An unmanned aerial vehicle automatic control method and system based on the Internet of Things

Abstract

The application provides an unmanned aerial vehicle automatic control method and system based on the Internet of Things, and relates to the technical field of digital processing.The method comprises the following steps: obtaining flight log information; constructing a three-dimensional fitting model to generate a three-dimensional flight modeling result, and obtaining a first flight estimation result in combination with flight task execution progress; obtaining automatic control data; calling environment parameter index information; inputting target unmanned aerial vehicle setting parameters and target unmanned aerial vehicle attitude data into a flight estimation model, taking the environment parameter index information as correction data to obtain a second flight estimation result, and combining the first flight estimation result to assist in correcting the automatic control data.The technical problem that the low precision of automatic control data in the execution of unmanned aerial vehicle tasks leads to the inability to guarantee the execution efficiency of unmanned aerial vehicle tasks is solved, the precision of automatic control data in the execution of unmanned aerial vehicle tasks is improved, the automatic control data is optimized, and the execution efficiency of unmanned aerial vehicle tasks is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of digital processing technology, and more specifically to an automated control method and system for unmanned aerial vehicles (UAVs) based on the Internet of Things (IoT). Background Technology

[0002] With the continuous development of Internet of Things (IoT) technology, drones are widely used in various dangerous, simple, and repetitive scenarios. Drones can be divided into military and civilian types. Common reconnaissance drones and target drones are military drones, while drones used for aerial photography, wildlife observation, and infectious disease monitoring are civilian drones.

[0003] During drone missions, battery limitations mean that if a drone must fly against the wind, its time in the air will be even shorter. Battery life and capacity limits the need for energy replenishment during drone missions. Therefore, the actual flight route planning needs to rationally plan the energy replenishment points. Currently, drone flight route planning is unreasonable, and the route planning and mission execution progress cannot meet the needs of flight mission execution.

[0004] Existing technologies suffer from low precision in automated control data during drone missions, which leads to a failure to guarantee mission execution efficiency. Summary of the Invention

[0005] This application provides an IoT-based automated control method and system for unmanned aerial vehicles (UAVs), which solves the technical problem of low accuracy of automated control data in UAV mission execution, leading to unreliable mission execution efficiency. By combining the flight mission execution progress, it achieves the technical effect of improving the accuracy of automated control data in UAV mission execution, optimizing automated control data, and ensuring the execution efficiency of UAV missions.

[0006] In view of the above problems, this application provides an automated control method and system for unmanned aerial vehicles based on the Internet of Things.

[0007] In a first aspect, this application provides an IoT-based automated control method for unmanned aerial vehicles (UAVs). The method is applied to an automated control system for a UAV, which is communicatively connected to a data acquisition device. The method includes: acquiring flight log information via a flight data recorder of the target UAV; constructing a three-dimensional fitting model, inputting the flight log information into the three-dimensional fitting model, performing simulation modeling, and fitting to generate a three-dimensional flight modeling result; using the three-dimensional flight modeling result and combining it with the flight mission execution progress of the target UAV to make a prediction, obtaining a first flight prediction result; acquiring automated control data of the target UAV, wherein the automated control data includes target UAV setting parameters, target UAV spatial position information, and target UAV attitude data; using the data acquisition device to input the target UAV spatial position information, performing data network retrieval, and retrieving environmental parameter index information; inputting the target UAV setting parameters and the target UAV attitude data into the trained flight prediction model, using the environmental parameter index information as correction data, and obtaining a second flight prediction result; and using the first flight prediction result and the second flight prediction result to perform auxiliary correction on the automated control data of the target UAV.

[0008] Secondly, this application provides an IoT-based unmanned aerial vehicle (UAV) automated control system, wherein the system is communicatively connected to a data acquisition device, and the system includes: a log information acquisition unit, which acquires flight log information through the flight data recorder of the target UAV; a simulation modeling unit, which constructs a three-dimensional fitting model, inputs the flight log information into the three-dimensional fitting model, performs simulation modeling, and generates a three-dimensional flight modeling result; an execution progress prediction unit, which estimates the execution progress of the target UAV's flight mission based on the three-dimensional flight modeling result, and obtains a first flight prediction result; and a control data acquisition unit, which acquires the automated control data of the target UAV. The system includes: automated control data comprising target UAV setting parameters, target UAV spatial location information, and target UAV attitude data; a location information input unit, used to input the target UAV spatial location information through the data acquisition device, perform data network retrieval, and retrieve environmental parameter index information; a prediction result acquisition unit, used to input the target UAV setting parameters and target UAV attitude data into a trained flight prediction model, using the environmental parameter index information as correction data, and obtain a second flight prediction result; and an auxiliary correction unit, used to perform auxiliary correction on the automated control data of the target UAV based on the first flight prediction result and the second flight prediction result.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages: By employing a flight data recorder to acquire flight log information from the target UAV; constructing a 3D fitting model; inputting the flight log information into the 3D fitting model to generate a 3D flight modeling result; combining this with the flight mission execution progress of the target UAV to obtain a first flight prediction result; acquiring automated control data; recording spatial position information; performing data network retrieval to retrieve environmental parameter index information; inputting the target UAV's set parameters and attitude data into the trained flight prediction model; using the environmental parameter index information as correction data to obtain a second flight prediction result; and combining this with the first flight prediction result to assist in correcting the automated control data of the target UAV. This embodiment of the application, by combining flight mission execution progress, achieves the technical effect of improving the accuracy of automated control data in UAV mission execution, optimizing automated control data, and ensuring the execution efficiency of UAV missions. Attached Figure Description

[0010] Figure 1 This is a flowchart illustrating an IoT-based automated control method for unmanned aerial vehicles (UAVs) according to this application. Figure 2 This is a schematic diagram illustrating the process of obtaining model index correlation error alerts in an IoT-based drone automated control method according to this application. Figure 3 This is a schematic diagram of the structure of an IoT-based unmanned aerial vehicle (UAV) automated control system according to this application.

[0011] Figure labeling: 11 Log information acquisition unit, 12 Simulation modeling unit, 13 Execution progress estimation unit, 14 Control data acquisition unit, 15 Location information input unit, 16 Estimation result acquisition unit, 17 Auxiliary correction unit. Detailed Implementation

[0012] This application provides an IoT-based automated control method and system for unmanned aerial vehicles (UAVs), which solves the technical problem of low accuracy of automated control data in UAV mission execution, leading to unreliable mission execution efficiency. By combining the flight mission execution progress, it achieves the technical effect of improving the accuracy of automated control data in UAV mission execution, optimizing automated control data, and ensuring the execution efficiency of UAV missions.

[0013] Example 1 like Figure 1As shown, this application provides an IoT-based automated control method for unmanned aerial vehicles (UAVs), wherein the method is applied to an automated control system for UAVs, the system being communicatively connected to a data acquisition device, and the method includes: S100: Obtain flight log information through the target drone's flight data recorder; S200: Construct a three-dimensional fitting model, input the flight log information into the three-dimensional fitting model, perform simulation modeling, and fit to generate a three-dimensional flight modeling result; S300: Based on the three-dimensional flight modeling results, combined with the flight mission execution progress of the target UAV, a first flight prediction result is obtained; Specifically, the flight data recorder, also known as the black box, is used to record data information during the flight process of the target UAV. The flight log information is the data information recorded by the flight data recorder. A three-dimensional fitting model is constructed. The three-dimensional fitting model is a real-scene stereoscopic three-dimensional model. The x-axis of the three-dimensional fitting model represents latitude, the y-axis represents longitude, and the z-axis represents altitude. The flight log information is input into the three-dimensional fitting model to perform real-scene reconstruction simulation modeling and generate a three-dimensional flight modeling result. The flight mission is the UAV flight mission, which can be reconnaissance, recording data information of a certain area, or other related tasks. Based on the three-dimensional flight modeling result, the flight mission execution progress of the target UAV is estimated. If the flight mission is to record data information of a certain area, the flight mission execution progress can be determined by combining the data information collection progress of the certain area, that is, the ratio of the area where data information collection has been completed to the total area of ​​the area. By combining the three-dimensional flight modeling result with the flight mission execution progress, the flight efficiency of the target UAV is evaluated to obtain a first flight prediction result, providing data support to ensure the stability of the first flight prediction result.

[0014] Furthermore, the step S200 of constructing a three-dimensional fitting model, inputting the flight log information into the three-dimensional fitting model, performing simulation modeling, and fitting to generate a three-dimensional flight modeling result includes: S210: Obtain the historical flight missions and historical spatial location information of the target UAV through the flight log information; S220: Obtain a set of historical energy supply point location information through the historical spatial location information; S230: Through information interaction analysis, analyze the historical flight mission and the historical energy supply point location information set to reconstruct the historical flight route information; S240: Construct a three-dimensional fitting model, import the historical flight route information into the three-dimensional fitting model, perform simulation modeling, and fit to generate a three-dimensional flight modeling result.

[0015] Specifically, by retrieving signals from historical data, flight log information is retrieved to obtain the target UAV's historical flight missions and historical spatial location information. This historical spatial location information is then integrated with refueling point location information to obtain a historical refueling point location information set. The elements in this set represent multiple refueling point locations. Using the historical flight missions as the target, information interaction analysis is performed using the historical spatial location information to determine the distribution of these multiple refueling point locations within the historical spatial location information, thus reconstructing the target UAV's historical flight path information. A three-dimensional fitting model is then constructed, and the historical flight path information is imported into this model for simulation modeling to obtain three-dimensional flight modeling results. This provides technical support for ensuring the data accuracy in the three-dimensional flight modeling results.

[0016] S400: Acquire the automated control data of the target UAV, wherein the automated control data includes target UAV setting parameters, target UAV spatial position information, and target UAV attitude data; S500: The spatial location information of the target UAV is recorded through the data acquisition device, and the data network is searched to retrieve environmental parameter index information; S600: Input the target UAV setting parameters and the target UAV attitude data into the trained flight prediction model, use the environmental parameter index information as correction data, and obtain the second flight prediction result; Specifically, the data storage unit of the UAV automated control system retrieves the automated control data of the target UAV. The data acquisition device has a built-in network retrieval engine that can input the spatial location information of the target UAV and retrieve environmental parameter information through historical time points corresponding to the spatial location information of the target UAV. The environmental parameter information includes, but is not limited to, environmental temperature, environmental wind speed, and environmental wind direction. The flight prediction model is based on the BP backpropagation model. The target UAV's set parameters and attitude data are input into the input port of the flight prediction model. The environmental parameter information is used as correction data. The output of the flight prediction model and the environmental parameter information are normalized. The results obtained from the normalization process are weighted using objective weighting methods such as entropy weighting to correct the output of the flight prediction model. The result of the weighted calculation is determined as the second flight prediction result. This avoids the error in the training data of the flight prediction model, which could lead to the model extracting error indicators, causing the flight prediction model to overfit and resulting in limited output accuracy.

[0017] Furthermore, the flight prediction model includes: S610: Based on the target UAV type parameters, perform information retrieval and extract the automated control dataset. The subset of the automated control dataset includes setting information of UAVs of the same type, spatial location information of UAVs of the same type, and attitude information of UAVs of the same type. S620: Through the data acquisition device, the spatial location information of the same type of UAV is recorded, data network retrieval is performed, and environmental parameter index information of the same type is retrieved; S630: Input the setting information of the same type of UAV and the attitude information of the same type of UAV into the input training port of the bp algorithm model, use the environmental parameter index information of the same type as correction data to train the model, and construct the flight prediction model after the model output is stable.

[0018] Furthermore, the step S630, which involves training the model and constructing the flight prediction model after the model output stabilizes, further includes: S631: Based on the aforementioned automated control dataset, perform data labeling to obtain a training subset, a validation subset, and a test subset; S632: Input the training subset into the training port in sequence to train the model, and obtain the flight training model after the model training is completed; S633: The flight training model is verified and tested using the verification subset and the test subset to obtain the test error value; S634: Determine whether the test error value is less than the preset error value. If the test error value is less than the preset error value, the evaluation is passed and the flight prediction model is determined.

[0019] Specifically, using big data and targeting drone type parameters (drone model) as the retrieval criterion, information retrieval is performed to extract the automated control dataset. Through the network retrieval engine built into the data acquisition device, the spatial location of similar drones and the time information of similar drones in those spatial locations are entered for data network retrieval. This retrieves similar environmental parameter information. By performing data labeling and classification, the automated control dataset is divided into training, validation, and test subsets. The data types of the training, validation, and test subsets are consistent. The elements of the training subset are the setting parameters and attitude data of similar drones. The setting information of similar drones includes multiple setting parameters and attitude information packages. The algorithm includes attitude data from multiple similar UAVs. The settings and attitude information of these UAVs are input into the training port of the backpropagation (BP) algorithm model. Environmental parameter information of the same type is used as correction data for model training. After training, a flight training model is obtained. The flight training model is then validated and tested using the validation subset and the test subset. A test error value is obtained, which is the average error between the outputs of the validation subset and the test subset. The test error value is compared to a preset error value. If the test error value is less than the preset error value, the output of the flight prediction model meets the requirements of the UAV automated control system, the validation and evaluation are passed, and the flight prediction model is determined, providing a model basis for data processing.

[0020] Furthermore, such as Figure 2 As shown, to determine the test error value and the preset error value, step S634 further includes: S634-1: If the test error value is not less than the preset error value, determine the index error set by comparing the test error value with the preset error value, wherein the elements in the index error set are distributed in ascending order of absolute error value; S634-2: Fit index parameters based on the index error set and construct an error function; S634-3: Extract correlation features based on error function to obtain index correlation error data; S634-4: By comparing the error data associated with the indicators with the preset error value, an error assessment is performed to obtain a model indicator-associated error alert.

[0021] Specifically, the test error value is compared with a preset error value. If the test error value is not less than the preset error value, it indicates that the flight training model needs further optimization training. The validation subset and test subset are input into the flight training model one by one. A first test error value is determined by the output results of the first element of the validation subset and the first element of the test subset. This process is repeated to obtain multiple test error values. An index error set is determined by comparing these multiple test error values ​​with the preset error value. The elements in the index error set are distributed in ascending order of absolute error value, where the absolute error value is the difference between the multiple test error values ​​and the preset error value. Based on the index error set, the automated control dataset... A multi-dimensional coordinate system is established, and the indicator error set is input. By mapping it to the indicator parameters of the automated control dataset, indicator parameter error fitting analysis is performed to construct an error function. The parameter mapping distance corresponds to the correlation degree corresponding to the correlation analysis. Associated features are extracted to obtain indicator association error data, which are associated indicators with high correlation to the indicator errors. Error evaluation can be performed by combining preset error values ​​to obtain model indicator association error alerts. These alerts indicate that the model's training data is abnormal and requires further optimization of the flight training model. The model is then validated again using a validation subset and a test subset, providing technical support to ensure the rationality of the flight prediction model.

[0022] S700: The automated control data of the target UAV is corrected by using the first flight prediction result and the second flight prediction result.

[0023] Furthermore, by using the first flight prediction result and the second flight prediction result, the automated control data of the target UAV is auxiliaryly corrected. Step S700 also includes: S710: Obtain a comprehensive flight assessment result by combining the first flight prediction result and the second flight prediction result; S720: Compare the comprehensive flight evaluation results with the automated control data of the target UAV to obtain the comparison results; S730: The comparison results are sent to the control terminal of the UAV automated control system via a communication connection to assist the control administrator in correcting the automated control data.

[0024] Specifically, the first flight prediction result and the second flight prediction result are normalized. The coefficient of variation (COP) method is used to weight the normalized results. The COP method is an objective weighting method that directly utilizes the information contained in the normalized results to calculate their weights. After determining the weights, the weights of the normalized results from the first and second flight prediction results are calculated sequentially to obtain a comprehensive flight evaluation result. The comprehensive flight evaluation result is then compared with the automated control data to obtain a comparison result, which is a correlation comparison result of common indicators. The comparison result is sent to the control terminal of the UAV automated control system to assist control administrators in correcting the automated control data. This provides technical support for improving the accuracy of the UAV automated control system's control data and reduces the control complexity of the UAV automated control system.

[0025] In summary, the IoT-based automated control method and system for unmanned aerial vehicles provided in this application have the following technical advantages: This application provides an IoT-based drone automation control method and system. By combining the flight log information obtained from the target drone's flight data recorder, constructing a 3D fitting model, inputting the flight log information into the 3D fitting model to generate a 3D flight modeling result, and combining this with the target drone's flight mission execution progress to obtain a first flight prediction result, automated control data is acquired, spatial position information is entered, and environmental parameter information is retrieved through data network retrieval, and the target drone's set parameters and attitude data are input into the trained flight prediction model. The environmental parameter information is used as correction data to obtain a second flight prediction result, which, combined with the first flight prediction result, assists in correcting the target drone's automated control data. This application achieves the technical effect of improving the accuracy of automated control data in drone mission execution, optimizing automated control data, and ensuring the execution efficiency of drone missions by combining the flight mission execution progress with the data acquisition of flight log information from the target drone's flight log recorder.

[0026] By using an automated control dataset, data is labeled to obtain training, validation, and test subsets. The training subset is then input into the training port sequentially to train the model and obtain a flight training model. The flight training model is then validated and tested using the validation and test subsets to obtain test error values. If the test error value is less than a preset error value, the evaluation is passed, and the flight prediction model is determined, providing a model basis for data processing.

[0027] By employing a comprehensive flight evaluation method that combines the first and second flight prediction results with the target UAV's automated control data for comparison, and sending the comparison results to the control terminal of the UAV's automated control system, the system assists control administrators in correcting the automated control data. This provides technical support for improving the accuracy of control data in UAV automated control systems and reduces the control complexity of UAV automated control systems.

[0028] Example 2 Based on the same inventive concept as the IoT-based drone automated control method described in the foregoing embodiments, such as Figure 3 As shown, this application provides an IoT-based automated control system for unmanned aerial vehicles (UAVs), wherein the system is communicatively connected to a data acquisition device, and the system includes: Log information acquisition unit 11, the log information acquisition unit 11 is used to acquire flight log information through the flight data recorder of the target UAV; Simulation modeling unit 12 is used to construct a three-dimensional fitting model, input the flight log information into the three-dimensional fitting model, perform simulation modeling, and fit to generate a three-dimensional flight modeling result. The execution progress estimation unit 13 is used to estimate the flight mission execution progress of the target UAV by combining the three-dimensional flight modeling results with the flight mission execution progress of the target UAV, and obtain the first flight estimation result. The control data acquisition unit 14 is used to acquire the automated control data of the target UAV, wherein the automated control data includes the target UAV setting parameters, the target UAV spatial position information and the target UAV attitude data; The location information input unit 15 is used to input the spatial location information of the target UAV through the data acquisition device, perform data network retrieval, and retrieve environmental parameter index information. The prediction result acquisition unit 16 is used to input the target UAV setting parameters and the target UAV attitude data into the trained flight prediction model, use the environmental parameter index information as correction data, and obtain the second flight prediction result. The auxiliary correction unit 17 is used to perform auxiliary correction on the automated control data of the target UAV based on the first flight prediction result and the second flight prediction result.

[0029] Furthermore, the system includes: A spatial location information acquisition unit is used to acquire the historical flight missions and historical spatial location information of the target UAV through the flight log information. A location information set acquisition unit, the unit being used to acquire a historical energy supply point location information set through the historical spatial location information; An interactive analysis unit is used to analyze the historical flight mission and the historical energy resupply point location information set through information interaction analysis, and to reconstruct the historical flight route information. The three-dimensional flight modeling result fitting unit is used to construct a three-dimensional fitting model, import the historical flight route information into the three-dimensional fitting model, perform simulation modeling, and fit to generate a three-dimensional flight modeling result.

[0030] Furthermore, the system includes: An automated control dataset extraction unit is used to perform information retrieval based on the target UAV type parameters and extract an automated control dataset. A subset of the automated control dataset includes setting information of UAVs of the same type, spatial location information of UAVs of the same type, and attitude information of UAVs of the same type. The location information input unit is used to input the spatial location information of the same type of UAV through the data acquisition device, perform data network retrieval, and retrieve environmental parameter index information of the same type. The model training unit is used to input the setting information and attitude information of the same type of UAV into the input training port of the BP algorithm model, and use the environmental parameter index information of the same type as correction data to train the model. After the model output is stable, the flight prediction model is constructed.

[0031] Furthermore, the system includes: A data labeling unit is used to label data based on the automated control dataset and obtain a training subset, a validation subset, and a test subset. A training model acquisition unit is used to sequentially input the training subset into the training port for model training, and acquire the flight training model after the model training is completed. A verification and testing unit is used to verify and test the flight training model through the verification subset and the test subset, and obtain the test error value. An error value evaluation unit is used to determine whether the test error value is less than the preset error value. If the test error value is less than the preset error value, the evaluation is passed and the flight prediction model is determined.

[0032] Furthermore, the system includes: An absolute value distribution unit is used to determine an index error set by comparing the test error value with the preset error value if the test error value is not less than the preset error value. The elements in the index error set are distributed in ascending order of absolute error value. A parameter fitting unit is used to fit index parameters based on the index error set and construct an error function. A correlation feature extraction unit is used to extract correlation features based on an error function to obtain index correlation error data. An error assessment unit is used to assess errors by comparing the index-related error data with the preset error value, and to obtain model index-related error alerts.

[0033] Furthermore, the system includes: A comprehensive flight evaluation result acquisition unit is used to acquire a comprehensive flight evaluation result by combining the first flight prediction result and the second flight prediction result. A data comparison unit is used to compare the comprehensive flight evaluation result with the automated control data of the target UAV and obtain the comparison result. A data correction unit is used to send the comparison results to the control terminal of the UAV automated control system via a communication connection, so as to assist the control and management personnel in correcting the automated control data.

[0034] This specification and accompanying drawings are merely illustrative examples of this application, and various modifications and combinations can be made thereto without departing from the spirit and scope of this application. If such modifications and variations fall within the scope of the claims and their equivalents, this application intends to include these modifications and variations.

Claims

1. An automated control method for unmanned aerial vehicles (UAVs) based on the Internet of Things (IoT), characterized in that, The method is applied to an automated control system for an unmanned aerial vehicle (UAV), wherein the system is communicatively connected to a data acquisition device, and the method includes: Obtain flight log information using the target drone's flight data recorder; The flight log information is used to obtain the target UAV's historical flight missions and historical spatial location information; The historical spatial location information is used to obtain a set of historical energy supply point location information, and the elements in the set of historical energy supply point location information are multiple energy supply point location information. Using the historical flight mission as the target, information interaction analysis is performed through the historical spatial location information to determine the distribution information of the multiple energy supply point locations in the historical spatial location information, and the historical flight route information of the target UAV is restored; A three-dimensional fitting model is constructed, the historical flight route information is imported into the three-dimensional fitting model, simulation modeling is performed, and a three-dimensional flight modeling result is generated through fitting. The target UAV performs a flight mission to record data information of a certain area. The progress of the flight mission is determined by the ratio of the area where data information has been collected to the total area of ​​the area. The flight efficiency of the target UAV is evaluated by comparing the 3D flight modeling results with the flight mission execution progress, and a first flight prediction result is obtained. Acquire the automated control data of the target drone, wherein the automated control data includes the target drone's set parameters, the target drone's spatial position information, and the target drone's attitude data; The spatial location information of the target UAV is recorded through the data acquisition device, and the data is retrieved through network retrieval to obtain environmental parameter information. The target UAV's set parameters and attitude data are input into the trained flight prediction model, and the environmental parameter index information is used as correction data to obtain a second flight prediction result. A comprehensive flight assessment result is obtained by combining the first flight prediction result and the second flight prediction result; The comprehensive flight evaluation results are compared with the automated control data of the target UAV to obtain the comparison results, which are the correlation comparison results of common indicators. The comparison results are sent to the control terminal of the UAV automated control system via a communication connection to assist the control administrator in correcting the automated control data.

2. The method as described in claim 1, characterized in that, The flight prediction model is constructed in the following way: Based on the target UAV type parameters, information retrieval is performed to extract the automated control dataset. The subset of the automated control dataset includes setting information of UAVs of the same type, spatial location information of UAVs of the same type, and attitude information of UAVs of the same type. The data acquisition device is used to input the spatial location information of the same type of UAV, perform data network retrieval, and retrieve environmental parameter index information of the same type. The setting information and attitude information of the same type of UAV are input into the input training port of the BP algorithm model. The environmental parameter index information of the same type is used as correction data to train the model. After the model output is stable, the flight prediction model is constructed.

3. The method as described in claim 2, characterized in that, The flight prediction model is determined through the following steps: Based on the aforementioned automated control dataset, data labeling is performed to obtain training subsets, validation subsets, and test subsets; The training subset is sequentially input into the input training port for model training. After the model training is completed, the flight training model is obtained. The flight training model is validated and tested using the validation subset and the test subset to obtain test error values; The test error value is compared with the preset error value. If the test error value is less than the preset error value, the evaluation is passed, and the flight prediction model is determined.

4. The method as described in claim 3, characterized in that, If the test error value is not less than the preset error value, the method further includes: The test error value and the preset error value are used to determine the index error set, and the elements in the index error set are distributed in order of increasing absolute error value; Based on the aforementioned index error set, index parameters are fitted to construct an error function; Based on the error function, correlation feature extraction is performed to obtain index correlation error data; Error assessment is performed by comparing the error data associated with the indicators with the preset error values ​​to obtain model indicator-related error alerts.

5. An automated control system for unmanned aerial vehicles (UAVs) based on the Internet of Things (IoT), characterized in that: The system is communicatively connected to a data acquisition device, and the system includes: A log information acquisition unit is used to acquire flight log information through the flight data recorder of the target UAV; A spatial location information acquisition unit is used to acquire the historical flight missions and historical spatial location information of the target UAV through the flight log information. A location information set acquisition unit is used to acquire a historical energy supply point location information set through the historical spatial location information, wherein the elements in the historical energy supply point location information set are multiple energy supply point location information. An interactive analysis unit is used to perform information interactive analysis on the historical flight mission as the target, through the historical spatial location information, to determine the distribution information of the multiple energy supply point locations in the historical spatial location information, and to restore the historical flight route information of the target UAV. A three-dimensional flight modeling result fitting unit is used to construct a three-dimensional fitting model, import the historical flight route information into the three-dimensional fitting model, perform simulation modeling, and fit to generate a three-dimensional flight modeling result. The execution progress estimation unit is used to determine the flight mission execution progress based on the ratio of the area where data information has been collected to the total area of ​​the area when the flight mission performed by the target UAV is to record data information of a certain area. The execution progress estimation unit is used to evaluate the flight efficiency of the target UAV by comparing the three-dimensional flight modeling results with the flight mission execution progress, and obtain a first flight prediction result. A control data acquisition unit is used to acquire the automated control data of the target UAV, wherein the automated control data includes target UAV setting parameters, target UAV spatial position information, and target UAV attitude data; The location information input unit is used to input the spatial location information of the target UAV through the data acquisition device, perform data network retrieval, and retrieve environmental parameter index information. The prediction result acquisition unit is used to input the target UAV setting parameters and the target UAV attitude data into the trained flight prediction model, and use the environmental parameter index information as correction data to obtain a second flight prediction result. A comprehensive flight evaluation result acquisition unit is used to acquire a comprehensive flight evaluation result by combining the first flight prediction result and the second flight prediction result. A data comparison unit is used to compare the comprehensive flight evaluation results with the automated control data of the target UAV to obtain comparison results, wherein the comparison results are the correlation comparison results of common indicators. A data correction unit is used to send the comparison results to the control terminal of the UAV automated control system via a communication connection, so as to assist the control administrator in correcting the automated control data.

Images