Flight quality comprehensive evaluation method based on time sequence fusion model

Abstract

The invention belongs to the technical field of flight quality evaluation, and discloses a time sequence fusion model-based flight quality comprehensive evaluation method, which comprises the following steps of: obtaining a flight subject completion degree through a ratio of an actual flight subject completion content item to a flight subject action element; establishing a key evaluation index of each flight action according to the flight subject, and obtaining the flight quality accuracy of each evaluation index based on the improved process capability index model; setting an evaluation index influence factor according to the severity level of the flight subject evaluation index to the flight; obtaining evaluation scores of all flight actions according to the flight quality accuracy and the influence factors, and obtaining a single-sortie flight quality evaluation score in combination with the flight subject completion degree; and according to the historical sortie evaluation score of the pilots of the fleet, based on a time sequence influence attenuation fusion formula, obtaining the comprehensive evaluation score of the flight quality of the pilots and the fleet.

Description

Technical Field

[0001] The present invention relates to the technical field of flight quality assessment, and in particular to a flight quality comprehensive assessment method based on a time series fusion model. Background Art

[0002] Flight quality assessment is an important part of flight training and an important part of training pilots and improving their flight training level. Through flight quality assessment, pilots can find existing problems, understand equipment performance, improve flight capabilities, and maximize the efficiency of equipment. Flight quality assessment mainly adopts manual evaluation. Flight instructors evaluate and score based on the flight outline through on-site observation. There is a certain degree of subjectivity. It is difficult to make an objective and fair scientific evaluation of flight training. It is time-consuming, labor-intensive and inefficient. Therefore, combined with flight subjects, the flight quality comprehensive evaluation algorithm based on the time series fusion model using flight parameter data is of great significance for evaluating flight quality and improving flight level.

[0003] At present, the research on flight quality assessment is mainly based on unified flight parameters or physiological and psychological parameters. According to single-value judgment, multiple indicators are integrated to conduct single-flight flight quality assessment. The differences in assessment indicators of different flight subjects are not considered, which leads to certain limitations. At the same time, there is a lack of comprehensive assessment of pilots' flight quality using historical flight assessment results.

[0004] Therefore, how to provide a comprehensive flight quality assessment method to obtain more objective and comprehensive assessment results, so as to more accurately grasp the shortcomings of flight capabilities and carry out targeted flight training, has become a problem that needs to be solved urgently. Summary of the invention

[0005] In view of this, an embodiment of the present invention provides a flight quality comprehensive evaluation method based on a time series fusion model to solve the problem that the flight capability evaluation method in the prior art is not objective and scientific enough.

[0006] The embodiment of the present invention provides a flight quality comprehensive evaluation method based on a time series fusion model, comprising:

[0007] The degree of completion of the flight subject is obtained by comparing the actual flight subject completion content items to the flight subject action elements;

[0008] Establish key evaluation indicators for each flight action according to the flight subject, and obtain the flight quality accuracy of each evaluation indicator based on the improved process capability index model;

[0009] According to the severity level of the flight subject assessment indicators, set the assessment indicator impact factor;

[0010] The evaluation scores of all flight actions are obtained based on the flight quality accuracy and influencing factors, and then the flight quality evaluation scores of a single flight are obtained in combination with the completion of the flight subjects;

[0011] According to the historical sortie evaluation scores of the fleet pilots and based on the time series impact attenuation fusion formula, the comprehensive evaluation scores of the pilots and fleet flight quality are obtained.

[0012] Optionally, it also includes:

[0013] Divide the flight phase into general phase and special phase; select key flight parameters and establish a key indicator system for flight evaluation based on the status and characteristics of the general phase and the special phase;

[0014] Among them, the flight maneuvers in the general stage include: taxiing takeoff, climbing, cruising, descending and approach landing; the flight maneuvers in the special stage include: turning, diving, jumping and rolling.

[0015] Optionally, the key flight parameters for taxiing takeoff include: speed, angle of attack, pitch angle and altitude; the key evaluation indicators for taxiing takeoff include: maximum taxiing speed, heading angle stability, nose wheel rotation rate and landing gear retraction height;

[0016] The key flight parameters for climbing include: pitch angle, heading angle, rate of ascent and descent, retracting and extending landing gear, and retracting and extending flaps; the key evaluation indicators for climbing include: minimum pitch angle, maximum roll angle, and maximum speed;

[0017] The key flight parameters of cruise include heading angle, roll angle, altitude, speed and overload; the key evaluation indicators of cruise include maximum speed, maximum pitch angle and maximum altitude;

[0018] The key flight parameters of descent include: pitch angle, roll angle, heading angle, lift rate and flap retraction; the key evaluation indicators of descent include: maximum descent rate and maximum roll angle;

[0019] The key flight parameters for approach and landing include: speed and angle of attack; the key evaluation indicators for approach and landing include: minimum pitch angle for approach, maximum landing speed and maximum roll angle;

[0020] The key flight parameters of a turn include: pitch angle, heading angle, rate of ascent, speed and altitude; the key evaluation indicators of a turn include: maximum speed, maximum bank angle and minimum altitude;

[0021] The key flight parameters of a dive include: overload, angle of attack, pitch angle, speed and altitude; the key evaluation indicators of a dive include: maximum descent rate, maximum altitude and maximum speed;

[0022] The key flight parameters of the jump include: angle of attack, speed, altitude, pitch angle and overload; the key evaluation indicators of the jump include: maximum pitch angle, maximum slope and minimum speed;

[0023] The key flight parameters of roll include: pitch angle, roll angle, speed and altitude; the key evaluation indicators of roll include: maximum pitch angle, minimum altitude and maximum speed.

[0024] Optionally, key evaluation indicators for each flight action are established according to the flight subject, and the flight quality accuracy of each evaluation indicator is obtained based on the improved process capability index model, including:

[0025] For each key evaluation indicator, a first range, a second range, a third range and a fourth range are set based on the corresponding quality standard value; the first range corresponds to the first coefficient, the second range corresponds to the second coefficient, the third range corresponds to the third coefficient, and the fourth range corresponds to the fourth coefficient;

[0026] The calculation formula for the quality accuracy of the key evaluation indicator value is:

[0027]

[0028] Where N represents the number of key evaluation indicators in each flight phase; i represents the i-th key evaluation indicator; y∈Y, x∈X, respectively represent flight subject X and Y flight actions; A i，y，x k represents the quality accuracy of the i-th key evaluation indicator of the y-th flight action in flight subject x; t , k t+1 represents coefficients, where t={1, 2, 3}, k1 is the first coefficient, k2 is the second coefficient, k3 is the third coefficient, and k4 is the fourth coefficient;

[0029] LU t and LD t Respectively represent the upper and lower limits of the first range / second range / third range / fourth range; LU3>LU2>LU1>S i >LD1>LD2>LD3, S i is the quality standard value of the i-th key evaluation indicator; (LD1, LU1) is the first range, (LD2, LD1) and (LU1, LU2) are the second range, (LD3, LD2) and (LU2, LU3) are the third range, and the others are the fourth range.

[0030] Optionally, according to the severity level of the flight subject evaluation index, the evaluation index influencing factors are set, including:

[0031] The impact is divided into four hazard severity levels: catastrophic, severe, minor and mild;

[0032] Set the impact factor formula to:

[0033]

[0034] Among them, f r is the impact index, r={1, 2, 3, 4}, the disaster level index f1=1, the severe level index f2=2, the mild level index f3=3, and the slight level index f4=4.

[0035] Optionally, the evaluation scores of all flight actions are obtained based on the flight quality accuracy and influencing factors, and then the flight quality evaluation scores of a single flight are obtained in combination with the completion of the flight subjects, including:

[0036] Get a single key indicator evaluation score:

[0037]

[0038] Get the quality assessment score of a single action:

[0039]

[0040] Get the flight quality assessment score for a single flight:

[0041]

[0042] in, Q represents the number of flight action elements in flight subject x, and q represents the actual flight completion status.

[0043] Optionally, according to the historical sortie evaluation scores of the fleet pilots, based on the time series impact attenuation fusion formula, the comprehensive evaluation scores of the pilot and fleet flight quality are obtained, including:

[0044] According to the historical evaluation scores of each pilot, a stability factor is set one by one, and a time series impact attenuation fusion formula is established in combination with the attenuation factor to calculate the flight quality evaluation score of each pilot in the current fleet corresponding to a single flight;

[0045] The stability factor is:

[0046]

[0047] Among them, σ a represents the standard deviation of pilot a’s historical flight evaluation scores.

[0048] Optionally, the timing impact attenuation fusion formula is:

[0049]

[0050] Among them, F arepresents the comprehensive evaluation score of pilot a based on his j flights, S n represents the nth flight quality evaluation score, λ an is the stability factor of pilot a’s nth flight, and γ is the attenuation factor.

[0051] Optionally, the attenuation factor γ is set to 0.002.

[0052] Optionally, the influence index of the maximum descent rate of approach is set to 4; the influence index of the maximum speed of straight taxiing is set to 1; the influence index of the touchdown speed is set to 2; the influence index of the retracted landing gear height is set to 1; the influence index of the maximum climb speed is set to 3; the influence index of the maximum deviation of the landing roll heading angle is set to 2; and the influence index of the flap extension speed is set to 1.

[0053] Beneficial effects of the embodiments of the present invention:

[0054] 1. This embodiment proposes a flight quality comprehensive evaluation method based on a time series fusion model. By establishing a flight quality evaluation system and influencing factors for each flight subject, using flight parameter data, and taking completion, accuracy, and stability as the evaluation basis, a scientific and objective quantitative evaluation of the flight quality of each flight subject is achieved. At the same time, combined with the time series fusion model, a comprehensive quality evaluation of the overall flight conditions of pilots and fleets is performed. The algorithm can also provide methods and means for further analysis and research in the future, and explore the differences in pilot flight control capabilities, flight training effects, and the overall flight level of the fleet, and can provide support for carrying out targeted flight training and improving overall flight capabilities.

[0055] 2. According to different flight subjects and different stages, a flight quality evaluation index system and influencing factors were created, and the evaluation indicators are more accurate.

[0056] 3. The embodiment of the present invention proposes a flight quality assessment method using flight parameter data, which integrates various indicator parameters and conducts an overall assessment from the three dimensions of completion, accuracy, and stability of the flight quality assessment indicators. The assessment method is more scientific.

[0057] 4. Through the time series fusion model, using the historical flight quality assessment results, a comprehensive assessment of pilots and fleets is conducted, making the assessment more comprehensive. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings, which are schematic and should not be construed as limiting the present invention in any way. In the accompanying drawings:

[0059] Figure 1 A flow chart of a method for comprehensive evaluation of flight quality based on a time series fusion model in an embodiment of the present invention is shown;

[0060] Figure 2 It shows the flight sorties of each pilot of a certain type of aircraft in a certain fleet;

[0061] Figure 3 It shows the proportion of each flight subject in a fleet;

[0062] Figure 4 The accuracy of the P8 flight D subject evaluation indicators for pilots of a certain fleet is shown;

[0063] Figure 5 It shows the changing trend of the maximum airspeed evaluation index of the pilot P1 of a certain fleet during the whole year of flying subject A;

[0064] Figure 6 It shows the changing trend of the landing gear retracting height evaluation index of pilot P1 of a certain fleet during the whole year of flying subject A;

[0065] Figure 7 It shows the changing trend of the evaluation index of the nose wheel pitch angle change rate of pilot P1 of a certain fleet during the whole year of flying subject A;

[0066] Figure 8 It shows the changing trend of the evaluation index of the level drift time of pilot P1 of a certain fleet throughout the year in the A subject flight;

[0067] Fig. 9 The chart shows the change trend of the pitch angle evaluation index of the takeoff from the ground for nearly 15 sorties of D flight by 5 pilots of a certain fleet;

[0068] Fig.10 It shows the flight quality assessment results of nearly 30 sorties of 8 pilots in a fleet. DETAILED DESCRIPTION

[0069] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0070] like Figure 1 As shown, an embodiment of the present invention provides a flight quality comprehensive evaluation method based on a time series fusion model, comprising:

[0071] Step S10, obtaining the flight subject completion degree through the ratio of the actual flight subject completion content items to the flight subject action elements.

[0072] The flight quality assessment index system is mainly established based on the requirements and standards of different flight subjects, and a flight subject action library is created according to the requirements of different flight subjects. Combined with the flight subject action library standards, the flight is divided into different stages, and the parameters and corresponding weights for assessing quality at different stages have different emphases, thereby creating a flight subject action standard library. The core original parameters involved include: track indicators (indicated airspeed, ground speed, radio altitude, pressure altitude, etc.), attitude indicators (magnetic heading angle, pitch angle, roll angle, sideslip angle, etc.), flap and aileron position data, wheel load and landing gear upper and lower lock data, etc.

[0073] According to the requirements and standards of different flight subjects, an evaluation index system is created based on their completion and accuracy. Completion refers to the degree to which the flight content items are completed based on the flight subject action library and the actual completion of the flight content items.

[0074] For flight subjects, according to the flight syllabus and manual, sort out the flight action elements Q, and according to the actual flight completion situation q, the completion degree of a single flight subject is calculated as follows:

[0075]

[0076] Step S20, establishing key evaluation indicators for each flight action according to the flight subject, and obtaining the flight quality accuracy of each evaluation indicator based on the improved process capability index model.

[0077] Accuracy refers to the evaluation of the accuracy of corresponding flight control based on the standard flight subject action library and the actual flight action parameter values.

[0078] For each key evaluation indicator, a first range, a second range, a third range and a fourth range are set based on the corresponding quality standard value; the first range corresponds to the first coefficient, the second range corresponds to the second coefficient, the third range corresponds to the third coefficient, and the fourth range corresponds to the fourth coefficient;

[0079] The calculation formula for the quality accuracy of the key evaluation indicator value is:

[0080]

[0081] Where N represents the number of key evaluation indicators in each flight phase; i represents the i-th key evaluation indicator; y∈Y, x∈X, respectively represent flight subject X and Y flight actions; A i，y，x k represents the quality accuracy of the i-th key evaluation indicator of the y-th flight action in flight subject x; t , k t+1 represents coefficients, where t={1, 2, 3}, k1 is the first coefficient, k2 is the second coefficient, k3 is the third coefficient, and k4 is the fourth coefficient;

[0082] LU t and LD t Respectively represent the upper and lower limits of the first range / second range / third range / fourth range; LU3>LU2>LU1>S i >LD1>LD2>LD3, S i is the quality standard value of the i-th key evaluation indicator; (LU1, LU1) is the first range, (LD2, LD1) and (LU1, LU2) are the second range, (LD3, LD2) and (LU2, LU3) are the third range, and the others are the fourth range.

[0083] Step S30, setting the evaluation index influencing factor according to the flight subject evaluation index and the severity level of the flight.

[0084] According to the impact degree of flight action indicators (according to GJB / Z99-1997 "System Safety Engineering Manual", the impact level is divided into four danger severity levels: I (catastrophic), II (serious), III (mild), and IV (slight)), the impact factor is established.

[0085] The impact level is divided into four hazard severity levels: catastrophic, severe, mild and slight.

[0086] Set the impact factor formula to:

[0087]

[0088] Among them, f r is the impact index, r={1, 2, 3, 4}, the disaster level index f1=1, the severe level index f2=2, the mild level index f3=3, and the slight level index f4=4.

[0089] In a specific implementation, the influence index of the maximum descent rate of approach is set to 4; the influence index of the maximum speed of straight taxiing is set to 1; the influence index of the touchdown speed is set to 2; the influence index of the retracted landing gear height is set to 1; the influence index of the maximum climb speed is set to 3; the influence index of the maximum deviation of the landing roll heading angle is set to 2; and the influence index of the flap extension speed is set to 1.

[0090] Step S40, obtaining evaluation scores of all flight actions according to the flight quality accuracy and influencing factors, and then obtaining a single-flight flight quality evaluation score in combination with the flight subject completion.

[0091] Get a single key indicator evaluation score:

[0092]

[0093] Get the quality assessment score of a single action:

[0094]

[0095] Get the flight quality assessment score for a single flight:

[0096]

[0097] in, Q represents the number of flight action elements in flight subject x, and q represents the actual flight completion status.

[0098] This embodiment proposes a flight quality assessment method using flight parameter data, which integrates various indicator parameters and conducts an overall assessment from three dimensions: the completion, accuracy, and stability of the flight quality assessment indicators. The assessment method is more scientific.

[0099] Step S50, according to the fleet pilots' historical sortie evaluation scores and based on the time series impact attenuation fusion formula, obtain the pilots' and fleets' flight quality comprehensive evaluation scores.

[0100] The above steps can be used to calculate the flight quality evaluation of each flight subject, and then the overall flight level of the pilot can be comprehensively evaluated based on the historical flight quality evaluation scores, so as to achieve a comprehensive evaluation of flight quality. The fusion of scores at each stage should be affected by the time domain. As time goes by, the impact of historical scores on the pilot's existing flight ability varies. From now to the past, the impact trend should gradually decay over time; at the same time, due to the volatility and differences in the flight level of pilots at different times and stages, the stability of the flight level is also an indicator that needs to be comprehensively considered. Therefore, the comprehensive evaluation algorithm of the time series fusion model can more reasonably and accurately evaluate the actual flight level of pilots and fleets.

[0101] In a specific embodiment, the stability factor is:

[0102]

[0103] Among them, σ a represents the standard deviation of pilot a’s historical flight evaluation scores.

[0104] The timing impact attenuation fusion formula is:

[0105]

[0106] Among them, F a represents the comprehensive evaluation score of pilot a based on his j flights, S n represents the nth flight quality evaluation score, λ an is the stability factor of pilot a’s nth flight, and γ is the attenuation factor. In a specific embodiment, the attenuation factor γ is set to 0.002.

[0107] Through the time series fusion model, historical flight quality assessment results are used to conduct a comprehensive assessment of pilots and fleets, making the assessment more comprehensive.

[0108] Taking a certain type of aircraft in a certain fleet as an example, this type of aircraft is divided into different flight phases, and combined with the requirements and standards of different flight subjects, evaluation indicators are screened to construct a flight quality evaluation indicator system.

[0109] According to the aircraft and flight characteristics of this type of aircraft, the flight phase is divided into general phase and special phase. The general phase includes taxiing and takeoff, climbing, cruising, descending, approach and landing; the special phase is divided into turning, diving, jumping, rolling, etc. according to the characteristics of different flight subjects. According to the status and characteristics of different flight phases, key parameters are screened and key indicators for flight index evaluation are established, as shown in Table 1. According to different flight subjects, different phases are divided, and a flight quality evaluation index system and influencing factors are created, and the evaluation indicators are more accurate.

[0110] Then, based on flight requirements and standards, a standard library of flight quality assessment indicators is established, as shown in Table 2.

[0111]

[0112]

[0113] Table 1 Flight quality evaluation indicators

[0114]

[0115] Table 2 Flight quality evaluation index standard library

[0116] According to the safety impact degree and expert knowledge of different flight evaluation indicators, they are divided into four danger severity levels: catastrophic, serious, mild, and slight. An evaluation indicator matrix is ​​created, as shown in Table 3.

[0117]

[0118] Table 3 Influencing factors of flight quality evaluation indicators

[0119] The overall flight situation of a certain type of aircraft in a certain fleet in a certain year is shown in Table 4, among which the flight sorties of each pilot are shown in Table 4. Figure 2 As shown, the proportion of each flight subject is Figure 3 shown.

[0120]

[0121] Table 4 Overall situation of a certain fleet and a certain aircraft type in a certain year

[0122] The fleet has a total of 5 F-type aircraft, corresponding to 8 pilots. They flew 612 sorties throughout the year, taking 735 hours, covering 10 flight subjects. Among them, P8 pilots flew the most sorties, a total of 93 sorties; P1 pilots flew the least, 65 sorties; D flight subject had the most sorties, accounting for 12.42%; A flight subject had the least sorties, accounting for 7.84%.

[0123] Through the flight quality assessment algorithm proposed in this embodiment, the flight quality assessment was carried out from the perspectives of pilots and the fleet using the flight data of the F-type aircraft in the fleet throughout the year. The flight quality assessment of flight subjects, flight quality assessment indicator trends, fleet flight quality assessment indicator comparison, and comprehensive flight quality assessment were analyzed and mined, which can provide a theoretical basis and data support for the subsequent improvement of flight quality.

[0124] Pilot P8 flew 93 sorties throughout the year, flying 10 flight subjects. One of the sorties was subject D, and he completed 9 specific actions required by the subject. The completion degree of this sortie was 1. According to the specific requirements and standards of flight subject D, a total of 50 evaluation indicators are set for this subject. Based on the flight quality accuracy calculation method proposed in this paper, the accuracy of the evaluation indicators of pilot P8's flight subject D can be obtained as follows: Figure 4 It is obvious that when pilot P8 flew the D subject of this sortie, the three evaluation indicators of touchdown speed, maximum deviation of landing run heading angle and selected landing configuration height were very poor, while the three evaluation indicators of takeoff speed, maximum taxiing speed and takeoff pitch angle were good, indicating that the pilot had a good takeoff but a poor landing, and needed targeted training to improve landing-related flight control capabilities.

[0125] Pilot P1 flew 65 sorties throughout the year, including 20 sorties in subject A. Some evaluation indicators were taken to analyze historical flight trends, such as Figures 5 to 8 As shown. Evaluate pilot P1 flight A subject data, by Figure 5 It can be found that when the pilot was flying, the maximum airspeed index gradually increased, and the control speed became better and better; however, when the landing gear was retracted during takeoff, the altitude fluctuated and showed a downward trend, and his control ability needs to be improved; the rate of change of the pitch angle when lifting the front wheel fluctuated greatly, which was very unstable and required more practice to improve; the level drift time was well controlled in the first half of the year, but showed a significant decline in the second half of the year, and still needs to be improved.

[0126] Analyze the data of the last 15 sorties of the five pilots flying D subjects and compare the flight quality evaluation indicators, such as Fig. 9 As shown. Fig. 9It can be found that the takeoff pitch angle indicators of pilots P6 and P4 fluctuated relatively greatly during the takeoff phase; although the indicator of pilot P5 showed an upward trend, the overall indicator was poor, and the pilot's control level of the takeoff pitch angle needs to be improved; the indicator of pilot P3 showed a downward trend, with good performance in the early stage and a sharp decline in the later stage, and practice should be strengthened; the indicator of pilot P2 was generally stable, and the indicator value was high, and the performance was good, which should be maintained.

[0127] Based on the above evaluation of the quality indicators of the fleet's F-type aircraft throughout the year, combined with the completion rate of a single flight and the indicator influencing factor matrix, the flight quality evaluation score of each flight can be calculated. The 10 flight subjects and the latest 3 flight data of each subject of 8 pilots are taken for evaluation.

[0128] As shown in Table 5.

[0129]

[0130]

[0131] Table 5 Flight quality evaluation of each pilot, each subject, and each sortie

[0132] From Table 5, we can find that pilot P2 ranks first and pilot P7 ranks last. The historical sortie evaluation scores of pilots P1 and P7 fluctuate greatly and are unstable, with relatively low total scores, and they need more practice to improve; pilots P3, P4, and P6 are relatively stable, with overall scores in the middle, and there is room for improvement; pilots P2, P5, and P8 are relatively stable, among which pilots P2 and P5 have relatively high historical sortie evaluation scores, which are outstanding and can be maintained.

[0133] According to the evaluation results, 0 to 60 is unqualified, 60 to 70 is qualified, 70 to 85 is good, and 85 to 100 is excellent. Fig.10 It can be seen that the final evaluation levels of the 8 pilots of the fleet's F-type aircraft are: P1 qualified, P2 excellent, P3 good, P4 good, P5 excellent, P6 good, P7 unqualified, and P8 good. The flight evaluation excellent rate of the pilots of the fleet's F-type aircraft is 25%, the unqualified rate is 12.5%, and the scores are generally above average. The comprehensive score is 76.47, which is good overall. It is necessary to continue to maintain the advantages, make up for the disadvantages, use the advantages to lead the disadvantages, practice more, and improve steadily.

[0134] This embodiment proposes a flight quality comprehensive evaluation method based on a time series fusion model. By establishing a flight quality evaluation system and influencing factors for each flight subject, using flight parameter data, and taking completion, accuracy, and stability as the evaluation basis, a scientific and objective quantitative evaluation of the flight quality of each flight subject is achieved. At the same time, combined with the time series fusion model, a comprehensive quality evaluation of the overall flight conditions of pilots and fleets is performed. The algorithm can also provide methods and means for further analysis and research in the future, and explore the differences in pilot flight control capabilities, flight training effects, and the overall flight level of the fleet, and can provide support for conducting targeted flight training and improving overall flight capabilities.

[0135] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A flight quality comprehensive evaluation method based on a time series fusion model, characterized in that: include: The degree of completion of the flight subject is obtained by comparing the actual flight subject completion content items to the flight subject action elements; Establish several key evaluation indicators for each flight action according to the flight subject, and obtain the flight quality accuracy of each evaluation indicator based on the improved process capability index model; According to the severity level of the flight subject assessment indicators, set the assessment indicator impact factor; Obtaining evaluation scores for all the flight actions according to the flight quality accuracy and influencing factors, and then obtaining a single-flight flight quality evaluation score in combination with the flight subject completion degree; According to the historical sortie evaluation scores of the fleet pilots and based on the time series impact attenuation fusion formula, the comprehensive evaluation scores of the pilots and fleet flight quality are obtained.

2. The flight quality comprehensive evaluation method based on the time series fusion model according to claim 1 is characterized in that: Also includes: The flight phases are divided into general phases and special phases; according to the states and characteristics of the general phases and the special phases, key flight parameters are screened and a key indicator system for flight evaluation is established; Among them, the flight actions in the general stage include: taxiing takeoff, climbing, cruising, descending and approach landing; the flight actions in the special stage include: turning, diving, jumping and rolling.

3. The flight quality comprehensive evaluation method based on the time series fusion model according to claim 2 is characterized in that: The key flight parameters of the taxiing takeoff include: speed, angle of attack, pitch angle and altitude; the key evaluation indicators of the taxiing takeoff include: maximum taxiing speed, heading angle stability, nose wheel rotation rate and landing gear retraction height; The key flight parameters of the climb include: pitch angle, heading angle, rate of ascent and descent, retracting and extending landing gear, and retracting and extending flaps; the key evaluation indicators of the climb include: minimum pitch angle, maximum roll angle, and maximum speed; The key flight parameters of the cruise include: heading angle, roll angle, altitude, speed and overload; the key evaluation indicators of the cruise include: maximum speed, maximum pitch angle and maximum altitude; The key flight parameters of the descent include: pitch angle, roll angle, heading angle, lift rate and flap retraction; the key evaluation indicators of the descent include: maximum descent rate and maximum roll angle; The key flight parameters of the approach and landing include: speed and angle of attack; the key evaluation indicators of the approach and landing include: minimum pitch angle of approach, maximum landing speed and maximum roll angle; The key flight parameters of the turn include: pitch angle, heading angle, rate of ascent, speed and altitude; the key evaluation indicators of the turn include: maximum speed, maximum slope and minimum altitude; The key flight parameters of the dive include: overload, angle of attack, pitch angle, speed and altitude; the key evaluation indicators of the dive include: maximum descent rate, maximum altitude and maximum speed; The key flight parameters of the jump include: angle of attack, speed, altitude, pitch angle and overload; the key evaluation indicators of the jump include: maximum pitch angle, maximum slope and minimum speed; The key flight parameters of the roll include: pitch angle, roll angle, speed and altitude; the key evaluation indicators of the roll include: maximum pitch angle, minimum altitude and maximum speed.

4. The flight quality comprehensive evaluation method based on the time series fusion model according to claim 2 is characterized in that: Establish key evaluation indicators for each flight action according to the flight subject, and obtain the flight quality accuracy of each evaluation indicator based on the improved process capability index model, including: For each of the key evaluation indicators, a first range, a second range, a third range and a fourth range are set based on the corresponding quality standard value; the first range corresponds to a first coefficient, the second range corresponds to a second coefficient, the third range corresponds to a third coefficient, and the fourth range corresponds to a fourth coefficient; The calculation formula for the quality accuracy of the key evaluation index value is: Wherein, N represents the number of key evaluation indicators in each flight phase; i represents the i-th key evaluation indicator; y∈Y, x∈X, respectively represent flight subject X and Y flight actions; A i，y，x k represents the quality accuracy of the i-th key evaluation indicator of the y-th flight action in flight subject x; t , k t+1 represents coefficients, wherein t={1, 2, 3}, k1 is the first coefficient, k2 is the second coefficient, k3 is the third coefficient, and k4 is the fourth coefficient; LU t and LD t Respectively represent the upper and lower limits of the first range / second range / third range / fourth range; LU3>LU2>LU1>S i >LD1>LD2>LD3, S i is the quality standard value of the i-th key evaluation indicator; (LD1, LU1) is the first range, (LD2, LD1) and (LU1, LU2) are the second range, (LD3, LD2) and (LU2, LU3) are the third range, and the others are the fourth range.

5. The flight quality comprehensive evaluation method based on the time series fusion model according to claim 4 is characterized in that: According to the severity level of the flight subject assessment indicators, set the assessment indicator influencing factors, including: The impact is divided into four hazard severity levels: catastrophic, severe, minor and mild; Set the impact factor formula to: Among them, f r is the impact index, r={1, 2, 3, 4}, the disaster level index f1=1, the severe level index f2=2, the mild level index f3=3, and the slight level index f4=4.

6. The flight quality comprehensive evaluation method based on the time series fusion model according to claim 5 is characterized in that: The evaluation scores of all the flight actions are obtained according to the flight quality accuracy and influencing factors, and the flight quality evaluation score of a single flight is obtained in combination with the completion degree of the flight subjects, including: Get a single key indicator evaluation score: Get the quality assessment score of a single action: Get the flight quality assessment score for a single flight: in, Q represents the number of flight action elements in flight subject x, and q represents the actual flight completion status.

7. The flight quality comprehensive evaluation method based on the time series fusion model according to claim 1 is characterized in that: According to the fleet pilots’ historical sortie evaluation scores and the time series impact attenuation fusion formula, the pilots’ and fleets’ flight quality comprehensive evaluation scores are obtained, including: According to the historical evaluation scores of each pilot, a stability factor is set one by one, and a time series impact attenuation fusion formula is established in combination with the attenuation factor to calculate the flight quality evaluation score of each pilot in the current fleet corresponding to a single flight; Wherein, the stability factor is: Among them, σ a represents the standard deviation of pilot a’s historical flight evaluation scores.

8. The flight quality comprehensive evaluation method based on the time series fusion model according to claim 7 is characterized in that: The timing impact attenuation fusion formula is: Among them, F a represents the comprehensive evaluation score of pilot a based on his j flights, S n represents the nth flight quality evaluation score, λ an is the stability factor of pilot a’s nth flight, and γ is the attenuation factor.

9. The flight quality comprehensive evaluation method based on the time series fusion model according to claim 8 is characterized in that: The attenuation factor γ is set to 0.

002.

10. The flight quality comprehensive evaluation method based on the time series fusion model according to claim 5 is characterized in that: Set the influence index of the maximum descent rate during approach to 4; set the influence index of the maximum speed during straight taxiing to 1; set the influence index of the touchdown speed to 2; set the influence index of the retracted landing gear height to 1; set the influence index of the maximum climb speed to 3; set the influence index of the maximum deviation of the landing roll heading angle to 2; set the influence index of the flap extension speed to 1.

Images