equations, the system identifies the list of parameters that must be input for the calculation of the physical law or equation. If all parameters have data source support, the system is identified as having strong physical interpretability. If any parameter lacks data source support or does not have a corresponding physical law or equation, the system is identified as having weak physical interpretability.

[0074] ④ Timeliness requirement: The collection time of multi-source meteorological monitoring data from the target monitoring station is determined based on the request type received from the business interface, and then the timeliness requirement level is automatically marked. Real-time: less than or equal to 5 minutes; Near real-time: 5 minutes to 1 hour; Offline: More than one hour.

[0075] Finally, the processed data and the corresponding first multidimensional feature vector are output.

[0076] Step 32: Multidimensional Decision Path Selection After determining the first multi-dimensional feature vector, a built-in decision tree combining rules and statistical models is used for classification. The decision logic follows a priority strategy: first, determine the degree of missing data; second, analyze the spatial topology; third, evaluate physical interpretability; and finally, match timeliness requirements. The engine has multiple pre-defined correction paths, such as: Path 1: If it is "isolated missing + dense network + strong physicality + real-time", then trigger "nearby station spatial interpolation + rapid correction of physical constraints"; Path 2: If it is "continuous missing + complex terrain + strong physicality + offline", then trigger "WRF downscaling simulation + physical information neural network (PINN) fusion correction"; Path 3: If it is “partially missing + isolated site + weak physicality + near real-time”, then “satellite data replacement + LSTM time series prediction filling” will be triggered.

[0077] The decision engine outputs one or more optimal correction technique path codes by matching feature vectors with a predefined rule base in real time.

[0078] Step 33: Dynamic Decision Making and Model Scheduling Based on the path encoding, it is responsible for scheduling and configuring the corresponding computing resources. Internally, it maintains a model pool containing physical models, statistical models, and AI models. The scheduler instantiates the corresponding model based on the path encoding and allocates computing resources according to the timeliness level: for "real-time" requirements, it allocates GPU resources and calls lightweight models to ensure latency; for "offline" requirements, it calls high-precision CPU clusters for deep computation.

[0079] Load the corresponding correction technology configuration from the preset model library; wherein the correction technology configuration defines at least one model or method and specifies the collaboration paradigm and execution parameters between models.

[0080] Step 34: Correct the execution module Based on the scheduling model instance, collaboration paradigm, execution parameters, and resource quota, the model and parameters allocated by the scheduling module are loaded, and the processed data is jointly corrected according to the specified collaboration paradigm and execution parameters; after the joint correction is completed, the final corrected meteorological monitoring data is output.

[0081] Example 3 This invention also provides an adaptive correction system for meteorological monitoring data of new energy sources, such as... Figure 3 As shown, it includes: The processing unit is used to preprocess the multi-source meteorological monitoring data collected from the target monitoring stations to obtain the processed data; The determination unit is used to determine the path code corresponding to the processed data using a pre-established multidimensional decision tree model; The correction unit is used to correct the processed data according to the path coding and scheduling corresponding models to obtain the final corrected meteorological monitoring data. The multidimensional decision tree model is constructed using multi-source meteorological monitoring data from historical monitoring stations.

[0082] Furthermore, the processing unit includes: The first acquisition module is used to clean the multi-source meteorological monitoring data of the target monitoring station to obtain the processed data.

[0083] Further, the defined units include: The second acquisition module is used to extract features from the processed data to obtain the first multidimensional feature vector; The third acquisition module is used to take the first multidimensional feature vector as input to the multidimensional decision tree model and output path encoding.

[0084] Furthermore, the second acquisition module includes: The quantification submodule is used to quantify the degree of data missing in the processed data using a missing rate statistical method. The first determination submodule is used to determine the distance matrix and elevation difference between the target monitoring station and other stations using the GIS platform, and to determine the spatial topology relationship of the processed data based on the distance matrix and elevation difference. The second determination submodule is used to determine the physical interpretability strength indicator of the processed data based on a preset meteorological and physical knowledge base. The third determination submodule is used to determine the timeliness requirement level based on the collection time of the processed data; The fourth determination submodule is used to identify the degree of data missingness, spatial topological relationships, physical interpretability strength, and timeliness requirement level, forming the first multidimensional feature vector.

[0085] Furthermore, the system also includes: a building unit for building a multidimensional decision tree model; the building unit includes: The fourth acquisition module is used to clean the multi-source meteorological monitoring data from historical monitoring stations to obtain cleaned data. The fifth acquisition module is used to extract features from the cleaned data to obtain the second multidimensional feature vector; The training module is used to train the initial decision tree model using the second multidimensional feature vector and the preset decision logic to obtain a multidimensional decision tree model. The preset decision-making logic includes: path direction and its corresponding path code.

[0086] Furthermore, the fifth acquisition module includes: The second quantification submodule is used to quantify the degree of data missing in the cleaned data using the missing rate statistical method. The fifth submodule is used to determine the distance matrix and elevation difference between historical monitoring stations and other stations using the GIS platform, and to determine the spatial topology of the cleaned data based on the distance matrix and elevation difference. The sixth submodule is used to determine the physical interpretability strength indicator of the cleaned data based on a preset meteorological and physical knowledge base. The seventh submodule is used to determine the timeliness requirement level based on the collection time of the cleaned data; The eighth determination submodule is used to identify the degree of data missingness, spatial topological relationships, physical interpretability strength, and timeliness requirement level, forming the second multidimensional feature vector.

[0087] Furthermore, the second quantization submodule is specifically used for: Calculate the missing rate of the cleaned data; If the missing rate is less than the first preset threshold, the data missing level is isolated missing; if the missing rate is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, the data missing level is partial missing; if the missing rate is greater than the second preset threshold, the data missing level is continuous missing.

[0088] Furthermore, the fifth determining submodule is specifically used for: If the distance between a historical monitoring station and its nearest other station is greater than a first distance threshold, then the spatial topology is that of an isolated station. If the number of other stations that are less than or equal to the second distance threshold from the historical monitoring station is greater than or equal to the preset number, then the spatial topology is a dense monitoring network. If the standard deviation of elevation calculated with the historical monitoring station as the center within the grid is greater than the third distance threshold, then the spatial topology is a complex terrain area.

[0089] Furthermore, the sixth determining submodule is specifically used for: Determine whether the meteorological elements in the cleaned data have corresponding physical laws or equations in the preset meteorological physics knowledge base. If they do, determine the list of parameters that must be input for the calculation corresponding to the physical law or equation; if they do not, mark the physical interpretability as weak. After determining the parameter list, it is determined whether each parameter in the parameter list is supported by a data source. If each parameter is supported by a data source, the physical interpretability strength is marked as strong; otherwise, the physical interpretability strength is marked as weak.

[0090] Furthermore, the seventh defining submodule is specifically used for: If the collection time is less than or equal to the first time threshold, the timeliness requirement level is real-time; if the collection time is greater than the first time threshold but less than the second time threshold, the timeliness requirement level is near real-time; if the collection time is greater than or equal to the second time threshold, the timeliness requirement level is offline.

[0091] Furthermore, the multidimensional decision tree model includes: root node, second-level nodes, third-level nodes, and leaf nodes; The root node represents the degree of data missing, the second-level nodes represent the spatial topological relationships, the third-level nodes represent the physical interpretability strength indicators, and the leaf nodes represent the timeliness requirement level.

[0092] Furthermore, the quantization submodule is specifically used for: Calculate the missing rate of the processed data; If the missing rate is less than the first preset threshold, the data missing level is isolated missing; if the missing rate is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, the data missing level is partial missing; if the missing rate is greater than the second preset threshold, the data missing level is continuous missing.

[0093] Furthermore, the first defined submodule is specifically used for: If the distance between the target monitoring station and its nearest other station is greater than a first distance threshold, the spatial topology is that of an isolated station. If the number of other stations that are less than or equal to the second distance threshold from the target monitoring station is greater than or equal to the preset number, then the spatial topology is a dense monitoring network. If the standard deviation of elevation calculated with the target monitoring station as the center within the grid where the target monitoring station is located is greater than the third distance threshold, then the spatial topology is a complex terrain area.

[0094] Furthermore, the second determining submodule is specifically used for: Determine whether the meteorological elements in the processed data have corresponding physical laws or equations in the preset meteorological physics knowledge base. If they do, determine the list of parameters that must be input for the calculation corresponding to the physical law or equation; if they do not, mark the physical interpretability as weak. After determining the parameter list, it is determined whether each parameter in the parameter list is supported by a data source. If each parameter is supported by a data source, the physical interpretability strength is marked as strong; otherwise, the physical interpretability strength is marked as weak.

[0095] Furthermore, the third determining submodule is specifically used for: If the collection time is less than or equal to the first time threshold, the timeliness requirement level is real-time; if the collection time is greater than the first time threshold but less than the second time threshold, the timeliness requirement level is near real-time; if the collection time is greater than or equal to the second time threshold, the timeliness requirement level is offline.

[0096] Furthermore, the correction unit includes: The scheduling module is used to schedule the correction technology configuration corresponding to the path encoding from the preset model library; wherein, the correction technology configuration defines at least one model or method and specifies the collaboration paradigm and execution parameters between models; The correction module is used to correct the processed data using correction technology to obtain the final corrected meteorological monitoring data.

[0097] Furthermore, the formula for calculating the missing rate includes: n=N_missing / N_total In the above formula, or The missing rate, N_missing This represents the number of missing data points in the processed data. N_total This represents the total amount of data after processing.

[0098] It is understood that the system embodiments provided above correspond to the method embodiments described above, and the specific details can be referred to each other, which will not be repeated here.

[0099] It is understood that the same or similar parts in the above embodiments can be referred to each other, and the contents not described in detail in some embodiments can be referred to the same or similar contents in other embodiments.

[0100] Example 4 like Figure 4 As shown, the present invention also provides an electronic device, which may be a computer device, a microcontroller device, a smart mobile device, etc. The electronic device in this embodiment may include a processor, a memory, a transceiver component, etc. The memory, processor, and transceiver component are connected via a bus; the memory can be used to store executable programs, and an exemplary executable program may include instructions; the processor is used to execute the instructions stored in the memory. The memory can also be used to store data, which can be accessed and / or modified when instructions are executed.

[0101] The processor may be a Central Processing Unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and it is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the storage medium to realize the corresponding method flow or corresponding function, so as to realize the steps of the adaptive correction method for new energy meteorological monitoring data in the above embodiments.

[0102] Example 5 Based on the same inventive concept, this invention also provides a readable storage medium, specifically an electronic device readable storage medium (Memory). This readable storage medium is a memory device within an electronic device used to store programs and data. It is understood that the storage medium here can include both the built-in storage medium of the electronic device and extended storage media supported by the electronic device. The storage medium provides storage space, which stores the terminal's operating system. Furthermore, this storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more executable programs (including program code). It should be noted that the storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device. Loading and executing one or more instructions stored in the storage medium by the processor can implement the steps of the adaptive correction method for new energy meteorological monitoring data in the above embodiments.

[0103] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0104] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure one One or more processes and / or boxes Figure one A device that provides the functions specified in one or more boxes.

[0105] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure one One or more processes and / or boxes Figure one The function specified in one or more boxes.

[0106] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure one One or more processes and / or boxes Figure one The steps of the function specified in one or more boxes.

[0107] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. An adaptive correction method for meteorological monitoring data of new energy sources, characterized in that, include: The multi-source meteorological monitoring data collected from the target monitoring stations are preprocessed to obtain the processed data. The path code corresponding to the processed data is determined using a pre-established multidimensional decision tree model. Based on the model corresponding to the path coding scheduling, the processed data is corrected to obtain the final corrected meteorological monitoring data. The multidimensional decision tree model is constructed using multi-source meteorological monitoring data from historical monitoring stations.

2. The method according to claim 1, characterized in that, The preprocessing of multi-source meteorological monitoring data collected from the target monitoring stations includes: The multi-source meteorological monitoring data from the target monitoring station are cleaned to obtain the processed data.

3. The method according to claim 1, characterized in that, The step of determining the path code corresponding to the processed data using a pre-established multidimensional decision tree model includes: Feature extraction is performed on the processed data to obtain a first multidimensional feature vector; Using the first multidimensional feature vector as the input to the multidimensional decision tree model, the path code is output.

4. The method according to claim 1, characterized in that, The step of extracting features from the processed data to obtain a first multidimensional feature vector includes: The missing data rate statistical method is used to quantify the degree of data missing in the processed data. The distance matrix and elevation difference between the target monitoring station and other stations are determined using a GIS platform, and the spatial topology relationship of the processed data is determined based on the distance matrix and the elevation difference. Based on a pre-set meteorological physics knowledge base, determine the physical interpretability strength indicator of the processed data; The timeliness requirement level is determined based on the collection time of the processed data. The degree of data missing, the spatial topological relationship, the physical interpretability strength indicator, and the timeliness requirement level constitute the first multidimensional feature vector.

5. The method according to claim 1, characterized in that, The process of establishing the multidimensional decision tree model includes: Data cleaning was performed on the multi-source meteorological monitoring data from the historical monitoring stations to obtain cleaned data; Feature extraction is performed on the cleaned data to obtain a second multidimensional feature vector; Using the second multidimensional feature vector and the preset decision logic, the initial decision tree model is trained to obtain the multidimensional decision tree model; The preset decision logic includes: path direction and its corresponding path code.

6. The method according to claim 5, characterized in that, The step of extracting features from the cleaned data to obtain a second multidimensional feature vector includes: The missing data rate statistical method is used to quantify the degree of missing data in the cleaned data; The distance matrix and elevation difference between the historical monitoring station and other stations are determined using a GIS platform, and the spatial topology of the cleaned data is determined based on the distance matrix and elevation difference. Based on a pre-set meteorological physics knowledge base, determine the physical interpretability strength indicator of the cleaned data; The timeliness requirement level is determined based on the collection time of the cleaned data. The degree of data missing, the spatial topological relationship, the physical interpretability strength indicator, and the timeliness requirement level constitute the second multidimensional feature vector.

7. The method according to claim 6, characterized in that, The multidimensional decision tree model includes: root node, second-level nodes, third-level nodes, and leaf nodes; The root node represents the degree of data missing, the second-level nodes represent the spatial topological relationships, the third-level nodes represent the physical interpretability strength indicators, and the leaf nodes represent the timeliness requirement level.

8. The method according to claim 4, characterized in that, The method of using missing rate statistics to quantify the degree of data missing in the processed data includes: Calculate the missing rate of the processed data; If the missing rate is less than a first preset threshold, the data missing degree is isolated missing; if the missing rate is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, the data missing degree is partial missing; if the missing rate is greater than the second preset threshold, the data missing degree is continuous missing.

9. The method according to claim 4, characterized in that, Determining the spatial topology of the processed data based on the distance matrix and the elevation difference includes: If the distance between the target monitoring station and its nearest other station is greater than a first distance threshold, then the spatial topology is an isolated station. If the number of other stations located less than or equal to the second distance threshold from the target monitoring station is greater than or equal to a preset number, then the spatial topology is a dense monitoring network. If the standard deviation of elevation calculated with the target monitoring station as the center within the grid where the target monitoring station is located is greater than the third distance threshold, then the spatial topology is a complex terrain area.

10. The method according to claim 4, characterized in that, The step of determining the physical interpretability strength indicator of the processed data based on a preset meteorological physics knowledge base includes: Determine whether the meteorological elements in the processed data have corresponding physical laws or equations in a preset meteorological physics knowledge base. If they do, determine the list of parameters that must be input for the calculation corresponding to the physical law or equation. If they do not, mark the physical interpretability as weak. After determining the parameter list, it is determined whether each parameter in the parameter list is supported by a data source. If each parameter is supported by a data source, the physical interpretability strength is marked as strong; otherwise, the physical interpretability strength is marked as weak.

11. The method according to claim 4, characterized in that, The step of determining the timeliness requirement level based on the collection time of the processed data includes: If the acquisition time is less than or equal to the first time threshold, the timeliness requirement level is real-time; if the acquisition time is greater than the first time threshold and less than the second time threshold, the timeliness requirement level is near real-time; if the acquisition time is greater than or equal to the second time threshold, the timeliness requirement level is offline.

12. The method according to claim 4, characterized in that, The step of correcting the processed data according to the model corresponding to the path coding scheduling to obtain the final corrected meteorological monitoring data includes: From a pre-defined model library, a correction technique configuration corresponding to the path encoding is scheduled; wherein the correction technique configuration defines at least one model or method and specifies the collaboration paradigm and execution parameters between models; The processed data is corrected using the correction technology to obtain the final corrected meteorological monitoring data.

13. The method according to claim 8, characterized in that, The formula for calculating the missing rate includes: η = N_missing / N_total In the above formula, η The missing rate, N_missing This represents the number of missing data points in the processed data. N_total This represents the total amount of data after processing.

14. An adaptive correction system for meteorological monitoring data of new energy sources, characterized in that, include: The processing unit is used to preprocess the multi-source meteorological monitoring data collected from the target monitoring stations to obtain the processed data; The determining unit is used to determine the path code corresponding to the processed data using a pre-established multidimensional decision tree model; The correction unit is used to correct the processed data according to the model corresponding to the path coding scheduling to obtain the final corrected meteorological monitoring data. The multidimensional decision tree model is constructed using multi-source meteorological monitoring data from historical monitoring stations.

15. The system according to claim 14, characterized in that, The processing unit includes: The first acquisition module is used to clean the multi-source meteorological monitoring data of the target monitoring station to obtain the processed data.

16. The system according to claim 14, characterized in that, The determining unit includes: The second acquisition module is used to extract features from the processed data to obtain a first multidimensional feature vector; The third acquisition module is used to output the path code by taking the first multidimensional feature vector as the input of the multidimensional decision tree model.

17. The system according to claim 16, characterized in that, The second acquisition module includes: The first quantification submodule is used to quantify the degree of data missing in the processed data using a missing rate statistical method. The first determining submodule is used to determine the distance matrix and elevation difference between the target monitoring station and other stations using a GIS platform, and to determine the spatial topology relationship of the processed data based on the distance matrix and the elevation difference. The second determining submodule is used to determine the physical interpretability strength identifier of the processed data based on a preset meteorological and physical knowledge base. The third determining submodule is used to determine the timeliness requirement level based on the collection time of the processed data; The fourth determining submodule is used to construct the first multidimensional feature vector by considering the degree of data missing, the spatial topological relationship, the physical interpretability strength identifier, and the timeliness requirement level.

18. The system according to claim 14, characterized in that, Also includes: A building unit is used to build the multidimensional decision tree model; The establishment unit includes: The fourth acquisition module is used to clean the multi-source meteorological monitoring data from the historical monitoring stations to obtain cleaned data. The fifth acquisition module is used to extract features from the cleaned data to obtain a second multidimensional feature vector; The training module is used to train the initial decision tree model using the second multidimensional feature vector and the preset decision logic to obtain the multidimensional decision tree model. The preset decision logic includes: path direction and its corresponding path code.

19. The system according to claim 18, characterized in that, The fifth acquisition module includes: The second quantification submodule is used to quantify the degree of data missing in the cleaned data using the missing rate statistical method. The fifth determination submodule is used to determine the distance matrix and elevation difference between the historical monitoring station and other stations using the GIS platform, and to determine the spatial topology relationship of the cleaned data based on the distance matrix and the elevation difference; The sixth determination submodule is used to determine the physical interpretability strength identifier of the cleaned data based on a preset meteorological and physical knowledge base; The seventh determination submodule is used to determine the timeliness requirement level based on the collection time of the cleaned data; The eighth determining submodule is used to construct the second multidimensional feature vector based on the degree of data missing, the spatial topological relationship, the physical interpretability strength identifier, and the timeliness requirement level.

20. The system according to claim 15, characterized in that, The multidimensional decision tree model includes: root node, second-level nodes, third-level nodes, and leaf nodes; The root node represents the degree of data missing, the second-level nodes represent the spatial topological relationships, the third-level nodes represent the physical interpretability strength indicators, and the leaf nodes represent the timeliness requirement level.

21. The system according to claim 17, characterized in that, The quantization submodule is specifically used for: Calculate the missing rate of the processed data; If the missing rate is less than a first preset threshold, the data missing degree is isolated missing; if the missing rate is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, the data missing degree is partial missing. If the missing rate is greater than the second preset threshold, then the degree of data missing is continuous missing.

22. The system according to claim 17, characterized in that, The first determining submodule is specifically used for: If the distance between the target monitoring station and its nearest other station is greater than a first distance threshold, then the spatial topology is an isolated station. If the number of other stations located less than or equal to the second distance threshold from the target monitoring station is greater than or equal to a preset number, then the spatial topology is a dense monitoring network. If the standard deviation of elevation calculated with the target monitoring station as the center within the grid where the target monitoring station is located is greater than the third distance threshold, then the spatial topology is a complex terrain area.

23. The system according to claim 17, characterized in that, The second determining submodule is specifically used for: Determine whether the meteorological elements in the processed data have corresponding physical laws or equations in the preset meteorological physics knowledge base. If they do, determine the list of parameters that must be input for the calculation corresponding to the physical law or equation. If it does not exist, the physical interpretability strength is marked as weak; After determining the parameter list, it is determined whether each parameter in the parameter list is supported by a data source. If each parameter is supported by a data source, the physical interpretability strength is marked as strong; otherwise, the physical interpretability strength is marked as weak.

24. The system according to claim 17, characterized in that, The third determining submodule is specifically used for: If the collection time is less than or equal to the first time threshold, the timeliness requirement level is real-time; if the collection time is greater than the first time threshold and less than the second time threshold, the timeliness requirement level is near real-time. If the collection time is greater than or equal to the second time threshold, then the timeliness requirement level is offline.

25. The system according to claim 17, characterized in that, The correction unit includes: The scheduling module is used to schedule the correction technology configuration corresponding to the path code from a preset model library; wherein the correction technology configuration defines at least one model or method and specifies the collaboration paradigm and execution parameters between models; The correction module is used to correct the processed data using the correction technology to obtain the final corrected meteorological monitoring data.

26. The system according to claim 21, characterized in that, The formula for calculating the missing rate includes: η = N_missing / N_total In the above formula, η The missing rate, N_missing This represents the number of missing data points in the processed data. N_total This represents the total amount of data after processing.

27. An electronic device, characterized in that, include: At least one processor and memory; The memory and processor are connected via a bus; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, the adaptive correction method for new energy meteorological monitoring data as described in any one of claims 1 to 13 is implemented.

28. A readable storage medium, characterized in that, It contains an execution program, which, when executed, implements the adaptive correction method for new energy meteorological monitoring data as described in any one of claims 1 to 13.

Images