A real-time flight training evaluation system for pilots

Abstract

The present invention relates to the technical field of electronic digital data processing, and specifically provides a real-time flight training evaluation system for pilots, including: a data monitoring module, used to monitor the instrument panel data of the tested flight trainee during the flight simulation process; a data acquisition module, used to collect the flight data of the tested flight trainee when the instrument panel data is not within the preset standard threshold range; an automatic identification module, used to train the SVDD model based on the pre-collected flight teacher eye movement data, and construct a hypersphere; and based on the hypersphere, determine whether there is an error in the flight operation of the tested flight trainee, if so, trigger the warning module; the warning module is used to send a warning signal, and analyze the flight data of the tested flight trainee with errors in the flight operation, obtain the analysis results, and complete the real-time flight training evaluation of the tested flight trainee. The technical solution of the present invention can improve the training efficiency of flight trainees.

Description

Technical Field

[0001] The invention belongs to the technical field of electronic digital data processing, and in particular relates to a real-time flight training evaluation system for pilots. Background Art

[0002] During flight training, the take-off and landing pattern is one of the important links of pilots' operating skills. Novice pilots often show irregular data and trajectories in take-off and landing pattern training, which is mainly manifested in unstable control of flight altitude, heading, and speed, and unsmooth coordination between turns and segments. These problems may lead to unsafe flights and do not comply with standard operating procedures. In order to improve the driving level of trainees in take-off and landing pattern flights, it is necessary to correct and improve them through scientific training methods and effective guidance strategies.

[0003] The training methods for novice pilots in the prior art often focus on analyzing their psychological state and operating state during flight learning or testing by collecting brain waves or eye movement data, and formulating the next teaching and training plan from the analysis results. In the actual flight training process, the flight instructor can also point out the student's operational deviations in real time and give corrective suggestions through radio communication or direct guidance in the aircraft, especially by setting flight parameter alarms in advance (such as altitude or speed exceeding the limit) to help students adjust their operations in time. The above process, whether it is data analysis or guidance for novice pilots, is mainly carried out by manpower, and its reliability will be unstable due to factors of human intervention. In addition, the analysis of the state of students during flight in the prior art is all analyzed using the full-course student flight data. Students are generally in a normal operating state in the early stage of flight simulation, which leads to certain interference in the analysis of the nodes where students actually make flight operation errors.

[0004] In order to further improve the training efficiency for novice pilots, ensure the objectivity in the training and data analysis process, and reduce the data interference in the normal flight phase, a real-time flight training evaluation system for pilots is proposed. Summary of the invention

[0005] In order to solve the problems existing in the prior art, the present invention provides a real-time flight training evaluation system for pilots, which can make efficient, timely and accurate judgments on the flight operations of the tested flight trainees.

[0006] To achieve the above object, the present invention provides the following solutions:

[0007] A real-time flight training and evaluation system for pilots, comprising: a data monitoring module, a data acquisition module, an automatic identification module and a warning module;

[0008] The data monitoring module is used to monitor the instrument panel data of the flight trainee during the flight simulation process, and when the instrument panel data is not within the preset standard threshold range, the data acquisition module and the automatic identification module are triggered;

[0009] The data acquisition module is used to collect flight data of the tested flight trainee with a timestamp of T1-T2; wherein the flight data includes a flight real-scene map, a flight trajectory map, a flight hotspot map, and eye movement data in a focus area of ​​interest; the focus area of ​​interest includes an attitude indicator and external visual, an altimeter, a heading indicator, an airspeed indicator, a vertical speed indicator, and a turn indicator; based on the gaze time, the number of gazes, the number of glances, and the glance time of the tested flight trainee on the focus area of ​​interest, the eye movement data of the tested flight trainee with a timestamp of T1-T2 is obtained; wherein T1 is the start timestamp of flight data acquisition since the data acquisition module is triggered, and T2 is the end timestamp of flight data acquisition;

[0010] The automatic identification module is used to train the SVDD model based on the pre-collected eye movement data of the flight instructor to construct a hypersphere; and based on the hypersphere, determine whether the flight operation of the tested flight trainee is wrong, and if so, trigger the warning module;

[0011] The warning module is used to send out warning signals and collect and analyze the data of the tested flight trainee with flight operation errors from the time when the data collection module is triggered to the time when the flight simulation training ends. The timestamp is T1-T E flight data to obtain analysis results; the flight instructor completes the real-time flight training evaluation of the tested flight trainee based on the analysis results; wherein, T E It is the end timestamp of the entire flight data.

[0012] Preferably, the instrument panel data includes altitude data, speed data, heading data and slope data; the preset standard threshold range includes an altitude range of 2050~2150 ft, a speed range of 75~85 knots, a heading range of -5~5°, and a slope range of 0~5°.

[0013] Preferably, the data acquisition module also collects the timestamps of the real flight missions of the flight instructor and the flight trainee during the same flight simulation process as T0-T E The flight instructor’s full flight data includes flight real-scene pictures, flight trajectory pictures, flight hot spots pictures, and full eye movement data in the areas of interest; T0 is the starting timestamp of the full flight data.

[0014] Preferably, the automatic identification module includes:

[0015] An eye movement data capture unit is used to capture the full eye movement data of the flight instructor and obtain the eye movement data of the flight instructor with a timestamp of T1-T2;

[0016] The hypersphere construction unit is used to obtain the hypersphere in the SVDD model based on the flight instructor's eye movement data with timestamp T1-T2 as the training data of the SVDD model; determine whether the eye movement data of the tested flight trainee is within the hypersphere, and if so, there is no error in the flight operation of the tested flight trainee; if not, there is an error in the flight operation of the tested flight trainee, triggering the warning module.

[0017] Preferably, the warning module includes an analysis unit and a warning unit;

[0018] The analysis unit is used to compare and analyze the flight real scene map, flight trajectory map and flight heat map of the tested flight trainee and the flight instructor who have errors in flight operation after the warning module is triggered, generate a curve graph of the percentage of the tested flight trainee's flight instrument interest area use time period, and generate a statistical bar graph of the tested flight trainee's gaze time, gaze number, glance number and glance time on the interest area to obtain analysis results;

[0019] The warning unit is used to project the warning signal, the analysis result and the real-time instrument panel data of the flight trainee during the flight simulation process onto the teaching observation panel of the flight instructor.

[0020] Preferably, the analysis unit further comprises:

[0021] A timestamp reading subunit is used to read the starting timestamp T1 of the flight data of the flight trainee who has an error in flight operation;

[0022] The supplementary data segment acquisition subunit is used to intercept the flight instructor's full flight data with the start timestamp T0 of the flight instructor's full flight data as the starting point and the start timestamp T1 of the flight student's flight data as the end point, and use the flight instructor's flight data with the acquired timestamp T0-T1 as the supplementary data segment;

[0023] The data supplement subunit is used to place the supplementary data segment with timestamp T0-T1 into the data segment with timestamp T1-T2. E Before the flight data of the tested flight trainee is obtained, the timestamp is T0-T E The full flight data of the pilot students being tested;

[0024] Detection subunit, used to detect timestamp T0-T EWhether the length of the full flight data of the tested flight student is consistent with that of the flight instructor. If so, the full flight data of the tested flight student is analyzed; if not, a manual warning is triggered, indicating a data reading error.

[0025] Preferably, in the warning unit, the warning signal includes a buzzer beeping, a teaching observation panel displaying the words "flight operation error" and a flashing warning light.

[0026] Preferably, the teaching observation panel is an HMI panel or a computer screen for human-computer interaction.

[0027] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention proposes a real-time flight training evaluation system for pilots, including a data monitoring module, a data acquisition module, an automatic identification module and a warning module. The data monitoring module collects various flight data in real time through the background data of the simulated flight device. When the flight data exceeds the set standard threshold, the data acquisition module is triggered to start collecting the flight data of the tested flight trainee. This method of triggering data acquisition after setting the standard threshold can effectively reduce the interference of the normal flight stage data when judging the flight operation of the tested flight trainee, and can save the computing resources of the system. By using the data acquisition module to collect the eye movement data of the tested flight trainee in the area of ​​interest, setting the data acquisition length and starting point, and processing the collected data, the automatic identification module identifies whether the current flight operation of the tested flight trainee is qualified through the SVDD method. This method can not only realize the judgment of the flight operation result of the tested flight trainee by the automatic identification module, but also ensure the objectivity of the flight operation judgment of the tested flight trainee. Finally, the results are transmitted to the warning module. After the warning module is triggered by the automatic recognition module, the warning part and the analysis part of the warning module project the student's analysis results and real-time flight status onto the flight teacher's teaching observation panel, so as to assist the teacher in providing real-time targeted flight guidance to the student and ensure that the teacher can implement guidance to the student. The full analysis results obtained by the analysis part can also realize the playback and comparison of the student's complete flight process, so that the teacher can also realize statistical analysis after the flight test, which is convenient for the subsequent formulation of targeted improvement plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0029] Figure 1 The present invention is a flowchart of the implementation of the real-time flight training evaluation system for pilots according to an embodiment of the present invention. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0032] Embodiment 1

[0033] like Figure 1 As shown, in this embodiment, the VR simulated flight program of AW109 is used for attitude training, the flight simulator training is introduced, and the take-off and landing route flight training is carried out after the training and data is collected.

[0034] A real-time flight training and evaluation system for pilots, comprising: a data monitoring module, a data acquisition module, an automatic identification module and a warning module;

[0035] The data monitoring module is used to monitor the instrument panel data of the flight trainee during the flight simulation process. When the instrument panel data is not within the preset standard threshold range, the data acquisition module and the automatic identification module are triggered; a further implementation method is that the instrument panel data includes altitude data, speed data, heading data and slope data; the preset standard threshold range includes an altitude range of 2050~2150 ft, a speed range of 75~85 knots, a heading range of -5~5°, and a slope range of 0~5°. Specifically, when the data monitoring module detects in real time that any value of the altitude, speed, heading, and slope is not within the preset standard threshold range, the data acquisition module is triggered to start collecting the flight data of the flight trainee.

[0036] A data acquisition module is used to collect flight data of the tested flight trainee with a timestamp of T1-T2; wherein the flight data includes a flight real-scene map, a flight trajectory map, a flight hotspot map, and eye movement data in a focus area of ​​interest; the focus area of ​​interest includes an attitude indicator and external visual, an altimeter, a heading indicator, an airspeed indicator, a vertical speed indicator, and a turn indicator; based on the fixation time, the number of fixations, the number of glances, and the glance time of the tested flight trainee on the focus area of ​​interest, the eye movement data of the tested flight trainee with a timestamp of T1-T2 is obtained; wherein T1 is the start timestamp of flight data collection since the data acquisition module is triggered, and T2 is the end timestamp of flight data collection;

[0037] In this embodiment, when one or more of the real-time monitoring data of altitude, speed, heading and slope exceed the preset standard threshold range, it is considered that the current flight of the flight trainee needs to be monitored and guided. At this time, the timestamp T1 at the current time is recorded to trigger the data acquisition module and the automatic identification module.

[0038] Specifically, the data acquisition module collects the flight data of the flight trainee being tested, takes the time point when the data acquisition module is triggered to start running as the data recording starting point, takes 3 minutes as the data collection time length to collect the flight data of the flight trainee being tested, and processes the collected data, recording the deadline as T2, T2=T1+3 (min); while processing and saving the data, mark the timestamp nodes of the starting and ending stages of the data. Use the flight data recording system to save the flight heat map and trajectory map.

[0039] In this embodiment, the data acquisition module also collects the timestamps of the real flight missions of the flight instructor and the flight trainee during the flight simulation process, which are T0-T E The flight instructor’s full flight data includes flight real-scene pictures, flight trajectory pictures, flight hot spots pictures, and full eye movement data in the areas of interest; T0 is the starting timestamp of the full flight data.

[0040] The automatic recognition module is used to train the SVDD model based on the pre-collected eye movement data of the flight instructor and construct a hypersphere; and based on the hypersphere, determine whether there is an error in the flight operation of the tested flight student, and if so, trigger the warning module; a further implementation method is that in the automatic recognition module, the pre-collected eye movement data of the flight instructor is the full-process eye movement data of the real flight mission that is the same as the flight simulation process of the tested flight student, wherein the timestamp of the full-process eye movement data is T0-T E Specifically, before the data monitoring module triggers the data acquisition module and the automatic recognition module, the training set backup data of the automatic recognition module is first collected. In this embodiment, the full eye movement data of three flight instructors in the academy who repeatedly perform the same flight mission with the tested flight trainees on the racecourse to simulate the task allocation in real take-off and landing are collected as the training set backup data of the SVDD model.

[0041] Among them, the eye movement data of the flight instructor in the whole process of focusing on the interest area includes: setting the flight instrument interest area where the pilot's attention is distributed as the attitude instrument and external visual, altimeter, heading indicator, airspeed indicator, vertical speed indicator and turn indicator; extracting the numerical indicator data such as the gaze time, number of gazes, number of glances, and glance time of the flight instructor in each interest area after data processing; using the flight data recording system to save the flight heat map and trajectory map. For the flight instructor's flight data collection as the data of the entire flight, the starting timestamp of the entire data is marked as T0, and the timestamp at the end is marked as T E .

[0042] The above-mentioned data processing method for the tested flight students and flight instructors is to use the flight data recording system to save the numerical data on each instrument panel, and use the eye tracker and video playback technology to save the data on the area of ​​interest, the real-life flight map, the trajectory map and the heat map.

[0043] A further embodiment is that the automatic identification module comprises:

[0044] The eye movement data capture unit is used to capture the flight instructor's full eye movement data to obtain the flight instructor's eye movement data with timestamp T1-T2; specifically, the eye movement data of the flight instructor in the entire interest area is collected under the same flight mission, and all time nodes corresponding to the eye movement data are retained; after the data acquisition module is started and triggered, the eye movement data of the flight instructor at the corresponding time point is triggered at the same time, the eye movement data of the flight instructor is automatically captured, and corresponding data processing is performed, that is, the data of the T1-T2 time period in the flight instructor's full time period data is captured to ensure that the data length of the training set is consistent with the data length of the test set.

[0045] The hypersphere construction unit is used to obtain the hypersphere in the SVDD model based on the flight instructor's eye movement data with timestamp T1-T2 as the training data of the SVDD model; determine whether the eye movement data of the tested flight trainee is within the hypersphere, and if so, there is no error in the flight operation of the tested flight trainee; if not, there is an error in the flight operation of the tested flight trainee, triggering the warning module.

[0046] In this embodiment, the steps of constructing the SVDD model hypersphere and using the hypersphere to judge the flight operation of the flight trainee are as follows:

[0047] First, define a hypersphere with a radius of r and a center of o. The training set samples constructed by the flight instructor’s eye movement data contained in the hypersphere are , the total number of target class samples is M, and the structural error of the hypersphere is defined as:

[0048] , (1)

[0049] The minimization constraint condition for defining the distance from each sample point to the center of the sphere o is:

[0050] , (2)

[0051] Then the relaxation factor ξ is introduced i , let a small number of target class samples be distributed outside the hypersphere to minimize the hypersphere. The minimization problem formula is expressed as:

[0052] , (3)

[0053] In the formula, C is the penalty parameter. In order to find the optimal solution of the hypersphere that meets the conditions, it is necessary to introduce the Lagrange multiplier, and the constructed Lagrange equation is as follows:

[0054] , (4)

[0055] Where β and γ are Lagrange multipliers for r, o, ξ in formula (4) i Taking partial derivatives we can get:

[0056] , (5)

[0057] Substituting formula (5) into formula (6), we can get the optimization function:

[0058] , (6)

[0059] The samples on the boundary and outside of the hypersphere are called support vectors and are defined as y s , define the center of the hypersphere as:

[0060] , (7)

[0061] The radius of the hypersphere can be obtained by the distance from any support vector on the hypersphere boundary to the center o: , (8)

[0062] For the new sample y consisting of the eye movement data of the tested flight trainees z The distance d to the center o of the hypersphere is calculated as:

[0063] , (9)

[0064] Determine the relationship between d and r. If d is less than or equal to r, the sample of the flight trainee being tested is located inside or on the surface of the hypersphere, and the flight operation is determined to be qualified. If d is greater than r, the sample of the flight trainee being tested is located outside the hypersphere, and the flight operation is determined to be unqualified. The greater the value of d exceeds the value of r, the greater the operational error of the trainee.

[0065] In this embodiment, in the step where the automatic identification module identifies whether the current flight operation of the tested flight trainee is qualified by the SVDD method, if the calculated data point of the current tested flight trainee is located in the hypersphere, the current flight operation is determined to be qualified, the system does not trigger the automatic identification module and deletes the currently collected data segment of the tested flight trainee and jumps back to the data monitoring module to continue to re-monitor the data on the instrument panel. When the data acquisition module and the automatic identification module are triggered again next time, the time node corresponding to the timestamp of starting to collect data T1 is recalibrated, and the above data collection and flight operation judgment process is repeated. This method can not only improve the accuracy of the judgment of the student's operation, but also reduce the manpower consumption of the teacher's guidance, and avoid the waste of the teacher's guidance attention due to the influence of the normal flight operation segment data on the accuracy of the analysis model judgment, or the misjudgment of an operation that is not enough to cause a flight operation error.

[0066] The warning module is used to send out warning signals and collect and analyze the data of the tested flight trainee with flight operation errors from the time when the data collection module is triggered to the time when the flight simulation training ends. The timestamp is T1-T E flight data to obtain analysis results; the flight instructor completes the real-time flight training evaluation of the tested flight trainee based on the analysis results; wherein, T E In this embodiment, after the warning module is triggered, the data acquisition module is used to collect the flight trainee’s data from the time when the data acquisition module is triggered to the time when the flight simulation training ends. E The flight data collected by the analysis unit is time stamped as T1-T E In particular, in the analysis unit, this embodiment analyzes the flight data with timestamps T1-T E The flight data was supplemented, and further analysis was done on the supplemented flight data of the flight trainees.

[0067] In this embodiment, the evaluation is, on the one hand, the embodiment of the warning signal of the warning module, and on the other hand, the analysis unit displays the entire instructor data and the synthesized flight trainee flight data through a visual display. The training evaluation is manually evaluated by manually observing the entire data view.

[0068] A further implementation is that the warning module includes an analysis unit and a warning unit;

[0069] an analysis unit, for comparing and analyzing the flight real scene map, flight trajectory map and flight heat map of the tested flight trainee and the flight instructor who have errors in flight operation after the self-warning module is triggered, generating a curve graph of the percentage of the tested flight trainee's flight instrument interest area usage time period, and generating a statistical bar graph of the tested flight trainee's gaze time, gaze number, glance number and glance time on the interest area, to obtain analysis results;

[0070] The warning unit is used to project warning signals, analysis results and real-time instrument panel data of the flight student being tested during the flight simulation process onto the flight instructor's teaching observation panel, so as to assist the instructor in providing real-time targeted flight guidance to the student.

[0071] A further embodiment is that in the warning unit, the warning signal includes a buzzer beeping, a teaching observation panel displaying the words "flight operation error" and a flashing warning light, and the position of the simulation device corresponding to the student who made the operation error and the student information can be displayed accordingly.

[0072] A further implementation method is that the teaching observation panel is an HMI panel or a computer screen for human-computer interaction to facilitate manual error correction, data storage, playback and other experimental operations.

[0073] In the present embodiment, since the flight eye movement data collected by the flight trainee is not the full data of the complete flight process, directly comparing it with the full data of the flight instructor will result in excessive deviation. Therefore, in the present embodiment, in order to reduce this deviation, the flight data of the flight trainee in the front section that has not been detected by the system as a flight operation error is supplemented with the standard data of the flight instructor.

[0074] A further embodiment is that the analysis unit further comprises:

[0075] A timestamp reading subunit is used to read the starting timestamp T1 of the flight data of the flight trainee who has an error in flight operation;

[0076] The supplementary data segment acquisition subunit is used to intercept the flight instructor's full flight data with the start timestamp T0 of the flight instructor's full flight data as the starting point and the start timestamp T1 of the flight student's flight data as the end point, and use the flight instructor's flight data with the acquired timestamp T0-T1 as the supplementary data segment;

[0077] The data supplement subunit is used to place the supplementary data segment with timestamp T0-T1 into the data segment with timestamp T1-T2. E Before the flight data of the tested flight trainee is obtained, the timestamp is T0-T E The full flight data of the pilot students being tested;

[0078] Detection subunit, used to detect timestamp T0-T E Whether the length of the full flight data of the tested flight student is consistent with that of the flight instructor. If so, the full flight data of the tested flight student is analyzed; if not, a manual warning is triggered, indicating a data reading error.

[0079] After supplementing the data of the flight trainee, the analysis unit of the warning module analyzes and compares the supplemented data of the flight trainee with the full flight data of the flight instructor, and analyzes the trainee's data performance and trajectory deviation during flight through visual analysis of the flight trajectory.

[0080] In addition, the VR flight simulation program can be used to observe the current control input from an external perspective. Using this external perspective, teachers can also effectively help flight trainees recognize the effects of control inputs such as pitch up, pitch down, and bank changes on the helicopter when providing guidance.

[0081] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A real-time flight training evaluation system for pilots, characterized in that: include: Data monitoring module, data acquisition module, automatic identification module and warning module; The data monitoring module is used to monitor the instrument panel data of the flight trainee during the flight simulation process, and when the instrument panel data is not within the preset standard threshold range, the data acquisition module and the automatic identification module are triggered; The data acquisition module is used to collect flight data of the tested flight trainee with a timestamp of T1-T2; wherein the flight data includes a flight real-scene map, a flight trajectory map, a flight hotspot map, and eye movement data in a focus area of ​​interest; the focus area of ​​interest includes an attitude indicator and external visual, an altimeter, a heading indicator, an airspeed indicator, a vertical speed indicator, and a turn indicator; based on the gaze time, the number of gazes, the number of glances, and the glance time of the tested flight trainee on the focus area of ​​interest, the eye movement data of the tested flight trainee with a timestamp of T1-T2 is obtained; wherein T1 is the start timestamp of flight data acquisition since the data acquisition module is triggered, and T2 is the end timestamp of flight data acquisition; The automatic identification module is used to train the SVDD model based on the pre-collected eye movement data of the flight instructor to construct a hypersphere; and based on the hypersphere, determine whether the flight operation of the tested flight trainee is wrong, and if so, trigger the warning module; The process of constructing the SVDD model hypersphere and the steps of using the hypersphere to judge the flight operation of the flight trainees are as follows: First, define a hypersphere with a radius of r and a center of o. The training set samples constructed by the flight teacher's eye movement data contained in the hypersphere are , the total number of target class samples is M, and the structural error of the hypersphere is defined as: ， （1） The minimization constraint condition for defining the distance from each sample point to the center of the sphere o is: ， （2） Then the relaxation factor ξ is introduced i , let the target class samples be distributed outside the hypersphere to minimize the hypersphere. The minimization problem formula is expressed as: ， （3） In the formula, C is the penalty parameter; in order to find the optimal solution of the hypersphere that meets the conditions, it is necessary to introduce the Lagrange multiplier, and the constructed Lagrange equation is as follows: ，（4） Where β and γ are Lagrange multipliers for r, o, ξ in formula (4) i Taking partial derivatives we get: ， （5） Substituting formula (5) into formula (6), we can get the optimization function: ， （6） The samples on the boundary and outside of the hypersphere are called support vectors and are defined as y s , define the center of the hypersphere as: ， （7） The radius of the hypersphere is obtained by the distance from any support vector on the hypersphere boundary to the center o: ， （8） For the new sample y consisting of the eye movement data of the tested flight trainees z The distance d to the center o of the hypersphere is calculated as: ， （9） Determine the magnitude relationship between d and r. If d is less than or equal to r, the sample of the flight trainee being tested is located inside or on the surface of the hypersphere, and the flight operation is determined to be qualified. If d is greater than r, the sample of the flight trainee being tested is located outside the hypersphere, and the flight operation is determined to be unqualified. The greater the value of d exceeds the value of r, the greater the operational error of the flight trainee being tested. The warning module is used to send out warning signals and collect and analyze the data of the tested flight trainee with flight operation errors from the time when the data collection module is triggered to the time when the flight simulation training ends. The timestamp is T1-T E flight data to obtain analysis results; the flight instructor completes the real-time flight training evaluation of the tested flight trainee based on the analysis results; wherein, T E It is the end timestamp of the entire flight data.

2. The real-time flight training evaluation system for pilots according to claim 1, characterized in that: The instrument panel data includes altitude data, speed data, heading data and slope data; the preset standard threshold range includes an altitude range of 2050~2150 ft, a speed range of 75~85 knots, a heading range of -5~5°, and a slope range of 0~5°.

3. The real-time flight training evaluation system for pilots according to claim 1, characterized in that: The data acquisition module also collects the timestamps of the real flight missions of the flight instructor and the flight trainee during the flight simulation process, which are T0-T E The flight instructor’s full flight data includes flight real-scene pictures, flight trajectory pictures, flight hot spots pictures, and full eye movement data in the areas of interest; T0 is the starting timestamp of the full flight data.

4. The real-time flight training evaluation system for pilots according to claim 3, characterized in that: The automatic identification module comprises: An eye movement data capture unit is used to capture the flight instructor's full eye movement data to obtain the flight instructor's eye movement data with a timestamp of T1-T2; The hypersphere construction unit is used to obtain the hypersphere in the SVDD model based on the flight instructor's eye movement data with timestamp T1-T2 as the training data of the SVDD model; determine whether the eye movement data of the tested flight trainee is within the hypersphere, and if so, there is no error in the flight operation of the tested flight trainee; if not, there is an error in the flight operation of the tested flight trainee, triggering the warning module.

5. The real-time flight training evaluation system for pilots according to claim 3, characterized in that: The warning module includes an analysis unit and a warning unit; The analysis unit is used to compare and analyze the flight real scene map, flight trajectory map and flight heat map of the tested flight trainee and the flight instructor who have errors in flight operation after the warning module is triggered, generate a curve graph of the percentage of the tested flight trainee's flight instrument interest area use time period, and generate a statistical bar graph of the tested flight trainee's gaze time, gaze number, glance number and glance time on the interest area to obtain analysis results; The warning unit is used to project the warning signal, the analysis result and the real-time instrument panel data of the flight trainee during the flight simulation process onto the teaching observation panel of the flight instructor.

6. The real-time flight training evaluation system for pilots according to claim 5, characterized in that: The analysis unit also includes: A timestamp reading subunit is used to read the starting timestamp T1 of the flight data of the tested flight trainee whose flight operation has errors; The supplementary data segment acquisition subunit is used to intercept the flight instructor's full flight data with the start timestamp T0 of the flight instructor's full flight data as the starting point and the start timestamp T1 of the flight student's flight data as the end point, and use the flight instructor's flight data with the acquired timestamp T0-T1 as the supplementary data segment; The data supplement subunit is used to place the supplementary data segment with timestamp T0-T1 into the data segment with timestamp T1-T2. E Before the flight data of the tested flight trainee is obtained, the timestamp is T0-T E The full flight data of the pilot students being tested; Detection subunit, used to detect timestamp T0-T E Whether the length of the full flight data of the tested flight student is consistent with that of the flight instructor. If so, the full flight data of the tested flight student is analyzed; if not, a manual warning is triggered, indicating a data reading error.

7. The real-time flight training evaluation system for pilots according to claim 6, characterized in that: In the warning unit, the warning signal includes the beeping of a buzzer, the display of the words "flight operation error" on the teaching observation panel, and the flashing of a warning light.

8. The real-time flight training evaluation system for pilots according to claim 6, characterized in that: The teaching observation panel is an HMI panel or a computer screen for human-computer interaction.

Images