Training quality monitoring system for civil aviation based on digital twin model

Abstract

The invention relates to the technical field of civil aviation training, in particular to a civil aviation training quality monitoring system based on a digital twin model, and the system comprises a data collection unit; the state analysis unit is used for determining a training state according to the posture change value and the track dispersion of the training task; the selection analysis unit is used for determining a to-be-detected point selection mode according to the training state; the image division unit is used for determining a sub-image division mode according to the abrupt change pixel point distribution coefficient; the quality confirmation unit is used for determining an image quality value confirmation mode according to the analysis condition; the processing and analyzing unit is used for determining a processing mode according to the image quality value; the image quality of the flight training simulation scene can be guaranteed, and then the training quality is guaranteed.

Description

Technical Field

[0001] The present invention relates to the field of civil aviation training technology, and in particular to a civil aviation training quality monitoring system based on a digital twin model. Background Art

[0002] Traditional civil aviation training, primarily through hands-on practice and written materials, has numerous limitations, including high costs and significant safety risks. However, when using digital twin models to simulate real-world scenarios for flight crew training, the image quality of the simulated scenes cannot be guaranteed, resulting in training results that fail to meet user needs. Therefore, improving the image quality of simulated scenes using digital twin models to ensure training quality is a pressing technical challenge facing those skilled in the art.

[0003] Chinese Patent Publication No. CN106530894B discloses a virtual head-up display method and system for flight training devices, comprising: constructing a virtual scene based on pre-acquired map data; processing the image of the virtual scene using a digital image processing method to obtain saliency information of the scene image; constructing an augmented reality model using a pattern recognition algorithm in combination with the saliency information, and overlaying the augmented reality model onto the virtual scene; and embedding the head-up display image of the virtual scene into an existing visual system for display within a video scene. This technical solution presents the following technical issues: it cannot guarantee the image quality of the flight training simulation scene, cannot select different methods to detect the image quality of the virtual scene based on the actual scene, and cannot guarantee the quality of training. Summary of the Invention

[0004] To this end, the present invention provides a civil aviation training quality monitoring system based on a digital twin model to overcome the problems in the prior art that the image quality of flight training simulation scenes cannot be guaranteed, different methods cannot be selected according to actual scenes to detect the image quality of virtual scenes, and the training quality cannot be guaranteed.

[0005] To achieve the above objectives, the present invention provides a civil aviation training quality monitoring system based on a digital twin model, comprising: A data acquisition unit, used to collect training information; A state analysis unit connected to the data acquisition unit, for determining a training state based on a posture change value and a trajectory discreteness of a training task; a selection and analysis unit, connected to the data acquisition unit and the state analysis unit, respectively, for determining a method for selecting points to be detected according to the training state, wherein the method for selecting points to be detected is to determine an optimization method according to setting conditions or to uniformly select points to be detected, wherein the optimization method is to select points to be detected according to a combination of sections or a fuel consumption slope; An image division unit connected to the selection and analysis unit is used to determine a sub-image division method based on the distribution coefficient of the sudden pixel points. The sub-image division method is uniform division according to a set angle or division according to pixel combination; a quality confirmation unit, connected to the selection and analysis unit and the image segmentation unit, respectively, for determining an image quality value confirmation method based on the analysis conditions, wherein the image quality value confirmation method is to perform regional determination based on scene feature values ​​corresponding to the sub-images or to perform overall determination based on Gaussian feature values; A processing and analysis unit is connected to the quality confirmation unit and is used to determine a processing method based on the image quality value, such as sending a quality-qualified reminder to the user or determining a quality adjustment method based on the trajectory deviation value, such as adjusting the data transmission delay or compression reference value.

[0006] Furthermore, the state analysis unit determines the training state according to the posture change value and trajectory discreteness of the training task, and the training state includes: A first training state in which the posture change value is greater than or equal to a preset posture change value or the trajectory discreteness is less than a preset trajectory discreteness; A second training state in which the posture change value is less than a preset posture change value and the trajectory discreteness is greater than or equal to the preset trajectory discreteness.

[0007] Furthermore, the selection and analysis unit responds to the training status to determine the selection method of the points to be detected; The selection analysis unit responds to the first training state, and the method for selecting the points to be detected is to determine the selection optimization method according to the setting conditions; The selection analysis unit responds to the second training state, and the method of selecting the points to be detected is to uniformly select the points to be detected; When the method of selecting the points to be tested is to select the points to be tested evenly, the training time is divided into several equal parts, and the training time points corresponding to the equal parts are used as the points to be tested.

[0008] Furthermore, the selection analysis unit responds to the setting conditions to determine the selection optimization mode; The setting conditions for selecting the analysis unit response are that the node shear floating value is greater than or equal to the preset node shear floating value and the node distribution coefficient is greater than or equal to the preset node distribution coefficient, and the optimization method is to select the points to be tested according to the combined paragraph; The setting condition for selecting the analysis unit response is that the node shear floating value is less than the preset node shear floating value or the node distribution coefficient is less than the preset node distribution coefficient, and the optimization method is to select the test point according to the fuel consumption slope; When selecting the points to be tested based on the fuel consumption slope, the optimization method is to obtain a fuel consumption change curve with the training time as the horizontal coordinate and the fuel consumption as the vertical coordinate, and divide it into several curve sections with the change point as the dividing point. The fuel consumption slope corresponding to each curve section is detected, and the curve section with a fuel consumption slope less than the preset fuel consumption slope is recorded as a feature section, and a preset number of training time points in each feature section are selected as the points to be tested.

[0009] Furthermore, the selection and analysis unit selects the points to be detected according to the combined paragraphs, including: Performing similarity analysis on each track point in the estimated movement track. When performing similarity analysis on a track point in the estimated trajectory image, record the track point as the target track point, and record the track points that do not include the target track point and are not included in the similar combination as reference track points. A set consisting of each reference trajectory point and the target trajectory point that is located after the target trajectory point in the moving order and before the first reference trajectory point whose shear difference value with the target trajectory point is greater than a preset shear difference value is recorded as a similar combination; Similarity analysis is performed on the trajectory points that are not recorded as similar combinations until all reference trajectory points are recorded as similar combinations; Each similar combination with a combination reference value smaller than a preset combination reference value is recorded as a combination segment, and the number of to-be-detected points corresponding to the combination segment is determined according to the combination reference value; The number of points to be detected corresponding to a single combination paragraph is positively correlated with the combination reference value.

[0010] Furthermore, the image division unit determines a sub-image division method according to a sudden pixel distribution coefficient; If the mutation pixel distribution coefficient is greater than or equal to the preset mutation similarity distribution coefficient, the sub-image is divided evenly according to the set angle; If the mutation pixel distribution coefficient is less than the preset mutation similarity distribution coefficient, the sub-image is divided according to the pixel combination; The setting angle is negatively correlated with the mutation pixel distribution coefficient.

[0011] Furthermore, the image segmentation unit performs segmentation according to pixel combinations, wherein: The virtual image is evenly divided into several perspective images, and combination detection is performed on each perspective image. When combination detection is performed on a single perspective image, the perspective image is recorded as the target perspective image, and the perspective image that does not include the target perspective image and is not included in the pixel combination is recorded as the reference perspective image; The number of mutation pixels corresponding to the target perspective image and the number of mutation pixels corresponding to each reference perspective image located after the target perspective image in the clockwise direction are accumulated one by one until the total number of mutation pixels exceeds the preset total number of mutation pixels. The reference perspective images before the reference perspective image corresponding to the last accumulated number of mutation pixels and the target perspective image are recorded as a pixel combination; Perform combination detection on the viewing angle images that are not recorded in the pixel combination until all viewing angle images are recorded in the pixel combination; The segmentation is performed using the starting and ending edges corresponding to each pixel combination as the segmentation lines.

[0012] Further, the quality confirmation unit is responsive to the analysis conditions to determine an image quality value confirmation method; The analysis condition for the quality confirmation unit response is that the comprehensive change value is greater than or equal to the preset comprehensive change value or the percentage of sudden change pixels is greater than or equal to the preset percentage of sudden change pixels. The image quality value is confirmed by performing regional judgment based on the scene feature values ​​corresponding to the sub-image; When the image quality value confirmation method is regional determination, the virtual image is segmented according to the sub-image division method to obtain a number of sub-images, and the scene feature value corresponding to each sub-image is input into the regression model to determine the image quality value; The analysis condition for the quality confirmation unit response is that the comprehensive change value is less than the preset comprehensive change value and the percentage of sudden change pixels is less than the preset percentage of sudden change pixels. The image quality value is confirmed by overall judgment based on the Gaussian eigenvalue; When the image quality value confirmation method is overall judgment, the Gaussian eigenvalue corresponding to the virtual image is input into the regression model to determine the image quality value.

[0013] Furthermore, the processing and analysis unit determines a processing method according to the image quality value; If the image quality value is greater than or equal to the preset image quality value, the processing method is to send a quality qualified reminder to the user; If the image quality value is less than the preset image quality value, the processing method is to determine the quality adjustment method according to the trajectory deviation value.

[0014] Furthermore, the processing and analysis unit determines a quality adjustment method based on the trajectory deviation value; If the trajectory deviation value is greater than or equal to the preset trajectory deviation value, the quality adjustment method is to reduce the data transmission delay; If the trajectory deviation value is less than the preset trajectory deviation value, the quality adjustment method is to increase the compression reference value; The reduction value of the data transmission delay is positively correlated with the position deviation, and the increase value of the compression reference value is negatively correlated with the image similarity.

[0015] Compared with the prior art, the beneficial effect of the present invention lies in that, in the technical solution of the present invention, the training state is determined according to the attitude change value and the trajectory discreteness, and the training difficulty of the flight training is effectively reflected by the attitude change value and the trajectory discreteness, and then different methods of selecting the points to be detected are adaptively selected according to the training state, so that the selection of the method of selecting the points to be detected is more in line with the actual application scenario, avoiding the problem in the prior art that the detection is only based on the training process of the trainees and different methods cannot be selected according to the actual scenario to detect the image quality of the virtual scene, and also avoiding the problem of low efficiency and insufficient accuracy of image quality detection due to the inability to select suitable points to be detected when the training difficulty is high, which can ensure the accuracy and effectiveness of the image quality detection results, thereby ensuring the quality of training.

[0016] Furthermore, the present invention effectively reflects the changes in the estimated movement trajectory by setting conditions, and then adaptively selects different selection optimization methods according to the set conditions, so that the selection of the selection optimization method is more in line with the actual application scenario, avoiding the problem of poor detection efficiency caused by the inability to select suitable points to be detected when the changes in the estimated trajectory are highly complex, thereby improving the efficiency and accuracy of image quality detection.

[0017] Furthermore, the present invention effectively reflects the distribution of features in the virtual image through the mutation pixel distribution coefficient, and then adaptively selects different sub-image division methods according to the mutation pixel distribution coefficient, so that the selection of sub-image division method is more in line with the actual application scenario, avoiding the problem of poor accuracy of image quality value confirmation leading to poor accuracy of image quality evaluation, thereby ensuring the quality of flight training. Furthermore, the present invention effectively reflects the flight changes corresponding to the points to be detected and the characteristics of the simulated images corresponding to the points to be detected through analysis conditions, and then adaptively selects different image quality value confirmation methods according to the analysis conditions, so that the selection of image quality value confirmation methods is more in line with actual application scenarios, avoiding the problem of poor accuracy of image quality evaluation due to a single image quality value confirmation method, ensuring that the accuracy of image quality value judgment can be improved when the flight changes are complex or the simulated images have many characteristics, thereby ensuring the quality of flight training.

[0018] Furthermore, the present invention effectively reflects the image quality of the simulation scene through the image quality value, and then adaptively selects different processing methods, so that the selection of processing method is more in line with the actual application scenario, avoiding the problem of poor training quality when the image quality of the simulation scene is poor, and then improving the image quality of the simulation scene according to the quality adjustment method, thereby ensuring the quality of flight training.

[0019] Furthermore, the present invention effectively reflects the degree of deviation between the flight trajectory and the estimated trajectory during the learning and training process through the trajectory deviation value, and then adaptively selects different quality adjustment methods according to the trajectory deviation value. When the trajectory deviation value is large, the data transmission delay is reduced and adjusted to ensure the real-time and accuracy of the flight trajectory data, thereby improving the image quality; when the trajectory deviation value is small, the compression reference value is increased and adjusted to help retain more detailed information, thereby maintaining the accuracy of the digital twin model and improving the image quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a unit connection diagram of the civil aviation training quality monitoring system based on the digital twin model of the present invention; Figure 2 This is a flow chart of the present invention for determining a method for selecting points to be detected based on a training state; Figure 3 This is a flow chart of the present invention for determining and selecting an optimization method based on setting conditions; Figure 4 This is a flow chart of a method for determining an image quality value according to analysis conditions according to the present invention. DETAILED DESCRIPTION

[0021] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention.

[0022] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0023] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside", and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0024] Furthermore, it should be noted that, in the description of the present invention, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0025] See also Figures 1 to 4 As shown, the present invention provides a civil aviation training quality monitoring system based on a digital twin model, comprising: A data acquisition unit, used to collect training information; A state analysis unit connected to the data acquisition unit, for determining a training state based on a posture change value and a trajectory discreteness of a training task; a selection and analysis unit, connected to the data acquisition unit and the state analysis unit, respectively, for determining a method for selecting points to be detected according to the training state, wherein the method for selecting points to be detected is to determine an optimization method according to setting conditions or to uniformly select points to be detected, wherein the optimization method is to select points to be detected according to a combination of sections or a fuel consumption slope; An image division unit connected to the selection and analysis unit is used to determine a sub-image division method based on the distribution coefficient of the sudden pixel points. The sub-image division method is uniform division according to a set angle or division according to pixel combination; a quality confirmation unit, connected to the selection and analysis unit and the image segmentation unit, respectively, for determining an image quality value confirmation method based on the analysis conditions, wherein the image quality value confirmation method is to perform regional determination based on scene feature values ​​corresponding to the sub-images or to perform overall determination based on Gaussian feature values; A processing and analysis unit is connected to the quality confirmation unit and is used to determine a processing method based on the image quality value, such as sending a quality-qualified reminder to the user or determining a quality adjustment method based on the trajectory deviation value, such as adjusting the data transmission delay or compression reference value.

[0026] The application scenario of the present invention is to detect image quality in a simulation scene during civil aviation training using a digital twin model. The simulation scene is obtained by inputting flight training data into the digital twin model. The simulation scene includes simulation images corresponding to several training time points, wherein the training time points are set by the user. A method for setting training time points is provided. In the order of training time from early to late, each 1 second is recorded as a training time point, that is, a simulation image is generated every 1 second. The training information includes several flight training data related to the training task. The training tasks include but are not limited to flight training in complex weather conditions, night flight training, navigation training, and mountain flight training. The flight training data include but are not limited to aircraft position, speed, acceleration, moving direction, light intensity, air pressure, and scene images. The scene images are captured images of the aircraft and the environment. The flight training data is input into the digital twin model to obtain an estimated movement trajectory. The estimated movement trajectory corresponds to the predicted aircraft position corresponding to each training time point. This is content that is easy for a person skilled in the art to understand and will not be described in detail.

[0027] The present invention provides several historical records, any of which records the attitude change value, trajectory discreteness, node shear floating value, node distribution coefficient, shear difference value and combination reference value during at least one historical process of civil aviation training, and each historical record corresponds to a qualified mark, which records whether the civil aviation training effect meets user requirements, and the qualified mark can be recorded manually.

[0028] Specifically, the state analysis unit determines the training state according to the posture change value and trajectory discreteness of the training task, and the training state includes: A first training state in which the posture change value is greater than or equal to a preset posture change value or the trajectory discreteness is less than a preset trajectory discreteness; A second training state in which the posture change value is less than a preset posture change value and the trajectory discreteness is greater than or equal to the preset trajectory discreteness.

[0029] The standard deviations of the rotational angular rates on the aircraft's pitch, roll, and yaw axes are detected respectively, and the maximum value among the standard deviations is recorded as the attitude change value. The rotational angular rates corresponding to the pitch, roll, and yaw axes are measured by gyroscopes. This is a content that is easily understood by those skilled in the art and will not be described in detail here. The trajectory dispersion is confirmed by detecting the predicted aircraft position at each training time point and recording the standard deviation of the shortest distance from each predicted aircraft position to the reference baseline as the trajectory dispersion. The reference baseline is the line connecting the predicted aircraft position corresponding to the starting training time point and the predicted aircraft position corresponding to the ending training time point. The starting training time point and the ending training time point are the first training time point and the last training time point in the training time, respectively, in descending order. The values ​​of the preset posture change value and the preset trajectory discreteness can be determined by the user according to the actual application scenario. The higher the user's demand for the image quality detection effect, the smaller the values ​​of the preset posture change value and the preset trajectory discreteness. Provide a value of the preset posture change value and the preset trajectory discreteness, detect the historical records of the evenly selected points to be detected, and record the average value of the posture change values ​​that can meet the user's needs as the preset posture change value, and record the average value of the trajectory discreteness that can meet the user's needs as the preset trajectory discreteness.

[0030] Specifically, the selection and analysis unit responds to the training state to determine the method for selecting the points to be detected; The selection analysis unit responds to the first training state, and the method for selecting the points to be detected is to determine the selection optimization method according to the setting conditions; The selection analysis unit responds to the second training state, and the method of selecting the points to be detected is to uniformly select the points to be detected; When the method of selecting the points to be tested is to select the points to be tested evenly, the training time is divided into several equal parts, and the training time points corresponding to the equal parts are used as the points to be tested.

[0031] Among them, when the training time is divided into several equal parts, a division method is provided. The training time is divided into φ equal parts. The number of equal division points that can be obtained is (φ-1). The value of φ can be determined by the user according to actual needs. The greater the user's detection accuracy of image quality, the larger the value of φ. A value of φ is provided, and φ is 30.

[0032] Specifically, the selection analysis unit responds to the setting conditions to determine the selection optimization mode; The setting conditions for selecting the analysis unit response are that the node shear floating value is greater than or equal to the preset node shear floating value and the node distribution coefficient is greater than or equal to the preset node distribution coefficient, and the optimization method is to select the points to be tested according to the combined paragraph; The setting condition for selecting the analysis unit response is that the node shear floating value is less than the preset node shear floating value or the node distribution coefficient is less than the preset node distribution coefficient, and the optimization method is to select the test point according to the fuel consumption slope; When selecting the points to be tested based on the fuel consumption slope, the optimization method is to obtain a fuel consumption change curve with the training time as the horizontal coordinate and the fuel consumption as the vertical coordinate, and divide it into several curve sections with the change point as the dividing point. The fuel consumption slope corresponding to each curve section is detected, and the curve section with a fuel consumption slope less than the preset fuel consumption slope is recorded as a feature section, and a preset number of training time points in each feature section are selected as the points to be tested.

[0033] The setting conditions include a first setting condition and a second setting condition. The first setting condition is that the node shear floating value is greater than or equal to the preset node shear floating value and the node distribution coefficient is greater than or equal to the preset node distribution coefficient. The second setting condition is that the node shear floating value is less than the preset node shear floating value or the node distribution coefficient is less than the preset node distribution coefficient. After obtaining the estimated movement trajectory, a three-dimensional rectangular coordinate system is established with the predicted aircraft position corresponding to the starting training time point as the origin. The predicted aircraft position in the estimated movement trajectory corresponding to each training time point excluding the starting training time point is recorded as a trajectory point. The vector angle corresponding to each trajectory point is detected. The method for confirming the vector angle corresponding to a single trajectory point is to record the trajectory point as the target trajectory point for a trajectory point, and to record the vector angle corresponding to the trajectory point adjacent to the target trajectory point and located before the target trajectory point in the movement order. The vector corresponding to the target trajectory point The angle ε between them is recorded as the vector angle, ], the vector corresponding to each trajectory point is tangent to the moving trajectory, and the direction of the vector is the same as the moving direction. Trajectory points with vector angles greater than the preset vector angle are recorded as trajectory nodes. The node shear floating value is the standard deviation of the vector angle corresponding to each trajectory node. The value of the preset vector angle can be determined by the user based on the actual application scenario. A preset vector angle value of 60° is provided. The movement order is the order in which each trajectory point is reached.

[0034] The node distribution coefficient is the average value of the reference minimum distances corresponding to each trajectory node. For a single trajectory node, this trajectory node is recorded as the target trajectory node, and the other trajectory nodes except the target trajectory node are recorded as reference trajectory nodes. The shortest distance from the target trajectory node to each reference trajectory node is detected, and the minimum value of each shortest distance is recorded as the reference minimum distance.

[0035] The values ​​of the preset node shear floating value and the preset node distribution coefficient can be determined by the user according to the actual application scenario. The larger the values ​​of the preset node shear floating value and the preset node distribution coefficient are, the higher the user's demand for selecting the point to be tested according to the fuel consumption slope. A value of the preset node shear floating value and the preset node distribution coefficient is provided, and the historical records of selecting the point to be tested according to the fuel consumption slope are detected. The average value of the node shear floating values ​​corresponding to the historical records that can meet the user's needs is recorded as the preset node shear floating value, and the average value of the node distribution coefficients corresponding to the historical records that can meet the user's needs is recorded as the preset node distribution coefficient; Recording a curve section whose fuel consumption slope deviation is less than a preset fuel consumption slope deviation as a characteristic section includes: performing a characteristic analysis on each curve section, wherein, when performing the characteristic analysis on a curve section, recording the curve section as a target curve section, and recording all other curve sections except the target curve section as reference curve sections; recording a reference curve section whose fuel consumption slope deviation from the target curve section is less than the preset fuel consumption slope deviation and the target curve section as a characteristic section, and continuing to perform characteristic analysis on curve sections that are not recorded as characteristic sections until all curve sections are recorded as characteristic sections.

[0036] Fuel consumption is the amount of fuel consumed by the aircraft at each training time point. For a curve segment, the fuel consumption slope corresponding to the curve segment = (fuel consumption corresponding to the segment end point - fuel consumption corresponding to the segment start point) / (training time corresponding to the segment end point - training time corresponding to the segment start point). The segment start point and segment end point are the first and last training time points of a curve segment, respectively. For any two curve sections, the larger value of the fuel consumption slopes corresponding to the two curve sections is recorded as k1, and the smaller value is recorded as k2. The fuel consumption slope deviation = |k1-k2| / k1; the values ​​of the preset fuel consumption slope deviation and the preset number can be determined by the user according to the actual application scenario. The greater the user's similarity requirement for each trajectory point in the combined section, the smaller the value of the preset fuel consumption slope deviation and the larger the value of the preset number. A preset fuel consumption slope deviation and preset number are provided. The fuel consumption slope deviation is 20%, and the preset number is 10% of the number of training time points contained in a single feature section.

[0037] Specifically, the selection and analysis unit selects the points to be detected according to the combined paragraphs, including: Performing similarity analysis on each track point in the estimated movement track. When performing similarity analysis on a track point in the estimated trajectory image, record the track point as the target track point, and record the track points that do not include the target track point and are not included in the similar combination as reference track points. A set consisting of each reference trajectory point and the target trajectory point that is located after the target trajectory point in the moving order and before the first reference trajectory point whose shear difference value with the target trajectory point is greater than a preset shear difference value is recorded as a similar combination; Similarity analysis is performed on the trajectory points that are not recorded as similar combinations until all reference trajectory points are recorded as similar combinations; Each similar combination with a combination reference value smaller than a preset combination reference value is recorded as a combination segment, and the number of to-be-detected points corresponding to the combination segment is determined according to the combination reference value; The number of points to be detected corresponding to a single combination paragraph is positively correlated with the combination reference value.

[0038] Among them, each similar combination whose combination reference value is less than a preset combination reference value is recorded as a combination paragraph, including: performing combination analysis on each similar combination, when performing combination analysis on a similar combination, recording the similar combination as a target similar combination, and recording other similar combinations other than the target similar combination as reference similar combinations; recording the reference similar combination whose combination reference value with the target similar combination is less than the preset combination reference value and the target similar combination as a combination paragraph, and continuing to perform combination analysis on similar combinations that have not been recorded as combination paragraphs until each similar combination is recorded as a combination paragraph.

[0039] The shear difference value is the absolute value of the difference between the vector angles of the two trajectory points. The combined reference value is the absolute value of the difference between the combined means corresponding to the two combined sections. The combined mean is the average value of the vector angles of the corresponding trajectory points in a combined section.

[0040] The values ​​of the preset shear difference value and the preset combination reference value can be determined by the user according to the actual application scenario. The greater the user's similarity requirement for each trajectory point in the combination paragraph, the smaller the values ​​of the preset shear difference value and the preset combination reference value are. A preset shear difference value and a preset combination reference value are provided. The historical records of the points to be detected are selected according to the combination paragraph. The average value of the shear difference values ​​corresponding to the historical records that can meet the user's requirements is recorded as the preset shear difference value, and the average value of the combination reference values ​​corresponding to the historical records that can meet the user's requirements is recorded as the preset combination reference value.

[0041] After determining the number of points to be detected corresponding to a combined paragraph according to the combined reference value, if the number of points to be detected corresponding to a combined paragraph is a, a trajectory points are randomly selected in the combined paragraph, and the training time points corresponding to each trajectory point are used as the points to be detected.

[0042] Specifically, the image division unit determines the sub-image division method according to the mutation pixel distribution coefficient; If the mutation pixel distribution coefficient is greater than or equal to the preset mutation similarity distribution coefficient, the sub-image is divided evenly according to the set angle; If the mutation pixel distribution coefficient is less than the preset mutation similarity distribution coefficient, the sub-image is divided according to the pixel combination; The setting angle is negatively correlated with the mutation pixel distribution coefficient.

[0043] Each point to be detected corresponds to a virtual image. In the present invention, the virtual image is a panoramic image. The panoramic image is mapped into a spherical mode under the actual viewing scene through spherical projection technology. The sphere is centered at the viewer's position. This is easy to understand for those skilled in the art and will not be described in detail. Set the angle to the angle of the sector of a single sub-image projected onto the equatorial plane, where the equatorial plane is a circle with the equator as its outline and the center of the sphere as its center. When the sub-image division method is to divide the sphere evenly according to the set angle, the equator is divided into equal parts according to the number of division points. The lines connecting the spheres at each division point are used as the dividing lines to divide the virtual image. The number of division points = 360° / set angle. For a single pixel, if the difference between the minimum pixel value of the pixel and its adjacent pixels is greater than the preset difference, the pixel is considered a mutation pixel. The preset difference value can be set by the user according to the actual situation. The greater the user's requirement for the accuracy of confirming the proportion of mutation pixels, the smaller the preset difference value. A preset difference value is provided, which is the minimum value of the minimum pixel value difference between each mutation pixel and the corresponding adjacent pixels in the historical records; The mutation pixel distribution coefficient is the average value of the reference distances corresponding to each mutation pixel. For a single mutation pixel, the mutation pixel is recorded as the target mutation pixel, and the other mutation pixels other than the target mutation pixel are recorded as reference mutation pixels. The shortest distance from the target mutation pixel to each reference mutation pixel is detected, and the average value of each shortest distance is recorded as the reference distance. The value of the preset mutation similarity distribution coefficient can be determined by the user according to the actual application scenario. The greater the user's requirement for the accuracy of the image quality value, the smaller the value of the preset mutation similarity distribution coefficient. A value of the preset mutation similarity distribution coefficient is provided. The historical records evenly divided according to the set angle are detected, and the average value of the mutation similarity distribution coefficients corresponding to the historical records that can meet the user's needs is recorded as the preset mutation similarity distribution coefficient.

[0044] Specifically, the image segmentation unit performs segmentation according to pixel combinations, wherein: The virtual image is evenly divided into several perspective images, and combination detection is performed on each perspective image. When combination detection is performed on a single perspective image, the perspective image is recorded as the target perspective image, and the perspective image that does not include the target perspective image and is not included in the pixel combination is recorded as the reference perspective image; The number of mutation pixels corresponding to the target perspective image and the number of mutation pixels corresponding to each reference perspective image located after the target perspective image in the clockwise direction are accumulated one by one until the total number of mutation pixels exceeds the preset total number of mutation pixels. The reference perspective images before the reference perspective image corresponding to the last accumulated number of mutation pixels and the target perspective image are recorded as a pixel combination; Perform combination detection on the viewing angle images that are not recorded in the pixel combination until all viewing angle images are recorded in the pixel combination; The segmentation is performed using the starting and ending edges corresponding to each pixel combination as the segmentation lines.

[0045] The virtual image is evenly divided into a number of perspective images, including: dividing the equator of the sphere into n equal parts along the equator, and dividing the virtual image using the lines connecting the spheres at the equal division points as dividing lines. The value of n can be determined by the user according to actual needs. Here, a value of n is provided, which is 360. The number of mutation pixels is the total number of mutation pixels in a single perspective image; the total number of mutation pixels is the sum of the number of mutation pixels corresponding to each perspective image; The value of the preset total number of mutation pixels can be determined by the user according to the actual application scenario. The greater the user's requirement for the accuracy of the image quality value, the smaller the value of the preset total number of mutation pixels. A value of the preset total number of mutation pixels is provided, and the historical records of segmentation based on pixel combinations are detected. The average value of the total number of mutation pixels corresponding to the historical records that can meet the user's requirements is recorded as the preset total number of mutation pixels. For a single pixel combination, the first dividing line of the first viewing angle image in the clockwise direction is the starting edge corresponding to the pixel combination, and the second dividing line of the last viewing angle image in the clockwise direction is the ending edge corresponding to the pixel combination.

[0046] Specifically, the quality confirmation unit determines an image quality value confirmation method in response to the analysis conditions; The analysis condition for the quality confirmation unit response is that the comprehensive change value is greater than or equal to the preset comprehensive change value or the percentage of sudden change pixels is greater than or equal to the preset percentage of sudden change pixels. The image quality value is confirmed by performing regional judgment based on the scene feature values ​​corresponding to the sub-image; When the image quality value confirmation method is regional determination, the virtual image is segmented according to the sub-image division method to obtain a number of sub-images, and the scene feature value corresponding to each sub-image is input into the regression model to determine the image quality value; The analysis condition for the quality confirmation unit response is that the comprehensive change value is less than the preset comprehensive change value and the percentage of sudden change pixels is less than the preset percentage of sudden change pixels. The image quality value is confirmed by overall judgment based on the Gaussian eigenvalue; When the image quality value confirmation method is overall judgment, the Gaussian eigenvalue corresponding to the virtual image is input into the regression model to determine the image quality value.

[0047] The analysis conditions include a first analysis condition and a second analysis condition. The first analysis condition is that the comprehensive change value is greater than or equal to a preset comprehensive change value or the percentage of sudden change pixels is greater than or equal to a preset percentage of sudden change pixels. The second analysis condition is that the comprehensive change value is less than the preset comprehensive change value and the percentage of sudden change pixels is less than the preset percentage of sudden change pixels. Each point to be tested corresponds to a virtual image. The method for confirming the comprehensive change value and the percentage of sudden pixel points is as follows: for a single point to be tested, record it as the target point to be tested. The comprehensive change value corresponding to the target point to be tested = the fuel consumption slope corresponding to the target point to be tested + the vector angle corresponding to the target point to be tested. The virtual image corresponding to the target point to be tested is recorded as the target image. The percentage of sudden pixel points = the number of sudden pixel points in the target image / the total number of pixels in the target image. The values ​​of the preset comprehensive change value and the preset mutation pixel ratio can be determined by the user according to the actual application scenario. The greater the user's demand for the accuracy of the image quality value, the smaller the values ​​of the preset comprehensive change value and the preset mutation pixel ratio are. A preset comprehensive change value and the preset mutation pixel ratio are provided. The image quality value confirmation method is the historical records of the overall judgment. The average value of the comprehensive change values ​​corresponding to the historical records that can meet the user's needs is recorded as the preset comprehensive change value, and the average value of the mutation pixel ratio corresponding to the historical records that can meet the user's needs is recorded as the preset mutation pixel ratio.

[0048] The scene feature value is confirmed in the following way: for a single sub-image, the sub-image is separated into three independent color channels of R, G, and B, and the first pixel mean corresponding to each channel is calculated. The first pixel mean corresponding to a single channel is the average value of the pixel values ​​corresponding to each pixel point in the single channel. For a single channel, the pixel point with the pixel value equal to the first pixel mean is recorded as the center pixel point, and the pixel points adjacent to the center pixel point are recorded as neighboring points. If the pixel value corresponding to the neighboring point is greater than or equal to the first pixel mean, it is recorded as 1; if the pixel value corresponding to the neighboring point is less than the first pixel mean, it is recorded as 0. The binary number obtained by concatenating the comparison results of each channel in the order of RGB is recorded as the scene feature value; The method for confirming the Gaussian eigenvalue is to perform Gaussian smoothing on the virtual image, and then obtain the first-order derivative images of the smoothed image in the x-direction and the y-direction through a differential filter. This is easy to understand for those skilled in the art and will not be described in detail. For a single first-order derivative image, the pixel point whose pixel value is the second pixel mean is recorded as the center pixel point, and the pixel points adjacent to the center pixel point are recorded as neighboring points. If the pixel value corresponding to the neighboring point is greater than or equal to the second pixel mean, it is recorded as 1; if the pixel value corresponding to the neighboring point is less than the second pixel mean, it is recorded as 0. The binary number obtained by connecting the comparison results of each first-order derivative image in series is recorded as the Gaussian eigenvalue; The scene feature values ​​or Gaussian feature values ​​are input into the regression model for training and prediction to obtain the image quality value. This is easily understood by those skilled in the art and will not be described in detail.

[0049] Specifically, the processing analysis unit determines a processing method according to the image quality value; If the image quality value is greater than or equal to the preset image quality value, the processing method is to send a quality qualified reminder to the user; If the image quality value is less than the preset image quality value, the processing method is to determine the quality adjustment method according to the trajectory deviation value.

[0050] Among them, the value of the preset image quality value can be determined by the user according to the actual application scenario. The higher the user's demand for training quality, the larger the value of the preset image quality value. A value of the preset image quality value is provided, and the average value of the image quality values ​​corresponding to the historical records that can meet the user's needs is recorded as the preset image quality value.

[0051] Specifically, the processing and analysis unit determines a quality adjustment method based on the trajectory deviation value; If the trajectory deviation value is greater than or equal to the preset trajectory deviation value, the quality adjustment method is to reduce the data transmission delay; If the trajectory deviation value is less than the preset trajectory deviation value, the quality adjustment method is to increase the compression reference value; The reduction value of the data transmission delay is positively correlated with the position deviation, and the increase value of the compression reference value is negatively correlated with the image similarity.

[0052] Among them, the movement route of the aircraft during the student training process is recorded as the actual movement route, the position deviation corresponding to each training time point is detected, and the average value of each position deviation is recorded as the trajectory deviation value; the position deviation is the shortest distance from the position of the aircraft in the actual movement route corresponding to a training time point to the position of the aircraft in the estimated movement route; the value of the preset trajectory deviation value can be determined by the user according to the actual application scenario. The smaller the value of the preset trajectory deviation value, the greater the user's demand for reducing the data transmission delay. A preset trajectory deviation value is provided, and the historical records of increasing the compression reference value are detected. The average value of the trajectory deviation values ​​corresponding to the historical records that can meet the user's needs is recorded as the preset trajectory deviation value; The data transmission delay is the time it takes to transmit flight training data to the digital twin model. It is understandable that the flight training data needs to be compressed after acquisition to improve data transmission efficiency. However, due to the presence of scene images in the flight data, the resolution is reduced after compression of the scene images, resulting in a decrease in data quality, which in turn leads to a lower image quality value for the simulated image. The compression reference value = compressed data capacity / original data capacity. The compressed data capacity is the memory capacity occupied by the compressed flight training data, measured in GB. The original data capacity is the memory capacity occupied by the uncompressed flight training data, measured in GB.

[0053] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

[0054] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that the present invention is susceptible to various modifications and variations. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A civil aviation training quality monitoring system based on a digital twin model, characterized by: include: A data acquisition unit, used to collect training information; A state analysis unit connected to the data acquisition unit, for determining a training state based on a posture change value and a trajectory discreteness of a training task; a selection and analysis unit, connected to the data acquisition unit and the state analysis unit, respectively, for determining a method for selecting points to be detected according to the training state, wherein the method for selecting points to be detected is to determine an optimization method according to setting conditions or to uniformly select points to be detected, wherein the optimization method is to select points to be detected according to a combination of sections or a fuel consumption slope; An image division unit connected to the selection and analysis unit is used to determine a sub-image division method based on the distribution coefficient of the sudden pixel points. The sub-image division method is uniform division according to a set angle or division according to pixel combination; a quality confirmation unit, connected to the selection and analysis unit and the image segmentation unit, respectively, for determining an image quality value confirmation method based on the analysis conditions, wherein the image quality value confirmation method is to perform regional determination based on scene feature values ​​corresponding to the sub-images or to perform overall determination based on Gaussian feature values; A processing and analysis unit is connected to the quality confirmation unit and is used to determine a processing method based on the image quality value, such as sending a quality-qualified reminder to the user or determining a quality adjustment method based on the trajectory deviation value, such as adjusting the data transmission delay or compression reference value.

2. The civil aviation training quality monitoring system based on digital twin model according to claim 1 is characterized in that: The state analysis unit determines the training state according to the posture change value and trajectory discreteness of the training task, and the training state includes: A first training state in which the posture change value is greater than or equal to a preset posture change value or the trajectory discreteness is less than a preset trajectory discreteness; A second training state in which the posture change value is less than a preset posture change value and the trajectory discreteness is greater than or equal to the preset trajectory discreteness.

3. The civil aviation training quality monitoring system based on digital twin model according to claim 2 is characterized in that: The selection and analysis unit responds to the training status to determine the method for selecting the points to be detected; The selection analysis unit responds to the first training state, and the method for selecting the points to be detected is to determine the selection optimization method according to the setting conditions; The selection analysis unit responds to the second training state, and the method of selecting the points to be detected is to uniformly select the points to be detected; When the method of selecting the points to be tested is to select the points to be tested evenly, the training time is divided into several equal parts, and the training time points corresponding to the equal parts are used as the points to be tested.

4. The civil aviation training quality monitoring system based on digital twin model according to claim 3 is characterized in that: The selection analysis unit responds to the setting conditions to determine the selection optimization mode; The setting conditions for selecting the analysis unit response are that the node shear floating value is greater than or equal to the preset node shear floating value and the node distribution coefficient is greater than or equal to the preset node distribution coefficient, and the optimization method is to select the points to be tested according to the combined paragraph; The setting condition for selecting the analysis unit response is that the node shear floating value is less than the preset node shear floating value or the node distribution coefficient is less than the preset node distribution coefficient, and the optimization method is to select the test point according to the fuel consumption slope; When selecting the points to be tested based on the fuel consumption slope, the optimization method is to obtain a fuel consumption change curve with the training time as the horizontal coordinate and the fuel consumption as the vertical coordinate, and divide it into several curve sections with the change point as the dividing point. The fuel consumption slope corresponding to each curve section is detected, and the curve section with a fuel consumption slope less than the preset fuel consumption slope is recorded as a feature section, and a preset number of training time points in each feature section are selected as the points to be tested.

5. The civil aviation training quality monitoring system based on digital twin model according to claim 4 is characterized in that: The selection and analysis unit selects the points to be detected according to the combined paragraphs, including: Performing similarity analysis on each track point in the estimated movement track. When performing similarity analysis on a track point in the estimated trajectory image, the track point is recorded as the target track point, and the track points that do not include the target track point and are not included in the similar combination are recorded as reference track points. A set consisting of each reference trajectory point and the target trajectory point that is located after the target trajectory point in the moving order and before the first reference trajectory point whose shear difference value with the target trajectory point is greater than a preset shear difference value is recorded as a similar combination; Similarity analysis is performed on the trajectory points that are not recorded as similar combinations until all reference trajectory points are recorded as similar combinations; Each similar combination with a combination reference value smaller than a preset combination reference value is recorded as a combination segment, and the number of to-be-detected points corresponding to the combination segment is determined according to the combination reference value; The number of points to be detected corresponding to a single combination paragraph is positively correlated with the combination reference value.

6. The civil aviation training quality monitoring system based on digital twin model according to claim 5 is characterized in that: The image division unit determines a sub-image division method according to a sudden pixel distribution coefficient; If the mutation pixel distribution coefficient is greater than or equal to the preset mutation similarity distribution coefficient, the sub-image is divided evenly according to the set angle; If the mutation pixel distribution coefficient is less than the preset mutation similarity distribution coefficient, the sub-image is divided according to the pixel combination; The setting angle is negatively correlated with the mutation pixel distribution coefficient.

7. The civil aviation training quality monitoring system based on digital twin model according to claim 6 is characterized in that: The image segmentation unit performs segmentation according to pixel combinations, wherein: The virtual image is evenly divided into several perspective images, and combination detection is performed on each perspective image. When combination detection is performed on a single perspective image, the perspective image is recorded as the target perspective image, and the perspective image that does not include the target perspective image and is not included in the pixel combination is recorded as the reference perspective image; The number of mutation pixels corresponding to the target perspective image and the number of mutation pixels corresponding to each reference perspective image located after the target perspective image in the clockwise direction are accumulated one by one until the total number of mutation pixels exceeds the preset total number of mutation pixels. The reference perspective images before the reference perspective image corresponding to the last accumulated number of mutation pixels and the target perspective image are recorded as a pixel combination; Perform combination detection on the viewing angle images that are not recorded in the pixel combination until all viewing angle images are recorded in the pixel combination; The segmentation is performed using the starting and ending edges corresponding to each pixel combination as the segmentation lines.

8. The civil aviation training quality monitoring system based on digital twin model according to claim 7 is characterized in that: The quality confirmation unit determines an image quality value confirmation method in response to the analysis condition; The analysis condition for the quality confirmation unit response is that the comprehensive change value is greater than or equal to the preset comprehensive change value or the percentage of sudden change pixels is greater than or equal to the preset percentage of sudden change pixels. The image quality value is confirmed by performing regional judgment based on the scene feature values ​​corresponding to the sub-image; When the image quality value confirmation method is regional determination, the virtual image is segmented according to the sub-image division method to obtain a number of sub-images, and the scene feature value corresponding to each sub-image is input into the regression model to determine the image quality value; The analysis condition for the quality confirmation unit response is that the comprehensive change value is less than the preset comprehensive change value and the percentage of sudden change pixels is less than the preset percentage of sudden change pixels. The image quality value is confirmed by overall judgment based on the Gaussian eigenvalue; When the image quality value confirmation method is overall judgment, the Gaussian eigenvalue corresponding to the virtual image is input into the regression model to determine the image quality value.

9. The civil aviation training quality monitoring system based on digital twin model according to claim 8 is characterized in that: The processing analysis unit determines a processing method according to the image quality value; If the image quality value is greater than or equal to the preset image quality value, the processing method is to send a quality qualified reminder to the user; If the image quality value is less than the preset image quality value, the processing method is to determine the quality adjustment method according to the trajectory deviation value.

10. The civil aviation training quality monitoring system based on digital twin model according to claim 9 is characterized in that: The processing and analyzing unit determines a quality adjustment method according to the trajectory deviation value; If the trajectory deviation value is greater than or equal to the preset trajectory deviation value, the quality adjustment method is to reduce the data transmission delay; If the trajectory deviation value is less than the preset trajectory deviation value, the quality adjustment method is to increase the compression reference value; The reduction value of the data transmission delay is positively correlated with the position deviation, and the increase value of the compression reference value is negatively correlated with the image similarity.

Images