Robust Meta-Aerodynamic Modeling and Prediction Method and Medium for Fixed-Wing Aircraft

Abstract

The present invention relates to a robust element flight aerodynamic modeling and prediction method and medium for a fixed-wing aircraft, which obtains discrete aerodynamic data sets under different state conditions and establishes an aerodynamic common basis function for the fixed-wing aircraft. Aiming at the disturbances of the true aerodynamic force and the true aerodynamic moment relative to the nominal aerodynamic model, on the basis of the aerodynamic common basis function, a robust common basis function is augmented to obtain an augmented common basis function. Using the augmented common basis function as the basis, the aerodynamic force and the aerodynamic moment of the fixed-wing aircraft are approximated to obtain a robust element flight aerodynamic model. Compared with the prior art, it lays a good foundation for effectively identifying the influence of wind and accurately predicting the aerodynamic force online, enables the robust element flight aerodynamic model to have better migration ability and engineering applicability, and thus improves the accuracy of predicting the aerodynamic force and the aerodynamic moment of the fixed-wing aircraft under strong wind conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of aerodynamic force prediction of aircraft, and particularly relates to a robust meta-flight aerodynamic force modeling and prediction method and medium for a fixed-wing aircraft. Background Art

[0002] Wind-resistant flight is the key and difficulty of the all-weather autonomous flight of current fixed-wing unmanned aerial vehicles. Due to the limitation of environmental interference factors such as wind, fixed-wing unmanned aerial vehicles are often referred to as "good weather aircraft". To achieve precise control of the maneuvering flight of unmanned aerial vehicles and improve the safety and reliability of unmanned aerial vehicles flying in windy environments, it is necessary to accurately predict the aerodynamic force (moment) of unmanned aerial vehicles under wind interference conditions in real time. Considering costs, fixed-wing unmanned aerial vehicles usually do not adopt overly expensive precision airflow sensor devices. Therefore, it is difficult to obtain accurate angle-of-attack and sideslip angle information in real time. Moreover, traditional aerodynamic models that describe the variation law of the aerodynamic force (moment) of an aircraft using physical variables such as the angle-of-attack and sideslip angle for the motion of an aircraft in the airflow coordinate system are not conducive to online migration applications during the flight of unmanned aerial vehicles. Therefore, for the real-time and accurate prediction of the aerodynamic force (moment) of fixed-wing unmanned aerial vehicles, the development of algorithms, especially advanced algorithms, is extremely urgent and necessary.

[0003] Regarding the online aerodynamic modeling and prediction of aircraft under strong wind interference, scholars first proposed the "Neural-fly" method for rotary-wing aircraft. Deep meta-learning with the Generative Adversative Nets (GAN) architecture was used to train a common basis function neural network model, and a common aerodynamic basis function model for rotary-wing aircraft under different wind conditions was established, thus realizing real-time aerodynamic modeling and transfer application under unknown wind conditions. Currently, for fixed-wing aircraft, considering more variables such as the angular velocity and rudder deflection angle of the aircraft, the deep meta-learning method is used to establish a common basis function neural network model for the aerodynamic force and aerodynamic moment of the aircraft under different wind conditions, which can better predict the aerodynamic force and aerodynamic moment under wind disturbance conditions. Aiming at the problems that the black-box neural network common basis function model in "Neural-fly" is difficult to be efficiently trained, fully tested and reliably verified in practical applications, scholars further developed a meta-flight aerodynamic modeling method for fixed-wing aircraft considering wind interference by using the multivariate function variable decomposition technology, and obtained an aerodynamic common basis function model for the aerodynamic force and aerodynamic moment of the aircraft that is determined by the characteristics of the fixed-wing aircraft itself and is common under different wind conditions, which can better achieve accurate prediction of the aerodynamic force and aerodynamic moment of the aircraft under wind disturbance conditions. However, due to the influence of factors such as engine jet flow and propeller slip flow in actual flight, there may be differences between the real aerodynamic force and real aerodynamic moment received by the aircraft and the results obtained through CFD numerical calculation and wind tunnel test. The aerodynamic common basis function of the fixed-wing aircraft obtained by the above method still has the problem of low reliability in online transfer application in practical applications. Summary of the Invention

[0004] The technical problem to be solved by this application is to provide a robust meta-flight aerodynamic modeling and prediction method for fixed-wing aircraft, which has the characteristic of improving the online prediction accuracy of the aerodynamic force and aerodynamic moment of fixed-wing aircraft under strong wind conditions.

[0005] In a first aspect, in one embodiment, a robust meta-flight aerodynamic modeling method for a fixed-wing aircraft is provided, including:

[0006] Obtain a discrete aerodynamic data set under different state conditions, and establish an aerodynamic common basis function of the fixed-wing aircraft;

[0007] Regarding the perturbations of the real aerodynamic force and real aerodynamic moment relative to the nominal aerodynamic model, on the basis of the aerodynamic common basis function, augment a common basis function with robustness to obtain an augmented common basis function;

[0008] Using the augmented common basis function as the basis, approximate the aerodynamic force and aerodynamic moment of the fixed-wing aircraft to obtain a robust meta-flight aerodynamic model.

[0009] In a second aspect, in one embodiment, a robust meta-flight aerodynamic force prediction method for a fixed-wing aircraft is provided, including:

[0010] Determine the wind disturbance action coefficient;

[0011] Obtain the motion variables of the fixed-wing aircraft relative to the earth, where the motion variables include flight altitude, ground speed vector, Euler attitude angle, angular velocity, and rudder deflection angle;

[0012] Input the motion variables and the wind disturbance action coefficient into the robust meta-flight aerodynamic model for aerodynamic force and aerodynamic moment prediction;

[0013] The robust meta-flight aerodynamic model is modeled based on the fixed-wing aircraft robust meta-flight aerodynamic force modeling method described in any one of the above embodiments.

[0014] In a third aspect, in one embodiment, a computer-readable storage medium is provided, in which a program is stored, and the program can be loaded and executed by a processor to perform the fixed-wing aircraft robust meta-flight aerodynamic force modeling method and / or the fixed-wing aircraft robust meta-flight aerodynamic force prediction method described in any one of the above embodiments.

[0015] The beneficial effects of the present invention are:

[0016] The augmented aerodynamic common basis function constructed in the robust meta-flight aerodynamic force modeling can accurately approximate the aerodynamic force and aerodynamic moment of the fixed-wing aircraft under the perturbation of the aerodynamic model, which lays a good foundation for effectively identifying the influence of the wind and accurately predicting the aerodynamic force online, enables the robust meta-flight aerodynamic model to have better migration ability and engineering applicability, and thus improves the accuracy of predicting the aerodynamic force and aerodynamic moment of the fixed-wing aircraft under strong wind conditions. Description of the Drawings

[0017] Figure 1 is a schematic flow chart of the fixed-wing aircraft robust meta-flight aerodynamic force modeling method according to an embodiment of the present application;

[0018] Figure 2 is Figure 1 a schematic flow chart of the method according to an embodiment of step S10 in the present application;

[0019] Figure 3 is a schematic flow chart of the fixed-wing aircraft robust meta-flight aerodynamic force prediction method according to an embodiment of the present application;

[0020] Figure 4 is a schematic diagram of the pitch moment prediction error of the robust meta-flight aerodynamic model in the discrete state according to an embodiment of the present application;

[0021] Figure 5Schematic diagram of the predicted pitch moment of the robust meta-flight aerodynamic model in the continuous state of an embodiment of the present application;

[0022] Figure 6 Schematic diagram of the prediction error of the pitch moment of the robust meta-flight aerodynamic model in the continuous state of an embodiment of the present application. Detailed implementation manners

[0023] The present invention will be further described in detail below in conjunction with the accompanying drawings through specific implementation manners. Similar elements in different implementation manners are labeled with related similar element numbers. In the following implementation manners, many detailed descriptions are provided to enable a better understanding of the present application. However, those skilled in the art can easily recognize that some of the features can be omitted in different situations, or can be replaced by other elements, materials, or methods. In some cases, some operations related to the present application are not shown or described in the specification to avoid overwhelming the core part of the present application with excessive descriptions. For those skilled in the art, it is not necessary to describe these related operations in detail, and they can fully understand the related operations based on the descriptions in the specification and the general technical knowledge in the art.

[0024] In addition, the features, operations, or characteristics described in the specification can be combined in any appropriate manner to form various implementation manners. At the same time, the steps or actions in the method description can also be reordered or adjusted in a manner obvious to those skilled in the art. Therefore, the various sequences in the specification and drawings are only for clearly describing a certain embodiment and do not mean that they are necessary sequences, unless it is stated that a certain sequence must be followed.

[0025] The serial numbers assigned to the components herein, such as "first", "second", etc., are only used to distinguish the described objects and do not have any sequential or technical meaning.

[0026] For the convenience of explaining the inventive concept of the present application, the aerodynamic prediction technology of fixed-wing aircraft is briefly described below.

[0027] The coordinate systems related to the meta-flight aerodynamic modeling of a fixed-wing aircraft include the body coordinate system, the airflow coordinate system, and the ground coordinate system. The aircraft body coordinate system S b is fixed to the aircraft, and the origin o b is located at the center of mass of the aircraft, and o b the x b axis is in the aircraft's symmetry plane and points to the nose, and o b the y b axis is perpendicular to the aircraft's symmetry plane and points to the right side of the fuselage, and o b the z b axis is in the aircraft's symmetry plane and points downward along the fuselage. The airflow coordinate system Sa Fixed to the aircraft, with the origin o a Located at the center of mass of the aircraft, o a The x a axis coincides with the airspeed, o a The z a axis is located in the aircraft's symmetric plane and is perpendicular to o a the x a axis and points downward towards the fuselage, o a The y a axis is perpendicular to o a the x a and z a plane, and its direction is determined by the right-hand rule. The ground coordinate system S g is fixed to the ground, with the origin o g located at a certain point on the ground, o g The x g axis points in a certain direction in the horizontal plane, o g The z g axis is perpendicular to the horizontal plane and points towards the center of the earth, o g The y g axis is determined by the right-hand rule.

[0028] In the currently developed technologies, in the aerodynamic modeling of fixed-wing aircraft considering wind interference and based on the common aerodynamic basis functions, two main methods have been developed. One is the "neuron flight" method that uses neural networks to model the common basis functions and obtains the common aerodynamic basis function models of the aerodynamic forces and moments of the aircraft under different wind conditions through deep meta-learning methods; the other is the meta-flight aerodynamic modeling method for fixed-wing aircraft considering wind interference, which decomposes the traditional aerodynamic model into multiple variables and obtains the common aerodynamic basis function models of the aerodynamic forces and moments of the aircraft determined by the characteristics of the fixed-wing aircraft itself and applicable under different wind conditions. Both of these methods construct the common aerodynamic basis functions based on the discrete-state aerodynamic force (moment) datasets obtained by means such as CFD numerical calculations and wind tunnel tests, laying a good foundation for the online prediction of the aerodynamic forces and moments of fixed-wing aircraft under strong wind conditions.

[0029] However, the applicant has found in the research that due to the influence of factors such as engine jet flow and propeller slipstream in actual flight, there may be differences between the actual aerodynamic forces and moments received by the aircraft and the results obtained through CFD numerical calculations and wind tunnel tests, thereby affecting the reliability of the common basis function models obtained by the above two methods in actual online migration applications.

[0030] In view of this, the present application provides a method and medium for robust meta-flight aerodynamic modeling and prediction of a fixed-wing aircraft. For an aerodynamic model perturbed within a certain range, the developed augmented common basis function model is used to effectively approximate and fit the aerodynamic forces and moments of the aircraft, thereby enhancing the robustness of the "neuron flight" aerodynamic model and the meta-flight aerodynamic model, and further improving the accuracy of online prediction of the aerodynamic forces and moments of the aircraft under strong wind conditions, laying a good foundation for the migration application of online prediction of aerodynamic forces and moments during the flight of a fixed-wing aircraft in a real wind field.

[0031] In an embodiment of the present application, a method for robust meta-flight aerodynamic modeling of a fixed-wing aircraft is provided. Please refer to Figure 1 , including:

[0032] Step S10: Obtain a discrete aerodynamic data set under different state conditions, and establish an aerodynamic common basis function of the fixed-wing aircraft.

[0033] In an embodiment, establishing the aerodynamic common basis function of the fixed-wing aircraft can include two methods, including:

[0034] Method 1: Use the discrete aerodynamic data set under different state conditions to establish a traditional aerodynamic model represented in the body coordinate system of the fixed-wing aircraft, perform Taylor expansion on the aerodynamic forces and moments in the aerodynamic model with respect to the wind speed vector represented in the ground coordinate system, compare the approximation accuracy of the Taylor expansion expressions at different orders to the original function, and obtain the expansion orders of the aerodynamic forces and aerodynamic moments that meet the preset accuracy requirements, thereby determining the common basis functions of the aerodynamic force and aerodynamic moment components.

[0035] Method 2: Use the discrete aerodynamic data set under different state conditions, based on different wind condition disturbances, expand the discrete aerodynamic data set to obtain multiple labeled data sets, and based on deep meta-learning, use a neural network model to model and train the aerodynamic common basis function, thereby obtaining a neural network model of the common basis function of the aerodynamic force and aerodynamic moment components.

[0036] In an embodiment, please refer to Figure 2 , step S10 may include:

[0037] Step S101: Obtain a discrete aerodynamic data set under different state conditions.

[0038] Step S102: Based on the discrete aerodynamic data set under different state conditions, establish an aerodynamic coefficient and aerodynamic moment coefficient model of the fixed-wing aircraft.

[0039] In an embodiment, step S102 includes:

[0040] ,

[0041] ,

[0042] Among them, , and respectively represent the axial force coefficient, lateral force coefficient, and normal force coefficient of the aircraft, that is, the aerodynamic force coefficients of the aircraft. , and respectively represent the rolling moment coefficient, pitching moment coefficient, and yawing moment coefficient of the aircraft, that is, the aerodynamic moment coefficients of the aircraft. represents the angle of attack of the aircraft, represents the sideslip angle of the aircraft, represents the Mach number of the aircraft; , and respectively represent the deflection angles of the elevator, aileron, and rudder of the aircraft; , and respectively represent the static rolling moment coefficient, static pitching moment coefficient, and static yawing moment coefficient of the aircraft; p, q, and r respectively represent the rolling angular velocity, pitching angular velocity, and yawing angular velocity; , , , and respectively represent the rolling moment dynamic derivative coefficient related to the rolling angular velocity, the rolling moment dynamic derivative coefficient related to the yawing angular velocity, the pitching moment dynamic derivative coefficient related to the pitching angular velocity, the yawing moment dynamic derivative coefficient related to the rolling angular velocity, and the yawing moment dynamic derivative coefficient related to the yawing angular velocity; and respectively represent the lateral reference length and longitudinal reference length of the aircraft; V represents the speed amplitude of the aircraft.

[0043] Step S103, based on the aerodynamic force coefficient and aerodynamic moment coefficient models, establish the traditional aerodynamic model of the fixed-wing aircraft.

[0044] In one embodiment, step S103 includes:

[0045] ,

[0046] Among them, , and respectively represent the axial force, lateral force, and normal force of the aircraft, that is, the aerodynamic force of the aircraft. , and They respectively represent the rolling moment, pitching moment, and yaw moment of the aircraft, that is, the aerodynamic moment of the aircraft. Q represents the dynamic pressure, , represents the air density, which is related to the altitude of the aircraft and is a function of the aircraft altitude. S represents the reference area of the aircraft, represents the diagonalization operation operator, represents the 3×3 identity matrix, represents the 3×3 zero matrix.

[0047] Step S104: Based on the traditional aerodynamic model, obtain the aerodynamic common basis functions of the fixed-wing aircraft.

[0048] In one embodiment, based on the traditional aerodynamic model obtained above, the aerodynamic forces and moments in the aerodynamic model are Taylor-expanded with respect to the wind speed vector expressed in the ground coordinate system. By comparing the approximation accuracies of the Taylor expansion expressions at different orders to the original function, the expansion orders of the aerodynamic forces and the expansion orders of the aerodynamic moments that meet the preset accuracy requirements are obtained, thereby determining the common basis functions of the aerodynamic force and aerodynamic moment components.

[0049] Among them, represents the wind speed vector expressed in the ground coordinate system, , and represent the velocity components of the wind speed vector in the x, y, and z directions of the ground coordinate system, and T represents the transpose.

[0050] In one embodiment, the aerodynamic common basis function vector includes:

[0051] ,

[0052] Among them, x represents the motion variable of the aircraft relative to the earth, represents the aerodynamic common basis function vector of the i-th aerodynamic force and aerodynamic moment component, represents the j-th aerodynamic common basis function of the i-th component, 1 ≤ j ≤ N i , N i represents the number of aerodynamic common basis functions of the i-th aerodynamic force and aerodynamic moment component, .

[0053] Make the modeled nominal aerodynamic model satisfy

[0054] .

[0055] Among them, , represents the pseudo-dynamic pressure, , represents the velocity vector of the aircraft relative to the earth in the ground coordinate system, represents the vector modulus operation, represents the standard atmospheric density corresponding to the sea level, 、 and represent the velocity components of the aircraft's ground speed vector in the x, y, and z directions of the ground coordinate system. represents the wind disturbance coefficient function related only to the wind speed variable represented in the ground coordinate system, ; represents the common basis function matrix related only to the motion variable x of the aircraft relative to the earth, , represents the common basis function vector for the axial force, represents the common basis function vector for the lateral force, represents the common basis function vector for the normal force, represents the common basis function vector for the rolling moment, represents the common basis function vector for the pitching moment, represents the common basis function vector for the yawing moment. , h represents the height of the aircraft, represents the Euler attitude angle, , 、 and represent the roll angle, pitch angle, and yaw angle respectively. represents the angular velocity, , represents the rudder deflection angle, .

[0056] Step S20, for the perturbations of the true aerodynamic force and true aerodynamic moment relative to the nominal aerodynamic model, on the basis of the aerodynamic common basis functions, augment the robust common basis functions to obtain the augmented common basis functions.

[0057] The following takes any one component of the aerodynamic force and aerodynamic moment received during the actual flight of a fixed-wing aircraft as an example to illustrate the method of obtaining the augmented common basis functions.

[0058] In one embodiment, the pitching moment is taken as an example for illustration.

[0059] Assume that the pitching moment received during the actual flight of a fixed-wing aircraft is:

[0060] ,

[0061] where, represents the pitching moment actually received by the aircraft, represents the true pitch moment coefficient, or , 、 、 and represent constant perturbation coefficients, represents the nominal pitch moment coefficient.

[0062] Those skilled in the art can understand that the relationship between the true pitch moment coefficient and the nominal pitch moment coefficient is not limited to the above two cases, and there can be other cases.

[0063] In case, the actual pitch moment of the aircraft can be further written as:

[0064] ,

[0065] where, , , .

[0066] Based on the nominal aerodynamic model and the common basis functions for pitch moment obtained previously, there is .

[0067] Therefore, the actual pitch aerodynamic moment of the fixed-wing aircraft can be expressed as:

[0068] .

[0069] Furthermore, the augmented common basis function for pitch moment is obtained as:

[0070] .

[0071] where x represents the motion variable of the aircraft relative to the ground, represents the augmented common basis function vector for pitch moment, represents the common basis function vector for pitch moment, , represents the standard atmospheric density corresponding to sea level, 、 and represent the velocity components of the aircraft's ground speed vector in the x, y, and z directions of the ground coordinate system, represents the pseudo-dynamic pressure, , represents the ground speed vector of the aircraft relative to the ground represented in the ground coordinate system, represents the vector modulus operation, and T represents the transpose operation.

[0072] In In the case of, the actual pitch aerodynamic moment of the fixed-wing aircraft can be expressed as:

[0073] .

[0074] Then the augmented common basis function for the pitch moment is:

[0075] .

[0076] Step S30: Using the augmented common basis function as the basis, approximate the aerodynamic forces and aerodynamic moments of the fixed-wing aircraft to obtain a robust element flight aerodynamic model.

[0077] In one embodiment, step S30 includes:

[0078] ,

[0079] wherein, , , and respectively represent the dimensionless axial force coefficient, lateral force coefficient and normal force coefficient of the aircraft, , and respectively represent the dimensionless roll moment coefficient, pitch moment coefficient and yaw moment coefficient of the aircraft; represents the wind disturbance action coefficient function to be determined online, ; represents the augmented common basis function matrix that is only related to the motion variable x of the aircraft relative to the earth, ; represents the wind speed vector expressed in the ground coordinate system, represents the augmented common basis function vector for the axial force, represents the augmented common basis function vector for the lateral force, represents the augmented common basis function vector for the normal force, represents the augmented common basis function vector for the roll moment, represents the augmented common basis function vector for the yaw moment.

[0080] The aerodynamic common basis function constructed in the obtained robust element flight aerodynamic modeling can accurately approximate the aerodynamic forces and aerodynamic moments of the fixed-wing aircraft under the perturbation of the aerodynamic model, which creates good conditions for effectively identifying the influence of the wind and accurately predicting the aerodynamic force online, making the robust element flight aerodynamic model have better migration ability and engineering applicability.

[0081] Based on the obtained robust meta-flight aerodynamic model, the aerodynamic force and moment of a fixed-wing aircraft can be predicted online in real time.

[0082] Based on the fixed-wing aircraft robust meta-flight aerodynamic force modeling method of any of the above embodiments, for the aerodynamic model with disturbances within a certain range, the developed augmented common basis function model can effectively approximate and fit the aerodynamic force and moment of the aircraft, thereby improving the robustness of the "neuron flight" aerodynamic model and the meta-flight aerodynamic model, obtaining a more robust robust meta-flight aerodynamic model, and further improving the accuracy of online prediction of the aerodynamic force and moment of the aircraft under strong wind conditions, laying a good foundation for the migration application of online prediction of the aerodynamic force and moment during the flight of a fixed-wing aircraft in a real wind field.

[0083] In one embodiment of the present application, a method for predicting the robust meta-flight aerodynamic force of a fixed-wing aircraft is provided, including:

[0084] Step S100, determining the wind disturbance action coefficient.

[0085] In one embodiment, determining the wind disturbance action coefficient includes any one of the following methods:

[0086] Method 1: Numerically differentiate the speed and angular velocity variables of the aircraft, solve the aerodynamic force and moment of the aircraft using the aircraft dynamics equation, and use the solved aerodynamic force and moment as observables. Based on the robust meta-flight aerodynamic model, the least squares method is used to determine the wind disturbance action coefficient.

[0087] Method 2: Using the speed and angular velocity variables of the aircraft as observables, based on the robust meta-flight aerodynamic model and the aircraft motion dynamics equation, regarding the wind disturbance action coefficient as a state variable, an extended Kalman filter is constructed to determine the wind disturbance action coefficient.

[0088] Step S200, obtaining the motion variables of the fixed-wing aircraft relative to the ground, where the motion variables include flight altitude, ground speed vector, Euler attitude angle, angular velocity, and rudder deflection angle.

[0089] Step S30, inputting the motion variables and the wind disturbance action coefficient into the robust meta-flight aerodynamic model for predicting the aerodynamic force and moment. Among them, the robust meta-flight aerodynamic model is modeled based on the fixed-wing aircraft robust meta-flight aerodynamic force modeling method of any of the above embodiments.

[0090] To facilitate the description of the prediction effect, the robust meta-flight aerodynamic force modeling is carried out and verified for a certain fixed-wing UAV below.

[0091] Since traditional aerodynamic modeling is an existing technology, the traditional aerodynamic moment model of a certain publicly available fixed-wing aircraft is used for modeling here. It is assumed that the actual aerodynamic moment coefficients , , and the nominal aerodynamic moment coefficients , , have a constant + proportional relationship, specifically

[0092] ,

[0093] ,

[0094] .

[0095] In the robust meta-flight aerodynamic modeling, first, the common basis function of the computational aerodynamic force is established through the meta-flight aerodynamic modeling technology, and the common basis function of the aerodynamic moment of the fixed-wing aircraft with the altitude , ground speed vector , Euler attitude angle , angular velocity , and rudder deflection angle as inputs (uniformly represented by the variable x) is obtained. To balance the approximation accuracy and generalization performance, the common basis function vectors , , are all taken as 20-dimensional. Further, the given is extended to , and to obtain the augmented common basis function vectors , and with a dimension of 25. During flight, the augmented common basis function is used as the basis, and combined with the wind disturbance action coefficient determined by the identification method, a robust meta-flight aerodynamic model of the fixed-wing aircraft is constructed to predict the aerodynamic moment received by the fixed-wing aircraft.

[0096] The developed robust meta-flight aerodynamic model is verified, considering two cases, one is the discrete state case, and the other is the continuous state case.

[0097] For the discrete state case, the aerodynamic moment prediction in the case of randomly given aircraft states is considered. The wind field speed in the environment is set to 20 m / s, and the direction vector is in the earth coordinate system. The altitude of the aircraft is h = 900 m, the Euler attitude angle is randomly given, and the representation of the ground speed vector of the aircraft relative to the earth in the body coordinate system is given as , , Subject to a uniform random distribution in the interval [-5, 5] m / s, the ground speed of the aircraft is determined by conversion calculation through Euler attitude angles. The angular velocities of the three axes of the aircraft all follow a uniform distribution in the interval [-1, 1] rad / s, and the elevator deflection angle , rudder deflection angle , aileron deflection angle Subject to a uniform distribution in the interval [-25, 25] deg. Based on the least squares method, the aerodynamic moment measurement data of the first 500 states are taken as identification data to determine the wind disturbance action coefficient, and then the aerodynamic moments of the last 1500 states are predicted. There is noise interference at the 40% level of the true value in the measurement data.

[0098] Please refer to Figure 4 , which shows the prediction error of the pitching aerodynamic moment of the aircraft. In the figure represents the number of discretizations, represents the pitching moment error, where the identification segment is the error between the measured pitching moment and the true pitching moment, and the prediction segment is the error between the predicted pitching moment and the true pitching moment. It can be seen from the figure that the predicted value based on the robust meta-flight aerodynamic model greatly reduces the measurement error, and the result is closer to the true value.

[0099] For the continuous state case, consider the aerodynamic moment prediction of the ideal forced pitching motion of the aircraft. The wind field speed in the environment is set to 20 m / s, and the direction is given as in the ground coordinate system. The altitude of the aircraft is h = 900 m, and the pitching attitude angle changes regularly at deg. The roll angle and yaw angle are 0 deg. The corresponding angular velocities of the three axes are determined by the attitude angle changes. The representation of the velocity of the aircraft relative to the earth in the ground coordinate system is given as m / s. The elevator control law is deg, the rudder deflection angle is = 10 deg, the aileron deflection angle is = 0 deg, and the data sampling time interval is . Based on the least squares method, the aerodynamic moment measurement data of the first 10 s are taken as identification data to determine the wind disturbance action coefficient, and then the aerodynamic moment data of the last 40 s are predicted. There is noise interference at the 40% level of the true value in the measurement data.

[0100] Please refer to Figure 5 and Figure 6 , Figure 5 which shows the prediction result of the pitching aerodynamic moment, Figure 6 which shows the prediction error of the pitching aerodynamic moment. In the figure represents time, Indicates the pitch moment error, where the identification segment is the error between the measured pitch moment and the true pitch moment, and the prediction segment is the error between the predicted pitch moment and the true pitch moment. The calculation results show that the predicted value is in good agreement with the true value, demonstrating the effectiveness of the robust meta-flight aerodynamic model.

[0101] In one embodiment of the present application, a computer-readable storage medium is provided. A program is stored on the storage medium, and the stored program includes the robust meta-flight aerodynamic modeling method and / or the robust meta-flight aerodynamic prediction method for a fixed-wing aircraft in any of the above embodiments that can be loaded and processed by a processor.

[0102] Those skilled in the art can understand that all or part of the functions of the above methods can be implemented in a hardware manner or in a computer program manner. When all or part of the functions in the above embodiments are implemented in a computer program manner, the program can be stored in a computer-readable storage medium, which can include: read-only memory, random access memory, magnetic disk, optical disk, hard disk, etc. The above functions are implemented by a computer executing the program. For example, the program is stored in the memory of the device, and when the processor executes the program in the memory, the above all or part of the functions can be implemented. In addition, when all or part of the functions in the above embodiments are implemented in a computer program manner, the program can also be stored in a storage medium such as a server, another computer, magnetic disk, optical disk, flash drive or mobile hard disk, downloaded or copied and saved to the memory of the local device, or the system of the local device is updated. When the processor executes the program in the memory, the above all or part of the functions in the above embodiments can be implemented.

[0103] The above uses specific examples to elaborate on the present invention, which is only used to help understand the present invention and is not intended to limit the present invention. For those skilled in the art of the present invention, based on the idea of the present invention, several simple deductions, deformations or substitutions can also be made.

Claims

1. A robust meta-flight aerodynamic modeling method for fixed-wing aircraft, characterized in that Including: Obtain discrete aerodynamic data sets under different state conditions, and establish an aerodynamic common basis function for a fixed-wing aircraft; For the perturbations of the true aerodynamic force and the true aerodynamic moment relative to the nominal aerodynamic model, on the basis of the aerodynamic common basis function, augment the common basis function with robustness to obtain an augmented common basis function, including: For the perturbations of the aerodynamic force and the aerodynamic moment suffered by a fixed-wing aircraft during actual flight relative to the nominal aerodynamic model, the method for obtaining the augmented common basis function includes: Let the pitch moment suffered by a fixed-wing aircraft during actual flight be: Among them, represents the pitching moment actually received by the aircraft, Q represents the dynamic pressure, , represents the air density, V represents the speed amplitude of the aircraft, represents the longitudinal reference length of the aircraft; represents the true pitching moment coefficient, or , , , and represent constant perturbation coefficients, represents the nominal pitching moment coefficient, q represents the roll angular velocity, represents the elevator deflection angle of the aircraft; then the augmented common basis function for the pitching moment is obtained, including: Or, Among them, \(x\) represents the motion variable of the aircraft relative to the ground, represents the augmented common basis function vector with respect to the pitching moment, represents the common basis function vector with respect to the pitching moment, , represents the standard atmospheric density corresponding to sea level, , and represent the velocity components of the aircraft's ground speed vector in the three directions of the ground coordinate system \(x\), \(y\), and \(z\), represents the pseudo-dynamic pressure, , represents the aircraft's ground speed vector expressed in the ground coordinate system, represents the vector modulus operation, and \(T\) represents the transpose operation; Using the augmented common basis function as the basis, approximate the aerodynamic force and the aerodynamic moment of the fixed-wing aircraft to obtain a robust meta-flight aerodynamic model.

2. The robust meta-flight aerodynamic modeling method for a fixed-wing aircraft according to claim 1, wherein, The obtaining of the discrete aerodynamic data sets under different state conditions and the establishment of the aerodynamic common basis function for the fixed-wing aircraft include any one of the following methods: Method 1: Use the discrete aerodynamic data sets under different state conditions to establish a traditional aerodynamic model expressed in the body coordinate system of the fixed-wing aircraft, perform Taylor expansion on the aerodynamic force and the aerodynamic moment in the aerodynamic model with respect to the wind speed vector expressed in the ground coordinate system, compare the approximation accuracy of the Taylor expansion expressions at different orders to the original function, and obtain the expansion orders of the aerodynamic force and the aerodynamic moment that meet the preset accuracy requirements, so as to determine the common basis function of the aerodynamic force and the aerodynamic moment components; Method 2: Use the discrete aerodynamic data sets under different state conditions, based on different wind condition disturbances, expand the discrete aerodynamic data sets to obtain multiple labeled data sets, and based on deep meta-learning, use a neural network model to model and train the aerodynamic common basis function, so as to obtain a neural network model of the common basis function of the aerodynamic force and the aerodynamic moment components.

3. The robust meta-flight aerodynamic modeling method for fixed-wing aircraft according to claim 1, characterized in that The approximating the aerodynamic force and the aerodynamic moment of the fixed-wing aircraft with the augmented common basis function as the basis to obtain a robust meta-flight aerodynamic model includes: Among them, , and respectively represent the axial force, lateral force, and normal force of the aircraft; , and respectively represent the rolling moment, pitching moment, and yaw moment of the aircraft; S represents the reference area of the aircraft, , represents the diagonalization operation operator, represents the lateral reference length of the aircraft; , and respectively represent the dimensionless axial force coefficient, lateral force coefficient, and normal force coefficient of the aircraft, , and respectively represent the dimensionless rolling moment coefficient, pitching moment coefficient, and yaw moment coefficient of the aircraft; represents the wind disturbance action coefficient function, ; represents the augmented common basis function matrix that is only related to the motion variable x of the aircraft relative to the ground, ; represents the wind speed vector expressed in the ground coordinate system, represents the augmented common basis function vector regarding the axial force, represents the augmented common basis function vector regarding the lateral force, represents the augmented common basis function vector regarding the normal force, represents the augmented common basis function vector regarding the rolling moment, represents the augmented common basis function vector regarding the yaw moment.

4. A robust meta-flight aerodynamic prediction method for fixed-wing aircraft, characterized in that, Including: Determine the wind disturbance action coefficient; Obtain the motion variables of the fixed-wing aircraft relative to the ground, and the motion variables include flight altitude, ground speed vector, Euler attitude angle, angular velocity, and rudder deflection angle; Input the motion variables and the wind disturbance action coefficient into the robust meta-flight aerodynamic model for aerodynamic force and aerodynamic moment prediction; The robust meta-flight aerodynamic model is modeled based on the fixed-wing aircraft robust meta-flight aerodynamic force modeling method described in any one of claims 1 to 3.

5. The robust meta-flight aerodynamic prediction method for a fixed-wing aircraft according to claim 4, wherein The determination of the wind disturbance action coefficient includes any one of the following methods: Method 1: Numerically differentiate the speed and angular velocity variables of the aircraft, use the aircraft dynamics equation to solve the aerodynamic force and the aerodynamic moment of the aircraft, and use the solved aerodynamic force and aerodynamic moment as the observed quantities. Based on the robust meta-flight aerodynamic model, use a least squares method to determine the wind disturbance action coefficient; Method 2: Use the speed and angular velocity variables of the aircraft as the observed quantities. Based on the robust meta-flight aerodynamic model and the aircraft motion dynamics equation, regard the wind disturbance action coefficient as a state variable, and construct an extended Kalman filter to determine the wind disturbance action coefficient.

6. A computer-readable storage medium, characterized in that, A program is stored in the medium, and the program can be loaded and executed by a processor to perform the robust meta-flight aerodynamic modeling method for a fixed-wing aircraft as described in any one of claims 1 to 3 and / or the robust meta-flight aerodynamic prediction method for a fixed-wing aircraft as described in any one of claims 4 to 5.

Images