A method and system for managing training information of unmanned aerial vehicle operators

Abstract

The present invention relates to the technical field of drone operator training data processing. The present invention provides a drone operator training information management method and system, which solves the problem of low accuracy of periodic change analysis in drone operator training data processing. The present invention acquires and analyzes drone flight data and personnel operation records, and combines historical data to build and optimize a time series time model to achieve dynamic simulation and prediction of the training process. The present invention can track training results in real time, generate personalized training plans, and dynamically adjust to adapt to the progress of trainees. The present invention also provides a drone operator training information management system including a server end, a background control end, a drone end, and a user mobile terminal. Through data classification, feature extraction, model construction, and error analysis, training items are generated, which effectively improves training efficiency and effect and optimizes resource allocation.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle operator training data processing, and in particular to a method and system for managing unmanned aerial vehicle operator training information. Background Art

[0002] In the UAV flight training scenario, the digital management method of UAV operator training information is a comprehensive strategy aimed at improving training efficiency, optimizing resource allocation, and real-time tracking and analysis of training results. Using advanced information technology, an integrated UAV flight training management platform is built, which covers the entire process of training plan formulation, execution, monitoring, evaluation and feedback.

[0003] The UAV flight training management platform has functions such as training course management, student information management, flight data recording and analysis, online teaching videos and simulation training. Through the built-in sensors and flight control system of the UAV, it collects flight data in real time including flight altitude, speed, attitude and route, and records students' operating behavior, reaction speed, error rate and other indicators during training.

[0004] The UAV flight training management platform uses big data analysis and artificial intelligence algorithms to conduct in-depth mining and analysis of collected flight data and student performance data. The UAV flight training management platform evaluates the training results of students through data analysis, identifies the strengths and weaknesses of students, and provides a basis for personalized training plans. According to the evaluation results, timely feedback is provided to students, and the training plan and teaching methods are adjusted to improve the training effect. According to the actual situation of the students and the evaluation results, personalized training plans are formulated to meet the different needs of students. During the training process, according to the progress and feedback of the students, the training plan is dynamically adjusted to ensure the achievement of training goals.

[0005] In order to improve the predictability of training results, the UAV flight training management platform can dynamically simulate and predict the training process of trainees by building a time series model. The prediction capability enables the platform to identify challenges and bottlenecks that trainees may encounter in advance, so as to adjust the training plan in a targeted manner to avoid problems and improve the overall training effect. Building a time series model requires processing and analyzing a large amount of time series data. The processing and analysis process is affected by many factors, such as training intensity, trainee status, and external environment. The influence of multiple factors leads to a decrease in the accuracy of analyzing the periodic changes of trainees in UAV flight training. Summary of the invention

[0006] In view of the shortcomings of the prior art, the present invention provides a method and system for managing training information of unmanned aerial vehicle operators, which solves the problem that building a time series model requires processing and analyzing a large amount of time series data, and the processing and analysis process is affected by various factors, such as training intensity, student status, external environment, etc. The influence of various factors leads to reduced accuracy in analyzing the periodic changes of students in unmanned aerial vehicle flight training.

[0007] In order to solve the above technical problems, the specific technical solutions of the present invention are as follows:

[0008] In a first aspect, the present invention provides a method for managing training information of unmanned aerial vehicle operators, comprising:

[0009] Step S101, obtaining UAV flight data and personnel training operation records, the UAV flight data includes flight altitude, speed, attitude, and flight trajectory, and the personnel training operation records include personnel training operation behavior, personnel training reaction speed, and personnel training error rate;

[0010] Step S102, classifying the UAV flight data and the personnel training operation records, obtaining the classification results of the UAV flight data and the personnel training operation records, establishing a data processing partition according to the classification results of the UAV flight data and the personnel training operation records, establishing a UAV operator training information management information processing mirror system on the cloud server, pre-processing the corresponding UAV flight data and the personnel training operation records in the data processing partition and extracting data features, and obtaining flight training stability features, operation accuracy features, and reaction speed features;

[0011] Step S103, retrieve the unmanned operation historical data, construct a first unmanned aerial vehicle personnel flight training time series model based on linear regression through the unmanned aerial vehicle historical data and flight training stability characteristics, operation accuracy characteristics and reaction speed characteristics, and optimize the first unmanned aerial vehicle personnel flight training time series model based on linear regression in real time, use the unmanned aerial vehicle flight data and personnel training operation records in the unmanned aerial vehicle operator training information management information processing mirror system to train the random forest model to obtain a second unmanned aerial vehicle personnel flight training time series model, optimize the second unmanned aerial vehicle personnel flight training time series model in real time in the unmanned aerial vehicle operator training information management information processing mirror system, compare the first unmanned aerial vehicle personnel flight training time series model based on linear regression with the second unmanned aerial vehicle personnel flight training time series model for real-time data processing results, and obtain a data comparison result. If the data comparison result is within a preset error range, the first unmanned aerial vehicle personnel flight training time series model based on linear regression takes effect;

[0012] Step S104, substituting the UAV flight data and the personnel training operation records into the first UAV personnel flight training time series model based on linear regression, obtaining the periodic change results of the UAV personnel flight training, collecting the real-time information of the UAV flight training, comparing the real-time information of the UAV flight training with the periodic change results of the UAV personnel flight training, and obtaining the real-time UAV personnel flight training error;

[0013] Step S105, analyzing the real-time UAV personnel flight training error to obtain the time information corresponding to the real-time UAV personnel flight training error, establishing a corresponding relationship between the time information corresponding to the real-time UAV personnel flight training error and the UAV flight training subject, and generating UAV flight training items.

[0014] Furthermore, the method for managing training information of drone operators provided by the present invention includes, in step S101:

[0015] Detect the integrity of UAV flight data and personnel training operation records. The detection content includes flight altitude, speed, attitude, flight trajectory, personnel training operation behavior, personnel training reaction speed and personnel training error rate, and obtain the original data detection results. The original data detection results include qualified detection data and missing data information. The data types of the original data detection results are compared with the historical data of UAV flight data and the historical data of personnel training operation records. If the data type comparison is consistent, the UAV flight data and personnel training operations match the historical UAV flight training subjects. If the data type comparison is inconsistent, retrieve the data with differences between the original data detection results and the historical data of UAV flight data and the historical data of personnel training operation records, and identify the data types to obtain the difference UAV flight training subject information;

[0016] A new data detection category is established based on the differential UAV flight training subject information, and the data corresponding to the new data detection category and the data corresponding to the original data detection result are substituted into the preset digital twin model to run, and the historical digital twin image of the UAV flight training personnel operation is obtained;

[0017] Obtain the coordinate information of the drone flight in the historical digital twin images of the drone flight training personnel's operations, and use the known coordinates of the drone flight training basic points to verify the accuracy of the coordinate information of the drone flight in the historical digital twin images of the drone flight training personnel's operations. If the coordinate information of the drone flight in the historical digital twin images of the drone flight training personnel's operations is verified, the drone flight data and personnel training operation records will take effect and be marked as usable data.

[0018] Furthermore, the method for managing the training information of drone operators provided by the present invention includes, in step S102:

[0019] Verify whether the UAV flight data and personnel training operation records have been classified according to the predetermined classification standards, including data type, training subject and timestamp, and pre-process the classified data, including data cleaning, data standardization and data conversion;

[0020] Receive the demand for training effect evaluation, determine the type of features to be extracted based on the demand for training effect evaluation, and use the feature time series analysis method to extract the selected features from the preprocessed data;

[0021] Use historical data or known training effect evaluation results to test the effectiveness of the extracted features. According to the test results, adjust the feature extraction strategy. Adjusting the feature extraction strategy includes modifying the feature type and optimizing the feature extraction algorithm.

[0022] Furthermore, the method for managing the training information of the drone operator provided by the present invention includes, in step S103:

[0023] The real-time acquired UAV flight data and the personnel training operation records are simultaneously input into a first UAV personnel flight training sequential time model and a second UAV personnel flight training sequential time model based on linear regression;

[0024] The first UAV personnel flight training time series time model and the second UAV personnel flight training time series time model based on linear regression respectively make predictions based on the input data to generate corresponding flight training periodic change results, and the corresponding flight training periodic change results include the first flight training periodic change results and the second flight training periodic change results;

[0025] Comparing the first flight training periodic change result and the second flight training periodic change result, and calculating an error between the first flight training periodic change result and the second flight training periodic change result;

[0026] An error index is defined to quantify the difference between the periodic change result of the first flight training and the periodic change result of the second flight training, and the data processing result error of the first UAV personnel flight training time series time model and the second UAV personnel flight training time series time model based on real-time linear regression is calculated, and the data processing result error is compared with a preset error range;

[0027] If the error of the real-time data processing result is within the preset error range, it is determined that the first UAV personnel flight training time series time model based on linear regression is effective;

[0028] If the error exceeds the preset range, the first UAV personnel flight training sequential time model and the second UAV personnel flight training sequential time model based on linear regression are optimized, and the above steps are repeated for verification.

[0029] Furthermore, the method for managing the training information of the drone operator provided by the present invention includes, in step S104:

[0030] Before comparing the real-time information of UAV flight training with the periodic change results of UAV personnel flight training, establish parallel data processing groups, establish data group correspondence between the real-time information of UAV flight training and the periodic change results of UAV personnel flight training and the data processing parallel groups, establish data group processing tasks based on the data group correspondence, and execute the data processing tasks to perform group and parallel preprocessing on the real-time information of UAV flight training and the periodic change results of UAV personnel flight training.

[0031] In the second aspect, a drone operator training information management system is provided, which applies the drone operator training information management method as described above, including: a server end, a background control end, a drone end and a user mobile terminal, wherein the server end establishes communication connections with the background control end, the drone end and the user mobile terminal respectively, and the server end analyzes and processes the data generated by the background control end, the drone end and the user mobile terminal, generates drone flight training items and transmits them to the background control end, the drone end and the user mobile terminal, and the server end includes:

[0032] A data acquisition unit is used to acquire UAV flight data and personnel training operation records. The UAV flight data includes flight altitude, speed, attitude, and flight trajectory. The personnel training operation records include personnel training operation behavior, personnel training reaction speed, and personnel training error rate.

[0033] A data classification unit is used to classify the UAV flight data and the personnel training operation records, obtain the classification results of the UAV flight data and the personnel training operation records, establish a data processing partition according to the classification results of the UAV flight data and the personnel training operation records, establish a UAV operator training information management information processing mirror system on the cloud server, pre-process the corresponding UAV flight data and personnel training operation records in the data processing partition and extract data features to obtain flight training stability features, operation accuracy features and reaction speed features;

[0034] A model building unit is used to retrieve unmanned operation historical data, construct a first unmanned aerial vehicle personnel flight training time series time model based on linear regression through unmanned aerial vehicle historical data and flight training stability characteristics, operation accuracy characteristics and reaction speed characteristics, and optimize the first unmanned aerial vehicle personnel flight training time series time model based on linear regression in real time, train a random forest model using unmanned aerial vehicle flight data and personnel training operation records in an unmanned aerial vehicle operator training information management information processing mirror system to obtain a second unmanned aerial vehicle personnel flight training time series time model, optimize the second unmanned aerial vehicle personnel flight training time series time model in real time in an unmanned aerial vehicle operator training information management information processing mirror system, compare the first unmanned aerial vehicle personnel flight training time series time model based on linear regression with the second unmanned aerial vehicle personnel flight training time series time model for real-time data processing results, and obtain a data comparison result. If the data comparison result is within a preset error range, the first unmanned aerial vehicle personnel flight training time series time model based on linear regression takes effect;

[0035] An error generating unit is used to substitute the UAV flight data and the personnel training operation records into a first UAV personnel flight training time series time model based on linear regression, obtain the periodic change result of the UAV personnel flight training, collect the real-time information of the UAV flight training, compare the real-time information of the UAV flight training with the periodic change result of the UAV personnel flight training, and obtain the real-time UAV personnel flight training error;

[0036] The result generating unit is used to analyze the real-time UAV personnel flight training error, obtain the time information corresponding to the real-time UAV personnel flight training error, establish a corresponding relationship between the time information corresponding to the real-time UAV personnel flight training error and the UAV flight training subject, and generate the UAV flight training items to be trained.

[0037] Furthermore, the drone operator training information management system and data acquisition unit provided by the present invention are also used for:

[0038] The training information data detection unit detects the integrity of the UAV flight data and the personnel training operation records. The detection content includes flight altitude, speed, attitude, flight trajectory, personnel training operation behavior, personnel training reaction speed and personnel training error rate, and obtains the original data detection result. The original data detection result includes qualified detection data and missing data information. The original data detection result is compared with the historical data of the UAV flight data and the historical data of the personnel training operation record. If the data type comparison is consistent, the UAV flight data and the personnel training operation match the historical UAV flight training subjects. If the data type comparison is inconsistent, the original data detection result and the historical data of the UAV flight data and the historical data of the personnel training operation record are retrieved and the data type is identified to obtain the difference UAV flight training subject information;

[0039] The digital image simulation unit establishes new data detection categories according to the differential UAV flight training subject information, substitutes the data corresponding to the new data detection categories and the data corresponding to the original data detection results into the preset digital twin model for operation, and obtains the historical digital twin images of the UAV flight training personnel's operation;

[0040] The data logic detection unit obtains the coordinate information of the drone flight in the historical digital twin image of the drone flight training personnel's operation, and uses the known coordinates of the drone flight training basic points to verify the accuracy of the coordinate information of the drone flight in the historical digital twin image of the drone flight training personnel's operation. If the coordinate information of the drone flight in the historical digital twin image of the drone flight training personnel's operation is verified, the drone flight data and personnel training operation records will take effect and be marked as usable data.

[0041] Furthermore, the data classification unit of the drone operator training information management system provided by the present invention is also used for:

[0042] The data feature verification unit verifies whether the UAV flight data and personnel training operation records have been classified according to the predetermined classification standards, including data type, training subject and timestamp, and pre-processes the classified data, including data cleaning, data standardization and data conversion;

[0043] The data feature extraction unit receives the demand for training effect evaluation, determines the type of features to be extracted according to the demand for training effect evaluation, and extracts the selected features from the preprocessed data using a feature time series analysis method;

[0044] The data feature verification unit uses historical data or known training effect evaluation results to verify the validity of the extracted features. According to the test results, the feature extraction strategy is adjusted. The adjustment of the feature extraction strategy includes modifying the feature type and optimizing the feature extraction algorithm.

[0045] Furthermore, the UAV operator training information management system and the model building unit provided by the present invention are also used for:

[0046] A data output unit, which simultaneously inputs the real-time acquired UAV flight data and the personnel training operation records into a first UAV personnel flight training sequential time model and a second UAV personnel flight training sequential time model based on linear regression;

[0047] A data prediction unit, which predicts the first UAV personnel flight training time series time model and the second UAV personnel flight training time series time model based on linear regression based on the input data respectively, and generates corresponding flight training periodic change results, wherein the corresponding flight training periodic change results include the first flight training periodic change results and the second flight training periodic change results;

[0048] A real-time data comparison unit compares the first flight training periodic change result and the second flight training periodic change result, and calculates an error between the first flight training periodic change result and the second flight training periodic change result;

[0049] An error definition unit defines an error index for quantifying the difference between the first flight training periodic change result and the second flight training periodic change result, calculating the data processing result error of the first UAV personnel flight training time series time model and the second UAV personnel flight training time series time model based on real-time linear regression, and comparing the data processing result error with a preset error range;

[0050] The model error judgment unit determines that the first UAV personnel flight training timing model based on linear regression is effective if the error of the real-time data processing result is within a preset error range; if the error exceeds the preset range, the first UAV personnel flight training timing model and the second UAV personnel flight training timing model based on linear regression are optimized, and the above steps are repeated for verification.

[0051] Furthermore, the result generating unit of the UAV operator training information management system provided by the present invention is also used for:

[0052] The data parallel processing grouping unit establishes data processing parallel groups before comparing the real-time information of UAV flight training with the periodic change results of UAV personnel flight training, establishes a data grouping correspondence between the real-time information of UAV flight training and the periodic change results of UAV personnel flight training and the data processing parallel groups, establishes data grouping processing tasks based on the data grouping correspondence, and executes the data processing tasks to perform grouping and parallel preprocessing on the real-time information of UAV flight training and the periodic change results of UAV personnel flight training.

[0053] Beneficial effects of the present invention: By constructing an integrated UAV flight training management platform, the present invention realizes the full process automation management of training plan formulation, execution, monitoring, evaluation and feedback, and significantly improves the training efficiency of UAV operators. Using advanced information technology, UAV flight data and trainee operation records are collected and analyzed in real time, and training plans and teaching methods are intelligently adjusted according to the analysis results to ensure that resources are allocated reasonably and efficiently.

[0054] By collecting drone flight data and personnel training operation records in real time, the present invention can instantly track and analyze the training results of trainees, helping the coaching team and trainees to understand the training progress and existing problems in a timely manner. By constructing a time series time model, the present invention can dynamically simulate and predict the training process of trainees, identify challenges and bottlenecks that trainees may encounter in advance, and thus adjust the training plan in a targeted manner to avoid problems and improve the overall training effect.

[0055] In terms of data processing, the present invention uses technical means such as data preprocessing, feature extraction, and parallel processing to improve the accuracy and efficiency of data processing, and provide a solid data foundation for subsequent model construction and data analysis. Through the communication connection between the server side, the background control side, the drone side, and the user's mobile terminal, the present invention realizes information sharing and collaborative work, allowing the coaching team, trainees, and managers to obtain the latest training information and data anytime and anywhere, and jointly promote the smooth development of training work.

[0056] To sum up, the beneficial effects of the present invention are not only reflected in improving the training efficiency and effectiveness of UAV operators, but also in comprehensively improving the overall level of UAV flight training by optimizing resource allocation, real-time tracking and analysis of training results, improving data processing accuracy and efficiency, realizing personalized training plans, and supporting dynamic adjustment of training plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0058] Figure 1 A schematic diagram of a method flow chart provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described in conjunction with the specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. The technical solutions provided by the embodiments of the present invention are described in detail below in conjunction with the drawings.

[0060] In order to better understand the purpose of the present invention, the present invention is described in further detail below.

[0061] First, see Figure 1 The present invention provides a method for managing training information of unmanned aerial vehicle operators, comprising:

[0062] Step S101, obtaining UAV flight data and personnel training operation records, the UAV flight data includes flight altitude, speed, attitude, and flight trajectory, and the personnel training operation records include personnel training operation behavior, personnel training reaction speed, and personnel training error rate;

[0063] Flight altitude: records the altitude of the drone during flight, which is one of the important indicators for evaluating drone performance and students' control ability.

[0064] Speed: Record the flight speed of the drone, including horizontal speed and vertical speed.

[0065] Attitude: Record the flight attitude of the drone, such as pitch angle, yaw angle, roll angle, etc.

[0066] Flight trajectory: Record the actual flight path of the drone, including the starting point, end point, and key points along the way.

[0067] Personnel training operation behavior: Detailed records of trainees’ specific operation behaviors during the process of controlling the drone, such as remote control operation, software interface operation, etc., will help understand the trainees’ operation habits and skill levels.

[0068] Personnel training reaction speed: evaluates trainees’ reaction speed when receiving instructions or encountering emergencies.

[0069] Personnel training error rate: records the number and types of errors made by trainees during training.

[0070] By comprehensively acquiring UAV flight data and personnel training operation records, the present invention provides a comprehensive and accurate information basis for subsequent data analysis and model building, which not only helps to evaluate the training effect of trainees, but also provides strong support for the formulation of personalized training plans. At the same time, through in-depth analysis and mining of these data, the present invention can also discover problems and challenges in the training process, providing a reference for the improvement and optimization of training methods.

[0071] Step S102, classifying the UAV flight data and the personnel training operation records, obtaining the classification results of the UAV flight data and the personnel training operation records, establishing a data processing partition according to the classification results of the UAV flight data and the personnel training operation records, establishing a UAV operator training information management information processing mirror system on the cloud server, pre-processing the corresponding UAV flight data and the personnel training operation records in the data processing partition and extracting data features, and obtaining flight training stability features, operation accuracy features, and reaction speed features;

[0072] In step S102, the present invention classifies and processes the UAV flight data and personnel training operation records in detail to improve the accuracy and efficiency of subsequent analysis and model building. The specific execution process of step S102 is as follows:

[0073] Classify the UAV flight data and personnel training operation records. The classification criteria may include data type (such as sensor data, operation logs, etc.), timestamp (specific time of recording data), training subjects (different flight training projects), etc. This step helps to organize a large amount of disorganized data into a structured form for subsequent processing.

[0074] Based on the classification results, different data processing partitions are established. Each partition is responsible for processing data of a specific type or a specific time period. The partition processing method can be carried out in parallel to improve the speed and efficiency of data processing.

[0075] A mirror system for information management and processing of drone operator training information is established on a cloud server. The mirror system can provide a stable and scalable computing environment to support large-scale data processing and model training. By establishing a mirror system, the present invention improves the consistency and repeatability of data processing.

[0076] In each data processing partition, the UAV flight data and personnel training operation records are preprocessed. The preprocessing steps may include data cleaning (removing outliers, duplicate values, etc.), data standardization (scaling the data to the same scale), and data conversion (converting the raw data into a format suitable for analysis).

[0077] Extract key features from the preprocessed data. In the present invention, flight training stability features, operation accuracy features and reaction speed features are mainly extracted. These features directly reflect the performance and ability level of trainees in flight training. The feature extraction method may include feature selection (selecting the most representative features from the original features) and feature construction (generating new features based on the original features).

[0078] In step S102, data preprocessing includes data cleaning, data standardization and data conversion. In the data standardization step, the Z-score standardization method is used to scale the data to the same level.

[0079] The Z-score normalization formula is:

[0080] Step S103, retrieve the unmanned operation historical data, construct a first unmanned aerial vehicle personnel flight training time series model based on linear regression through the unmanned aerial vehicle historical data and flight training stability characteristics, operation accuracy characteristics and reaction speed characteristics, and optimize the first unmanned aerial vehicle personnel flight training time series model based on linear regression in real time, use the unmanned aerial vehicle flight data and personnel training operation records in the unmanned aerial vehicle operator training information management information processing mirror system to train the random forest model to obtain a second unmanned aerial vehicle personnel flight training time series model, optimize the second unmanned aerial vehicle personnel flight training time series model in real time in the unmanned aerial vehicle operator training information management information processing mirror system, compare the first unmanned aerial vehicle personnel flight training time series model based on linear regression with the second unmanned aerial vehicle personnel flight training time series model for real-time data processing results, and obtain a data comparison result. If the data comparison result is within a preset error range, the first unmanned aerial vehicle personnel flight training time series model based on linear regression takes effect;

[0081] The present invention adopts a variety of technical means to construct and optimize the UAV personnel flight training time series time model, so that the UAV personnel flight training time series time model can accurately reflect the training effect and periodic changes of the trainees. The following is the detailed execution process of this step:

[0082] Retrieve historical data of UAV operation: retrieve a large amount of historical data of UAV operation from the database, including various records and performances of trainees in UAV flight training in the past.

[0083] Constructing the first UAV personnel flight training time series temporal model based on linear regression: Using UAV historical data and the flight training stability characteristics, operation accuracy characteristics, and reaction speed characteristics extracted from UAV flight data and personnel training operation records, the first UAV personnel flight training time series temporal model based on linear regression is constructed through time series analysis, aiming to capture the periodic changes in the training effect of trainees.

[0084] Real-time optimization of the first UAV personnel flight training sequential time model based on linear regression: During and after the model construction process, the first UAV personnel flight training sequential time model based on linear regression is continuously optimized in real time, including adjusting model parameters according to newly collected data, and using technical means such as online learning to enable the model to adapt to data changes and maintain accuracy and effectiveness.

[0085] Training the second UAV personnel flight training time series temporal model: In the UAV operator training information management information processing mirror system, the random forest model is trained using new UAV flight data and personnel training operation records. Random forest is an ensemble learning method that builds multiple decision trees and aggregates their results to make a final prediction. This step aims to obtain a different but equally effective prediction model than the first UAV personnel flight training time series temporal model based on linear regression.

[0086] Real-time optimization of the second UAV flight training time series model: Similar to the first UAV flight training time series model based on linear regression, the second UAV flight training time series model also needs to be optimized in real time to adapt to data changes. By continuously receiving new training data and updating model parameters, the second UAV flight training time series model can accurately reflect the training status of the trainees.

[0087] Comparison of real-time data processing results of the two models: Comparison of the real-time data processing results of the first UAV personnel flight training sequential time model based on linear regression and the second UAV personnel flight training sequential time model involves quantitative evaluation of the prediction results of the two models and calculation of the difference (i.e., error) between them.

[0088] Model validity judgment: If the real-time data processing results of the first UAV personnel flight training time series model based on linear regression are similar to the results of the second UAV personnel flight training time series model within the preset error range, the first UAV personnel flight training time series model based on linear regression is considered to be valid and reliable under the current data conditions. At this time, the first UAV personnel flight training time series model based on linear regression is effective and will be used for subsequent prediction and evaluation work.

[0089] Through the dual model construction, real-time optimization and result comparison process in step S103, the present invention not only improves the robustness and accuracy of model construction, but also further improves the reliability of the first UAV personnel flight training time series model based on linear regression by introducing the second UAV personnel flight training time series model as a verification benchmark. This dual model strategy effectively reduces the risk of deviation that may be caused by a single model and provides strong support for the accurate evaluation of the training effect of trainees.

[0090] In step S103, a time series time model is constructed by using the drone historical data and the extracted features. A linear regression model is used as the first drone personnel flight training time series time model based on linear regression, and its formula is expressed as:

[0091] The linear regression model formula is:

[0092] in, is the predicted result of the periodic change of flight training, is the intercept term, , , are the coefficients of stability, accuracy and reaction speed characteristics, respectively. , , are the extracted features respectively.

[0093] For the random forest model (second UAV personnel flight training time series model), it is an integrated learning method that does not rely on a single mathematical formula, but can integrate the prediction results of multiple decision trees by voting or averaging.

[0094] Step S104, substituting the UAV flight data and the personnel training operation records into the first UAV personnel flight training time series model based on linear regression, obtaining the periodic change results of the UAV personnel flight training, collecting the real-time information of the UAV flight training, comparing the real-time information of the UAV flight training with the periodic change results of the UAV personnel flight training, and obtaining the real-time UAV personnel flight training error;

[0095] In step S104, the present invention realizes real-time monitoring and evaluation of the flight training effect of UAV personnel through data processing and comparative analysis. The following is the detailed execution process of this step:

[0096] The newly collected UAV flight data and personnel training operation records are substituted into the first UAV personnel flight training time series model based on linear regression that has been in effect. The first UAV personnel flight training time series model based on linear regression is based on a large amount of historical data and real-time optimization, and can accurately reflect the periodic changes of trainees in flight training.

[0097] By calculating the first UAV personnel flight training time series model based on linear regression, the periodic change results of UAV personnel flight training are obtained. The periodic change results of UAV personnel flight training are a time series prediction, which shows the performance trends of trainees in different training stages, including changes in stability, accuracy, reaction speed, etc.

[0098] During the actual training process, the drone’s built-in sensors and control systems are used to collect flight training data in real time, including key indicators such as flight altitude, speed, attitude, trajectory, as well as the trainees’ operating behavior, reaction speed, and error rate.

[0099] The collected real-time information of UAV flight training is compared and analyzed with the periodic change results predicted by the first UAV personnel flight training time series model based on linear regression. By comparing the difference between the two, the gap between the actual performance of the trainees in the current training stage and the expected goal can be intuitively evaluated.

[0100] According to the results of comparative analysis, the real-time UAV personnel flight training error is calculated. The real-time UAV personnel flight training error value quantifies the difference between the actual performance of the trainees and the prediction of the first UAV personnel flight training time series time model based on linear regression, which is one of the important indicators for evaluating the training effect of trainees.

[0101] Through the execution of step S104, the present invention can grasp the performance and progress of the trainees in the UAV flight training in real time. When it is found that the training effect of the trainees is greatly deviated from the expectation or there is an obvious downward trend, measures can be taken immediately to intervene and adjust so that the training goal can be achieved smoothly. At the same time, these real-time error data also provide strong support for the subsequent optimization of training plans and the formulation of personalized training programs.

[0102] Step S105, analyzing the real-time UAV personnel flight training error to obtain the time information corresponding to the real-time UAV personnel flight training error, establishing a corresponding relationship between the time information corresponding to the real-time UAV personnel flight training error and the UAV flight training subject, and generating UAV flight training items.

[0103] In step S105, the present invention achieves accurate positioning of the training bottlenecks and weak links of trainees by analyzing the real-time UAV flight training errors, and further generates targeted UAV flight training items. The specific execution process of this step is as follows:

[0104] The real-time UAV personnel flight training error calculated in step S104 is analyzed in detail, including the size, trend and cause of the error. Through this step, it can be preliminarily determined which training links the trainees have problems or deficiencies.

[0105] Time information extraction: Extract the corresponding time information from the error data. The time information marks the specific moment or training stage when the error occurred, which helps to clarify when and where the trainees experienced poor training results.

[0106] Establishing the correspondence between subjects: Establishing the correspondence between the extracted time information and the UAV flight training subjects. This step is based on the training plan and the time arrangement of the actual training process to determine the training subject that the trainee is currently conducting or has completed at the time when the error occurs.

[0107] Generation of training items: Generate UAV flight training items based on the results of error analysis and time information. These items are customized for the deficiencies and weaknesses of trainees in specific training subjects, aiming to help trainees make up for their knowledge or skills deficiencies and improve the overall training effect.

[0108] Project feedback and adjustment: The generated training projects will be fed back to the relevant trainees or coaching team in a timely manner. At the same time, according to the actual feedback from the trainees and the progress of training, the content and difficulty of the training projects will be dynamically adjusted to improve the scientificity and effectiveness of the training plan.

[0109] Through the execution of step S105, the present invention not only realizes the accurate positioning and cause analysis of the trainee's training errors, but also provides a personalized improvement path for the trainees by generating targeted training items. This closed-loop training optimization mechanism helps to continuously improve the training effect and practical ability of the trainees, and provides a strong guarantee for the professional training of drone operators.

[0110] Specifically, the method for managing the training information of drone operators provided by the present invention includes, in step S101:

[0111] Detect the integrity of UAV flight data and personnel training operation records. The detection content includes flight altitude, speed, attitude, flight trajectory, personnel training operation behavior, personnel training reaction speed and personnel training error rate, and obtain the original data detection results. The original data detection results include qualified detection data and missing data information. The data types of the original data detection results are compared with the historical data of UAV flight data and the historical data of personnel training operation records. If the data type comparison is consistent, the UAV flight data and personnel training operations match the historical UAV flight training subjects. If the data type comparison is inconsistent, retrieve the data with differences between the original data detection results and the historical data of UAV flight data and the historical data of personnel training operation records, and identify the data types to obtain the difference UAV flight training subject information;

[0112] A new data detection category is established based on the differential UAV flight training subject information, and the data corresponding to the new data detection category and the data corresponding to the original data detection result are substituted into the preset digital twin model to run, and the historical digital twin image of the UAV flight training personnel operation is obtained;

[0113] Obtain the coordinate information of the drone flight in the historical digital twin images of the drone flight training personnel's operations, and use the known coordinates of the drone flight training basic points to verify the accuracy of the coordinate information of the drone flight in the historical digital twin images of the drone flight training personnel's operations. If the coordinate information of the drone flight in the historical digital twin images of the drone flight training personnel's operations is verified, the drone flight data and personnel training operation records will take effect and be marked as usable data.

[0114] Data integrity test content: including UAV flight data such as flight altitude, speed, attitude, flight trajectory, as well as personnel training operation records such as personnel training operation behavior, reaction speed, error rate, etc.

[0115] Purpose of detection: To ensure the collected data is complete and without omissions, and to lay the foundation for subsequent data processing and analysis.

[0116] After the test is completed, the results are divided into two parts: qualified test data and missing data information. Missing data are marked for possible subsequent data supplementation or special processing. The data types of the original data test results are compared with the historical data of drone flight data and personnel training operation records. Verify the consistency of current data and historical data to ensure data continuity and comparability.

[0117] If the data types are compared and consistent, it is considered that the UAV flight data and personnel training operations match the historical UAV flight training subjects and can be used directly for subsequent processing.

[0118] If the data types are inconsistent, further data will be retrieved to identify the data types of the discrepancies between the original data detection results and the historical data of the UAV flight data and the historical data of the personnel training operation records, so as to clarify the specific content and reasons for the differences.

[0119] Identify the differences and obtain the difference UAV flight training subject information. Establish a new data detection category based on the difference information to classify and process these special or abnormal data. Substitute the data corresponding to the new data detection category and the data corresponding to the original data detection result into the preset digital twin model for operation. The digital twin model is a virtualized simulation model that can simulate the actual scene and operation process of UAV flight training. Through the operation of the model, the historical digital twin image of the UAV flight training personnel's operation is obtained, which can intuitively display the key information such as the UAV flight trajectory and operation behavior.

[0120] The coordinate information of the drone flight is extracted from the historical digital twin images of the drone flight training personnel. The extracted coordinate information is verified using the known coordinates of the basic points of the drone flight training. The purpose of the verification is to ensure that the flight trajectory simulated by the digital twin model is consistent with the actual situation and to improve the accuracy and reliability of the data.

[0121] If the coordinate information is verified, the drone flight data and personnel training operation records are confirmed to be valid and accurate, and are marked as usable data for subsequent data processing and analysis.

[0122] If the verification fails, further inspection and adjustment of the digital twin model or original data is required.

[0123] Specifically, the method for managing the training information of drone operators provided by the present invention includes, in step S102:

[0124] Verify whether the UAV flight data and personnel training operation records have been classified according to the predetermined classification standards, including data type, training subject and timestamp, and pre-process the classified data, including data cleaning, data standardization and data conversion;

[0125] Receive the demand for training effect evaluation, determine the type of features to be extracted based on the demand for training effect evaluation, and use the feature time series analysis method to extract the selected features from the preprocessed data;

[0126] Use historical data or known training effect evaluation results to test the effectiveness of the extracted features. According to the test results, adjust the feature extraction strategy. Adjusting the feature extraction strategy includes modifying the feature type and optimizing the feature extraction algorithm.

[0127] Verify whether the UAV flight data and personnel training operation records have been classified according to the predetermined classification standards. The classification standards include data types (such as numerical type, classification type, etc.), training subjects (such as basic flying skills, emergency handling, etc.) and timestamps (used to identify the specific time of data recording). Classification helps with subsequent data management and processing and improves data processing efficiency.

[0128] A series of preprocessing operations are performed on the classified data, including data cleaning, data standardization, and data conversion.

[0129] Data cleaning: Remove or correct errors, outliers, and duplicate records in the data to ensure data accuracy and consistency.

[0130] Data standardization: Scale the data to the same level, eliminate the dimensional differences between different features, and improve the accuracy and stability of data analysis.

[0131] Data conversion: Convert data types according to analysis requirements, such as encoding categorical variables into numerical variables for mathematical calculations.

[0132] Receive specific requirements for training effect evaluation, which may come from the coaching team, trainees, or system administrators. Determine the type of features to be extracted based on the evaluation requirements. Features are directly related to training effects, such as flight stability, operational accuracy, and reaction speed. Use feature time series analysis to extract selected features from preprocessed data. Time series analysis can capture the laws of data changes over time, which is particularly important for predicting and evaluating trainees' training effects.

[0133] Feature validity test and strategy adjustment: Use historical data or known training effect evaluation results to test the validity of the extracted features. The purpose of the test is to confirm whether the extracted features truly reflect the training effect of the trainees and can be used for subsequent analysis and prediction. Adjust the feature extraction strategy based on the test results. If some features do not perform well in the test, it may be necessary to modify the feature type (such as replacing them with more representative features) or optimize the feature extraction algorithm (such as adjusting algorithm parameters, introducing new algorithms, etc.) to improve the validity and accuracy of the features.

[0134] Specifically, the method for managing the training information of drone operators provided by the present invention includes, in step S103:

[0135] The real-time acquired UAV flight data and the personnel training operation records are simultaneously input into a first UAV personnel flight training sequential time model and a second UAV personnel flight training sequential time model based on linear regression;

[0136] The first UAV personnel flight training time series time model and the second UAV personnel flight training time series time model based on linear regression respectively make predictions based on the input data to generate corresponding flight training periodic change results, and the corresponding flight training periodic change results include the first flight training periodic change results and the second flight training periodic change results;

[0137] Comparing the first flight training periodic change result and the second flight training periodic change result, and calculating an error between the first flight training periodic change result and the second flight training periodic change result;

[0138] An error index is defined to quantify the difference between the periodic change result of the first flight training and the periodic change result of the second flight training, and the data processing result error of the first UAV personnel flight training time series time model and the second UAV personnel flight training time series time model based on real-time linear regression is calculated, and the data processing result error is compared with a preset error range;

[0139] If the error of the real-time data processing result is within the preset error range, it is determined that the first UAV personnel flight training time series time model based on linear regression is effective;

[0140] If the error exceeds the preset range, the first UAV personnel flight training sequential time model and the second UAV personnel flight training sequential time model based on linear regression are optimized, and the above steps are repeated for verification.

[0141] In step S103, the present invention verifies and selects a more accurate and reliable model for subsequent flight training effect evaluation by comparing the prediction results of two independent time series models (the first UAV personnel flight training time series model and the second UAV personnel flight training time series model based on linear regression). The following is the specific execution process of this step:

[0142] The real-time UAV flight data and personnel training operation records are input into the two models at the same time. These data include flight data such as flight altitude, speed, attitude, flight trajectory, and personnel training operation records such as operation behavior, reaction speed, and error rate.

[0143] The first UAV crew flight training temporal time model based on linear regression and the second UAV crew flight training temporal time model receive input data respectively.

[0144] The two models make independent predictions based on the input data and generate corresponding flight training periodic change results. These results reflect the changing trend of students’ flight training effects at different time points.

[0145] The prediction results of the two models (the periodic change results of the first flight training and the periodic change results of the second flight training) are compared.

[0146] The difference between the two results is calculated, which is the prediction error. The calculation of the error involves a variety of statistical indicators, such as mean square error (MSE), mean absolute error (MAE), etc., to comprehensively evaluate the prediction performance of the two models.

[0147] The mean square error (MSE) formula is:

[0148] in, is the actual value, is the predicted value and n is the number of samples.

[0149] The mean absolute error (MAE) formula is:

[0150] in, is the actual value, is the predicted value and n is the number of samples.

[0151] Define error metrics to quantify the difference between the predictions of the two models. These metrics should accurately reflect the prediction accuracy and stability of the model.

[0152] Compare the error of the real-time calculated data processing result with the preset error range. The preset error range is set in advance according to actual needs and data characteristics to determine the acceptability of the model prediction results.

[0153] If the error of the real-time data processing result is within the preset error range, the first UAV personnel flight training time series model based on linear regression (or a model with better performance is selected according to the actual application scenario) is determined to be effective and used for subsequent flight training effect evaluation.

[0154] If the error exceeds the preset range, it means that the prediction performance of the current model is poor and the two models need to be further optimized. Optimization measures may include adjusting model parameters, introducing new feature variables, increasing the amount of training data, etc.

[0155] If it is found that the model prediction error exceeds the preset range, the first UAV personnel flight training sequential time model and the second UAV personnel flight training sequential time model based on linear regression will be optimized.

[0156] After the optimization is completed, repeat the above steps for verification until a model that meets the preset error requirements is found.

[0157] Through this step, the present invention ensures the accuracy and reliability of the selected model, and provides strong support for the subsequent generation of targeted UAV flight training projects. At the same time, by comparing the prediction results of the two models, the present invention also improves the scientificity and objectivity of model selection.

[0158] Specifically, the method for managing the training information of drone operators provided by the present invention includes, in step S104:

[0159] Before comparing the real-time information of UAV flight training with the periodic change results of UAV personnel flight training, establish parallel data processing groups, establish data group correspondence between the real-time information of UAV flight training and the periodic change results of UAV personnel flight training and the data processing parallel groups, establish data group processing tasks based on the data group correspondence, and execute the data processing tasks to perform group and parallel preprocessing on the real-time information of UAV flight training and the periodic change results of UAV personnel flight training.

[0160] In step S104, in order to improve data processing efficiency and accuracy, the present invention adopts a method of parallel data processing grouping to perform parallel preprocessing on the real-time information of UAV flight training and the periodic change results of UAV personnel flight training. The specific execution process of this step is as follows:

[0161] Since a large amount of real-time data and periodic change data are generated during UAV flight training, it is necessary to adopt a parallel processing strategy to speed up data processing and reduce processing time. Before comparing real-time information with periodic change results, multiple data processing parallel groups are first established. These groups can be divided according to data type, source or processing requirements.

[0162] The real-time information of UAV flight training and the periodic change results of UAV personnel flight training are matched with each data processing parallel group. This correspondence ensures that each group of data can be assigned to the appropriate processing unit for processing.

[0163] According to the corresponding relationship of data groups, corresponding data processing tasks are established for each parallel group. The tasks include pre-processing steps such as data cleaning, formatting, and standardization for subsequent analysis.

[0164] These processing tasks are performed in parallel on multiple processing units (such as multi-core CPUs or GPUs). Parallel processing can significantly reduce the overall processing time and improve data processing efficiency.

[0165] After completing the parallel processing, the data processing results of all groups are integrated together to form a complete preprocessed data set for subsequent comparative analysis.

[0166] Through this step, the present invention realizes efficient parallel preprocessing of the real-time information and periodic change results of UAV flight training, laying a solid foundation for subsequent data comparison analysis and error analysis. The parallel processing strategy not only improves the data processing speed, but also helps to improve the responsiveness and stability of the overall system.

[0167] In the second aspect, a drone operator training information management system is provided, which applies the drone operator training information management method as described above, including: a server end, a background control end, a drone end and a user mobile terminal, wherein the server end establishes communication connections with the background control end, the drone end and the user mobile terminal respectively, and the server end analyzes and processes the data generated by the background control end, the drone end and the user mobile terminal, generates drone flight training items and transmits them to the background control end, the drone end and the user mobile terminal, and the server end includes:

[0168] A data acquisition unit is used to acquire UAV flight data and personnel training operation records. The UAV flight data includes flight altitude, speed, attitude, and flight trajectory. The personnel training operation records include personnel training operation behavior, personnel training reaction speed, and personnel training error rate.

[0169] A data classification unit is used to classify the UAV flight data and the personnel training operation records, obtain the classification results of the UAV flight data and the personnel training operation records, establish a data processing partition according to the classification results of the UAV flight data and the personnel training operation records, establish a UAV operator training information management information processing mirror system on the cloud server, pre-process the corresponding UAV flight data and personnel training operation records in the data processing partition and extract data features to obtain flight training stability features, operation accuracy features and reaction speed features;

[0170] A model building unit is used to retrieve unmanned operation historical data, construct a first unmanned aerial vehicle personnel flight training time series time model based on linear regression through unmanned aerial vehicle historical data and flight training stability characteristics, operation accuracy characteristics and reaction speed characteristics, and optimize the first unmanned aerial vehicle personnel flight training time series time model based on linear regression in real time, train a random forest model using unmanned aerial vehicle flight data and personnel training operation records in an unmanned aerial vehicle operator training information management information processing mirror system to obtain a second unmanned aerial vehicle personnel flight training time series time model, optimize the second unmanned aerial vehicle personnel flight training time series time model in real time in an unmanned aerial vehicle operator training information management information processing mirror system, compare the first unmanned aerial vehicle personnel flight training time series time model based on linear regression with the second unmanned aerial vehicle personnel flight training time series time model for real-time data processing results, and obtain a data comparison result. If the data comparison result is within a preset error range, the first unmanned aerial vehicle personnel flight training time series time model based on linear regression takes effect;

[0171] An error generating unit is used to substitute the UAV flight data and the personnel training operation records into a first UAV personnel flight training time series time model based on linear regression, obtain the periodic change result of the UAV personnel flight training, collect the real-time information of the UAV flight training, compare the real-time information of the UAV flight training with the periodic change result of the UAV personnel flight training, and obtain the real-time UAV personnel flight training error;

[0172] The result generating unit is used to analyze the real-time UAV personnel flight training error, obtain the time information corresponding to the real-time UAV personnel flight training error, establish a corresponding relationship between the time information corresponding to the real-time UAV personnel flight training error and the UAV flight training subject, and generate the UAV flight training items to be trained.

[0173] Specifically, the drone operator training information management system and data acquisition unit provided by the present invention are also used for:

[0174] The training information data detection unit detects the integrity of the UAV flight data and the personnel training operation records. The detection content includes flight altitude, speed, attitude, flight trajectory, personnel training operation behavior, personnel training reaction speed and personnel training error rate, and obtains the original data detection result. The original data detection result includes qualified detection data and missing data information. The original data detection result is compared with the historical data of the UAV flight data and the historical data of the personnel training operation record. If the data type comparison is consistent, the UAV flight data and the personnel training operation match the historical UAV flight training subjects. If the data type comparison is inconsistent, the original data detection result and the historical data of the UAV flight data and the historical data of the personnel training operation record are retrieved and the data type is identified to obtain the difference UAV flight training subject information;

[0175] The digital image simulation unit establishes new data detection categories according to the differential UAV flight training subject information, substitutes the data corresponding to the new data detection categories and the data corresponding to the original data detection results into the preset digital twin model for operation, and obtains the historical digital twin images of the UAV flight training personnel's operation;

[0176] The data logic detection unit obtains the coordinate information of the drone flight in the historical digital twin image of the drone flight training personnel's operation, and uses the known coordinates of the drone flight training basic points to verify the accuracy of the coordinate information of the drone flight in the historical digital twin image of the drone flight training personnel's operation. If the coordinate information of the drone flight in the historical digital twin image of the drone flight training personnel's operation is verified, the drone flight data and personnel training operation records will take effect and be marked as usable data.

[0177] Specifically, the drone operator training information management system and data classification unit provided by the present invention are also used for:

[0178] The data feature verification unit verifies whether the UAV flight data and personnel training operation records have been classified according to the predetermined classification standards, including data type, training subject and timestamp, and pre-processes the classified data, including data cleaning, data standardization and data conversion;

[0179] The data feature extraction unit receives the demand for training effect evaluation, determines the type of features to be extracted according to the demand for training effect evaluation, and extracts the selected features from the preprocessed data using a feature time series analysis method;

[0180] The data feature verification unit uses historical data or known training effect evaluation results to verify the validity of the extracted features. According to the test results, the feature extraction strategy is adjusted. The adjustment of the feature extraction strategy includes modifying the feature type and optimizing the feature extraction algorithm.

[0181] Specifically, the UAV operator training information management system and the model building unit provided by the present invention are also used for:

[0182] A data output unit, which simultaneously inputs the real-time acquired UAV flight data and the personnel training operation records into a first UAV personnel flight training sequential time model and a second UAV personnel flight training sequential time model based on linear regression;

[0183] A data prediction unit, which predicts the first UAV personnel flight training time series time model and the second UAV personnel flight training time series time model based on linear regression based on the input data respectively, and generates corresponding flight training periodic change results, wherein the corresponding flight training periodic change results include the first flight training periodic change results and the second flight training periodic change results;

[0184] A real-time data comparison unit compares the first flight training periodic change result and the second flight training periodic change result, and calculates an error between the first flight training periodic change result and the second flight training periodic change result;

[0185] An error definition unit defines an error index for quantifying the difference between the first flight training periodic change result and the second flight training periodic change result, calculating the data processing result error of the first UAV personnel flight training time series time model and the second UAV personnel flight training time series time model based on real-time linear regression, and comparing the data processing result error with a preset error range;

[0186] The model error judgment unit determines that the first UAV personnel flight training timing model based on linear regression is effective if the error of the real-time data processing result is within a preset error range; if the error exceeds the preset range, the first UAV personnel flight training timing model and the second UAV personnel flight training timing model based on linear regression are optimized, and the above steps are repeated for verification.

[0187] Specifically, the UAV operator training information management system provided by the present invention, the result generation unit is also used for:

[0188] The data parallel processing grouping unit establishes data processing parallel groups before comparing the real-time information of UAV flight training with the periodic change results of UAV personnel flight training, establishes a data grouping correspondence between the real-time information of UAV flight training and the periodic change results of UAV personnel flight training and the data processing parallel groups, establishes data grouping processing tasks based on the data grouping correspondence, and executes the data processing tasks to perform grouping and parallel preprocessing on the real-time information of UAV flight training and the periodic change results of UAV personnel flight training.

[0189] The present invention effectively solves the complexity of processing and analyzing a large amount of time series data in the process of building a time series model, as well as the problem of reduced analysis accuracy caused by various factors (such as training intensity, student status, external environment, etc.). The specific solutions are as follows:

[0190] Multi-dimensional data collection and integration: The present invention comprehensively collects data from two dimensions: UAV flight data and personnel training operation records. Flight data includes flight altitude, speed, attitude, flight trajectory, etc., while personnel training operation records cover key indicators such as operation behavior, reaction speed, and error rate.

[0191] Data preprocessing and feature extraction: After obtaining the original data, the present invention performs preprocessing such as classification, cleaning, standardization and conversion on the data to ensure the quality and consistency of the data. Through feature extraction technology, key features such as flight training stability features, operation accuracy features, reaction speed features, etc. are extracted from the preprocessed data. These features are directly related to the training effect and periodic changes of the trainees.

[0192] Dual model construction and real-time optimization: The present invention constructs two time series models: the first is a model built based on UAV historical data and extracted key features; the second is a model trained in the UAV operator training information management information processing mirror system using the random forest algorithm.

[0193] By optimizing the two models in real time, the present invention can continuously adapt to changes in the training environment and data, and maintain the accuracy and effectiveness of the model. The present invention compares the real-time data processing results of the two models and determines the accuracy of the model by calculating the error. Only when the error is within the preset range will the first model be identified as the effective model.

[0194] This dual-model verification mechanism effectively reduces the deviation that may be caused by a single model and improves the accuracy of the analysis results. During the training process, the present invention can collect drone flight training information in real time, and dynamically compare it with the periodic change results, and timely discover and feedback training errors. Through in-depth analysis of the error, the present invention can establish a corresponding relationship between the time information corresponding to the error and the drone flight training subject, thereby generating targeted training items to be trained, providing a basis for personalized training plans.

[0195] By constructing a sequential time model, the present invention can dynamically simulate and predict the training process of trainees and identify potential problems and challenges in advance. According to the prediction results, the present invention can intelligently adjust the training plan and teaching methods to avoid problems and improve the overall training effect.

[0196] In summary, the present invention successfully solves the problems of data processing and analysis complexity and reduced analysis accuracy faced in the process of constructing a time series model through technical means and system design such as multi-dimensional data collection and integration, data preprocessing and feature extraction, dual model construction and real-time optimization, model verification and error control, and real-time data processing.

[0197] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations, and the above-described embodiments of the present invention do not constitute a limitation on the scope of protection of the present invention.

Claims

1. A method for managing training information of unmanned aerial vehicle operators, characterized in that: include: Step S101, obtaining UAV flight data and personnel training operation records, the UAV flight data includes flight altitude, speed, attitude, and flight trajectory, and the personnel training operation records include personnel training operation behavior, personnel training reaction speed, and personnel training error rate; Step S102, classifying the UAV flight data and the personnel training operation records, obtaining the classification results of the UAV flight data and the personnel training operation records, establishing a data processing partition according to the classification results of the UAV flight data and the personnel training operation records, establishing a UAV operator training information management information processing mirror system on the cloud server, pre-processing the corresponding UAV flight data and the personnel training operation records in the data processing partition and extracting data features, and obtaining flight training stability features, operation accuracy features, and reaction speed features; Step S103, retrieve the unmanned operation historical data, construct a first unmanned aerial vehicle personnel flight training time series model based on linear regression through the unmanned aerial vehicle historical data and flight training stability characteristics, operation accuracy characteristics and reaction speed characteristics, and optimize the first unmanned aerial vehicle personnel flight training time series model based on linear regression in real time, use the unmanned aerial vehicle flight data and personnel training operation records in the unmanned aerial vehicle operator training information management information processing mirror system to train the random forest model to obtain a second unmanned aerial vehicle personnel flight training time series model, optimize the second unmanned aerial vehicle personnel flight training time series model in real time in the unmanned aerial vehicle operator training information management information processing mirror system, compare the first unmanned aerial vehicle personnel flight training time series model based on linear regression with the second unmanned aerial vehicle personnel flight training time series model for real-time data processing results, and obtain a data comparison result. If the data comparison result is within a preset error range, the first unmanned aerial vehicle personnel flight training time series model based on linear regression takes effect; Step S104, substituting the UAV flight data and the personnel training operation records into the first UAV personnel flight training time series model based on linear regression, obtaining the periodic change results of the UAV personnel flight training, collecting the real-time information of the UAV flight training, comparing the real-time information of the UAV flight training with the periodic change results of the UAV personnel flight training, and obtaining the real-time UAV personnel flight training error; Step S105, analyzing the real-time UAV personnel flight training error to obtain the time information corresponding to the real-time UAV personnel flight training error, establishing a corresponding relationship between the time information corresponding to the real-time UAV personnel flight training error and the UAV flight training subject, and generating UAV flight training items.

2. The method for managing training information of unmanned aerial vehicle operators according to claim 1, characterized in that: In step S101, it includes: Detect the integrity of UAV flight data and personnel training operation records. The detection content includes flight altitude, speed, attitude, flight trajectory, personnel training operation behavior, personnel training reaction speed and personnel training error rate, and obtain the original data detection results. The original data detection results include qualified detection data and missing data information. The data types of the original data detection results are compared with the historical data of UAV flight data and the historical data of personnel training operation records. If the data type comparison is consistent, the UAV flight data and personnel training operations match the historical UAV flight training subjects. If the data type comparison is inconsistent, retrieve the data with differences between the original data detection results and the historical data of UAV flight data and the historical data of personnel training operation records, and identify the data types to obtain the difference UAV flight training subject information; A new data detection category is established based on the differential UAV flight training subject information, and the data corresponding to the new data detection category and the data corresponding to the original data detection result are substituted into the preset digital twin model to run, and the historical digital twin image of the UAV flight training personnel operation is obtained; Obtain the coordinate information of the drone flight in the historical digital twin images of the drone flight training personnel's operations, and use the known coordinates of the drone flight training basic points to verify the accuracy of the coordinate information of the drone flight in the historical digital twin images of the drone flight training personnel's operations. If the coordinate information of the drone flight in the historical digital twin images of the drone flight training personnel's operations is verified, the drone flight data and personnel training operation records will take effect and be marked as usable data.

3. The method for managing training information of unmanned aerial vehicle operators according to claim 1, characterized in that: In step S102, it includes: Verify whether the UAV flight data and personnel training operation records have been classified according to the predetermined classification standards, including data type, training subject and timestamp, and pre-process the classified data, including data cleaning, data standardization and data conversion; Receive the demand for training effect evaluation, determine the type of features to be extracted based on the demand for training effect evaluation, and use the feature time series analysis method to extract the selected features from the preprocessed data; Use historical data or known training effect evaluation results to test the effectiveness of the extracted features. According to the test results, adjust the feature extraction strategy. Adjusting the feature extraction strategy includes modifying the feature type and optimizing the feature extraction algorithm.

4. The method for managing training information of unmanned aerial vehicle operators according to claim 1, characterized in that: In step S103, it includes: The real-time acquired UAV flight data and the personnel training operation records are simultaneously input into a first UAV personnel flight training sequential time model and a second UAV personnel flight training sequential time model based on linear regression; The first UAV personnel flight training time series time model and the second UAV personnel flight training time series time model based on linear regression respectively make predictions based on the input data to generate corresponding flight training periodic change results, and the corresponding flight training periodic change results include the first flight training periodic change results and the second flight training periodic change results; Comparing the first flight training periodic change result and the second flight training periodic change result, and calculating an error between the first flight training periodic change result and the second flight training periodic change result; An error index is defined to quantify the difference between the periodic change result of the first flight training and the periodic change result of the second flight training, and the data processing result error of the first UAV personnel flight training time series time model and the second UAV personnel flight training time series time model based on real-time linear regression is calculated, and the data processing result error is compared with a preset error range; If the error of the real-time data processing result is within the preset error range, it is determined that the first UAV personnel flight training time series time model based on linear regression is effective; If the error exceeds the preset range, the first UAV personnel flight training sequential time model and the second UAV personnel flight training sequential time model based on linear regression are optimized, and the above steps are repeated for verification.

5. The method for managing training information of unmanned aerial vehicle operators according to claim 1, characterized in that: In step S104, it includes: Before comparing the real-time information of UAV flight training with the periodic change results of UAV personnel flight training, establish parallel data processing groups, establish data group correspondence between the real-time information of UAV flight training and the periodic change results of UAV personnel flight training and the data processing parallel groups, establish data group processing tasks based on the data group correspondence, and execute the data processing tasks to perform group and parallel preprocessing on the real-time information of UAV flight training and the periodic change results of UAV personnel flight training.

6. A UAV operator training information management system, using the UAV operator training information management method according to any one of claims 1 to 5, characterized in that: include: The server side, the backend control side, the drone side and the user's mobile terminal, respectively establish communication connections with the backend control side, the drone side and the user's mobile terminal, analyze and process the data generated by the backend control side, the drone side and the user's mobile terminal, generate drone flight training items and transmit them to the backend control side, the drone side and the user's mobile terminal, and the server side includes: A data acquisition unit is used to acquire UAV flight data and personnel training operation records. The UAV flight data includes flight altitude, speed, attitude, and flight trajectory. The personnel training operation records include personnel training operation behavior, personnel training reaction speed, and personnel training error rate. A data classification unit is used to classify the UAV flight data and the personnel training operation records, obtain the classification results of the UAV flight data and the personnel training operation records, establish a data processing partition according to the classification results of the UAV flight data and the personnel training operation records, establish a UAV operator training information management information processing mirror system on the cloud server, pre-process the corresponding UAV flight data and personnel training operation records in the data processing partition and extract data features to obtain flight training stability features, operation accuracy features and reaction speed features; A model building unit is used to retrieve unmanned operation historical data, construct a first unmanned aerial vehicle personnel flight training time series time model based on linear regression through unmanned aerial vehicle historical data and flight training stability characteristics, operation accuracy characteristics and reaction speed characteristics, and optimize the first unmanned aerial vehicle personnel flight training time series time model based on linear regression in real time, train a random forest model using unmanned aerial vehicle flight data and personnel training operation records in an unmanned aerial vehicle operator training information management information processing mirror system to obtain a second unmanned aerial vehicle personnel flight training time series time model, optimize the second unmanned aerial vehicle personnel flight training time series time model in real time in an unmanned aerial vehicle operator training information management information processing mirror system, compare the first unmanned aerial vehicle personnel flight training time series time model based on linear regression with the second unmanned aerial vehicle personnel flight training time series time model for real-time data processing results, and obtain a data comparison result. If the data comparison result is within a preset error range, the first unmanned aerial vehicle personnel flight training time series time model based on linear regression takes effect; An error generating unit is used to substitute the UAV flight data and the personnel training operation records into a first UAV personnel flight training time series time model based on linear regression, obtain the periodic change result of the UAV personnel flight training, collect the real-time information of the UAV flight training, compare the real-time information of the UAV flight training with the periodic change result of the UAV personnel flight training, and obtain the real-time UAV personnel flight training error; The result generating unit is used to analyze the real-time UAV personnel flight training error, obtain the time information corresponding to the real-time UAV personnel flight training error, establish a corresponding relationship between the time information corresponding to the real-time UAV personnel flight training error and the UAV flight training subject, and generate the UAV flight training items to be trained.

7. The UAV operator training information management system according to claim 6, characterized in that: The data acquisition unit is also used for: The training information data detection unit detects the integrity of the UAV flight data and the personnel training operation records. The detection content includes flight altitude, speed, attitude, flight trajectory, personnel training operation behavior, personnel training reaction speed and personnel training error rate, and obtains the original data detection result. The original data detection result includes qualified detection data and missing data information. The original data detection result is compared with the historical data of the UAV flight data and the historical data of the personnel training operation record. If the data type comparison is consistent, the UAV flight data and the personnel training operation match the historical UAV flight training subjects. If the data type comparison is inconsistent, the original data detection result and the historical data of the UAV flight data and the historical data of the personnel training operation record are retrieved and the data type is identified to obtain the difference UAV flight training subject information; The digital image simulation unit establishes new data detection categories according to the differential UAV flight training subject information, substitutes the data corresponding to the new data detection categories and the data corresponding to the original data detection results into the preset digital twin model for operation, and obtains the historical digital twin images of the UAV flight training personnel's operation; The data logic detection unit obtains the coordinate information of the drone flight in the historical digital twin image of the drone flight training personnel's operation, and uses the known coordinates of the drone flight training basic points to verify the accuracy of the coordinate information of the drone flight in the historical digital twin image of the drone flight training personnel's operation. If the coordinate information of the drone flight in the historical digital twin image of the drone flight training personnel's operation is verified, the drone flight data and personnel training operation records will take effect and be marked as usable data.

8. The UAV operator training information management system according to claim 6, characterized in that: Data classification unit, also used for: The data feature verification unit verifies whether the UAV flight data and personnel training operation records have been classified according to the predetermined classification standards, including data type, training subject and timestamp, and pre-processes the classified data, including data cleaning, data standardization and data conversion; The data feature extraction unit receives the demand for training effect evaluation, determines the type of features to be extracted according to the demand for training effect evaluation, and extracts the selected features from the preprocessed data using a feature time series analysis method; The data feature verification unit uses historical data or known training effect evaluation results to verify the validity of the extracted features. According to the test results, the feature extraction strategy is adjusted. The adjustment of the feature extraction strategy includes modifying the feature type and optimizing the feature extraction algorithm.

9. The UAV operator training information management system according to claim 6, characterized in that: Model building unit, also used for: A data output unit, which simultaneously inputs the real-time acquired UAV flight data and the personnel training operation records into a first UAV personnel flight training sequential time model and a second UAV personnel flight training sequential time model based on linear regression; A data prediction unit, which predicts the first UAV personnel flight training time series time model and the second UAV personnel flight training time series time model based on linear regression based on the input data respectively, and generates corresponding flight training periodic change results, wherein the corresponding flight training periodic change results include the first flight training periodic change results and the second flight training periodic change results; A real-time data comparison unit compares the first flight training periodic change result and the second flight training periodic change result, and calculates an error between the first flight training periodic change result and the second flight training periodic change result; An error definition unit defines an error index for quantifying the difference between the first flight training periodic change result and the second flight training periodic change result, calculating the data processing result error of the first UAV personnel flight training time series time model and the second UAV personnel flight training time series time model based on real-time linear regression, and comparing the data processing result error with a preset error range; The model error judgment unit determines that the first UAV personnel flight training timing model based on linear regression is effective if the error of the real-time data processing result is within a preset error range; if the error exceeds the preset range, the first UAV personnel flight training timing model and the second UAV personnel flight training timing model based on linear regression are optimized, and the above steps are repeated for verification.

10. The UAV operator training information management system according to claim 6, characterized in that: The result generation unit is also used to: The data parallel processing grouping unit establishes data processing parallel groups before comparing the real-time information of UAV flight training with the periodic change results of UAV personnel flight training, establishes a data grouping correspondence between the real-time information of UAV flight training and the periodic change results of UAV personnel flight training and the data processing parallel groups, establishes data grouping processing tasks based on the data grouping correspondence, and executes the data processing tasks to perform grouping and parallel preprocessing on the real-time information of UAV flight training and the periodic change results of UAV personnel flight training.

Images