Pilot driving simulation training method and device, electronic equipment and storage medium

Abstract

The invention relates to the technical field of driving simulation training, in particular to a pilot driving simulation training method and device, electronic equipment and a storage medium. Receiving at least one flight training subject corresponding to the target pilot; pilot behavior data and flight control data input by a target pilot for each flight training subject are acquired; acquiring flight parameter data of the target training aircraft based on the pilot behavior data and the flight control data change; and inputting the flight training subject, the flight parameter data, the pilot behavior data and the flight control data into a preset evaluation model, and outputting a target score corresponding to the target pilot. And the accuracy of the obtained target score corresponding to the target pilot is ensured. Because there is no need for a teacher to score a target pilot based on subjective judgments. According to the method, a scientific and objective standard is provided for evaluation of the target pilot, and an accurate target score is obtained in the training process of the target pilot driving model.

Description

Technical Field

[0001] The present invention relates to the technical field of driving simulation training, and in particular to a pilot driving simulation training method, device, electronic equipment and storage medium. Background Art

[0002] The civil aviation industry plays a fundamental and pioneering role. Pilots, as the linchpin of the civil aviation industry, have a critical role to play in ensuring flight safety and enhancing the international influence and competitiveness of my country's civil aviation industry.

[0003] Existing research indicates that approximately 70% of flight accidents are caused by pilot errors. During flight, pilots experiencing abnormal cognitive states such as spatial disorientation, cognitive impairment, high psychological stress, and excessive workload, as well as abnormal behavioral states such as operational errors, increase flight safety risks. In my country, to ensure that pilots' flying abilities consistently meet flight safety requirements, pilots undergo flight training twice a year using full-motion simulators.

[0004] Meanwhile, during the simulator training assessment phase, pilot evaluations are primarily based on subjective scores from flight instructors. During training, flight instructors not only monitor the various behaviors and cognitive dynamics of their trainees, but also undertake tasks such as setting up various flight scenarios and issuing flight instructions. Consequently, instructors often lose their full attention to their trainees, leading to minor anomalies being overlooked, ultimately impacting assessment results.

[0005] Therefore, how to ensure accurate scoring during pilot flight model training has become an urgent problem to be solved. Summary of the Invention

[0006] In view of this, the present invention provides a pilot driving simulation training method, device, electronic equipment and storage medium to solve the problem of how to ensure accurate scoring during the pilot driving model training process.

[0007] In a first aspect, the present invention provides a pilot driving simulation training method, the method comprising:

[0008] Receive at least one flight training subject corresponding to the target pilot;

[0009] Obtaining pilot behavior data and flight control data input by the target pilot for each flight training subject; the pilot behavior data includes: pilot hand position, hand movement trajectory, foot position, and foot movement trajectory; the flight control data includes at least one parameter of the aircraft control stick, throttle handle, trim handle, and flight control buttons;

[0010] Obtain flight parameter data of the target training aircraft based on changes in pilot behavior data and flight control data;

[0011] The flight training subjects, flight parameter data, pilot behavior data and flight control data are input into the preset evaluation model, and the target score corresponding to the target pilot is output.

[0012] The pilot driving simulation training method provided in an embodiment of the present application receives at least one flight training subject corresponding to a target pilot; obtains pilot behavior data and flight control data input by the target pilot for each flight training subject; obtains flight parameter data of the target training aircraft that changes based on the pilot behavior data and flight control data; inputs the flight training subject, flight parameter data, pilot behavior data, and flight control data into a preset evaluation model, and outputs a target score corresponding to the target pilot, thereby ensuring the accuracy of the target score obtained for the target pilot. This eliminates the need for instructors to score the target pilot based on subjective judgment. This method provides a scientific and objective standard for evaluating the target pilot, enabling accurate target scores to be obtained during the target pilot's driving model training process.

[0013] In an optional embodiment, the preset evaluation model includes an expert library model and an action evaluation model. Flight training subjects, flight parameter data, pilot behavior data, and flight control data are input into the preset evaluation model, and a target score corresponding to the target pilot is output, including:

[0014] Inputting flight training subjects, flight parameter data, pilot behavior data, and flight control data into the expert library model in the preset evaluation model;

[0015] The expert database model matches the flight training subjects with the matching items in the expert database to determine the standard behavior data and standard control data corresponding to the flight training subjects;

[0016] Comparing the standard behavior data and the standard control data with the pilot behavior data and the flight control data, respectively, and outputting initial scores corresponding to the pilot behavior data and the flight control data;

[0017] The pilot behavior data, flight control data, initial score and flight parameter data are input into the action evaluation model, and the target score corresponding to the pilot behavior data and flight control data is output.

[0018] The pilot driving simulation training method provided in the embodiments of the present application inputs flight training subjects, flight parameter data, pilot behavior data, and flight control data into an expert database model within a pre-set evaluation model. The expert database model then matches the flight training subjects with matching items in the expert database to determine standard behavior data and standard control data corresponding to the flight training subjects, thereby ensuring the accuracy of the determined standard behavior data and standard control data corresponding to the flight training subjects. The standard behavior data and standard control data are then compared with the pilot behavior data and flight control data, respectively, and initial scores corresponding to the pilot behavior data and flight control data are output, thereby ensuring the accuracy of the output initial scores corresponding to the flight maneuvers. The pilot behavior data, flight control data, initial scores, and flight parameter data are then input into the action evaluation model, which outputs target scores corresponding to the pilot behavior data and flight control data. This fully considers the actual conditions during flight. Furthermore, the action evaluation model performs a comprehensive assessment based on the flight parameter data, flight maneuvers, and initial scores, enabling a more comprehensive assessment of the impact of the target pilot maneuvers on flight safety, stability, and efficiency, thereby ensuring the accuracy of the output target scores corresponding to the flight maneuvers.

[0019] In an optional embodiment, the standard behavior data and the standard control data are compared with the pilot behavior data and the flight control data, respectively, and initial scores corresponding to the pilot behavior data and the flight control data are output, including:

[0020] Comparing the standard parameter data corresponding to the standard behavior data and the standard control data with the flight parameter data corresponding to the pilot behavior data and the flight control data to obtain parameter deviations;

[0021] and / or

[0022] Comparing the action execution sequence corresponding to the standard behavior data and the standard control data with the action execution sequence corresponding to the pilot behavior data and the flight control data, respectively, to obtain the action execution sequence deviation;

[0023] and / or

[0024] Comparing the execution time corresponding to each action in the standard behavior data and the standard control data with the execution time corresponding to each action in the pilot behavior data and the flight control data, respectively, to obtain the execution time deviation;

[0025] Based on the parameter deviation, and / or the action execution sequence deviation, and / or the execution time deviation, the initial scores corresponding to the pilot behavior data and the flight control data are output.

[0026] The pilot flight simulation training method provided in the embodiments of the present application compares the standard parameter data corresponding to standard behavior data and standard control data with the flight parameter data corresponding to pilot behavior data and flight control data to obtain parameter deviations. This ensures the accuracy of the obtained parameter deviations and accurately measures the differences between the target pilot's operations and the standards. By comparing actual flight parameters with standard parameters, the target pilot's accuracy in speed control, attitude adjustment, and other aspects during takeoff can be accurately determined. This precise assessment helps identify subtle issues in the target pilot's operations and provides precise guidance for subsequent training and improvement. By comparing the action execution sequence corresponding to the standard behavior data and standard control data with the action execution sequence corresponding to the pilot behavior data and flight control data, the action execution sequence deviation is determined, helping to ensure that the target pilot strictly follows the correct procedures. During flight, the execution sequence of each action has its own specific logic and safety requirements. By comparing the action execution sequence, it is possible to promptly identify whether the target pilot has made an action sequence error, avoid safety hazards caused by improper sequence, standardize flight operation procedures, and improve flight safety. Comparing the execution time of each maneuver in the standard behavioral and control data with the execution time of each maneuver in the pilot behavioral and flight control data yields execution time deviations, which can be used to assess the target pilot's ability to time maneuvers. Different flight maneuvers have varying timing requirements during different flight phases, and accurate timing is crucial for flight stability and safety. Analyzing execution time deviations can help target pilots optimize their operational rhythm, improve their time management skills, and ensure smooth and accurate flight operations. Based on parameter deviations, maneuver sequence deviations, and / or execution time deviations, initial scores for the pilot behavioral and flight control data are generated, providing a comprehensive and integrated basis for evaluating the target pilot's operational performance. A single evaluation metric may not fully reflect the target pilot's operational proficiency; however, a comprehensive consideration of multiple deviations can more objectively and accurately assess the target pilot's performance during a flight operation.

[0027] In an optional embodiment, the pilot behavior data, flight control data, initial score, and flight parameter data are input into the action evaluation model, and the target score corresponding to the pilot behavior data and flight control data is output, including:

[0028] Inputting pilot behavior data, flight control data, initial scores, and flight parameter data into the action evaluation model;

[0029] The action evaluation model identifies pilot behavior data, flight control data, and flight parameter data to determine the target training aircraft's flight safety, flight stability, flight efficiency, and degree of compliance with preset goals based on the pilot behavior data and flight control data.

[0030] The initial score is adjusted based on flight safety, flight stability, flight efficiency and compliance with preset goals, and the target score corresponding to the pilot behavior data and flight control data is output.

[0031] The pilot driving simulation training method provided in the embodiments of the present application inputs pilot behavior data, flight control data, initial scores, and flight parameter data into a maneuver evaluation model. The maneuver evaluation model identifies the pilot behavior data, flight control data, and flight parameter data, and determines the flight safety, flight stability, flight efficiency, and compliance with preset targets of the target training aircraft based on the pilot behavior data and flight control data. This model can comprehensively determine the impact of the flight maneuvers on flight safety, stability, efficiency, and compliance with preset targets. The initial scores are then adjusted based on flight safety, flight stability, flight efficiency, and compliance with preset targets, and a target score corresponding to the pilot behavior data and flight control data is output, which more accurately reflects the actual quality and effectiveness of the flight maneuvers. The initial scores are primarily based on a comparison of the flight maneuvers with standard behavior data and standard control data, while the target scores comprehensively consider the actual impact of the flight maneuvers on various aspects of flight, making the evaluation results more objective, comprehensive, and accurate. This provides a more reliable basis for training evaluation and skill assessment of target pilots, facilitates the selection of outstanding target pilots, and provides more targeted guidance for their career development.

[0032] In an optional embodiment, the method further includes:

[0033] Store flight training subjects, pilot behavior data, flight control data, and target scores in correspondence with target pilots;

[0034] Identify flight training subjects, pilot behavior data, flight control data, and target scores to determine the target pilot's operating habits and operational issues for each flight training subject;

[0035] Generate corresponding training plans for target pilots based on operating habits and operational problems.

[0036] The pilot simulation training method provided in the present application associates flight training subjects, pilot behavior data, flight control data, and target scores with target pilots, thereby creating a comprehensive profile of the target pilot. This profile comprehensively records the target pilot's operational performance in different flight training subjects and the corresponding comprehensive evaluation. Through this profile, training institutions or management personnel can gain a comprehensive and systematic understanding of the target pilot's flight performance, providing a solid data foundation for subsequent analysis and decision-making. The flight training subjects, pilot behavior data, flight control data, and target scores are identified to determine the target pilot's operational habits and operational issues for each flight training subject, ensuring the accuracy of the identified operational habits and operational issues for each flight training subject. Then, a training plan is generated for the target pilot based on these operational habits and operational issues. The training plan can include specialized practice sessions, focusing on precise control of operational ranges. For operational issues, the training plan can include detailed theoretical explanations and simulated operational exercises to strengthen the pilot's understanding and practical application of correct operational methods. This targeted training avoids blind training, greatly improving training efficiency and enabling the target pilot to improve his or her skills in the shortest possible time. Furthermore, unlike general training programs, the training plan for target pilots can tailor training content and progress to the specific circumstances of the target pilot. By resolving operational issues and optimizing operating habits, the target pilot's operational level can be significantly improved, providing a strong guarantee for flight safety. Operational issues are often potential hazards to flight safety. For example, improper landing operations can cause the aircraft to overrun the runway. Through targeted training, these issues are eliminated, ensuring that the target pilot can operate the aircraft in a standardized and accurate manner during actual flight, reducing the risk of flight accidents caused by human error, and protecting the lives and property of passengers and crew.

[0037] In an optional embodiment, flight training subjects, pilot behavior data, flight control data, and target scores are identified to determine the target pilot's operating habits and operational problems for each flight training subject, including:

[0038] Input flight training subjects, pilot behavior data, flight control data, and target scores into a pre-set behavior analysis model;

[0039] The preset behavior analysis model extracts features from flight training subjects, pilot behavior data, flight control data, and target scores, and outputs the target pilot's operating habits and operating problems for each flight training subject; wherein, the preset behavior analysis model is trained based on historical training data corresponding to multiple training target pilots and historical operating habits and historical operating problems corresponding to the historical training data; the historical training data includes: historical flight training subjects and historical pilot behavior data, historical flight control data, and historical target scores corresponding to the historical flight training subjects; the historical operating habits and historical operating problems are obtained based on expert review.

[0040] The pilot driving simulation training method provided in the embodiment of the present application inputs flight training subjects, pilot behavior data, flight control data, and target scores into a preset behavior analysis model; the preset behavior analysis model extracts features from the flight training subjects, pilot behavior data, flight control data, and target scores, and outputs the target pilot's operating habits and operating problems for each flight training subject, thereby ensuring the accuracy of the output of the target pilot's operating habits and operating problems for each flight training subject.

[0041] In an optional embodiment, a preset behavior analysis model extracts features from flight training subjects, pilot behavior data, flight control data, and target scores, and outputs the target pilot's operating habits and operational problems for each flight training subject, including:

[0042] Extract basic behavioral features from pilot behavior data;

[0043] Extracting basic operation characteristics from the flight control data; the basic operation characteristics include at least one of the operation force, operation angle, and operation duration of the operation component;

[0044] Conduct correlation analysis between target scores and basic behavioral characteristics and basic operational characteristics to extract score-related features;

[0045] Use sequence analysis algorithms to mine behavioral pattern features from pilot behavior data and flight control data;

[0046] Clustering the extracted basic behavior features, basic operation features, score-related features, and behavior pattern features with the historical features corresponding to the historical training data in the preset behavior analysis model; each cluster represents an operation habit pattern;

[0047] Based on the clustering results, determine the target pilots' operating habits for each flight training subject;

[0048] The extracted basic behavioral features, basic operational features, score-related features, and behavioral pattern features are compared with the historical features corresponding to the historical training data in the preset behavioral analysis model to determine the operational problems of the target pilots in various flight training subjects.

[0049] The pilot flight simulation training method provided in the present application extracts basic behavioral features from pilot behavior data, visually demonstrating the target pilot's behavior during flight operations. Extracting basic operational features from flight control data directly reflects the target pilot's ability to control aircraft control components. Accurate operational force, angle, and duration are crucial for ensuring flight safety and stability. Correlation analysis is performed between the target score and basic behavioral and operational features to extract score-related features. The target score is a quantitative assessment of the target pilot's overall performance in flight training subjects. This correlation analysis identifies which basic behavioral and operational features significantly influence the score. Using a sequence analysis algorithm, behavioral pattern features are mined from pilot behavior and flight control data. Once the target pilot's behavioral pattern features are identified, these patterns can be used to predict their operational behavior in specific scenarios. The extracted basic behavioral features, basic operational features, score-related features, and behavioral pattern features are clustered with historical features corresponding to historical training data in a pre-set behavioral analysis model. This approach accurately determines the target pilot's operational habits and how they compare to those of other target pilots. Based on the clustering results, the target pilot's operational habits for each flight training subject were determined, ensuring the accuracy of the identified target pilot's operational habits for each flight training subject. The extracted basic behavioral characteristics, basic operational characteristics, score-related characteristics, and behavioral pattern characteristics were compared with the historical characteristics corresponding to the historical training data in the preset behavioral analysis model to determine the target pilot's operational problems for each flight training subject, thereby clearly identifying the target pilot's operational problems.

[0050] In a second aspect, the present invention provides a pilot driving simulation training device, the device comprising:

[0051] A receiving module, configured to receive at least one flight training subject corresponding to a target pilot;

[0052] a first acquisition module for acquiring pilot behavior data and flight control data input by the target pilot for each flight training subject; the pilot behavior data including: pilot hand position, hand movement trajectory, foot position, and foot movement trajectory; and the flight control data including at least one parameter of the aircraft's control stick, throttle handle, trim handle, and flight control buttons;

[0053] The second acquisition module is used to obtain flight parameter data of the target training aircraft based on the pilot behavior data and flight control data changes;

[0054] The first output module is used to input flight training subjects, flight parameter data, pilot behavior data and flight control data into a preset evaluation model and output a target score corresponding to the target pilot.

[0055] The pilot driving simulation training device provided in an embodiment of the present application receives at least one flight training subject corresponding to a target pilot; obtains pilot behavior data and flight control data input by the target pilot for each flight training subject; obtains flight parameter data of the target training aircraft that changes based on the pilot behavior data and flight control data; inputs the flight training subject, flight parameter data, pilot behavior data, and flight control data into a preset evaluation model, and outputs a target score corresponding to the target pilot, thereby ensuring the accuracy of the target score obtained for the target pilot. This eliminates the need for instructors to score the target pilot based on subjective judgment. The above device provides a scientific and objective standard for evaluating the target pilot, enabling accurate target scores to be obtained during the target pilot's driving model training process.

[0056] In a third aspect, the present invention provides an electronic device comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to thereby perform the pilot driving simulation training method of the first aspect or any corresponding embodiment thereof.

[0057] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the pilot driving simulation training method of the first aspect or any corresponding embodiment thereof.

[0058] In a fifth aspect, the present invention provides a computer program product comprising computer instructions for causing a computer to execute the pilot driving simulation training method of the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0060] Figure 1is a flowchart of a pilot driving simulation training method according to an embodiment of the present invention;

[0061] Figure 2 is a schematic diagram of a pilot simulator according to an embodiment of the present invention;

[0062] Figure 3 is a flowchart of another pilot driving simulation training method according to an embodiment of the present invention;

[0063] Figure 4 is a flow chart of processing flight parameter data, pilot behavior data, and flight control data according to an embodiment of the present invention;

[0064] Figure 5 is a flowchart of another pilot driving simulation training method according to an embodiment of the present invention;

[0065] Figure 6 is a structural block diagram of a pilot driving simulation training device according to an embodiment of the present invention;

[0066] Figure 7 FIG. 4 is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0067] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are 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 those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.

[0068] It should be noted that the pilot driving simulation training method provided in the embodiments of the present application can be executed by a pilot driving simulation training device. The pilot driving simulation training device can be implemented as part or all of an electronic device through software, hardware, or a combination of software and hardware. The electronic device can be a server or a terminal. The server in the embodiments of the present application can be a single server or a server cluster consisting of multiple servers. The terminal in the embodiments of the present application can be a smartphone, personal computer, tablet computer, wearable device, intelligent robot, or other intelligent hardware device. In the following method embodiments, the execution subject is an electronic device as an example for description.

[0069] According to an embodiment of the present invention, an embodiment of a pilot driving simulation training method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0070] In this embodiment, a pilot driving simulation training method is provided, which can be used for the above-mentioned electronic equipment. Figure 1 FIG. 1 is a flow chart of a pilot driving simulation training method according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0071] Step S101: receiving at least one flight training subject corresponding to a target pilot.

[0072] Specifically, the electronic device can receive at least one flight training subject corresponding to the target pilot input by the instructor. The flight training subject may include aircraft information (e.g., aircraft weight and balance), position information (aircraft position, position movement), environmental information (weather, time, meteorological conditions, visual effects of the training scene), training content (preset aircraft failures, flight routes used in flight training), flight phases, predefined flight routes, and flight targets.

[0073] Step S102 : obtaining pilot behavior data and flight control data input by the target pilot for each flight training subject.

[0074] Among them, the pilot behavior data includes: pilot hand position, hand movement trajectory, foot position, and foot movement trajectory; flight control data includes at least one parameter of the aircraft control stick, throttle handle, trim handle, and flight control button.

[0075] Specifically, the electronic device can obtain pilot behavior data and flight control data based on high-precision sensors installed on the flight simulator.

[0076] For example, electronic devices can use optical sensors, inertial sensors, or electromagnetic induction sensors installed on a flight simulator to collect the hand position and movement trajectory of the target pilot. Specifically, optical sensors can determine the hand's position and movement trajectory by capturing reflective markings or feature points on the hand; inertial sensors measure the hand's acceleration and angular velocity, combining integration operations and other methods to infer the hand's position and motion state; and electromagnetic induction sensors detect changes in hand position by emitting and receiving electromagnetic field signals.

[0077] Electronic devices can use pressure sensors, position sensors, or angle sensors installed on the flight simulator's pedals to collect the target pilot's corresponding foot position and foot movement trajectory. Specifically, the pressure sensor can measure the pressure applied to the pedal by the foot, thereby reflecting the target pilot's rudder or brake operation strength; the position sensor and angle sensor are used to determine the foot position and pedal rotation angle, thereby obtaining the foot movement trajectory.

[0078] Electronic devices can collect aircraft control stick parameters using various sensors, such as displacement sensors and force sensors, mounted on the control stick. Specifically, displacement sensors measure the forward and backward, left and right movement of the stick, thereby detecting changes in its position. Force sensors detect the magnitude and direction of the force applied by the pilot when manipulating the stick. These sensors convert the collected signals into electrical or digital signals and transmit them to a data acquisition system for recording and processing.

[0079] Electronic devices can use an angle sensor or displacement sensor installed on the throttle handle to measure the rotation angle or movement distance of the throttle handle, thereby obtaining the change in throttle opening. Furthermore, this information can be combined with relevant engine parameters (such as speed and fuel flow) to more comprehensively understand the impact of throttle operation on engine output power.

[0080] The electronics can be based on micro switches or pressure sensors installed on flight control buttons to detect when the buttons are pressed and released, as well as the force of the press (if necessary). These signals are recorded and processed by a data acquisition system.

[0081] For example, Figure 2 The figure shows a schematic diagram of a flight simulator. The electronic equipment uses various sensors installed on the flight simulator to obtain pilot behavior data and flight control data input by the target pilot for various flight training subjects.

[0082] Step S103: obtaining flight parameter data of the target training aircraft based on changes in the pilot behavior data and flight control data.

[0083] The flight parameter data may include at least one of flight altitude, flight speed, heading angle, flight attitude, wing stress, and fuselage vibration frequency.

[0084] Specifically, electronic equipment can obtain flight altitude using devices such as barometric altimeters and radio altimeters. Electronic equipment can measure flight speed using a system consisting of a pitot tube (or pitot tube) and related sensors (such as static pressure ports and temperature sensors). Electronic equipment can measure heading angle using devices such as a magnetic compass, inertial navigation system (INS), and global positioning system (GPS). Electronic equipment can measure the aircraft's pitch angle (the angle of the aircraft's nose relative to the horizontal plane), roll angle (the aircraft's rotation angle around the longitudinal axis), and yaw angle (the aircraft's rotation angle around the vertical axis) using inertial sensors such as gyroscopes and accelerometers. These sensors can sense changes in the aircraft's attitude in real time and transmit the data to the avionics system for processing and display. Electronic equipment can use accelerometers or vibration sensors installed in key locations on the fuselage (such as the fuselage frame and engine pylons) to measure the vibration frequency and amplitude of the fuselage.

[0085] In step S104 , the flight training subjects, flight parameter data, pilot behavior data, and flight control data are input into a preset evaluation model, and a target score corresponding to the target pilot is output.

[0086] Specifically, the electronic device can input the collected flight training subjects, flight parameter data, pilot behavior data and flight control data into a preset evaluation model. The preset evaluation model extracts features of the flight training subjects, flight parameter data, pilot behavior data and flight control data, and outputs a target score corresponding to the target pilot.

[0087] This step will be described in detail below.

[0088] The pilot driving simulation training method provided in an embodiment of the present application receives at least one flight training subject corresponding to a target pilot; obtains pilot behavior data and flight control data input by the target pilot for each flight training subject; obtains flight parameter data of the target training aircraft that changes based on the pilot behavior data and flight control data; inputs the flight training subject, flight parameter data, pilot behavior data, and flight control data into a preset evaluation model, and outputs a target score corresponding to the target pilot, thereby ensuring the accuracy of the target score obtained for the target pilot. This eliminates the need for instructors to score the target pilot based on subjective judgment. This method provides a scientific and objective standard for evaluating the target pilot, enabling accurate target scores to be obtained during the target pilot's driving model training process.

[0089] In this embodiment, a pilot driving simulation training method is provided, which can be used for the above-mentioned electronic equipment. Figure 3 FIG. 1 is a flow chart of a pilot driving simulation training method according to an embodiment of the present invention. Figure 3 As shown, the process includes the following steps:

[0090] Step S201: receiving at least one flight training subject corresponding to a target pilot.

[0091] For details about this step, please refer to the above description of step S101 and will not be repeated here.

[0092] Step S202 : obtaining pilot behavior data and flight control data input by the target pilot for each flight training subject.

[0093] Among them, the pilot behavior data includes: pilot hand position, hand movement trajectory, foot position, and foot movement trajectory; flight control data includes at least one parameter of the aircraft control stick, throttle handle, trim handle, and flight control button.

[0094] For details about this step, please refer to the above description of step S102 and will not be repeated here.

[0095] Step S203: Acquire flight parameter data of the target training aircraft based on changes in the pilot behavior data and flight control data.

[0096] For details about this step, please refer to the above description of step S103 and will not be repeated here.

[0097] In step S204 , the flight training subjects, flight parameter data, pilot behavior data, and flight control data are input into a preset evaluation model, and a target score corresponding to the target pilot is output.

[0098] Specifically, the above step S204 may include the following steps:

[0099] Step S2041 , inputting flight training subjects, flight parameter data, pilot behavior data, and flight control data into the expert library model in the preset evaluation model.

[0100] Specifically, before inputting the flight training subjects, flight parameter data, pilot behavior data and flight control data into the expert library model in the preset evaluation model, the electronic device may process the flight parameter data, pilot behavior data and flight control data.

[0101] For example, Figure 4 As shown, the electronic device can organize the collected flight parameter data, pilot behavior data, and flight control data. Then, the organized flight parameter data, pilot behavior data, and flight control data are filtered and preprocessed. Finally, feature extraction is performed on the flight parameter data, pilot behavior data, and flight control data to obtain behavior features, control features, and flight parameter features.

[0102] The electronic equipment can input flight training subjects, behavioral characteristics, control characteristics and flight parameter characteristics into the expert library model in the preset evaluation model.

[0103] In step S2042 , the expert database model matches the flight training subject with each matching item in the expert database to determine the standard behavior data and standard control data corresponding to the flight training subject.

[0104] The process of generating the expert database is as follows: electronic equipment can obtain the operation records of senior target pilots, covering the operation details of different aircraft models, weather conditions, and flight phases; receive standardized action descriptions in flight training materials and manuals; receive aviation accident cases and corresponding analysis materials on correct operations; and introduce advanced flight simulation experimental data, which can accurately simulate the best operations in various complex flight scenarios.

[0105] The electronic equipment will then integrate the collected data. Taking flight training subjects as nodes, the related actions, applicable conditions, influencing factors, etc. are connected as edges. For example, under the "take-off" node, the action and condition nodes such as "the order of starting the engine" and "the taxiing speed corresponding to different runway lengths" are connected. Through the knowledge graph, the relationship between the knowledge of each flight training subject can be clearly presented, which facilitates the rapid retrieval and matching of the model, thereby generating an expert database. In addition, the latest aviation research results, changes in industry standards, and new flight experience summaries are regularly collected. For example, when a new aircraft model is put into use, its related flight training subjects and standard behavior data and standard control data are promptly included in the expert database; if the industry updates the operating standards of a certain flight training subject, the corresponding content in the expert database is quickly updated to ensure that the expert database always maintains timeliness and accuracy.

[0106] Specifically, for flight training subjects, the electronic equipment first extracts features from the subject. This not only captures surface features such as common types, mission objectives, and environmental conditions, but also uses deep learning algorithms to deeply mine these features. For example, for the "landing in adverse weather" flight training subject, deep learning analysis of meteorological data extracts deeper features such as cloud thickness trends and airflow instability. Aircraft status data is analyzed to extract stress distribution characteristics of the aircraft structure in adverse environments. These deep features more accurately describe the characteristics of the flight training subject.

[0107] The electronic device then calculates a first similarity between the flight training subject and each matching item in the expert database based on an introduced semantic similarity calculation method. Furthermore, the electronic device calculates a second similarity in the knowledge structure between the flight training subject and each matching item in the expert database, combining the similarities of nodes and edges in the knowledge graph. The electronic device averages the first and second similarities to obtain a target similarity. Matching items with a target similarity greater than a first preset similarity threshold are identified as target items, and the actions corresponding to the target items are identified as the standard behavioral data and standard control data corresponding to the flight training subject.

[0108] In an optional embodiment, if no matching items with a target similarity greater than a first preset similarity threshold are found, the expert database model sorts the items by target similarity from high to low and selects the closest matching items. A genetic algorithm is then used to optimize the standard behavior data and standard control data for these similar items. For example, for a new special flight training subject, the expert database model finds several similar items and then uses a genetic algorithm to perform crossover and mutation operations on the action sequences and operating parameters of these items to generate a set of optimized standard behavior data and standard control data for the flight training subject.

[0109] Step S2043 : Compare the standard behavior data and the standard control data with the pilot behavior data and the flight control data, respectively, and output initial scores corresponding to the pilot behavior data and the flight control data.

[0110] Specifically, the above step S2043 may include the following steps:

[0111] In step a1, the standard parameter data corresponding to the standard behavior data and the standard control data are compared with the flight parameter data corresponding to the pilot behavior data and the flight control data to obtain parameter deviations.

[0112] Specifically, the electronic device can accurately match standard behavior data with pilot behavior data based on flight phase, operational tasks, and other dimensions. For example, during landing, the standard hand operation trajectory is matched with the target pilot's actual hand movement trajectory within the same time interval to ensure the accuracy of the comparison.

[0113] Similarly, the electronic equipment also matches standard control data and flight control data based on the operating components and flight phases. For example, the standard throttle handle operating force is matched with the actual throttle handle operating force data during the takeoff acceleration phase.

[0114] Then, the electronic equipment extracts key standard parameters from the standard behavior data and standard control data, such as standard hand movement speed, standard joystick operation angle range, etc.; and extracts actual parameters from the flight parameter data corresponding to the pilot behavior data and flight control data, such as actual average hand movement speed, actual maximum joystick operation angle, etc.

[0115] For parameters that can be expressed numerically, such as control force, control angle, and flight speed, the difference between the standard parameter data and the actual flight parameter data is directly calculated. For example, if the standard control stick angle is 10° and the actual control angle is 12°, the angle deviation is 2°.

[0116] For some parameters that have a reasonable range, such as flight altitude, which should remain within a certain range during the cruise phase, the actual flight altitude is compared with the standard altitude range. If the actual altitude exceeds the standard range, the excess value or proportion is calculated as the range deviation. For example, if the standard cruise altitude range is 10,000 to 10,500 meters and the actual flight altitude is 10,600 meters, the altitude deviation is 100 meters.

[0117] For time-varying parameters, such as flight speed trends or hand movement trajectory trends, analyze the difference between the standard trend and the actual trend. Trend deviations can be quantified by calculating curve similarity and slope differences. For example, during a standard takeoff acceleration, the speed should increase evenly. If the actual speed increase fluctuates or is uneven, this deviation from the trend can be calculated.

[0118] Based on the above numerical differences, range deviations, and trend differences, the parameter deviations are calculated.

[0119] and / or

[0120] In step a2, the action execution sequence corresponding to the standard behavior data and the standard control data is compared with the action execution sequence corresponding to the pilot behavior data and the flight control data to obtain the action execution sequence deviation.

[0121] Specifically, the electronic equipment can preprocess the collected pilot behavior data to remove interference factors such as noise and outliers. The data is then analyzed based on a time series to extract the order in which the target pilot's hand and foot movements were executed. For example, by analyzing video data of hand movements, the order in which the target pilot operated the joystick, buttons, and other devices can be determined. Similar processing is performed on the flight control data, organizing the collected operation signals of each operating component in chronological order to extract the specific time point and execution order of each operation. For example, from the sensor data of the throttle handle, the specific time and order of gradual throttle application during takeoff can be determined.

[0122] The standard and actual action execution sequences are then divided according to flight phases, such as takeoff, cruising, and landing. Within each phase, the actions in the standard sequence are matched one-to-one with those in the actual sequence. For example, during takeoff, the "Set Navigation Parameters" action in the standard sequence is compared with the actual execution of that action.

[0123] By comparing the order of actions in the standard and actual execution sequences, discrepancies are detected, resulting in deviations in the execution sequence. If an action in the actual execution sequence is found to be earlier or later than the standard sequence, or if an action is missing or redundant, these discrepancies are recorded. For example, the standard sequence requires a final pre-takeoff check before taxiing to the runway, but in actual operation, the target pilot taxied first and then performed the check, resulting in a deviation in the execution sequence.

[0124] Optionally, to more accurately assess the extent of deviations in the execution sequence, deviations can be quantified. For example, a deviation scoring system could be established based on factors such as the number of deviations and their importance within the overall operational process. Deviating actions that are critical to flight safety would receive a higher score; deviations in the sequence of minor actions would receive a lower score.

[0125] and / or

[0126] In step a3, the execution time corresponding to each action in the standard behavior data and the standard control data is compared with the execution time corresponding to each action in the pilot behavior data and the flight control data to obtain the execution time deviation.

[0127] Specifically, the electronic device can compare the execution time corresponding to each action in the standard behavior data and the standard control data with the execution time corresponding to each action in the pilot behavior data and the flight control data according to each flight phase to obtain the execution time deviation.

[0128] For example, during takeoff, standard procedures dictate that the aircraft begin rotating after reaching a certain rollout speed. Let's assume this standard time is 30 seconds after the rollout begins. If the target pilot, in actual operation, only begins rotating 35 seconds after the rollout begins, a 5-second time discrepancy occurs, which may indicate that the target pilot misjudged the takeoff speed or made an operational delay. During the cruise phase, after the aircraft reaches cruising altitude, a series of system checks and parameter adjustments must be completed within a specific timeframe. Let's assume the standard specifies completion within 5 minutes of reaching cruising altitude. If the target pilot completes these operations 8 minutes after reaching cruising altitude, a 3-minute time discrepancy occurs, which may reflect a lack of proficiency or task prioritization during the cruise phase. During landing, for example, lowering the landing gear is typically required upon approach to a certain altitude and distance. Let's assume the standard is 2 nautical miles from the runway threshold and 1,000 feet. If the target pilot lowers the landing gear 1.5 nautical miles from the runway threshold and at an altitude of 800 feet, there will be time and space deviations, which may affect the aircraft's landing attitude and safety.

[0129] Step a4: outputting initial scores corresponding to the pilot behavior data and the flight control data based on the parameter deviation, and / or the action execution sequence deviation, and / or the execution time deviation.

[0130] Specifically, the electronic device quantifies the degree of impact of the numerical difference, range deviation, and trend difference included in the parameter deviation on flight safety and operational performance, and determines the weights corresponding to the numerical difference, range deviation, and trend difference. The electronic device then performs a weighted summation of the quantized numerical difference, range deviation, and trend difference to obtain the quantified parameter deviation.

[0131] The electronic device can also quantify the degree of impact of the action execution sequence deviation and / or execution time deviation on flight safety and operational performance to obtain a quantified action execution sequence deviation and / or a quantified execution time deviation.

[0132] Then, the electronic equipment can set weights for parameter deviation, action execution sequence deviation and execution time deviation respectively according to their relative importance to flight safety and operation quality.

[0133] Finally, the electronic device calculates a single score for each deviation item based on the quantified parameter deviation, action execution sequence deviation, execution time deviation, and the set weights. This is usually calculated using a linear or nonlinear function relationship. For example, for parameter deviation, a function can be set such that when the deviation value is 0, the single score is full (e.g., 100 points); as the deviation value increases, the score gradually decreases, and when the deviation value reaches a certain threshold, the single score is 0 points. The electronic device then sums the single scores of each deviation item according to the weights to obtain the initial score corresponding to the pilot behavior data and flight control data.

[0134] In step S2044 , the pilot behavior data, flight control data, initial score, and flight parameter data are input into the action evaluation model, and the target score corresponding to the pilot behavior data and flight control data is output.

[0135] Specifically, the above step S2044 may include the following steps:

[0136] Step b1: inputting pilot behavior data, flight control data, initial score and flight parameter data into the action evaluation model.

[0137] Specifically, the electronic device inputs the pilot behavior data, flight control data, initial score and flight parameter data into the action evaluation model.

[0138] In step b2, the action evaluation model identifies the pilot behavior data, flight control data, and flight parameter data to determine the flight safety, flight stability, flight efficiency, and conformity with the preset goals of the target training aircraft based on the pilot behavior data and flight control data.

[0139] Specifically, let the pilot behavior data be B = [b1, b2, …, bn], where bi represents the i-th behavior characteristic parameter; the flight control data be C = [c1, c2, …, cm], cj represents the j-th control characteristic parameter; the flight parameter data be F = [f1, f2, …, fk], fl represents the l-th flight state parameter.

[0140] The action evaluation model introduces the concept of fuzzy risk entropy. In traditional risk assessments, risk states are often simply categorized as "safe" or "dangerous," but risks in actual flight are characterized by ambiguity and uncertainty. Fuzzy risk entropy, based on fuzzy mathematical theory, quantifies the ambiguity of risk. Information entropy is a metric used to measure uncertainty, while fuzzy risk entropy reflects the magnitude of potential flight risks by calculating the degree of uncertainty in the risk state. Higher entropy values ​​indicate a more uncertain risk state and lower flight safety; lower entropy values ​​indicate a clearer risk state and higher flight safety.

[0141] The degree to which various flight parameter data, such as altitude, speed, and heading angle, deviate from standard parameters during flight directly impacts flight safety. By calculating the difference between these parameters and the standard and assigning different weights based on their impact on flight safety, parameter deviations are converted into numerical risk indicators. For example, a flight altitude deviation exceeding a safety threshold for a short period of time may be assigned a higher risk weight, as it could lead to a collision with an obstacle or entry into hazardous airspace.

[0142] Based on historical pilot behavior data, historical flight control data, and the current flight scenario, the system uses pre-configured machine learning models (such as Bayesian networks and hidden Markov models) to predict the probability of an operational anomaly. For example, in complex weather conditions, the probability of an operational error is calculated based on the pilot's previous operational records under similar conditions. The higher the probability of an operational anomaly, the greater the risk to the flight.

[0143] Then, parameter deviation and operational abnormality probability are incorporated into the fuzzy set. Each risk indicator corresponds to a fuzzy membership function, which is used to describe the degree to which the indicator belongs to different risk levels (such as low risk, medium risk, and high risk). Through fuzzy logic operations, multiple risk indicators are integrated into a comprehensive fuzzy risk set. Then, the entropy value of the fuzzy risk set is calculated based on the information entropy formula to obtain the fuzzy risk entropy. The calculation formula is as follows:

[0144] Among them, H represents the fuzzy risk entropy, p(x i ) is the membership degree of the i-th risk level in the fuzzy set.

[0145] Next, various risk factors related to flight safety are identified, such as the aforementioned parameter deviations and operational anomaly probability indicators, forming a risk factor set R = [r1, r2, …, rp]. Each risk factor is assigned a corresponding weight μs, reflecting its significance to flight safety. Weights can be determined using methods such as the Analytic Hierarchy Process (AHP), principal component analysis (PCA), or reinforcement learning algorithms. For example, reinforcement learning algorithms can dynamically adjust weights based on the correlation between risk factors and actual flight accidents, ensuring they better reflect actual circumstances.

[0146] According to the fuzzy risk entropy H and the weight μs of each risk factor, the flight safety score Ssafe is calculated as follows:

[0147] The calculation formula is:

[0148] Among them, α is the safety reference factor, μ sThe risk factor weights are dynamically optimized through reinforcement learning algorithms. For example, when the system detects an increase in the frequency of near-miss accidents caused by a certain type of operation, the weight μs of the corresponding risk factor will automatically increase.

[0149] In addition, during the flight, the various flight parameter data (such as flight altitude, speed, heading angle, attitude angle, etc.) do not change independently, and there are complex correlations between them. Multi-scale dynamic correlation analysis aims to capture the relationship between these parameters at different time scales. Different time scales mean different time ranges of concern. At short time scales (such as seconds), parameter changes may reflect the response of the pilot's immediate operations; at long time scales (such as minutes and hours), it more reflects the overall trend and stability of the flight process. By analyzing at multiple time scales, we can have a more comprehensive and in-depth understanding of the coordinated change patterns between flight parameters and judge the stability of the flight status.

[0150] The action evaluation model can calculate the correlation between the lth flight parameter data and the mth flight parameter data at the time scale τ based on the Pearson correlation coefficient method, which is expressed as the correlation coefficient ρlm(τ). The specific formula is as follows:

[0151]

[0152] Among them, x lt and y mt are the data of the lth and mth flight parameters at time point t, and are the average values ​​of the corresponding parameters in the sliding window, and n is the number of data points in the sliding window.

[0153] Then, based on the calculated correlation coefficient ρlm(τ), the flight stability score Sstable is calculated. The formula is as follows:

[0154]

[0155] Among them, the molecular part: represents the sum of the correlation coefficients for all flight parameters across all time scales τ and any pairwise combination. This method comprehensively summarizes the interrelationships between flight parameters at different time scales, reflecting the degree of coordinated variation within the flight parameter system as a whole. A larger sum indicates a stronger correlation between flight parameters and a closer coordinated variation between parameters at different time scales, indicating a more stable flight state overall.

[0156] Denominator: is a normalization factor, where k is the total number of flight parameters. This factor is introduced to eliminate the influence of the number of flight parameters on the calculated results, making stability scores calculated for different flight scenarios (involving different numbers of flight parameters) comparable. For example, if normalization is not performed, the sum of the results will increase as the number of flight parameters increases, but this does not necessarily mean that flight stability has improved. This normalization factor allows for reasonable adjustments to the scores.

[0157] β is the stability adjustment factor, used to adjust the overall range and sensitivity of the flight stability score. The value of β can be set and adjusted based on actual needs and experience. A larger β value makes the score more sensitive to changes in flight parameter correlations, while a smaller β value makes the score change more gradually, allowing for flexible control of the stability score calculation results in different evaluation scenarios.

[0158] Next, the action evaluation model constructs an energy-time dual-objective optimization function, comprehensively evaluating efficiency based on fuel consumption E and flight time T. The flight efficiency score Seff is: Among them, E std is the standard fuel consumption, T std is the standard flight time, γ is the efficiency weight coefficient, Estd, Tstd are the standard fuel consumption and standard flight time.

[0159] In flight training and missions, preset goals are often described in text form, such as "land smoothly on the designated runway after completing an instrument approach" and "perform cargo flight missions according to the planned route." Natural language processing (NLP) technology is used to parse the preset target text and descriptive records of the actual operation process. Basic processing steps such as word segmentation, part-of-speech tagging, and named entity recognition are used to convert the text into a structured form that can be processed by computers. For example, for the target text of "performing a precision approach under low visibility conditions," key entities and concepts such as "low visibility" and "precision approach" can be parsed.

[0160] Then, the action evaluation model can calculate the similarity between the descriptive record of the current operation process generated based on the pilot behavior data and flight control data and the preset target text based on semantic matching algorithms such as word vector models (such as Word2Vec, BERT, etc.) or knowledge graphs. Taking BERT as an example, it can learn deep semantic representations of words and sentences based on large-scale text pre-training. The preset target text and the current descriptive record are input into the BERT model respectively to obtain the corresponding semantic vector representation, and then the similarity value between the two semantic vectors is calculated by measurement methods such as cosine similarity and Euclidean distance, which is recorded as Sim. The closer the Sim value is to 1, the higher the consistency between the actual operation description and the preset target text at the semantic level; the closer it is to 0, the greater the semantic difference.

[0161] Then, the preset target parameter set G = [g1, g2, …, gq] contains the standard values ​​of key parameters in the flight mission, such as flight altitude, speed, route coordinates, etc. Δgi represents the degree of deviation between the i-th parameter in the actual flight and the preset target parameter gi. The calculation method varies depending on the parameter type:

[0162] For numerical parameters such as flight altitude and speed, absolute deviation or relative deviation is usually used for calculation. For example, absolute deviation Δgi = |xi-gi|, where xi is the actual parameter value; relative deviation Intuitively reflects the ratio of the actual value to the standard value.

[0163] For non-numeric parameters such as route name and mission type, the deviation is determined by judging whether the actual value is consistent with the preset value. If they are consistent, Δgi=0; otherwise, Δgi=1.

[0164] Flight missions often involve multiple parameters, and each Δgi reflects the deviation of a single parameter. This product combines the deviations of all parameters. When all parameters are uniform, the product is 1; if even one parameter deviates, the result is less than 1. The more deviating parameters there are and the greater the degree of deviation, the smaller the product, intuitively reflecting the degree of deviation between the overall parameters and the preset target.

[0165] Finally, the fit score formula The semantic similarity and task parameter deviation are integrated. Among them, Sim evaluates the consistency between the actual operation and the preset goal from the language description level. The deviation is measured at the level of specific parameter execution, and the multiplication of the two achieves a multi-dimensional characterization of the task execution. θ is the target weight coefficient, which is used to adjust the overall range and importance of the fit score. In different flight mission scenarios, the value of θ can be flexibly adjusted according to the characteristics of the mission. For example, for some special flight missions that require extremely high accuracy in the description of the operating process (such as aerial photography, which requires the shooting action to be completed strictly in accordance with a specific description), the value of θ can be increased so that semantic similarity occupies a larger proportion in the fit assessment; for conventional transport flight missions, if more attention is paid to the execution accuracy of the actual parameters, θ can be appropriately adjusted to balance the influence of semantic and parameter deviations, thereby achieving a comprehensive and accurate assessment of the quality of task execution.

[0166] Step b3: Adjust the initial score based on flight safety, flight stability, flight efficiency, and compliance with preset goals, and output the target score corresponding to the pilot behavior data and flight control data.

[0167] Specifically, electronic equipment can set weight information corresponding to flight safety, flight stability, flight efficiency, and the degree of compliance with preset goals. Among them, the setting of weights generally follows the following principles: flight safety is of vital importance and should be given the highest weight in most cases, because safety is the primary prerequisite for flight; flight stability is directly related to the comfort and safety of flight, and the weight is also relatively high; flight efficiency has an important impact on the operating costs and economic benefits of airlines, and the appropriate weight is determined according to the nature of the flight mission (for example, commercial flights pay more attention to efficiency); the degree of compliance with preset goals reflects the quality of the completion of the flight mission, and the weight is adjusted according to the importance and specificity of the mission.

[0168] Specifically, the electronic device can adjust the initial score based on flight safety, flight stability, flight efficiency, and the degree of compliance with preset targets and the corresponding weight information, and output the target score.

[0169] For example, let the initial score be 1 and the flight safety score be S. safe , flight stability score is S stable , the flight efficiency score is S eff , the score of fit with the preset goal is S goal , and the corresponding weights are W safe 、W stable 、W eff 、W goal .

[0170] The calculation formula of target score T is:

[0171] T=I×(1-W safe -W stable -W eff -W goal )+S safe ×W safe +S stable ×W stable +S eff ×W eff +S goal ×W goal .

[0172] Among them, I×(1-W safe -W stable -W eff -W goal ) represents the portion of the initial score that is not covered by the adjustment factors of each dimension, and the following items are the adjustment contributions of the evaluation results of each dimension to the score.

[0173] The pilot driving simulation training method provided in the embodiment of the present application inputs flight training subjects, flight parameter data, pilot behavior data, and flight control data into an expert library model in a preset evaluation model; the expert library model matches the flight training subjects with each matching item in the expert library, determines the standard behavior data and standard control data corresponding to the flight training subjects, and ensures the accuracy of the determined standard behavior data and standard control data corresponding to the flight training subjects.

[0174] The standard parameter data corresponding to the standard behavioral and control data are then compared with the flight parameter data corresponding to the pilot behavioral and control data to obtain parameter deviations. This ensures the accuracy of the obtained parameter deviations and accurately measures the difference between the target pilot's performance and the standard. This precise assessment helps identify subtle flaws in the target pilot's performance and provides precise guidance for subsequent training and improvement. The action execution sequence corresponding to the standard behavioral and control data is compared with the action execution sequence corresponding to the pilot behavioral and control data to obtain action execution sequence deviations, which helps ensure that the target pilot strictly follows the correct procedures. During flight, each action execution sequence has specific logic and safety requirements. By comparing action execution sequences, it is possible to promptly identify any action sequence errors by the target pilot, avoid safety hazards caused by improper sequencing, standardize flight operations, and improve flight safety. The execution time corresponding to each action in the standard behavioral and control data is compared with the execution time corresponding to each action in the pilot behavioral and control data to obtain execution time deviations, which can be used to assess the target pilot's ability to control action timing. Different flight actions have different timing requirements at different flight phases, and accurate timing control is crucial for flight stability and safety. By analyzing execution time deviations, target pilots can optimize their operational rhythm, improve their time management skills, and ensure the smoothness and accuracy of flight operations. Based on parameter deviations, and / or deviations in the execution sequence, and / or execution time deviations, the initial scores corresponding to the pilot's behavioral data and flight control data are output, providing a comprehensive and integrated basis for the target pilot's operational assessment. A single evaluation metric may not fully reflect the target pilot's operational level, but a comprehensive consideration of multiple deviations can more objectively and accurately evaluate the target pilot's performance in a flight operation. For example, even if the target pilot's parameter deviation is small, an incorrect execution sequence or unreasonable execution time will result in a low score, prompting the target pilot to comprehensively improve his or her operational skills.

[0175] Then, the pilot behavior data, flight control data, initial scores, and flight parameter data are input into the maneuver evaluation model. The maneuver evaluation model identifies the pilot behavior data, flight control data, and flight parameter data, and determines the target training aircraft's flight safety, flight stability, flight efficiency, and degree of compliance with preset objectives based on the pilot behavior data and flight control data. This allows for a comprehensive assessment of the impact of flight maneuvers on flight safety, stability, efficiency, and degree of compliance with preset objectives. For example, when evaluating a turning maneuver, not only is the standardization of the maneuver itself considered (partially reflected in the initial score), but flight parameter data (such as speed, altitude, and attitude changes) are also combined to determine whether the maneuver affects flight safety (such as whether there is a risk of stalling), flight stability (such as whether it causes excessive turbulence), flight efficiency (such as whether it consumes excessive fuel), and compliance with the preset route or operational objectives (such as whether the turn angle and position are accurately completed). This avoids the limitations of evaluating flight maneuvers from only a single dimension. The initial score is then adjusted based on flight safety, flight stability, flight efficiency, and alignment with pre-set goals, outputting a target score corresponding to the pilot's behavioral data and flight control data, more accurately reflecting the actual quality and effectiveness of the flight maneuvers. The initial score is primarily based on a comparison of the flight maneuvers with standard behavioral and control data, while the target score comprehensively considers the actual impact of the flight maneuvers on all aspects of flight, making the evaluation results more objective, comprehensive, and accurate. This provides a more reliable basis for training evaluations and skill assessments of target pilots, aiding in the selection of outstanding target pilots while also providing more targeted guidance for their career development.

[0176] In this embodiment, a pilot driving simulation training method is provided, which can be used for the above-mentioned electronic equipment. Figure 5 FIG. 1 is a flow chart of a pilot driving simulation training method according to an embodiment of the present invention. Figure 5 As shown, the process includes the following steps:

[0177] Step S301: receiving at least one flight training subject corresponding to a target pilot.

[0178] For details about this step, please refer to the above description of step S201 and will not be repeated here.

[0179] Step S302 : obtaining pilot behavior data and flight control data input by the target pilot for each flight training subject.

[0180] Among them, the pilot behavior data includes: pilot hand position, hand movement trajectory, foot position, and foot movement trajectory; flight control data includes at least one parameter of the aircraft control stick, throttle handle, trim handle, and flight control button.

[0181] For details about this step, please refer to the above description of step S202 and will not be repeated here.

[0182] Step S303: Acquire flight parameter data of the target training aircraft based on changes in the pilot behavior data and flight control data.

[0183] For details about this step, please refer to the above description of step S203 and will not be repeated here.

[0184] In step S304 , the flight training subjects, flight parameter data, pilot behavior data, and flight control data are input into a preset evaluation model, and a target score corresponding to the target pilot is output.

[0185] For details about this step, please refer to the above description of step S204 and will not be repeated here.

[0186] Step S305 , storing the flight training subjects, pilot behavior data, flight control data, and target scores in correspondence with the target pilots.

[0187] Specifically, the electronic device can store the pilot behavior data, flight control data, and target score corresponding to the target pilot in correspondence with the target pilot, thereby building a flight file corresponding to the target pilot.

[0188] Step S306 , identifying the flight training subjects, pilot behavior data, flight control data, and target scores, and determining the target pilot's operating habits and operating problems for each flight training subject.

[0189] Specifically, the above step S306 may include the following steps:

[0190] Step S3061: Inputting flight training subjects, pilot behavior data, flight control data, and target scores into a preset behavior analysis model.

[0191] Specifically, the electronic device may input flight training subjects, pilot behavior data, flight control data, and target scores into a preset behavior analysis model.

[0192] In step S3062, the preset behavior analysis model extracts features from the flight training subjects, pilot behavior data, flight control data, and target scores, and outputs the target pilot's operating habits and operating problems for each flight training subject.

[0193] Among them, the preset behavior analysis model is trained based on the historical training data corresponding to multiple training target pilots and the historical operating habits and historical operating problems corresponding to the historical training data; the historical training data includes: historical flight training subjects and historical pilot behavior data, historical flight control data and historical target scores corresponding to the historical flight training subjects; historical operating habits and historical operating problems are obtained based on expert review.

[0194] Specifically, the above step S3062 may include the following steps:

[0195] Step c1: extract basic behavioral features from pilot behavior data.

[0196] Among them, the basic behavior features may include hand behavior features and foot behavior features, wherein the hand behavior features include hand position features, hand movement trajectory features, and hand speed and acceleration features; the foot behavior features include foot position features, foot movement trajectory features, and foot speed and acceleration features.

[0197] Specifically, the pre-set behavioral analysis model analyzes the pilot's hand position data to extract information about the relative hand position within the cockpit. For example, recording the hand's position relative to key control components such as the joystick and throttle at different times can reveal the target pilot's starting position and range of hand movement during the operation. For example, during takeoff, the process and position changes of the hand from its initial position to a specific grip on the joystick can reveal the target pilot's operating habits and preparatory movements. The pre-set behavioral analysis model analyzes hand movement trajectory data to obtain characteristics such as trajectory length, direction, and curvature. For example, by calculating the length of the hand movement trajectory when operating the joystick, a longer trajectory may indicate that the target pilot made larger movements when adjusting the aircraft's attitude. The curvature of the trajectory can reflect the smoothness of the movement; a smaller curvature change indicates a smoother movement, while a smaller curvature change may indicate a less smooth movement. Furthermore, the trajectory's directional information can be extracted to understand the hand's movement trends in different directions, which is very helpful in determining the target pilot's operating habits for controlling the aircraft in different directions. The pre-set behavioral analysis model calculates the speed and acceleration of the hand movement based on the changes in hand position over time. Speed ​​characteristics can reflect the speed of a target pilot's operations. For example, in an emergency, hand movements may accelerate. Acceleration characteristics can reflect the speed of speed changes; larger accelerations may indicate that the target pilot is rapidly adjusting his or her maneuvers. By analyzing these characteristics, we can assess the target pilot's reaction speed and control ability in different situations.

[0198] Similarly, based on the hand behavior characteristics, the preset behavior analysis model extracts the foot behavior characteristics using the same method.

[0199] Step c2: extracting basic operation features from flight control data.

[0200] The basic operation feature includes at least one of the operation force, operation angle, and operation duration of the operation component.

[0201] Specifically, the preset behavior analysis model can obtain multiple valuable features by analyzing the operation force data. For example, the average operation force can reflect the approximate degree of force applied by the target pilot during a specific operation, which helps to evaluate whether the target pilot has appropriate control force on the operating components. Another example is the range of change in the operation force. A larger range of change may indicate that the target pilot's force is not stable enough during operation, or that the force is adjusted to a large extent when responding to different flight conditions. In addition, the rate of change of the operation force, that is, the change in force per unit time, can also be analyzed. This is very helpful for judging the target pilot's operational reaction speed and ability to respond to changes in flight status. In actual applications, if it is found that the operation force of a target pilot's joystick changes too drastically during takeoff, it may be necessary to provide targeted training to improve the smoothness of the operation.

[0202] Furthermore, when extracting control angle features using the pre-set behavioral analysis model, key indicators can be considered. For example, the maximum control angle reflects the maximum angle range a target pilot can achieve during a turn, providing valuable insights into their ability to adjust the aircraft's attitude. The control angle curve shows how the angle changes over time during a turn. By analyzing the curve's shape and slope, we can understand the smoothness and continuity of the target pilot's control. For example, large fluctuations in the control angle curve may indicate a lack of skill or nervousness during the turn. Furthermore, the average control angle across different control phases can be analyzed to compare the performance of different target pilots or the same target pilot under different circumstances. During training, if a target pilot's control stick angle consistently deviates from the standard range during a turn, targeted turning training and guidance can be provided.

[0203] Finally, a pre-set behavioral analysis model extracts operation duration characteristics. For example, average operation duration can reflect the typical time taken by target pilots to perform specific operations. If a target pilot's average operation duration is significantly higher than that of other target pilots, this may indicate that their operation speed is slow and that they need to improve their efficiency. The distribution of operation duration is also worth noting. For example, the standard deviation of operation duration can measure the degree of dispersion of operation duration. A smaller standard deviation indicates relatively stable operation duration, while a larger standard deviation indicates greater fluctuation in operation duration. Furthermore, the proportion of operation duration in different operation phases can be analyzed to understand whether the target pilot's time allocation in each phase is reasonable. During actual flight training, if a target pilot is found to spend too much time on landing operations, it may be necessary to examine whether there are problems with their operation process and conduct corresponding optimization training.

[0204] Step c3: perform correlation analysis on the target score with the basic behavior characteristics and basic operation characteristics to extract score-related features.

[0205] Specifically, the preset behavior analysis model can calculate the correlation coefficient between the target score and each basic behavior feature and basic operation feature based on the correlation analysis method. For example, the Pearson correlation coefficient between the hand operation force and the target score is calculated to measure the degree of linear correlation between the two. If the correlation coefficient is positive and the value is large, it means that an increase in hand operation force may be related to an increase in the target score; conversely, if the correlation coefficient is negative, it indicates that there is an inverse relationship between the two. In this way, features that are significantly correlated with the target score can be preliminarily screened out.

[0206] Then, based on the results of the above association analysis, the preset behavior analysis model will screen out features that are significantly correlated with the target score. These features are the score-related features.

[0207] Step c4: Use sequence analysis algorithms to mine behavioral pattern features in pilot behavior data and flight control data.

[0208] Specifically, the pre-set behavioral analysis model can use frequent sequence mining algorithms to identify pilot behavior data and flight control data, identifying patterns in the operational sequences frequently performed by target pilots during different flight phases or missions. For example, during takeoff, it may be discovered that target pilots frequently follow a sequence of "starting the engine -> checking the instruments -> pushing the throttle -> rotating the wheels."

[0209] In step c5, the extracted basic behavior features, basic operation features, score-related features, and behavior pattern features are clustered with each historical feature corresponding to the historical training data in the preset behavior analysis model.

[0210] Each category represents an operating habit pattern.

[0211] Specifically, the preset behavior analysis model can cluster the extracted basic behavior features, basic operation features, score-related features, and behavior pattern features with each historical feature corresponding to the historical training data in the preset behavior analysis model based on a preset clustering algorithm.

[0212] Among them, the preset clustering algorithm can be a K-Means algorithm, a hierarchical clustering algorithm, DBSCAN (density-based spatial clustering algorithm), etc. The embodiment of the present application does not specifically limit the preset clustering algorithm.

[0213] Step c6: Determine the target pilot's operating habits for each flight training subject based on the clustering results.

[0214] Specifically, the preset behavior analysis model determines the target pilot's operating habits for each flight training subject based on the clustering results.

[0215] In step c7, the extracted basic behavior features, basic operation features, score-related features, and behavior pattern features are compared with the historical features corresponding to the historical training data in the preset behavior analysis model to determine the operation problems of the target pilot in each flight training subject.

[0216] Specifically, the preset behavior analysis model can compare the extracted basic behavior features, basic operation features, scoring-related features, and behavior pattern features with each historical feature corresponding to the historical training data in the preset behavior analysis model, thereby obtaining comparison results corresponding to the basic behavior features, basic operation features, scoring-related features, and behavior pattern features. The comparison results can be the similarities between the basic behavior features, basic operation features, scoring-related features, and behavior pattern features and each historical feature corresponding to the historical training data. Then, based on the comparison results corresponding to the basic behavior features, basic operation features, scoring-related features, and behavior pattern features, at least one operational issue is determined whose similarity with the target pilot exceeds a preset threshold.

[0217] Step S307: Generate a training plan corresponding to the target pilot based on the operating habits and operating problems.

[0218] Specifically, electronic equipment can tailor training plans to target pilots' unique operating habits and specific operational challenges, ensuring that the content and intensity of training are tailored to their individual circumstances. For example, specialized fine-tuning training sessions are designed for pilots accustomed to large-scale control of the joystick. For pilots struggling with timing during flare-out, extensive simulated landing training is provided. Secondly, training programs progress from the basics to the complex. Simple, easily correctable operational issues are addressed first, such as correcting individual incorrect maneuvers within the correct operating procedures, before delving deeper into more complex issues, such as improving comprehensive operational capabilities in complex weather conditions. For example, in the early stages of flight training, the focus is on improving pilots' accuracy in reading and operating basic flight instruments. As training progresses, instrument flight training in complex weather conditions is gradually introduced. Finally, flight theory is closely integrated with practical flight training. Through theoretical learning, such as studying the forces involved in landing an aircraft in aerodynamic principles, pilots understand the scientific basis behind these operations, thereby better mastering practical skills. In practical training, the application of theoretical knowledge is continuously reinforced, such as by putting theoretical knowledge about landing timing and attitude control into practice during actual landing operations.

[0219] For example, specialized simulated flight training is designed to address the large amplitude and lack of precision in operational habits. Within a simulated environment, specific operational tasks are set, requiring pilots to precisely control the angle and force of the control components. For example, in stick control training, it is stipulated that under specific flight conditions, the control stick's operating angle error must be kept within a very small range, and operational accuracy is improved through repeated practice.

[0220] Correct irregularities in pilots' operating habits based on standard flight operating procedures. Create a detailed operating procedure manual containing standard operating steps and timelines for each flight phase, ensuring pilots strictly adhere to it during simulated and actual flight training. For example, during takeoff, clearly define the sequence and operating requirements for each step, from engine startup to the start of the rollout. Through repeated training, pilots develop the habit of standardized operation.

[0221] Finally, flight training in complex weather conditions, such as crosswinds and low visibility, is introduced as appropriate based on the pilots' training progress. These simulated environments comprehensively test and enhance pilots' operational skills, situational awareness, and decision-making abilities. For example, in crosswind conditions, pilots are trained to maintain a stable flight path by adjusting the stick, throttle, and rudder. Furthermore, simulated training for various emergency situations, such as engine failure and aircraft system malfunctions, cultivates pilots' calm response and proper operational habits in emergencies. Training emphasizes the execution of emergency procedures and the understanding of key operational nodes to enhance pilots' ability to handle emergencies.

[0222] The pilot flight simulation training method provided in the embodiments of the present application associates flight training subjects, pilot behavior data, flight control data, and target scores with a target pilot, thereby creating a comprehensive profile of the target pilot. This profile comprehensively records the target pilot's operational performance in various flight training subjects and the corresponding comprehensive evaluation. Through this profile, training institutions or management personnel can gain a comprehensive and systematic understanding of the target pilot's flight capabilities, providing a solid data foundation for subsequent analysis and decision-making. Flight training subjects, pilot behavior data, flight control data, and target scores are input into a pre-set behavioral analysis model. Extracting basic behavioral features from the pilot behavior data can intuitively demonstrate the target pilot's behavior during operation. Extracting basic operational features from the flight control data directly reflects the target pilot's ability to control aircraft control components. Accurate operational force, angle, and duration are crucial to ensuring flight safety and stability. Correlation analysis is performed between the target score and the basic behavioral and operational features to extract score-related features. The target score is a quantitative assessment of the target pilot's overall performance in the flight training subjects. This correlation analysis can identify which basic behavioral and operational features significantly influence the score. Using a sequence analysis algorithm, behavioral pattern features are mined from pilot behavior and flight control data. Once the target pilot's behavioral pattern features are identified, these patterns can be used to predict their operational behavior in specific scenarios. The extracted basic behavioral features, basic operational features, score-related features, and behavioral pattern features are clustered with the historical features corresponding to the historical training data in the pre-set behavioral analysis model. This approach accurately determines the target pilot's operational habits and their similarities and differences with those of other target pilots. Based on the clustering results, the target pilot's operational habits for each flight training subject are determined, ensuring the accuracy of the identified target pilot's operational habits for each flight training subject. The extracted basic behavioral features, basic operational features, score-related features, and behavioral pattern features are compared with the historical features corresponding to the historical training data in the pre-set behavioral analysis model to identify the target pilot's operational problems for each flight training subject, thereby clearly identifying any operational issues encountered by the target pilot.

[0223] Finally, a training plan tailored to the target pilot's operating habits and operational issues is generated. This training plan can include specialized practice sessions, focusing on developing precise control of the operating range. For operational issues (such as improper engine parameter adjustment during takeoff), the training plan can include detailed theoretical explanations and simulated operational exercises to strengthen the pilot's understanding and practical skills of correct operating methods. This targeted training avoids blind training, greatly improving training efficiency and enabling the target pilot to improve their skills in the shortest possible time. Furthermore, unlike general training plans, the training plan tailored to the target pilot's specific circumstances allows for customized training content and schedule. For example, for a target pilot with good operating habits but experiencing operational issues in complex weather conditions, the training plan can focus on simulated training in complex weather environments while also reinforcing their good operating habits. This helps the target pilot leverage their strengths, overcome weaknesses, and achieve personalized career development. By resolving operational issues and optimizing operating habits, the target pilot's operational skills can be significantly improved, providing a strong guarantee for flight safety. Operational issues often pose a potential risk to flight safety, such as improper landing operations that could cause the aircraft to overrun the runway. Through targeted training, these problems are eliminated to ensure that target pilots can operate the aircraft in a standardized and accurate manner during actual flights, reduce the risk of flight accidents caused by human operational errors, and protect the lives and property of passengers and crew members.

[0224] This embodiment also provides a pilot driving simulation training device for implementing the aforementioned embodiments and preferred implementations. Details already described will not be repeated. As used below, the term "module" may refer to a combination of software and / or hardware that implements a predetermined function. While the devices described in the following embodiments are preferably implemented using software, implementation using hardware, or a combination of software and hardware, is also possible and contemplated.

[0225] This embodiment provides a pilot driving simulation training device, such as Figure 6 Shown, including:

[0226] A receiving module 401 is configured to receive at least one flight training subject corresponding to a target pilot;

[0227] The first acquisition module 402 is configured to acquire pilot behavior data and flight control data input by the target pilot for each flight training subject. The pilot behavior data includes: pilot hand position, hand movement trajectory, foot position, and foot movement trajectory; the flight control data includes at least one parameter of the aircraft control stick, throttle handle, trim handle, and flight control buttons.

[0228] The second acquisition module 403 is used to acquire flight parameter data of the target training aircraft based on the pilot behavior data and flight control data changes;

[0229] The first output module 404 is used to input the flight training subjects, flight parameter data, pilot behavior data, and flight control data into a preset evaluation model, and output a target score corresponding to the target pilot.

[0230] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0231] The pilot driving simulation training device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0232] An embodiment of the present invention further provides an electronic device having the above Figure 5 The pilot flight simulation training device shown.

[0233] See also Figure 7 , Figure 7 is a structural diagram of an electronic device provided by an optional embodiment of the present invention, such as Figure 7 As shown, the electronic device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components utilize different buses to communicate with each other and can be installed on a common mainboard or installed in other ways as needed. The processor can process instructions executed in the electronic device, including instructions stored in or on the memory to display graphical information of a GUI on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple electronic devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 7 A processor 10 is taken as an example.

[0234] The electronic device further includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30 and the output device 40 may be connected via a bus or other means. Figure 7 The bus connection is taken as an example.

[0235] The input device 30 can receive input digital or character information and generate key signal input related to user settings and function control of the electronic device, such as a touch screen, a keypad, a mouse, a trackpad, a touch pad, an indicator stick, one or more mouse buttons, a trackball, a joystick, etc. The output device 40 can include a display device, an auxiliary lighting device (e.g., an LED), and a tactile feedback device (e.g., a vibration motor). The above-mentioned display device includes but is not limited to a liquid crystal display, a light emitting diode, a display, and a plasma display. In some optional embodiments, the display device can be a touch screen.

[0236] The embodiment of the present invention also provides a computer-readable storage medium. The above-mentioned method according to the embodiment of the present invention can be implemented in hardware, firmware, or implemented as a computer code that can be recorded in a storage medium, or implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that a computer, a processor, a microprocessor controller or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor or hardware, the method shown in the above embodiment is implemented.

[0237] Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention. Such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A pilot driving simulation training method, characterized in that: The method comprises: Receive at least one flight training subject corresponding to the target pilot; obtaining pilot behavior data and flight control data input by the target pilot for each of the flight training subjects; the pilot behavior data including: pilot hand position, hand movement trajectory, foot position, and foot movement trajectory; and the flight control data including at least one parameter of the aircraft control stick, throttle handle, trim handle, and flight control buttons; Acquiring flight parameter data of the target training aircraft based on changes in the pilot behavior data and flight control data; The flight training subject, the flight parameter data, the pilot behavior data, and the flight control data are input into a preset evaluation model, and a target score corresponding to the target pilot is output.

2. The method according to claim 1, characterized in that The preset evaluation model includes an expert library model and an action evaluation model. The flight training subject, the flight parameter data, the pilot behavior data, and the flight control data are input into the preset evaluation model, and the target score corresponding to the target pilot is output, including: Inputting the flight training subject, the flight parameter data, the pilot behavior data, and the flight control data into the expert library model in the preset evaluation model; The expert database model matches the flight training subject with each matching item in the expert database to determine the standard behavior data and standard control data corresponding to the flight training subject; comparing the standard behavior data and the standard control data with the pilot behavior data and the flight control data, respectively, and outputting initial scores corresponding to the pilot behavior data and the flight control data; The pilot behavior data, the flight control data, the initial score, and the flight parameter data are input into the action evaluation model, and the target score corresponding to the pilot behavior data and the flight control data is output.

3. The method according to claim 2, characterized in that Comparing the standard behavior data and the standard control data with the pilot behavior data and the flight control data, respectively, and outputting initial scores corresponding to the pilot behavior data and the flight control data, includes: comparing the standard parameter data corresponding to the standard behavior data and the standard control data with the flight parameter data corresponding to the pilot behavior data and the flight control data, respectively, to obtain parameter deviations; and / or comparing the action execution sequence corresponding to the standard behavior data and the standard control data with the action execution sequence corresponding to the pilot behavior data and the flight control data, respectively, to obtain an action execution sequence deviation; and / or comparing the execution time corresponding to each action in the standard behavior data and the standard control data with the execution time corresponding to each action in the pilot behavior data and the flight control data to obtain an execution time deviation; The initial scores corresponding to the pilot behavior data and the flight control data are output according to the parameter deviation, and / or the action execution sequence deviation, and / or the execution time deviation.

4. The method according to claim 2, characterized in that Inputting the pilot behavior data, the flight control data, the initial score, and the flight parameter data into the action evaluation model, and outputting the target score corresponding to the pilot behavior data and the flight control data, includes: inputting the pilot behavior data, the flight control data, the initial score, and the flight parameter data into the action evaluation model; The action evaluation model identifies the pilot behavior data, the flight control data, and the flight parameter data, and determines the flight safety, flight stability, flight efficiency, and conformity with preset goals of the target training aircraft based on the pilot behavior data and the flight control data; The initial score is adjusted based on the flight safety, flight stability, flight efficiency, and compliance with preset targets, and the target score corresponding to the pilot behavior data and the flight control data is output.

5. The method according to claim 1, characterized in that The method further comprises: storing the flight training subject, the pilot behavior data, the flight control data, and the target score in correspondence with the target pilot; identifying the flight training subjects, the pilot behavior data, the flight control data, and the target scores, and determining the target pilot's operating habits and operating problems for each of the flight training subjects; A training plan corresponding to the target pilot is generated based on the operating habits and the operating problems.

6. The method according to claim 5, characterized in that The identifying of the flight training subjects, the pilot behavior data, the flight control data, and the target scores to determine the target pilot's operating habits and operating problems for each of the flight training subjects includes: Inputting the flight training subject, the pilot behavior data, the flight control data, and the target score into a preset behavior analysis model; The preset behavior analysis model extracts features from the flight training subjects, the pilot behavior data, the flight control data, and the target scores, and outputs the target pilot's operating habits and operating problems for each flight training subject. The preset behavior analysis model is trained based on historical training data corresponding to multiple training target pilots and historical operating habits and historical operating problems corresponding to the historical training data. The historical training data includes historical flight training subjects and historical pilot behavior data, historical flight control data, and historical target scores corresponding to the historical flight training subjects. The historical operating habits and historical operating problems are obtained based on expert review.

7. The method according to claim 6, characterized in that The preset behavior analysis model extracts features from the flight training subjects, the pilot behavior data, the flight control data, and the target score, and outputs the target pilot's operating habits and operating problems for each flight training subject, including: extracting basic behavioral features from the pilot behavior data; Extracting basic operation characteristics from the flight control data; the basic operation characteristics include at least one of the operation force, operation angle, and operation duration of the operation component; Performing correlation analysis on the target score with the basic behavior characteristics and the basic operation characteristics to extract score-related features; Using a sequence analysis algorithm to mine behavioral pattern features in the pilot behavior data and the flight control data; Clustering the extracted basic behavior features, the basic operation features, the score-related features, and the behavior pattern features with the historical features corresponding to the historical training data in the preset behavior analysis model; wherein each cluster represents an operation habit pattern; determining, based on the clustering results, the operating habits of the target pilot for each of the flight training subjects; The extracted basic behavior features, the basic operation features, the score-related features, and the behavior pattern features are compared with each historical feature corresponding to the historical training data in the preset behavior analysis model to determine the operation problems of the target pilot for each of the flight training subjects.

8. A pilot driving simulation training device, characterized in that: The device comprises: A receiving module, configured to receive at least one flight training subject corresponding to a target pilot; a first acquisition module for acquiring pilot behavior data and flight control data input by the target pilot for each of the flight training subjects; the pilot behavior data including: pilot hand position, hand movement trajectory, foot position, and foot movement trajectory; and the flight control data including at least one parameter of the aircraft control stick, throttle handle, trim handle, and flight control buttons; a second acquisition module, configured to acquire flight parameter data of the target training aircraft based on changes in the pilot behavior data and flight control data; The first output module is used to input the flight training subject, the flight parameter data, the pilot behavior data and the flight control data into a preset evaluation model, and output a target score corresponding to the target pilot.

9. An electronic device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the pilot driving simulation training method according to any one of claims 1 to 7 by executing the computer instructions.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the pilot driving simulation training method according to any one of claims 1 to 7.

Images