Pilot training simulator based on big data and machine learning

Abstract

The present invention relates to the field of flight training simulators, specifically to a pilot training simulator based on big data and machine learning. This simulator collects pilot head movement data in real time, combines it with machine algorithms to dynamically correct the field of view, and reduces simulator viewing angle delay and distortion issues. Furthermore, by combining high-frequency and low-frequency motions of motion simulation components and simulated seats, it achieves a simulation effect closer to that of a real flight attitude. Supported by long-term flight simulation data, it generates simulation change curves using big data, which intuitively display training progress trends. This allows for the training effects of different pilots to be determined, helping to formulate personalized training plans and reducing the cost of pilot simulation training.

Description

Technical Field

[0001] The present invention relates to the field of flight training simulators, in particular to a pilot training simulator based on big data and machine learning. Background Art

[0002] The current model of flight training is basically to use computer-assisted training equipment to conduct classroom instruction, followed by integrated procedure trainer or flight training device training to conduct cockpit familiarization, equipment operation and flight procedure training;

[0003] Based on the training requirements of different aircraft types, flight training equipment mainly includes high-level training equipment such as full flight simulators (FFS) and flight training devices (FTDs), and low-level training equipment such as integrated procedure trainers (IPTs). FFSs are the most realistic trainers that simulate the entire cockpit of a specific aircraft type, including motion, visual, and sound systems. FTDs are FFSs that do not provide motion systems or external cockpit visual systems. IPTs use a touchscreen virtual cockpit and have the advantage of being low-cost compared to other training equipment.

[0004] Due to the high production and training costs of FTD in all aspects from research and development, procurement to operation and maintenance, the training time allowed for a single training is short, and there is a lack of training time to fully familiarize oneself with the cockpit, resulting in a long training time occupied by the simulator. Therefore, in order to maximize the efficiency of simulation training, it is necessary to collect pilot training data more comprehensively and scientifically during the pilot simulation training process to evaluate the training status of different pilots, so as to more reasonably arrange the pilot training schedule and training types, and improve the operating efficiency of flight simulation equipment at all stages. To this end, this application proposes a solution. Summary of the Invention

[0005] In the present invention, by analyzing flight operation data in a standardized manner, pilot operation defects can be quickly identified, reducing subjective errors in manual evaluation. Long-term flight simulation data is used as support, and simulation change curves are generated by big data. The simulation change curves intuitively display the training progress trend, thereby obtaining the training effects of different pilots, helping to formulate personalized training plans, and reducing the cost of pilot simulation training. In order to address the technical defects proposed in the background technology, a pilot training simulator based on big data and machine learning is now proposed.

[0006] The object of the present invention can be achieved by the following technical solution: A pilot training simulator based on big data and machine learning includes a fixed base, a motion simulation component is installed above the fixed base, and the top of the motion simulation component is movably connected to an outer cabin;

[0007] A simulated seat is installed inside the outer cabin, and an interactive display cabin and a simulated operating console are also installed inside the outer cabin. The simulated seat can simulate shaking and tilt adjustment. A camera component is integrated on the interactive display cabin. The motion simulation component can simulate the flight attitude of the aircraft by performing high-frequency reciprocating expansion and contraction.

[0008] The motion simulation component, interactive display cabin and simulation seat are all controlled by a pilot training simulation system;

[0009] The pilot training simulation system includes a motion acquisition module, a field of view transformation module, a dynamic output module, a simulation control module, and a statistical evaluation module. The motion acquisition module can acquire the head movements of the simulated trainee, obtain the field of view angle based on the acquisition results, and send the field of view angle to the field of view transformation module. The field of view transformation module calculates the field of view angle based on the field of view and corrects the display image on the interactive display cabin.

[0010] The simulation control module is used to collect flight simulation data, perform standardized analysis on the collected flight simulation data, and obtain simulation training evaluation results. The statistical evaluation module performs statistics on the simulation training evaluation results obtained each time, generates a simulation change curve based on the statistical results, realizes the evaluation of the training effect, and outputs the evaluation results of the training effect.

[0011] As a preferred embodiment of the present invention, the dynamic output module is connected to the simulation operating console, and obtains the flight simulation posture through the simulation operating console, and sends the flight simulation posture to the motion simulation component and the simulation seat. After obtaining the flight simulation posture, the motion simulation component analyzes the flight simulation posture to obtain low-frequency jitter and posture simulation, and controls the telescopic cylinder to complete the action within the set response time. The simulation seat analyzes according to the flight simulation posture to obtain a high-frequency jitter effect, and outputs the high-frequency jitter effect to realize jitter simulation.

[0012] As a preferred embodiment of the present invention, the motion acquisition module acquires the motion of the simulated trainee through a camera, imports the acquired images into a preset image processing model for processing, marks the head position of the simulated trainee from the images, and further marks the eyeball images in the head position to obtain the eyeball position;

[0013] The motion acquisition module uses the midpoint of the interactive display cabin in the vertical direction as the standard height and draws a horizontal line along the standard height. The motion acquisition module uses the midpoint of the interactive display cabin in the horizontal direction as the standard distance and draws a vertical line along the standard distance. The motion acquisition module records the horizontal line and the vertical line as the standard dividing line.

[0014] As a preferred embodiment of the present invention, after obtaining the eyeball position, the motion acquisition module connects the eyeball position and the horizontal line of the standard dividing line in a vertical plane to obtain a horizontal connecting line, connects the eyeball position and the numerical line of the standard dividing line in the horizontal plane to obtain a vertical connecting line, and then connects the eyeball position and the interactive display cabin along the normal direction to obtain a normal connecting line.

[0015] The motion acquisition module calculates the angle between the horizontal connecting line and the normal connecting line, records it as the horizontal angle, and records the angle between the vertical connecting line and the normal connecting line as the vertical angle. The motion acquisition module records the vertical angle and the horizontal angle as the field of view angle.

[0016] As a preferred embodiment of the present invention, after obtaining the field of view angle, the field of view transformation module first obtains an initial display screen, which is a panoramic screen, and displays a portion of the area directly in front through the interactive display cabin. The field of view transformation module adjusts the screen in opposite directions according to the horizontal angle and the vertical angle in the field of view angle, so that the screen area displayed by the interactive display cabin moves to simulate the real screen area.

[0017] As a preferred embodiment of the present invention, the flight simulation data acquired by the simulation control module includes a flight trajectory, a flight risk, and a risk response time. The simulation control module compares the flight trajectory with a preset trajectory to obtain an overlap rate between the flight trajectory and the preset trajectory.

[0018] The simulation control module counts the number of flight risk occurrences to obtain the total number of risks;

[0019] The simulation management and control module counts the risk response time to obtain the longest risk response time and the average risk response time, and makes independent threshold judgments on the longest risk response time and the average risk response time, and generates a risk response delay signal and a risk response slow signal according to the judgment results.

[0020] As a preferred embodiment of the present invention, the simulation management and control module records the trajectory overlap rate, the total number of risks, the longest risk response time and the average risk response time as simulation training evaluation data, and sends the simulation training evaluation data to the statistical evaluation module.

[0021] As a preferred embodiment of the present invention, the statistical evaluation module assigns different data in the simulation training evaluation data received each time to different evaluation curves to form a trajectory coincidence rate change curve, a total risk number change curve, a longest risk response time change curve, and an average risk response time change curve;

[0022] The statistical evaluation module fits the changing trends of different curves to obtain the changing trends of trajectory overlap rate, total risk times, longest risk response time and average risk response time, and records them for output as training effect evaluation.

[0023] Compared with the prior art, the present invention has the following beneficial effects:

[0024] 1. This invention addresses the issues of traditional simulator perspective delay and distortion by collecting real-time pilot head movement data and combining it with machine algorithms to dynamically correct the field of view. Furthermore, by combining high-frequency and low-frequency motion of motion simulation components and simulated seats, and linking data-driven dynamic output modules, the motion frequency limit of mechanical simulators is increased, achieving a simulation effect that is closer to the actual flight attitude.

[0025] 2. The present invention uses standardized analysis of flight operation data to quickly identify pilot operational deficiencies and reduce subjective errors in manual assessments. Long-term flight simulation data is used as support to generate simulation change curves based on big data. The simulation change curves intuitively display training progress trends, allowing for the training effects of different pilots to be determined. This helps develop personalized training plans, shortens the training cycle, and reduces the cost of pilot simulation training. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] To facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0027] Figure 1 This is a schematic diagram of the main structure of the present invention;

[0028] Figure 2 Schematic diagram of the interactive display cabin structure of the present invention.

[0029] Figure 3 is a system block diagram of the present invention;

[0030] Figure 4 It is a system flow chart of the present invention.

[0031] In the figure: 1. Outer cabin; 2. Motion simulation component; 3. Fixed base; 4. Interactive display cabin; 5. Simulation seat; 6. Simulation operating console. DETAILED DESCRIPTION

[0032] The following is a clear and complete description of the technical solutions of the present invention in conjunction with the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0033] Example 1: Please refer to Figure 1 - Figure 4 As shown, a pilot training simulator based on big data and machine learning includes a fixed base 3, a motion simulation component 2 is mounted above the fixed base 3, and an outer cabin 1 is movably connected to the top of the motion simulation component 2;

[0034] A simulated seat 5 is installed inside the outer cabin 1. An interactive display cabin 4 and a simulated operating console 6 are also installed inside the outer cabin 1. The simulated seat 5 can simulate shaking and tilt adjustment. A camera component is integrated on the interactive display cabin 4. The motion simulation component 2 can simulate the flight attitude of the aircraft by performing high-frequency reciprocating expansion and contraction.

[0035] The motion simulation component 2, the interactive display cabin 4 and the simulation seat 5 are all controlled by a pilot training simulation system, which includes a motion acquisition module, a field of view transformation module, a dynamic output module, a simulation control module and a statistical evaluation module.

[0036] The dynamic output module in the pilot training simulation system is connected to the simulation console 6, and obtains the flight simulation attitude through the simulation console 6, and sends the flight simulation attitude to the motion simulation component 2 and the simulation seat 5. After obtaining the flight simulation attitude, the motion simulation component 2 analyzes the flight simulation attitude to obtain low-frequency jitter and attitude simulation, and controls the telescopic cylinder to complete the action within the set response time. The simulation seat 5 analyzes the flight simulation attitude to obtain high-frequency jitter effect, and outputs the high-frequency jitter effect to realize jitter simulation.

[0037] Example 2: Please refer to Figure 1 - Figure 4 As shown, the pilot training simulation system includes a motion acquisition module, a field of view transformation module, a dynamic output module, a simulation control module, and a statistical evaluation module;

[0038] The motion acquisition module uses the midpoint of the interactive display cabin 4 in the vertical direction as the standard height and draws a horizontal line along the standard height. The motion acquisition module uses the midpoint of the interactive display cabin 4 in the horizontal direction as the standard distance and draws a vertical line along the standard distance. The motion acquisition module records the horizontal line and the vertical line as the standard dividing line.

[0039] The motion acquisition module collects the simulated trainee's motions through a camera, imports the collected images into a preset image processing model for processing, marks the simulated trainee's head position from the image, and further marks the eyeball image in the head position to obtain the eyeball position;

[0040] After obtaining the eyeball position, the motion acquisition module connects the eyeball position and the horizontal line of the standard dividing line in the vertical plane to obtain a horizontal connecting line, connects the eyeball position and the vertical line of the standard dividing line in the horizontal plane to obtain a vertical connecting line, and then connects the eyeball position and the interactive display cabin 4 along the normal direction to obtain a normal connecting line;

[0041] The motion acquisition module calculates the angle between the horizontal connecting line and the normal connecting line and records it as the horizontal angle, and records the angle between the vertical connecting line and the normal connecting line as the vertical angle. The motion acquisition module records the vertical angle and the horizontal angle as the field of view angle;

[0042] The motion acquisition module sends the field of view angle to the field of view transformation module, and the field of view transformation is calculated based on the field of view angle, specifically:

[0043] After obtaining the field of view angle, the field of view transformation module first obtains the initial display image, which is a panoramic image, and displays a portion of the area directly in front of the interactive display cabin 4. The field of view transformation module adjusts the image in opposite directions according to the horizontal angle and the vertical angle in the field of view angle, so that the image area displayed by the interactive display cabin 4 moves, simulating the real image area, and realizing display correction of the image on the interactive display cabin 4, thereby improving the authenticity of the image displayed on the interactive display cabin 4;

[0044] The simulation control module is used to collect flight simulation data and perform standardized analysis on the collected flight simulation data. The flight simulation data obtained by the simulation control module includes flight trajectory, flight risk, and risk response time. The simulation control module compares the flight trajectory with the preset trajectory to obtain the overlap rate between the flight trajectory and the preset trajectory;

[0045] The simulation control module counts the number of flight risks and obtains the total number of risks;

[0046] The simulation control module counts the risk response time, obtains the longest risk response time and the average risk response time, and makes independent threshold judgments on the longest risk response time and the average risk response time. If the longest risk response time is greater than the set threshold, a risk response delay signal is generated. If the longest risk response time is not greater than the set threshold, no response is made. If the average risk response time is greater than the set threshold, a risk response slow signal is generated. If the average risk response time is not greater than the set threshold, no response is made.

[0047] The simulation control module records the risk response delay signal and the risk response slow signal as the simulation training evaluation result, and directly outputs it after the training is completed;

[0048] The simulation control module records the trajectory overlap rate, total number of risks, longest risk response time, and average risk response time as simulation training evaluation data, and sends the simulation training evaluation data to the statistical evaluation module;

[0049] The statistical evaluation module collects statistics on the simulation training evaluation results obtained each time, and assigns different data in the simulation training evaluation data received each time to different evaluation curves, forming a trajectory overlap rate change curve, a total risk number change curve, a longest risk response time change curve, and an average risk response time change curve;

[0050] The statistical evaluation module fits the changing trends of different curves to obtain the changing trends of trajectory overlap rate, total risk times, longest risk response time and average risk response time. When the changing trend of trajectory overlap rate is positive, the changing trend of total risk times is negative, the changing trend of longest risk response time and the changing trend of average risk response time are negative, it indicates that the training effect is excellent and the pilot training performance has improved. When the changing trend of trajectory overlap rate is negative, the changing trend of total risk times is positive, the changing trend of longest risk response time and the changing trend of average risk response time are positive, it indicates that the training result is poor and the pilot training performance has declined, and it is recorded as training effect evaluation and output.

[0051] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to specific embodiments. Obviously, many modifications and variations are possible based on the contents of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A pilot training simulator based on big data and machine learning, comprising a fixed base (3), characterized in that: A motion simulation component (2) is installed above the fixed base (3), and the top of the motion simulation component (2) is movably connected to the outer cabin (1); A simulation seat (5) is installed inside the outer cabin (1), and an interactive display cabin (4) and a simulation operating console (6) are also installed inside the outer cabin (1). The simulation seat (5) can perform shaking simulation and tilt adjustment simulation. A camera component is integrated on the interactive display cabin (4). The motion simulation component (2) can simulate the flight attitude of the aircraft by performing high-frequency reciprocating expansion and contraction. The motion simulation component (2), the interactive display cabin (4) and the simulation seat (5) are all controlled by a pilot training simulation system; The pilot training simulation system includes a motion acquisition module, a field of view conversion module, a dynamic output module, a simulation control module and a statistical evaluation module. The motion acquisition module can collect the head motion of the simulated training personnel, obtain the field of view angle according to the acquisition result, and send the field of view angle to the field of view conversion module. The field of view conversion module calculates according to the field of view angle and performs display correction on the picture on the interactive display cabin (4); The simulation control module is used to collect flight simulation data, perform standardized analysis on the collected flight simulation data, and obtain simulation training evaluation results. The statistical evaluation module collects statistics on the simulation training evaluation results obtained each time, generates a simulation change curve based on the statistical results, evaluates the training effect, and outputs the evaluation results of the training effect; The simulation control module records the trajectory overlap rate, the total number of risks, the longest risk response time and the average risk response time as simulation training evaluation data, and sends the simulation training evaluation data to the statistical evaluation module; The statistical evaluation module assigns different data in the simulation training evaluation data received each time to different evaluation curves to form a trajectory overlap rate change curve, a total risk number change curve, a longest risk response time change curve, and an average risk response time change curve; The statistical evaluation module fits the changing trends of different curves to obtain the changing trends of trajectory overlap rate, total risk times, longest risk response time and average risk response time, and records them for output as training effect evaluation.

2. The pilot training simulator based on big data and machine learning according to claim 1, characterized in that: The dynamic output module is connected to the simulation console (6), and obtains the flight simulation posture through the simulation console (6), and sends the flight simulation posture to the motion simulation component (2) and the simulation seat (5). After obtaining the flight simulation posture, the motion simulation component (2) analyzes the flight simulation posture to obtain low-frequency jitter and posture simulation, and controls the telescopic cylinder to complete the action within a set response time. The simulation seat (5) analyzes the flight simulation posture to obtain a high-frequency jitter effect, and outputs the high-frequency jitter effect to achieve jitter simulation.

3. The pilot training simulator based on big data and machine learning according to claim 1, characterized in that: The motion acquisition module acquires the motion of the simulated trainee through a camera, imports the acquired images into a preset image processing model for processing, marks the head position of the simulated trainee from the images, and further marks the eyeball images in the head position to obtain the eyeball position; The motion acquisition module uses the midpoint of the interactive display cabin (4) in the vertical direction as a standard height and draws a horizontal line along the standard height. The motion acquisition module uses the midpoint of the interactive display cabin (4) in the horizontal direction as a standard distance and draws a vertical line along the standard distance. The motion acquisition module records the horizontal line and the vertical line as a standard dividing line.

4. The pilot training simulator based on big data and machine learning according to claim 3, characterized in that: After obtaining the eyeball position, the motion acquisition module connects the eyeball position and the horizontal line of the standard dividing line in a vertical plane to obtain a horizontal connecting line, connects the eyeball position and the vertical line of the standard dividing line in a horizontal plane to obtain a vertical connecting line, and then connects the eyeball position and the interactive display cabin (4) along the normal direction to obtain a normal connecting line; The motion acquisition module calculates the angle between the horizontal connecting line and the normal connecting line, records it as the horizontal angle, and records the angle between the vertical connecting line and the normal connecting line as the vertical angle. The motion acquisition module records the vertical angle and the horizontal angle as the field of view angle.

5. The pilot training simulator based on big data and machine learning according to claim 1, characterized in that: After obtaining the field of view angle, the field of view transformation module first obtains an initial display screen, which is a panoramic screen, and displays a portion of the area directly in front through the interactive display cabin (4). The field of view transformation module adjusts the screen in opposite directions according to the horizontal angle and the vertical angle in the field of view angle, so that the screen area displayed by the interactive display cabin (4) moves, simulating a real screen area.

6. The pilot training simulator based on big data and machine learning according to claim 1, characterized in that: The flight simulation data acquired by the simulation control module includes a flight trajectory, a flight risk, and a risk response time. The simulation control module compares the flight trajectory with a preset trajectory to obtain an overlap rate between the flight trajectory and the preset trajectory. The simulation control module counts the number of flight risk occurrences to obtain the total number of risks; The simulation management and control module counts the risk response time to obtain the longest risk response time and the average risk response time, and makes independent threshold judgments on the longest risk response time and the average risk response time, and generates a risk response delay signal and a risk response slow signal according to the judgment results.

Images