Confidence evaluation and integration method of meteorological prediction model for low-altitude air route risk control

Abstract

The application relates to a meteorological prediction model confidence evaluation and integration method for low-altitude air route risk control, which is applied to the technical field of meteorological prediction and comprises the following steps: acquiring prediction weather data, a meteorological prediction model and real weather data; generating prediction wind speed data and prediction precipitation data from the prediction weather data based on wind speed and precipitation; respectively matching the prediction wind speed data and the prediction precipitation data with the real weather data based on preset standards to generate wind speed data matching results and precipitation data matching results; respectively performing performance ranking on the meteorological prediction model based on the wind speed data matching results and the precipitation data matching results to generate a wind speed model evaluation list and a precipitation model evaluation list; selecting models in the wind speed model evaluation list and the precipitation model evaluation list as to-be-used wind speed models and to-be-used precipitation models; and constructing a target weather prediction model based on the to-be-used wind speed models and the to-be-used precipitation models. The application has the effect of improving the accuracy of meteorological prediction.

Description

Technical Field

[0001] This application relates to the technical field of meteorological forecasting, and in particular to a confidence assessment and integration method for meteorological forecasting models for low-altitude airway risk control. Background Technology

[0002] With the large-scale application of unmanned transport aircraft, the risk of unmanned transport aircraft crashing due to severe weather such as strong winds and heavy rain has become a core bottleneck for commercial promotion.

[0003] Currently, most meteorological forecasting models still have a basic accuracy rate of 90%-95%, but they all suffer from problems such as model simplification and lack of integrated optimization. The shortcomings of simplistic models are that some models have systematic biases, such as overestimating / underestimating wind speed, or missing risks, such as failing to identify rainfall, making it difficult to meet the high standards of air route safety. The lack of integrated optimization is due to the fact that existing technologies do not have a differentiated assessment system for wind speed and rainfall, and directly splicing models can easily lead to insufficient accuracy in identifying dangerous areas.

[0004] Therefore, there is an urgent need for a technical means to accurately predict weather conditions. Summary of the Invention

[0005] To improve the accuracy of weather forecasts, this application provides a method for confidence assessment and integration of weather forecast models for low-altitude airway risk control.

[0006] Firstly, this application provides a method for confidence assessment and integration of meteorological prediction models for low-altitude airway risk control, employing the following technical solution: A method for confidence assessment and integration of meteorological forecasting models for low-altitude airway risk control, comprising: Acquire forecast weather data, provide meteorological forecasting models for the forecast weather data, and provide real weather data; The predicted weather data is divided into subsets based on wind speed and precipitation to generate predicted wind speed data and predicted precipitation data. Based on preset standards, the predicted wind speed data and the predicted precipitation data are matched with the actual weather data to generate wind speed data matching results and precipitation data matching results. Based on the matching results of the wind speed data and the matching results of the precipitation data, the meteorological prediction models are ranked according to their performance, and a wind speed model evaluation list and a precipitation model evaluation list are generated. A predetermined number of models are selected from the wind speed model evaluation list and the precipitation model evaluation list as the wind speed model and precipitation model to be used; A target weather prediction model is constructed based on the wind speed model and precipitation model to be used.

[0007] By adopting the above technical solution, we first acquire multi-model predicted weather data, corresponding meteorological prediction models, and real weather data. Then, based on wind speed and precipitation, we divide the predicted weather data into predicted wind speed data and predicted precipitation data. Subsequently, we complete the accurate matching of the two types of predicted data with real weather data according to the preset standards of spatiotemporal synchronization to ensure the consistency of the evaluation benchmark. Then, we use specific performance indicators for wind speed and precipitation to rank the meteorological prediction models, generating accurate evaluation lists for wind speed models and precipitation models. Finally, we select a preset number of high-quality models as models to be used, and ultimately construct the target weather forecasting system. The meteorological forecasting model effectively addresses the systematic biases, missed reporting risks, and lack of targeted integration and optimization inherent in existing single meteorological forecasting models. Through factor-based differentiated evaluation, it achieves accurate identification of the wind speed and precipitation forecasting performance of different models. By leveraging the integration and optimization of superior models, it fully integrates the forecasting advantages of high-quality models, compensates for the performance shortcomings of single models, significantly reduces the bias in wind speed forecasting and the risk of missed reporting in precipitation forecasting, and ultimately improves the accuracy of meteorological forecasting. This provides highly reliable meteorological data support for the route planning of low-altitude unmanned transport aircraft, effectively ensuring route safety.

[0008] Optionally, the step of partitioning the predicted weather data into subsets based on wind speed and precipitation to generate predicted wind speed data and predicted precipitation data includes: Extract common data, wind speed data, and precipitation data from the predicted weather data; The common data is bound and integrated with the wind speed data to generate predicted wind speed data; The common data is bound and integrated with the precipitation data to generate predicted precipitation data.

[0009] Optionally, the step of matching the predicted wind speed data and the predicted precipitation data with the actual weather data based on preset standards to generate wind speed data matching results and precipitation data matching results includes: Based on the common data and the wind speed data, the real weather data is extracted to obtain the real wind speed data; Based on the preset standard, the predicted wind speed data is matched one-to-one with the actual wind speed data to generate wind speed data matching results. Based on the common data and the precipitation data, the real weather data is extracted to obtain the real precipitation data; Based on the preset standard, the predicted precipitation data is matched one-to-one with the actual precipitation data to generate precipitation data matching results.

[0010] Optionally, the step of ranking the meteorological prediction models based on the wind speed data matching results and the precipitation data matching results, and generating wind speed model evaluation lists and precipitation model evaluation lists, includes: Obtain wind speed assessment indicators and precipitation assessment indicators; Based on the wind speed evaluation index, calculate the performance data of each meteorological prediction model in the wind speed data matching result, and generate the wind speed model performance value. The meteorological prediction models are sorted according to the wind speed performance values ​​to generate a wind speed model evaluation list; Construct confusion matrix parameters based on the precipitation data matching results; Based on the precipitation assessment index and the confusion matrix parameters, the performance data of each meteorological prediction model in the precipitation data matching results are calculated to generate precipitation model performance values; The meteorological prediction models are sorted according to their precipitation performance values ​​to generate a precipitation model evaluation list.

[0011] Optionally, selecting a preset number of models from the wind speed model evaluation list and the precipitation model evaluation list as the wind speed model and precipitation model to be used includes: Based on preset optimization requirements, key indicators are selected for the wind speed assessment index and the precipitation assessment index to obtain the target wind speed index and the target precipitation index. Select a preset number of models corresponding to the target wind speed index from the wind speed model list as the wind speed models to be used. A preset number of models corresponding to the target precipitation index are selected from the precipitation model list as wind speed models to be used.

[0012] Optionally, the target weather prediction model includes a target wind speed model and a target precipitation model; the construction of the target weather prediction model based on the wind speed model and the precipitation model to be used includes: Obtain the wind speed data value of the target wind speed index in the wind speed model to be used and the precipitation data value of the target precipitation index in the precipitation model to be used; Calculate the average value of the wind speed data and the average value of the precipitation data respectively to obtain the wind speed prediction value and the precipitation prediction value; Define a wind speed model to be trained based on the wind speed model to be used; Define a precipitation model to be trained based on the precipitation model to be used; The wind speed model to be trained is trained based on the wind speed prediction value to generate the target wind speed model; The precipitation model to be trained is trained based on the precipitation forecast values ​​to generate the target precipitation model.

[0013] Optionally, after constructing the target weather prediction model based on the wind speed model and the precipitation model to be used, the method further includes: Based on the target wind speed model and the wind speed model to be used, wind speed prediction is performed, and wind speed prediction results are generated. Precipitation prediction is performed based on the target precipitation model and the precipitation model to be used, and precipitation prediction results are generated. Obtain the predicted date and the corresponding actual data for the predicted date; A comparison result of the wind speed model is generated based on the real data and the wind speed prediction results. A comparison result of precipitation models is generated based on the real data and the precipitation prediction results.

[0014] Secondly, this application provides a confidence assessment and integration device for meteorological prediction models for low-altitude airway risk control, employing the following technical solution: A confidence assessment and integration device for meteorological forecasting models for low-altitude airway risk control, comprising: The weather data acquisition module is used to acquire predicted weather data, provide meteorological prediction models for the predicted weather data, and provide real weather data. The prediction data partitioning module is used to partition the predicted weather data into subsets based on wind speed and precipitation, generating predicted wind speed data and predicted precipitation data. The matching result generation module is used to match the predicted wind speed data and the predicted precipitation data with the real weather data based on preset standards, and generate wind speed data matching results and precipitation data matching results. The evaluation list generation module is used to rank the performance of the meteorological prediction models based on the matching results of the wind speed data and the matching results of the precipitation data, and generate the wind speed model evaluation list and the precipitation model evaluation list respectively. The standby model selection module is used to select a preset number of models from the wind speed model evaluation list and the precipitation model evaluation list as the wind speed model and precipitation model to be used. The weather model building module is used to build a target weather prediction model based on the wind speed model to be used and the precipitation model to be used.

[0015] By adopting the above technical solution, we first acquire multi-model predicted weather data, corresponding meteorological prediction models, and real weather data. Then, based on wind speed and precipitation, we divide the predicted weather data into predicted wind speed data and predicted precipitation data. Subsequently, we complete the accurate matching of the two types of predicted data with real weather data according to the preset standards of spatiotemporal synchronization to ensure the consistency of the evaluation benchmark. Then, we use specific performance indicators for wind speed and precipitation to rank the meteorological prediction models, generating accurate evaluation lists for wind speed models and precipitation models. Finally, we select a preset number of high-quality models as models to be used, and ultimately construct the target weather forecasting system. The meteorological forecasting model effectively addresses the systematic biases, missed reporting risks, and lack of targeted integration and optimization inherent in existing single meteorological forecasting models. Through factor-based differentiated evaluation, it achieves accurate identification of the wind speed and precipitation forecasting performance of different models. By leveraging the integration and optimization of superior models, it fully integrates the forecasting advantages of high-quality models, compensates for the performance shortcomings of single models, significantly reduces the bias in wind speed forecasting and the risk of missed reporting in precipitation forecasting, and ultimately improves the accuracy of meteorological forecasting. This provides highly reliable meteorological data support for the route planning of low-altitude unmanned transport aircraft, effectively ensuring route safety.

[0016] Thirdly, this application provides an electronic device that adopts the following technical solution: An electronic device includes a processor coupled to a memory; The processor is used to execute a computer program stored in the memory, so that the electronic device executes the computer program of the method for confidence assessment and integration of meteorological prediction models for low-altitude airway risk control as described in any one of the first aspects.

[0017] Fourthly, this application provides a computer-readable storage medium, which adopts the following technical solution: A computer-readable storage medium storing a computer program capable of being loaded by a processor and executing the method for assessing and integrating confidence levels of meteorological forecasting models for low-altitude airway risk control as described in any one of the first aspects.

[0018] In summary, this application includes at least one of the following beneficial technical effects: First, multi-model weather forecast data, corresponding meteorological forecast models, and real weather data are acquired. Then, based on wind speed and precipitation, the forecast weather data is divided into forecast wind speed data and forecast precipitation data. Subsequently, according to a preset standard of spatiotemporal synchronization, the two types of forecast data are accurately matched with real weather data to ensure the consistency of the evaluation benchmark. Then, specific performance indicators for wind speed and precipitation are used to rank the meteorological forecast models, generating accurate evaluation lists for wind speed models and precipitation models. A preset number of high-quality models are selected as models to be used, and finally, the target weather forecast model is constructed. This process effectively solves the problems of systematic bias, missed reporting risk, and lack of targeted integration and optimization in existing single meteorological forecast models. Through factor-based differentiated evaluation, the performance of different models in wind speed and precipitation forecasts is accurately identified. By leveraging the integration and optimization of superior models, the predictive advantages of high-quality models are fully integrated, making up for the performance shortcomings of single models, significantly reducing the bias of wind speed forecasts and the risk of missed reporting in precipitation forecasts, and ultimately improving the accuracy of meteorological forecasts. This provides highly reliable meteorological data support for low-altitude unmanned transport aircraft route planning and effectively ensures route safety. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating a method for assessing and integrating the confidence level of a meteorological forecasting model for low-altitude airway risk control, as provided in an embodiment of this application.

[0020] Figure 2 This is a structural block diagram of a meteorological prediction model confidence assessment and integration device for low-altitude airway risk control provided in an embodiment of this application.

[0021] Figure 3 This is a structural block diagram of the electronic device provided in the embodiments of this application. Detailed Implementation

[0022] The present application will be further described in detail below with reference to the accompanying drawings.

[0023] This application provides a method for confidence assessment and integration of meteorological forecasting models for low-altitude airway risk control. This method can be executed by an electronic device, which can be a server or a terminal device. The server can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services. The terminal device can be a smartphone, tablet, desktop computer, etc., but is not limited to these.

[0024] Figure 1This is a flowchart illustrating a method for assessing and integrating the confidence level of a meteorological forecasting model for low-altitude airway risk control, as provided in an embodiment of this application.

[0025] like Figure 1 As shown, the main process of this method is described below (steps S101 to S106): Step S101: Obtain forecast weather data, provide a meteorological forecast model for the forecast weather data, and provide real weather data.

[0026] In this embodiment, weather data predicted by meteorological forecasting models and actual weather data are collected to obtain predicted weather data and actual weather data. The actual weather forecasting models used are also determined. There are ten meteorological forecasting models in total; therefore, the predicted weather data consists of the predicted values ​​given by these ten models, with a data size of 20,763,720. 6 10, meaning each prediction model corresponds to 20,763,720 rows. The data consists of 6 columns, representing 10 models. The predicted weather data includes x-axis coordinates, y-axis coordinates, date number, forecast time (hours), wind speed, rainfall, and model number. The x-axis and y-axis coordinates represent the coordinates of a point on the map, indicating where the weather forecast is performed. For each point, ten weather prediction models are used, resulting in ten sets of predicted data including wind speed and precipitation. Similarly, the actual weather data consists of real-world weather values. The data size is 20,763,720. 6, meaning each detection position corresponds to 20,763,720 rows. The data consists of 6 columns. The predicted weather data includes x-axis coordinates, y-axis coordinates, date number, forecast time (hour), wind speed, and rainfall. The x-axis and y-axis coordinates represent the coordinates of a point on the map, i.e., meteorological monitoring is performed at that point. During the forecasting process, monitoring is performed at each point to obtain monitoring data including wind speed and precipitation.

[0027] Step S102: Based on wind speed and precipitation, the predicted weather data is divided into subsets to generate predicted wind speed data and predicted precipitation data.

[0028] For step S102, common data, wind speed data, and precipitation data are extracted from the predicted weather data; the common data and wind speed data are bound and integrated to generate predicted wind speed data; the common data and precipitation data are bound and integrated to generate predicted precipitation data.

[0029] In this embodiment, the overall predicted weather data, which includes wind speed data and precipitation data, is subset-divided. That is, the wind speed data and precipitation data are split into predicted wind speed data used only to display wind speed and predicted precipitation data used only to display precipitation. In the process of splitting, the predicted weather data is first split, and common data such as x-axis coordinates, y-axis coordinates, date number, forecast time (hour), and model number are extracted, as well as separate wind speed data and separate precipitation data. The wind speed data and precipitation data are then bound to the common data to obtain the predicted wind speed data and predicted precipitation data.

[0030] Step S103: Based on preset standards, the predicted wind speed data and predicted precipitation data are matched with the actual weather data to generate wind speed data matching results and precipitation data matching results.

[0031] For step S103, real weather data is extracted based on common data and wind speed data to obtain real wind speed data; predicted wind speed data and real wind speed data are matched one-to-one according to preset standards to generate wind speed data matching results; real weather data is extracted based on common data and precipitation data to obtain real precipitation data; predicted precipitation data and real precipitation data are matched one-to-one according to preset standards to generate precipitation data matching results.

[0032] In this embodiment, when matching predicted wind speed data and predicted precipitation data with actual weather data, it is also necessary to perform data classification and extraction processing on the actual weather data. Since the actual weather data does not include the model number, the common data used in this step does not include the model number. That is, the common data and wind speed data are used to extract data from the actual weather data to obtain the actual wind speed data used for matching with the predicted wind speed data. Similarly, the common data and precipitation data are used to extract data from the actual weather data to obtain the actual precipitation data used for matching with the predicted precipitation data. Then, according to preset standards, namely x-axis coordinates, y-axis coordinates, date number, and forecast time, the predicted and actual data are matched to ensure that the comparison benchmark is consistent, and the final wind speed data matching result and precipitation data matching result are obtained.

[0033] Step S104: Based on the matching results of wind speed data and precipitation data, the meteorological prediction models are ranked according to their performance, and a wind speed model evaluation list and a precipitation model evaluation list are generated.

[0034] For step S104, wind speed assessment indicators and precipitation assessment indicators are obtained; based on the wind speed assessment indicators, the performance data of each meteorological prediction model in the wind speed data matching results is calculated to generate wind speed model performance values; the meteorological prediction models are sorted according to their wind speed performance values ​​to generate a wind speed model evaluation list; confusion matrix parameters are constructed based on the precipitation data matching results; based on the precipitation assessment indicators and confusion matrix parameters, the performance data of each meteorological prediction model in the precipitation data matching results is calculated to generate precipitation model performance values; the meteorological prediction models are sorted according to their precipitation performance values ​​to generate a precipitation model evaluation list.

[0035] In this embodiment, when conducting the evaluation, wind speed and precipitation have different evaluation indicators. The wind speed evaluation indicators include mean absolute error (MAE), root mean square error (RMSE), bias, and coefficient of determination (R²). When performing calculations, the mean absolute error (MAE) is calculated as: MAE = Σ|predicted value - actual value| / n, which measures the average absolute deviation between the predicted and actual values. The smaller the value, the more accurate the prediction. The root mean square error (RMSE) is calculated as: RMSE = √(Σ(predicted value - actual value)² / n), which penalizes larger prediction errors and is suitable for measuring overall prediction accuracy. The bias is calculated as: Bias = Σ(predicted value - actual value) / n, which measures the systematic bias of the prediction. A positive value indicates that the prediction is too high, and a negative value indicates that the prediction is too low. The coefficient of determination (R²) is calculated as: R² = [nΣxy - ΣxΣy]² / [(nΣx² - (Σx)²)(nΣy² - (Σy)²)], which measures the model's ability to explain data variation. The range is 0-1, and the closer to 1, the better the model fit. After the calculations are completed, four sets of wind speed model performance values ​​are obtained. The meteorological prediction models corresponding to the coefficient of determination are arranged in descending order, and the meteorological prediction models corresponding to other indicators are arranged in ascending order, thus obtaining a list of wind speed model evaluations.

[0036] Precipitation evaluation metrics include accuracy, precision, recall, F1 score, overall false alarm rate, false positive rate, and false negative rate. During processing, a confusion matrix parameter is first constructed based on the precipitation data matching results.

[0037] Confusion matrix parameters: TP (TruePositive): True positive - both actual and predicted rain; TN (True Negative): True negative - no rain actually occurred and no rain is predicted; FP (False Positive): False positive - no rain actually falls but rain is predicted; FN (False Negative): False negative - rain actually occurred but no rain was predicted; Next, the performance data of each meteorological prediction model in the precipitation data matching results were calculated using precipitation assessment indices and confusion matrix parameters, generating precipitation model performance values. In the calculations, accuracy was calculated as: Accuracy = (TP + TN) / (TP + TN + FP + FN), which is the proportion of correct predictions out of all predictions, measuring overall prediction performance; precision was calculated as: Precision = TP / (TP + FP), which is the proportion of predicted rainfall that actually occurred, measuring the reliability of the rainfall prediction; recall was calculated as: Recall = TP / (TP + FN), which is the proportion of actual rainfall that was correctly predicted, measuring the completeness of the rainfall prediction; and the F1 score was calculated as: F1 = 2 × (Precision × Recall) / (Precision + Recall), the harmonic mean of precision and recall, comprehensively evaluates classification performance. The overall false alarm rate (FAR) is calculated as: FAR = (FP + FN) / (TP + TN + FP + FN), which represents the proportion of incorrect predictions among all forecasts, reflecting the reliability of the prediction. The false positive rate / false alarm rate (FPR) is calculated as: FPR = FP / (FP + TN), which represents the proportion of predictions indicating rain when there is no rain, measuring the false warning rate. The false negative rate / missed prediction rate (FNR) is calculated as: FNR = FN / (TP + FN), which represents the proportion of predictions indicating no rain when there is rain, measuring the risk of missed predictions. The overall false alarm rate, false positive rate / false alarm rate, and false negative rate / missed prediction rate are sorted in ascending order of performance value, and other evaluation indicators are sorted in descending order of performance value to generate the final precipitation model evaluation list. Step S105: Select a preset number of models from the wind speed model evaluation list and the precipitation model evaluation list as the wind speed model and precipitation model to be used.

[0038] For step S105, key indicators are selected for wind speed assessment indicators and precipitation assessment indicators based on preset optimization requirements to obtain target wind speed indicators and target precipitation indicators; a preset number of models corresponding to the target wind speed indicators are selected from the wind speed model list as wind speed models to be used; a preset number of models corresponding to the target precipitation indicators are selected from the precipitation model list as wind speed models to be used.

[0039] In this embodiment, for the wind speed model, in practical applications, the deviation of predicted wind speed is of greater concern. Therefore, the mean absolute error, root mean square error, and coefficient of determination are taken as the target wind speed indicators, and the top three models in the wind speed model list corresponding to these three indicators are selected as the wind speed models to be used. For the precipitation model, in practical applications, the underreporting rate is of greater concern. The underreporting rate is taken as the target precipitation indicator, and the top three models in the precipitation model list corresponding to the underreporting rate are selected as the wind speed models to be used.

[0040] Step S106: Construct a target weather prediction model based on the wind speed model and precipitation model to be used.

[0041] For step S106, the target weather prediction model includes a target wind speed model and a target precipitation model; obtain the wind speed data values ​​of the target wind speed index in the wind speed model to be used and the precipitation data values ​​of the target precipitation index in the precipitation model to be used; calculate the average value of the wind speed data values ​​and the average value of the precipitation data values ​​respectively to obtain the wind speed prediction value and the precipitation prediction value; define the wind speed model to be trained based on the wind speed model to be used; define the precipitation model to be trained based on the precipitation model to be used; train the wind speed model to be trained based on the wind speed prediction value to generate the target wind speed model; train the precipitation model to be trained based on the precipitation prediction value to generate the target precipitation model.

[0042] In this embodiment, wind speed data corresponding to the target wind speed indicators in the selected wind speed models are extracted, and the average value of each target wind speed indicator is calculated to obtain the predicted wind speed value. Similarly, precipitation data corresponding to the target precipitation indicators in the selected precipitation models are extracted, and the average value of the target precipitation indicators is calculated to obtain the predicted precipitation value. A wind speed model to be trained is defined using the wind speed models to be used, and a precipitation model to be trained is defined using the precipitation models to be used. The predicted wind speed values ​​are used as the predicted values ​​for training the wind speed models to generate the target wind speed model. The predicted precipitation values ​​are used as the predicted values ​​for training the precipitation models to generate the target precipitation model.

[0043] It should be noted that for wind speed models, these metrics are commonly used to measure the difference between predicted and actual values, and are used to measure the overall accuracy of a prediction model in fitting real data. In practical applications, wind speed is more weighted towards bias, i.e., BIAS. Therefore, when defining a new model, the BIAS metric should be ignored initially, and the performance of 10 models should be measured using the other three metrics. Based on the actual calculation results, the average of the predicted values ​​of the top three models is taken to obtain a prediction value, which is defined as the prediction value of the wind speed model to be trained. Then, the BIAS metric is used to compare all models, including the newly defined model. According to the calculation results, the newly defined prediction value is found to perform best in BIAS and also best in the calculation of the other three metrics, thus completing the training of the new model and obtaining the target wind speed model. The precipitation model is similar, and will not be elaborated further here.

[0044] In this embodiment, wind speed prediction is performed based on the target wind speed model and the wind speed model to be used, and wind speed prediction results are generated; precipitation prediction is performed based on the target precipitation model and the precipitation model to be used, and precipitation prediction results are generated; the prediction date and the corresponding real data are obtained; wind speed model comparison results are generated based on the real data and the wind speed prediction results; and precipitation model comparison results are generated based on the real data and the precipitation prediction results.

[0045] After the model is built, two new models are used to make weather predictions, which yield wind speed and precipitation predictions. The two predictions are then compared with the actual data to generate two comparison results. Based on the comparison results, the most accurate wind speed and precipitation prediction data can be determined.

[0046] Figure 2 The structural block diagram of a meteorological prediction model confidence assessment and integration device 200 for low-altitude airway risk control provided in the application embodiment.

[0047] like Figure 2 As shown, the meteorological forecasting model confidence assessment and integration device 200 for low-altitude airway risk control mainly includes: The weather data acquisition module 201 is used to acquire forecast weather data, provide meteorological prediction models for forecast weather data, and real weather data. The prediction data partitioning module 202 is used to partition the predicted weather data into subsets based on wind speed and precipitation, and generate predicted wind speed data and predicted precipitation data. The matching result generation module 203 is used to match the predicted wind speed data and predicted precipitation data with the real weather data based on preset standards, and generate wind speed data matching results and precipitation data matching results. The evaluation list generation module 204 is used to rank the performance of meteorological prediction models based on the matching results of wind speed data and precipitation data, and generate evaluation lists for wind speed models and precipitation models respectively. The standby model selection module 205 is used to select a preset number of models from the wind speed model evaluation list and the precipitation model evaluation list as standby wind speed models and standby precipitation models. Weather model building module 206 is used to build a target weather prediction model based on the wind speed model and precipitation model to be used.

[0048] As an optional implementation of this embodiment, the prediction data segmentation module 202 is specifically used to extract common data, wind speed data and precipitation data from the prediction weather data; bind and integrate the common data and wind speed data to generate prediction wind speed data; bind and integrate the common data and precipitation data to generate prediction precipitation data.

[0049] As an optional implementation of this embodiment, the matching result generation module 203 is specifically used to extract data from real weather data based on common data and wind speed data to obtain real wind speed data; to perform one-to-one matching between predicted wind speed data and real wind speed data based on preset standards to generate wind speed data matching results; to extract data from real weather data based on common data and precipitation data to obtain real precipitation data; and to perform one-to-one matching between predicted precipitation data and real precipitation data based on preset standards to generate precipitation data matching results.

[0050] As an optional implementation of this embodiment, the evaluation list generation module 204 is specifically used to obtain wind speed evaluation indicators and precipitation evaluation indicators; calculate the performance data of each meteorological prediction model in the wind speed data matching results based on the wind speed evaluation indicators, and generate wind speed model performance values; sort the meteorological prediction models according to their wind speed performance values ​​to generate a wind speed model evaluation list; construct confusion matrix parameters based on the precipitation data matching results; calculate the performance data of each meteorological prediction model in the precipitation data matching results based on the precipitation evaluation indicators and confusion matrix parameters, and generate precipitation model performance values; sort the meteorological prediction models according to their precipitation performance values ​​to generate a precipitation model evaluation list.

[0051] As an optional implementation of this embodiment, the candidate model selection module 205 is specifically used to select key indicators for wind speed evaluation indicators and precipitation evaluation indicators based on preset optimization requirements, so as to obtain target wind speed indicators and target precipitation indicators; select a preset number of models corresponding to the target wind speed indicators from the wind speed model list as candidate wind speed models; and select a preset number of models corresponding to the target precipitation indicators from the precipitation model list as candidate wind speed models.

[0052] As an optional implementation of this embodiment, the weather model construction module 206 is specifically used to obtain the wind speed data value of the target wind speed index in the wind speed model to be used and the precipitation data value of the target precipitation index in the precipitation model to be used; calculate the average value of the wind speed data value and the average value of the precipitation data value respectively to obtain the wind speed prediction value and the precipitation prediction value; define the wind speed model to be trained based on the wind speed model to be used; define the precipitation model to be trained based on the precipitation model to be used; train the wind speed model to be trained based on the wind speed prediction value to generate the target wind speed model; and train the precipitation model to be trained based on the precipitation prediction value to generate the target precipitation model.

[0053] As an optional implementation of this embodiment, the meteorological prediction model evaluation and integration device 200 for low-altitude airway risk control further includes: The wind speed prediction generation module is used to predict wind speed based on the target wind speed model and the wind speed model to be used, and generate wind speed prediction results. The precipitation prediction generation module is used to predict precipitation based on the target precipitation model and the precipitation model to be used, and generate precipitation prediction results. The date data acquisition module is used to acquire the predicted date and the corresponding actual data. The wind speed comparison generation module is used to generate wind speed model comparison results based on real data and wind speed prediction results; The precipitation comparison generation module is used to generate precipitation model comparison results based on real data and precipitation prediction results.

[0054] In one example, the module in any of the above devices may be one or more integrated circuits configured to implement the above methods, such as one or more application-specific integrated circuits (ASICs), or one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs), or a combination of at least two of these integrated circuit forms.

[0055] For example, when modules in a device can be implemented via a processing element scheduler, the processing element can be a general-purpose processor, such as a central processing unit (CPU) or other processor capable of calling programs. Alternatively, these modules can be integrated together as a system-on-a-chip (SOC).

[0056] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described device and module can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0057] Figure 3 This is a structural block diagram of the electronic device 300 provided in an embodiment of this application.

[0058] like Figure 3 As shown, the electronic device 300 includes a processor 301 and a memory 302, and may further include one or more of an information input / output (I / O) interface 303, a communication component 304, and a communication bus 305.

[0059] The processor 301 controls the overall operation of the electronic device 300 to complete all or part of the steps of the aforementioned method for assessing and integrating the confidence level of a meteorological prediction model for low-altitude airway risk control. The memory 302 stores various types of data to support the operation of the electronic device 300. This data may include, for example, instructions for any application or method operating on the electronic device 300, as well as application-related data. The memory 302 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as one or more of Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0060] I / O interface 303 provides an interface between processor 301 and other interface modules, such as keyboards, mice, and buttons. These buttons can be virtual or physical. Communication component 304 is used for wired or wireless communication between electronic device 300 and other devices. Wireless communication includes Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, or 4G, or a combination thereof. Therefore, the corresponding communication component 304 may include a Wi-Fi component, a Bluetooth component, and an NFC component.

[0061] The electronic device 300 can be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to execute the confidence assessment and integration method for meteorological prediction models for low-altitude airway risk control given in the above embodiments.

[0062] The communication bus 305 may include a path for transmitting information between the aforementioned components. The communication bus 305 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The communication bus 305 can be divided into an address bus, a data bus, a control bus, etc.

[0063] Electronic device 300 may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), and in-vehicle terminals (such as in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers, and may also be servers.

[0064] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described method for assessing and integrating the confidence level of a meteorological prediction model for low-altitude airway risk control.

[0065] The computer-readable storage medium may include various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0066] The terms “comprising,” “including,” or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0067] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the foregoing application concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions claimed in this application.

Claims

1. A method for confidence assessment and integration of meteorological forecasting models for low-altitude airway risk control, characterized in that, include: Acquire forecast weather data, provide meteorological forecasting models for the forecast weather data, and provide real weather data; The predicted weather data is divided into subsets based on wind speed and precipitation to generate predicted wind speed data and predicted precipitation data. Based on preset standards, the predicted wind speed data and the predicted precipitation data are matched with the actual weather data to generate wind speed data matching results and precipitation data matching results. Based on the matching results of the wind speed data and the matching results of the precipitation data, the meteorological prediction models are ranked according to their performance, and a wind speed model evaluation list and a precipitation model evaluation list are generated. A predetermined number of models are selected from the wind speed model evaluation list and the precipitation model evaluation list as the wind speed model and precipitation model to be used; A target weather prediction model is constructed based on the wind speed model and precipitation model to be used.

2. The method according to claim 1, characterized in that, The step of partitioning the predicted weather data into subsets based on wind speed and precipitation to generate predicted wind speed data and predicted precipitation data includes: Extract common data, wind speed data, and precipitation data from the predicted weather data; The common data is bound and integrated with the wind speed data to generate predicted wind speed data; The common data is bound and integrated with the precipitation data to generate predicted precipitation data.

3. The method according to claim 2, characterized in that, The process of matching the predicted wind speed data and the predicted precipitation data with the actual weather data based on preset standards to generate wind speed data matching results and precipitation data matching results includes: Based on the common data and the wind speed data, the real weather data is extracted to obtain the real wind speed data; Based on the preset standard, the predicted wind speed data is matched one-to-one with the actual wind speed data to generate wind speed data matching results. Based on the common data and the precipitation data, the real weather data is extracted to obtain the real precipitation data; Based on the preset standard, the predicted precipitation data is matched one-to-one with the actual precipitation data to generate precipitation data matching results.

4. The method according to claim 1, characterized in that, The process of ranking the meteorological prediction models based on the matching results of the wind speed data and the matching results of the precipitation data, and generating evaluation lists for wind speed models and precipitation models, includes: Obtain wind speed assessment indicators and precipitation assessment indicators; Based on the wind speed evaluation index, calculate the performance data of each meteorological prediction model in the wind speed data matching result, and generate the wind speed model performance value. The meteorological prediction models are sorted according to the wind speed performance values ​​to generate a wind speed model evaluation list; Construct confusion matrix parameters based on the precipitation data matching results; Based on the precipitation assessment index and the confusion matrix parameters, the performance data of each meteorological prediction model in the precipitation data matching results are calculated to generate precipitation model performance values; The meteorological prediction models are sorted according to their precipitation performance values ​​to generate a precipitation model evaluation list.

5. The method according to claim 4, characterized in that, The step of selecting a preset number of models from the wind speed model evaluation list and the precipitation model evaluation list as the wind speed model and precipitation model to be used includes: Based on preset optimization requirements, key indicators are selected for the wind speed assessment index and the precipitation assessment index to obtain the target wind speed index and the target precipitation index. Select a preset number of models corresponding to the target wind speed index from the wind speed model list as the wind speed models to be used. A preset number of models corresponding to the target precipitation index are selected from the precipitation model list as wind speed models to be used.

6. The method according to claim 5, characterized in that, The target weather prediction model includes a target wind speed model and a target precipitation model; the construction of the target weather prediction model based on the wind speed model and precipitation model to be used includes: Obtain the wind speed data value of the target wind speed index in the wind speed model to be used and the precipitation data value of the target precipitation index in the precipitation model to be used; Calculate the average value of the wind speed data and the average value of the precipitation data respectively to obtain the wind speed prediction value and the precipitation prediction value; Define the wind speed model to be trained based on the wind speed model to be used; Based on the precipitation model to be used, define the precipitation model to be trained; The wind speed model to be trained is trained based on the wind speed prediction value to generate the target wind speed model; The precipitation model to be trained is trained based on the precipitation forecast values ​​to generate the target precipitation model.

7. The method according to claim 6, characterized in that, After constructing the target weather prediction model based on the wind speed model and the precipitation model to be used, the method further includes: Based on the target wind speed model and the wind speed model to be used, wind speed prediction is performed, and wind speed prediction results are generated. Precipitation prediction is performed based on the target precipitation model and the precipitation model to be used, and precipitation prediction results are generated. Obtain the predicted date and the corresponding actual data for the predicted date; A comparison result of the wind speed model is generated based on the real data and the wind speed prediction results. A comparison result of precipitation models is generated based on the real data and the precipitation prediction results.

8. A confidence assessment and integration device for meteorological forecasting models for low-altitude airway risk control, characterized in that, include: The weather data acquisition module is used to acquire predicted weather data, provide meteorological prediction models for the predicted weather data, and provide real weather data. The prediction data partitioning module is used to partition the predicted weather data into subsets based on wind speed and precipitation, generating predicted wind speed data and predicted precipitation data. The matching result generation module is used to match the predicted wind speed data and the predicted precipitation data with the real weather data based on preset standards, and generate wind speed data matching results and precipitation data matching results. The evaluation list generation module is used to rank the performance of the meteorological prediction models based on the matching results of the wind speed data and the matching results of the precipitation data, and generate the wind speed model evaluation list and the precipitation model evaluation list respectively. The standby model selection module is used to select a preset number of models from the wind speed model evaluation list and the precipitation model evaluation list as the wind speed model and precipitation model to be used. The weather model building module is used to build a target weather prediction model based on the wind speed model to be used and the precipitation model to be used.

9. An electronic device, characterized in that, Includes a processor, which is coupled to a memory; The processor is configured to execute a computer program stored in the memory, causing the electronic device to perform the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It includes a computer program or instructions that, when run on a computer, cause the computer to perform the method as described in any one of claims 1 to 7.

Images