Method for identifying and calculating longitudinal aerodynamic coefficient in open-loop state large angle-of-attack range
By combining optimization methods and genetic algorithms to identify the longitudinal aerodynamic coefficient of the UAV, the problem of discrepancies between the aerodynamic coefficient in flight tests and actual conditions was solved, achieving accurate aerodynamic parameter identification and enhanced model applicability over a wide angle-of-attack range.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- XIAN AIRCRAFT DESIGN INST OF AVIATION IND OF CHINA
- Filing Date
- 2025-12-31
- Publication Date
- 2026-04-10
AI Technical Summary
In existing technologies, the aerodynamic coefficients obtained from flight tests differ from those obtained in actual flight conditions, making it difficult to accurately identify the longitudinal aerodynamic coefficients of UAVs at high angles of attack.
A combinatorial optimization method was adopted, using a genetic algorithm to optimize the longitudinal aerodynamic model parameters. The flight state at a small angle of attack was simulated through a flight simulation model, the sum of squared errors was calculated, and the optimal aerodynamic parameters were identified in multiple sub-intervals. Finally, aerodynamic characteristic curves were plotted to verify the applicability and robustness of the model.
It improves the accuracy and efficiency of longitudinal aerodynamic coefficient identification for UAVs, enhances the applicability of the model under complex flight conditions, and is suitable for different types of UAVs.
Smart Images
Figure CN121835201A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application belongs to the technical field of unmanned aerial vehicle design, and particularly relates to a longitudinal aerodynamic coefficient identification calculation method in an open-loop state and a large angle of attack range. BACKGROUND
[0002] Aircraft aerodynamic coefficients are of great significance in aircraft design, performance evaluation and simulator simulation. Generally, there are three main ways to obtain aerodynamic coefficients: numerical calculation, wind tunnel test and flight test. However, the aerodynamic coefficients obtained by numerical calculation and wind tunnel test often have certain differences from the actual flight state. In contrast, the aerodynamic coefficients obtained by flight test are closer to the actual flight state of the aircraft. Therefore, this paper proposes a longitudinal aerodynamic coefficient parameter identification calculation method in an open-loop state and a large angle of attack range. SUMMARY
[0003] In order to solve the above problems, the application provides a longitudinal aerodynamic coefficient identification calculation method in an open-loop state and a large angle of attack range, to solve the problem that the aerodynamic coefficients obtained by the test in the prior art have differences from the actual flight state.
[0004] The technical scheme of the application is: a longitudinal aerodynamic coefficient identification calculation method in an open-loop state and a large angle of attack range, comprising:
[0005] Obtaining the main parameters of the longitudinal aerodynamic model, determining the collection requirements of the flight test data, and establishing a flight simulation model;
[0006] Simulating the flight state of the unmanned aerial vehicle under the condition of a small angle of attack range based on the flight simulation model, obtaining simulation data, comparing the simulation data with the test data in the actual flight test, calculating the error sum of squares of the two, and minimizing;
[0007] Dividing the entire angle of attack range into multiple subintervals, obtaining the optimal aerodynamic parameters of each subinterval based on the minimized error sum of squares, inputting the optimal aerodynamic parameters into the flight simulation model, comparing and verifying with the test data, and when the difference between the optimal aerodynamic parameters and the test data is within a set range, performing the next step;
[0008] Drawing an aerodynamic characteristic curve according to the optimal aerodynamic parameters, verifying the applicability of the flight simulation model, and optimizing the robustness of the flight simulation model.
[0009] Preferably, the collection requirements of the flight test data include:
[0010] Flight state parameters: flight speed, height and attitude angle of the unmanned aerial vehicle under different angle of attack ranges;
[0011] Flight test conditions: the angle of attack range, flight speed range and environmental parameters of the flight test are determined.
[0012] Preferably, the flight simulation model is established based on the dynamic equation of the UAV and the aerodynamic characteristic data, wherein the dynamic equation of the UAV includes a six-degree-of-freedom motion equation, an aerodynamic characteristic model and a control input model.
[0013] The aerodynamic characteristic data of the UAV includes a lift coefficient, a drag coefficient and a pitching moment coefficient.
[0014] Preferably, the flight state of the UAV in a small angle of attack range is simulated, including:
[0015] A series of flight angle of attack values are set, and the corresponding dynamic response is calculated under a given external excitation;
[0016] The control simulation conditions are consistent with the actual flight test conditions, including flight speed and height parameters.
[0017] Preferably, the error sum of squares of the two is calculated, and the calculation formula is:
[0018] ;
[0019] Wherein, i is the flight state parameter serial number, n is the total number of flight state parameters required for calculating SSE, f(t) is the flight state parameter change function with time, subscript exp represents flight test data, and sim represents flight simulation data.
[0020] Preferably, the method for minimizing the error sum of squares is:
[0021] A genetic algorithm is used as an optimization tool;
[0022] The initial population size, crossover probability and mutation probability of the genetic algorithm are set, and the fitness function is defined as the reciprocal of the error sum of squares;
[0023] In each generation, the error sum of squares corresponding to each individual is calculated, and selection, crossover and mutation operations are performed according to the fitness value; the process is repeated until the convergence condition is met, and the minimized error sum of squares is obtained.
[0024] Preferably, for each sub-interval, the flight state of the UAV in a small angle of attack range is simulated, and the minimized error sum of squares of the simulation data and the test data in the actual flight test is calculated as the optimal aerodynamic parameter.
[0025] Preferably, the optimal aerodynamic parameters of each region are integrated into a unified flight simulation model when compared and verified with test data; a comparison threshold is set, and by comparing the simulation results under different angle of attack ranges with the test data, when the difference is less than the comparison threshold, it is judged that the optimal aerodynamic parameters and the test data difference is within the set range.
[0026] Preferably, according to the optimized optimal aerodynamic parameters, the lift coefficient, drag coefficient and pitch moment coefficient curves of the unmanned aerial vehicle under different angle of attack ranges are drawn.
[0027] Preferably, according to the lift coefficient, drag coefficient and pitch moment coefficient curves under different angle of attack ranges, the lift coefficient, drag coefficient and pitch moment coefficient are extracted;
[0028] The extracted lift coefficient, drag coefficient and pitch moment coefficient are input into the flight simulation model, and under the same flight speed / height conditions, the longitudinal response of the unmanned aerial vehicle is simulated by changing the angle of attack input, and the simulation output is extracted;
[0029] The simulation output and the corresponding test data are plotted in the same coordinate system, and the fitting degree of the two is observed intuitively, and the applicability of the flight simulation model is determined according to the fitting degree.
[0030] Preferably, when the model accuracy of some regions is insufficient, the parameters of the genetic algorithm are further adjusted, and the least squares calculation is performed again.
[0031] The open-loop state large angle of attack range longitudinal aerodynamic coefficient identification calculation method of the application has the following advantages:
[0032] The parameter identification is performed in a combined optimization manner, and the optimization method is a genetic algorithm, which has low requirements for the initial value of the longitudinal aerodynamic coefficient in the aerodynamic model module;
[0033] The angle of attack change range in single flight simulation data and flight test data should not be too large, so as to ensure that the longitudinal aerodynamic coefficient in a small angle of attack range changes approximately linearly with the angle of attack;
[0034] The flight data under multiple different flight speeds (or flight angles of attack) are used for longitudinal aerodynamic coefficient parameter identification, and the longitudinal aerodynamic coefficient characteristics under a large angle of attack range are obtained;
[0035] It has universality and is suitable for longitudinal aerodynamic coefficient parameter identification of different types of unmanned aerial vehicles, improves the efficiency of parameter identification, and enhances the applicability of the model under complex flight conditions. BRIEF DESCRIPTION OF DRAWINGS
[0036] Fig. 1 It is a schematic diagram of the overall process of the application;
[0037] Fig. 2 Figure 1 is a schematic diagram of a flight path of an aircraft according to an embodiment of the present application. DETAILED DESCRIPTION
[0038] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described in more detail below with reference to the drawings in the embodiments of the present application. Identical or similar labels in the drawings represent identical or similar elements or elements with identical or similar functions throughout. The described embodiments are part of the embodiments of the present application, rather than all the embodiments of the present application. The embodiments described below with reference to the drawings are exemplary and are intended to explain the present application, and cannot be understood as a limitation of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative labor fall within the scope of protection of the present application. The embodiments of the present application will be described in detail below with reference to the drawings.
[0039] The first aspect of the present application provides a method for identifying and calculating longitudinal aerodynamic coefficients in a large angle-of-attack range in an open loop state, as shown in Figs. 1-2 , the main parameters of the longitudinal aerodynamic model and the parameters that should be included in the flight test data are first determined, and then the corresponding flight simulation model is established; a combined optimization method is used to identify and calculate the longitudinal aerodynamic model parameters in a small angle-of-attack range. First, flight simulation data under a given longitudinal aerodynamic model is calculated, then the sum of squared errors between the flight simulation data and the flight test data is calculated, and finally a genetic algorithm is used to optimize the calculation according to the sum of squared errors under different longitudinal aerodynamic model parameters, and the longitudinal aerodynamic model parameters in the small angle-of-attack range are finally obtained. The longitudinal aerodynamic model parameter solving and calculation in the small angle-of-attack range is repeated multiple times to obtain the longitudinal aerodynamic coefficient characteristics in the large angle-of-attack range.
[0040] Specifically, the following steps are included:
[0041] Step S100, the main parameters of the longitudinal aerodynamic model are obtained, the acquisition requirements of the flight test data are determined, and the flight simulation model is established.
[0042] Step S110, the key parameters involved in the longitudinal aerodynamic characteristics of the unmanned aerial vehicle need to be first determined. These parameters include but are not limited to the lift coefficient, the drag coefficient, the pitch moment coefficient, and the nonlinear aerodynamic characteristic parameters related to the angle of attack, etc.
[0043] The specific steps are as follows:
[0044] (1) According to the flight characteristics of the UAV, relevant aerodynamic documents or manuals are consulted to determine the standard form of the longitudinal aerodynamic model and the key parameters required. For example, the lift coefficient is usually expressed as a function of the angle of attack a, that is, (CL = CL0 + CLa a), where C L0 is the lift coefficient at the zero-lift angle of attack, and CLa is the slope of the lift curve.
[0045] (2) According to the actual flight characteristics of the UAV, further clarify the possible nonlinear terms or high-order effects (such as stall, airflow separation, etc.) in the model, and include them in the parameter list.
[0046] Step S120, determine the collection requirements of flight test data
[0047] (1) Flight state parameters: flight speed, height, attitude angle (pitch angle, roll angle, yaw angle) of the UAV at different angles of attack.
[0048] (2) Flight test conditions: clarify the angle of attack range (usually including small angle of attack and large angle of attack regions), flight speed range and environmental parameters (temperature, air pressure, etc.) of the flight test.
[0049] Step S200, simulate the flight state of the UAV in the small angle of attack range based on the flight simulation model, obtain the simulation data, compare the simulation data with the test data in the actual flight test, calculate the sum of squared errors of the two, and minimize.
[0050] Step 210: establish a flight simulation model based on the dynamic equation of the UAV and the aerodynamic characteristic data, wherein the dynamic equation of the UAV includes six-degree-of-freedom motion equation, aerodynamic characteristic model and control input model.
[0051] The aerodynamic characteristic data of the UAV includes lift coefficient, drag coefficient, lateral force coefficient, lift-drag ratio and directional moment coefficient.
[0052] Preferably, the flight state of the UAV in the small angle of attack range is simulated, including:
[0053] A series of flight angles of attack are set, and the corresponding dynamic response is calculated under given external excitation;
[0054] The control simulation conditions are consistent with the actual flight test conditions, including flight speed and height parameters.
[0055] Step 220: compare the aerodynamic load values obtained by simulation with the measured data in the actual flight test, and calculate the sum of squared errors (Sum of Squared Errors, SSE) of the two. The sum of squared errors of the two is calculated, and the calculation formula is:
[0056] ;
[0057] where i is the flight state parameter index, n is the total number of flight state parameters needed for SSE calculation, f(t) is the flight state parameter variation function with time, subscript exp represents flight test data, and sim represents flight simulation data.
[0058] Step 230: Optimize parameters to minimize the sum of squared errors:
[0059] To improve the accuracy of the model, key parameters in the longitudinal aerodynamic model need to be optimized. The specific method is as follows:
[0060] (1) Use Genetic Algorithm (GA) as the optimization tool. This algorithm can find the global optimal solution in complex nonlinear problems by simulating natural selection and genetic mechanisms.
[0061] (2) Set the initial population size, crossover probability, mutation probability, and other parameters of the genetic algorithm, and define the fitness function as the reciprocal of the sum of squared errors (SSE) (i.e. the higher the fitness, the better the model).
[0062] (3) Iterative optimization process: In each generation, calculate the (SSE) corresponding to each individual (i.e. a set of aerodynamic parameters), and perform selection, crossover and mutation operations according to the fitness value. Repeat this process until the convergence condition is met (such as the number of iterations reaches the upper limit or the fitness change is less than the threshold).
[0063] Step S300, divide the entire angle of attack range into multiple subintervals, obtain the optimal aerodynamic parameters for each subinterval based on the minimized sum of squared errors, input each optimal aerodynamic parameter into the flight simulation model, and compare and verify with the test data. When the difference between the optimal aerodynamic parameters and the test data is within the set range, execute the next step.
[0064] Step 310: Divide the angle of attack interval:
[0065] Since the aerodynamic characteristics of the UAV may differ significantly in different angle of attack regions (e.g. linear characteristics in small angle of attack regions, and stall or airflow separation phenomena in large angle of attack regions), the entire angle of attack range needs to be divided into multiple subintervals. The aerodynamic characteristics in each subinterval can be described by different models.
[0066] Step 320: Repeat the parameter identification process:
[0067] For each angle of attack subinterval, repeat the simulation, comparison and optimization process in step two to obtain the optimal aerodynamic parameters in that region. In particular, in the large angle of attack region, the following factors need to be considered:
[0068] (1) Introduction of nonlinear effects: For example, the lift coefficient may no longer have a simple linear relationship with the angle of attack, but may have a stall point or saturation phenomenon.
[0069] (2) Multivariable optimization: In the large angle of attack range, there may be coupling effects between aerodynamic parameters (such as the mutual influence of lift and drag changes), and multiple parameters need to be optimized simultaneously to obtain an accurate model.
[0070] Step 330: Verify the accuracy of the overall model:
[0071] After completing the parameter identification of all sub-intervals, the optimal parameters of each region are integrated into a unified flight simulation model. Set a comparison threshold, compare the simulation results under different angles of attack with the test data, and when the difference is less than the comparison threshold, determine that the optimal aerodynamic parameters are within the set range. Otherwise, recalculate the optimal aerodynamic parameters.
[0072] Step S400, according to each optimal aerodynamic parameter, draw the aerodynamic characteristic curve, verify the applicability of the flight simulation model, and optimize the robustness of the flight simulation model.
[0073] Step 410: Draw the aerodynamic characteristic curve:
[0074] According to the optimized aerodynamic parameters, draw the lift coefficient, drag coefficient, and pitch moment coefficient curves of the UAV under different angles of attack. These curves should accurately reflect the aerodynamic characteristics of the UAV, especially the change law in the small angle of attack and large angle of attack regions.
[0075] Step 420: Verify the applicability of the model:
[0076] Input the extracted lift coefficient, drag coefficient, and pitch moment coefficient into the flight simulation model, change the angle of attack input under the same flight speed / height conditions, simulate the longitudinal response of the UAV, and extract the simulation output;
[0077] By comparing the simulation results with the actual flight test data, the prediction accuracy of the model under different flight conditions is evaluated. In particular, in the large angle of attack range, the ability of the model to describe nonlinear phenomena such as stall and airflow separation needs to be focused on.
[0078] Step 430: Optimize the robustness of the model:
[0079] If the accuracy of the model in some regions is found to be insufficient, the parameters of the genetic algorithm can be further adjusted (such as increasing the population size or improving the crossover and mutation strategy), and the least squares calculation is performed again to improve the efficiency and accuracy of the optimization process.
[0080] In summary, the application has the following advantages:
[0081] The parameter identification is performed in a combined optimization manner, and the optimization method is a genetic algorithm, and the initial value of the longitudinal aerodynamic coefficient in the aerodynamic model module has a low requirement;
[0082] The change range of the angle of attack in the single flight simulation data and the flight test data cannot be too large, so as to ensure that the longitudinal aerodynamic coefficient in a small angle of attack range changes approximately linearly with the angle of attack;
[0083] The flight data under multiple different flight speeds (or flight angles of attack) are used for longitudinal aerodynamic coefficient parameter identification, and the longitudinal aerodynamic coefficient characteristics in a large angle of attack range are obtained;
[0084] It is universal, suitable for longitudinal aerodynamic coefficient parameter identification of different types of unmanned aerial vehicles, improves the parameter identification efficiency, and enhances the applicability of the model in complex flight conditions.
[0085] The above is only a specific embodiment of the application, but the protection scope of the application is not limited to this. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the application, which should be covered within the protection scope of the application. Therefore, the protection scope of the application should be subject to the protection scope of the claims.
Claims
1. A method for identifying and calculating longitudinal aerodynamic coefficient parameters in an open-loop state, characterized in that, include: Obtain the main parameters of the longitudinal aerodynamic model, determine the requirements for flight test data acquisition, and establish a flight simulation model; The flight state of the UAV under small angle of attack range is simulated based on the flight simulation model to obtain simulation data. The simulation data is compared with the test data in the actual flight test, the sum of squared errors between the two is calculated and minimized. The entire angle of attack range is divided into multiple sub-intervals. The optimal aerodynamic parameters for each sub-interval are obtained based on minimizing the sum of squared errors. The optimal aerodynamic parameters are input into the flight simulation model and compared with the experimental data. When the difference between the optimal aerodynamic parameters and the experimental data is within the set range, the next step is executed. Aerodynamic characteristic curves are plotted based on each optimal aerodynamic parameter to verify the applicability of the flight simulation model and optimize its robustness.
2. The method for identifying and calculating the longitudinal aerodynamic coefficient in the large angle-of-attack range in open-loop condition as described in claim 1, characterized in that, The requirements for collecting flight test data include: Flight status parameters: flight speed, altitude, and attitude angle of the UAV at different angles of attack; Flight test conditions: Define the angle of attack range, flight speed range, and environmental parameters for the flight test.
3. The method for identifying and calculating the longitudinal aerodynamic coefficient in the large angle-of-attack range in open-loop condition as described in claim 1, characterized in that, A flight simulation model is established based on the dynamic equations and aerodynamic characteristic data of the UAV. The dynamic equations of the UAV include six-degree-of-freedom motion equations, aerodynamic characteristic model and control input model. The aerodynamic characteristics data of the UAV include lift coefficient, drag coefficient, and pitch moment coefficient.
4. The method for identifying and calculating the longitudinal aerodynamic coefficient in the large angle-of-attack range in open-loop condition as described in claim 1, characterized in that, Simulate the flight state of a drone at a small angle of attack, including: Set a series of flight angle of attack values and calculate the corresponding dynamic response under a given external excitation; The control simulation conditions are consistent with the actual flight test conditions, including flight speed and altitude parameters.
5. The method for identifying and calculating the longitudinal aerodynamic coefficient in the large angle-of-attack range in open-loop condition as described in claim 2, characterized in that, The sum of squared errors between the two is calculated using the following formula: ； Where i is the sequence number of the flight state parameter, n is the total number of flight state parameters required to calculate SSE, f(t) is the function of the flight state parameter changing over time, the subscript exp represents flight test data, and sim represents flight simulation data.
6. The method for identifying and calculating the longitudinal aerodynamic coefficient in the large angle-of-attack range in open-loop condition as described in claim 5, characterized in that, The method for minimizing the sum of squared errors is as follows: Use genetic algorithms as optimization tools; Set the initial population size, crossover probability, and mutation probability for the genetic algorithm; and define the fitness function as the reciprocal of the sum of squared errors. In each generation, the sum of squared errors for each individual is calculated, and selection, crossover, and mutation operations are performed based on the fitness value; this process is repeated until the convergence condition is met, and the minimized sum of squared errors is obtained.
7. The method for identifying and calculating the longitudinal aerodynamic coefficient in the large angle-of-attack range in open-loop condition as described in claim 1, characterized in that, For each sub-interval, the flight state of the UAV in the small angle of attack range is simulated, and the minimum sum of squared errors between the simulation data and the test data in the actual flight test is calculated as the optimal aerodynamic parameters.
8. The method for identifying and calculating the longitudinal aerodynamic coefficient in the large angle-of-attack range in open-loop condition as described in claim 7, characterized in that, When comparing and verifying with experimental data, the optimal aerodynamic parameters of each region are integrated into a unified flight simulation model; a comparison threshold is set, and by comparing the simulation results with experimental data at different angles of attack, when the difference is less than the comparison threshold, it is determined that the difference between the optimal aerodynamic parameters and the experimental data is within the set range.
9. The method for identifying and calculating the longitudinal aerodynamic coefficient in the large angle-of-attack range in open-loop condition as described in claim 1, characterized in that, Based on the optimized aerodynamic parameters, the lift coefficient, drag coefficient, and pitch moment coefficient curves of the UAV at different angles of attack were plotted.
10. The method for identifying and calculating the longitudinal aerodynamic coefficient in the large angle-of-attack range in open-loop condition as described in claim 9, characterized in that, Based on the lift coefficient, drag coefficient, and pitching moment coefficient curves at different angles of attack, extract the lift coefficient, drag coefficient, and pitching moment coefficient. The extracted lift coefficient, drag coefficient, and pitch moment coefficient are input into the flight simulation model. Under the same flight speed / altitude conditions, the longitudinal response of the UAV is simulated by changing the angle of attack input, and the simulation output is extracted. Plot the simulation output and the corresponding experimental data on the same coordinate system to visually observe the degree of fit between the two, and determine the applicability of the flight simulation model based on the degree of fit.
11. The method for identifying and calculating the longitudinal aerodynamic coefficient in the large angle-of-attack range in open-loop condition as described in claim 6, characterized in that, When the model accuracy is insufficient in certain regions, the parameters of the genetic algorithm are further adjusted, and the minimum calculation of the sum of squared errors is performed again.