Flight parameter data driven reproduction method
By modifying the six-degree-of-freedom flight simulation program and data-driven methods, the problem of flight test result discrepancies caused by model differences in flight control law design was solved, enabling rapid and automatic data analysis and optimization design, thus improving safety and efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHENGDU AIRCRAFT DESIGN INST OF AVIATION IND CORP OF CHINA
- Filing Date
- 2025-12-27
- Publication Date
- 2026-05-05
AI Technical Summary
In the design of flight control laws, existing technologies suffer from discrepancies between aerodynamic models and actual aircraft, leading to differences between aircraft response and theoretical results during test flights. This affects flight safety and results in low data analysis efficiency.
By modifying the six-degree-of-freedom flight simulation program, adding a flight parameter input interface, and extracting characteristic motion segments from historical flight parameters, the program uses identification methods to match and import these segments into the simulation program for automatic driving simulation, enabling a comparison between actual test flight results and simulation results.
It improves data analysis efficiency, enabling rapid and automatic comparison of flight test results with simulation results, guiding the optimization design of flight control laws, and ensuring flight safety.
Smart Images
Figure CN121978985A_ABST
Abstract
Description
Technical Field
[0001] This application belongs to the field of flight control technology, and in particular relates to a flight parameter data-driven reproduction method. Background Technology
[0002] When designing flight control laws, it is necessary to first establish a six-degree-of-freedom flight simulation program, which includes flight control laws and aircraft body models (including actuator models, aerodynamic models, mass moment of inertia models, engine models, six-degree-of-freedom kinematic models, sensor models, etc.). Among them, the aerodynamic model is obtained through wind tunnel tests, and the engine model is obtained through bench tests. Due to the limitations of test conditions, the aerodynamic model usually has certain differences from the real aircraft, and the differences may be large under certain conditions. These differences will affect the design of flight control laws, causing the aircraft's response during actual test flights to differ from the theoretical results, which may seriously endanger flight safety.
[0003] When analyzing flight parameters for test flights, it is often necessary to simulate and reproduce the test flight actions using the same control inputs as the test flight mission. This allows for comparison of the actual test flight results with the simulation results, determining whether the aircraft's response during the test flight is consistent with the theoretical simulation. If there are discrepancies, further analysis can be conducted to identify the differences between the model parameters and the actual aircraft, guiding the optimization design of the flight control law. Summary of the Invention
[0004] The purpose of this invention is to provide a rapid flight parameter data reproduction method based on a six-degree-of-freedom flight simulation program. This method rapidly extracts data segments of interest from a complete flight parameter record using identification techniques and automatically imports them into the flight simulation program for simulation. The program uses the same control inputs or state information as the test flight mission to simulate and reproduce the test flight actions, comparing the differences between the actual test flight results and the simulation results. This determines whether the aircraft's response during the test flight is consistent with the theoretical simulation, achieving consistency between the ground and air, improving data analysis efficiency, and further identifying differences between model parameters and the actual aircraft through analysis, thus guiding the optimization design of flight control laws.
[0005] This application provides a flight parameter data-driven reproduction method, the method comprising: Modify the six-degree-of-freedom flight simulation program; Feature actions are manually extracted from historical flight data records, and representative data segments are extracted and saved as action 1 template, action 2 template, ... action n template; Select an action template, match the recorded flight parameter data with the action template using identification methods, then extract and segment the matching data segments into data segment 1, data segment 2, ... data segment n; Manually configure driver options SWGDAS1 and SWGDAS2 according to the driver intent; The modified six-degree-of-freedom flight simulation program is called repeatedly, and the simulation is driven according to the drive options SWGDAS1 and SWGDAS2. The simulation results are recorded in the output result file as output result 1, output result 2, ..., output result n. Compare data segment 1 and output result 1, data segment 2 and output result 2, ..., data segment n and output result n in sequence to create a graph, and then output it to a document.
[0006] Preferably, the modification of the six-degree-of-freedom flight simulation program includes: Branches are set at the control stick control commands, throttle stick control commands, sensor acquisition signals, and control surface positions, respectively, and flight parameter input interfaces are added to allow subsequent use of flight parameters recorded from test flights for driving.
[0007] Preferably, after manually extracting feature actions from historical flight data parameters and saving representative data segments as action 1 template, action 2 template, ... action n template, the method further includes: Save the extracted Action 1 template, Action 2 template, ... Action n template to the template library for easy reuse later.
[0008] Preferably, the selection of the action template involves matching the recorded flight parameter data with the action template using identification methods, including: Select an action template from the template library, and match the recorded flight parameter data with the action template using identification methods.
[0009] Preferably, the step of selecting an action template from the template library and matching the recorded flight parameter data with the action template using identification methods includes: Select an action template from the template library, and use interpretation rules, image recognition, and fast Fourier transform methods to obtain feature information based on the type of the identified object. Then match the recorded flight parameter data with the action template.
[0010] Preferably, the drive option SWGDAS1 has the following meaning: SWGDAS1=1 indicates that the drive is operated solely using pilot stick commands; SWGDAS1=2 indicates that the pilot's stick control commands and sensor-acquired signals are used simultaneously for driving; SWGDAS1=3 indicates that the drive is directly driven by the control surface position.
[0011] Preferably, the drive option SWGDAS2 has the following meaning: SWGDAS2=0 indicates that the pilot does not use stick control commands or control surface positions for propulsion. SWGDAS2=0 indicates that the pilot's throttle control commands are not used for propulsion; SWGDAS2=1 indicates that the drive is performed using the pilot's throttle control commands.
[0012] Preferably, data segment 1, data segment 2, ... data segment n are used to drive the modified simulation program.
[0013] The beneficial effects of this invention are: This invention presents a rapid, driven reproduction method for flight parameter data based on a six-degree-of-freedom flight simulation program. This method, on the one hand, rapidly extracts data segments of interest from a complete flight parameter recording using identification techniques; on the other hand, it automatically imports the extracted data segments into the flight simulation program for driven simulation. The simulation reproduces the flight test actions using the same control inputs or state information as the test flight mission, allowing for comparison of the differences between the actual test flight results and the simulation results. This method, based on automated batch control, improves data analysis efficiency. This invention has been successfully applied to the analysis of flight parameter data from multiple sorties of various aircraft types. Attached Figure Description
[0014] Figure 1 A schematic diagram of the modification of a six-degree-of-freedom flight simulation driver program; Figure 2 A screenshot of the template for identifying flight parameter data; Figure 3 This is a flowchart of the flight parameter data simulation driven process based on a six-degree-of-freedom flight simulation program. Figure 4 This is a schematic diagram of a six-degree-of-freedom flight simulation program; Figure 5 A schematic diagram of a six-degree-of-freedom flight simulation program with an added drive interface; Figure 6 A schematic diagram of the control stick operation command x; Figure 7 This is a schematic diagram showing the position of the control surfaces; Figure 8 This is a schematic diagram of the sensor acquiring the signal y. Figure 9 A schematic diagram of flight parameter data and identification results; Figure 10 This is a schematic diagram of the first group of pulse-driven results; Figure 11 This is a schematic diagram of the second group of pulse-driven results; Figure 12 This is a schematic diagram of the third group of pulse drive results. Detailed Implementation
[0015] The following is a description of the embodiments and appendices. Figure 1 - Appendix Figure 12The present invention will be described in further detail, but the embodiments of the present invention are not limited thereto.
[0016] This application provides a method for rapid flight parameter data reproduction, which quickly extracts data segments of interest from a complete flight parameter record using identification methods, automatically imports them into a flight simulation program for simulation, and then plots and exports the comparison curves to a document, achieving a consistent comparison between ground and air, and improving data analysis efficiency.
[0017] The technical solution of the present invention is as follows: Step 1: Modify the existing six-degree-of-freedom flight simulation program by setting branches at the control stick control commands, throttle stick control commands, sensor signal acquisition, and control surface positions, and adding flight parameter input interfaces for subsequent use of flight parameters recorded from test flights; The meanings of the driver option SWGDAS are as follows: SWGDAS=0 indicates that the pilot does not use stick control commands or control surface positions for propulsion. SWGDAS=1 indicates that the drive is operated solely using pilot stick commands; SWGDAS=2 indicates that the pilot's stick control commands and sensor-acquired signals are used simultaneously for driving; SWGDAS=3 indicates that the drive is directly driven by the control surface position.
[0018] Step 2: Template Extraction: Manually extract feature actions from historical flight data parameters, extract representative data segments and save them as Action 1 template, Action 2 template, ... Action n template. Save the extracted templates to the template library for easy reuse later. Step 3: Select an action template from the template library, and match the recorded flight parameter data with the action template through identification methods (such as interpretation rules, image recognition, fast Fourier transform, etc., which can be used to obtain feature information according to the type of identification object). Then, extract the matching data segments and store them as data segment 1, data segment 2, ... data segment n, which are used to drive the simulation program modified in Step 1. Step 4: Manually configure the driver options SWGDAS according to the driver intent; Step 5: Repeatedly call the modified six-degree-of-freedom flight simulation program, perform drive simulation according to the drive option SWGDAS, and record the simulation results into output result 1, output result 2, ..., output result n; Step 6: Compare data segment 1 and output result 1, data segment 2 and output result 2, ... data segment n and output result n in sequence to create a graph, and output it to a document.
[0019] This invention provides a method for rapid drive-driven reproduction of flight parameter data based on a six-degree-of-freedom flight simulation program. On the one hand, the method uses identification means to quickly extract data segments of interest from a complete flight parameter record; on the other hand, it automatically imports the extracted data segments into the flight simulation program for drive-driven simulation. The test flight actions are simulated and reproduced using the same control inputs or state information as the test flight mission, so as to compare the differences between the actual test flight results and the simulation results.
[0020] This method, based on automated batch processing, improves data analysis efficiency. This invention has been successfully applied to the analysis of flight parameter data from multiple sorties of various aircraft types.
[0021] In other embodiments of this application, the methods provided are as follows: Step 1: Create a tool using a third-party tool (such as MATLAB's Simulink toolbox) such as... Figure 4 The six-degree-of-freedom flight simulation program shown includes a control law model Gcl(s), an actuator model Gact(s), a six-degree-of-freedom motion model Gac(s), and a sensor model Gsens(s), where s is the Laplace operator; Sensor acquires and outputs signals: y=Gsens(s)*Gac(s)*Gact(s)*Gcl(s)*(xy) Where x represents the joystick control command.
[0022] In this example, each model is instantiated as follows:
[0023] Step 2: Add a flight test data driving interface to the six-degree-of-freedom flight simulation program created in Step 1. Figure 5 The green section includes pilot stick commands, sensor signals, and control surface positions. Drive signals are selected via the SWGDAS drive option. If SWGDAS=0, then the control stick control command x_mix=x, the sensor acquisition signal y_mix=y, and the control surface position de_mix=de; If SWGDAS=1, then the control stick control command x_mix=drv.x, the sensor acquisition signal y_mix=y, and the control surface position de_mix=de; If SWGDAS=2, then the control stick control command x_mix=drv.x, the sensor acquisition signal y_mix=drv.y, and the control surface position de_mix=de; If SWGDAS=3, then the control stick control command x_mix=x, the sensor acquisition signal y_mix=y, and the control surface position de_mix=drv.de.
[0024] Among them, drv.x is the joystick control command taken from the flight parameter data; drv.y is the sensor acquisition signal taken from the flight parameter data; drv.de contains the control surface positions taken from flight parameter data.
[0025] Figure 5 To enable the addition of a drive interface to the six-DOF flight simulation program, the drive data drv.x, drv.y, and drv.de are input through the From Workspace module, and the simulation results out.x, out.y, and out.de are output through the To Workspace module.
[0026] Step 3: Construct a set of historical flight parameters: Use the Signal Editor module of the Simulink toolbox to construct parameters as follows. Figure 6 The control stick operation command x shown contains three sets of pulse commands and one set of square wave commands. The specific values are shown in Table 1.
[0027] The first group of pulses has a maximum amplitude of 1 and a minimum amplitude of -1; the second group of pulses is a mirror image of the first group of pulses; the third group of pulses has a maximum amplitude of 1.2 and a minimum amplitude of -0.8; the square wave command has a maximum amplitude of 1 and a minimum amplitude of -1.
[0028] Figure 7 and Figure 8 To use the joystick control command x in Simulink to control the box Figure 4 To distinguish them from subsequent simulation results, white noise interference was added before the sensor model to simulate the disturbance of actual flight parameters, based on the simulated control surface position de and sensor acquisition signal y.
[0029] x, de, and y are all output to the variables fly.x, fly.de, and fly.y via Simulink's To Workspace module as historical flight parameters; Table 1 Control stick commands x
[0030] Step 4: Template Extraction: Manually extract feature actions from the historical flight parameters fly.x. Extract the first set of pulse data (data between 0 and 3 seconds) and save it as the pulse action template impulse. Extract the last set of square wave data (data between 11 and 14.5 seconds) and save it as the square wave action template step. Save the extracted templates to the template library file template.mat. The steps in Matlab are as follows: 4.1 Extracting pulse template data: index = fly.tout > 0 & fly.tout < 3; impulse.tout = fly.tout(index); impulse.data=fly.x.data(index); 4.2 Extracting Square Wave Template Data: index=fly.tout>11&fly.tout<13.5; step.tout = fly.tout(index); step.data = fly.x.data(index); 4.3 Save the data to the template library file template.mat: save('template.mat','impulse','step'); Step 5: Select an action template from the template library. Use identification methods (such as interpretation rules, image recognition, fast Fourier transform, etc., can be used to obtain feature information based on the type of object being identified; this example uses the findsignal function in MATLAB) to match the recorded flight parameter data fly.x with the action template impulse. Then, extract the matching data segments from fly and store them as flycut{1}, flycut{2}, ..., flycut{n}. Taking pulse recognition as an example, the findsignal function in MATLAB is used for action recognition: 5.1 Select the action template "impulse" from the template library; load('template.mat','impulse') 5.2 Match the recorded flight parameter data fly.x with the motion template impulse to obtain the start position, end position and error of the data; [istart1,istop1,dist1]=findsignal(fly.x.data,impulse.data,"MaxNumSegments",10) 5.3 Match the recorded flight parameter data fly.x with the mirror image of the motion template impulse to obtain the start position, end position and error of the data; [istart2,istop2,dist2]=findsignal(fly.x.data,-impulse.data,"MaxNumSegments",10) 5.4 Combine the results of the two identification steps in steps 5.2 and 5.3; istart = [istart1' , istart2' ]'; istop = [istop1', istop2']'; dist = [dist1', dist2']'; 5.5 Sort the data in ascending order of error to obtain the start position newistart, end position newistop, and error newdist. In this example, we take the data with an error dist less than 5. [dist_sort,index_sort] = sort(dist); istart_sort = istart(index_sort); istop_sort= istop(index_sort); index=dist_sort<5; newdist = dist_sort(index); newistart = istart_sort(index); newistop = istop_sort(index); 5.6 Extract the identification results and save the data segments sequentially to flycut{1}, flycut{2}, ..., flycut{n}; n = length(newdist); for i=1:n flycut{i}.x.time =fly.tout(newistart(i):newistop(i)); flycut{i}.x.signals.values=fly.x.data(newistart(i):newistop(i)); flycut{i}.x.signals.dimensions=1; flycut{i}.de.time =fly.tout(newistart(i):newistop(i)); flycut{i}.de.signals.values=fly.de.data(newistart(i):newistop(i)); flycut{i}.de.signals.dimensions=1; flycut{i}.y.time =fly.tout(newistart(i): newistop(i)); flycut{i}.y.signals.values=fly.y.data(newistart(i):newistop(i)); flycut{i}.y.signals.dimensions=1; end Step 6: Manually configure the drive option SWGDAS according to the drive intention. Taking the drive using only the pilot's stick control commands as an example, set SWGDAS=1. Step 7: Input the data saved in Step 5 through the From Workspace module, and run the six-DOF flight simulation program created in Step 2 with the added drive interface in a loop. Record the simulation results in simout{1}, simout{2}, ..., simout{n} files. Input the drive data drv.x, drv.y, and drv.de through the From Workspace module, and output the simulation results out.x, out.y, and out.de through the To Workspace module. 7.1 Get the number of loops nsim = length(flycut); 7.2 Loop through and acquire drive data, and run the flight simulation program to obtain simulation results. for i=1:nsim drv.x = flycut{i}.x; drv.de = flycut{i}.de; drv.y = flycut{i}.y; simout{i}=sim('model_drv',[flycut{i}.x.time(1) flycut{i}.x.time(end)]); end Note: model_drv is the name of the six-degree-of-freedom flight simulation program after the driver interface is added.
[0031] Step 8: Compare the data segments flycut{1}, flycut{2}, ..., flycut{n} with the output results simout{1}, simout{2}, ..., simout{n} in sequence to create a graph, and output it to a document.
[0032] 8.1 Plot the flight parameter data constructed in step 3 and the identification result data obtained in step 5 (see...) Figure 9 The plotting results are then output to a Word document via the MATLAB-Word interface. plot(fly.tout, fly.x.data) hold on for i=1:nsim plot(flycut{i}.x.time,flycut{i}.x.signals.values,'r'); end 8.2 Draw a comparison chart of the flycut data segment obtained through identification in step 5 and the simout simulation result obtained through driving simulation in step 7 (see...). Figures 10-12 The plotting results are then output to a Word document via the MATLAB-Word interface. for i=1:nsim figure subplot(3,1,1) plot(flycut{i}.x.time,flycut{i}.x.signals.values); hold on plot(simout{i}.tout,simout{i}.x.data,'r'); xlabel('time(s)') legend({'flight parameter data x','driving result x'}); subplot(3,1,2) plot(flycut{i}.de.time,flycut{i}.de.signals.values); hold on plot(simout{i}.tout,simout{i}.de.data,'r'); xlabel('time(s)') legend({'flight parameter data de','driving result de'}); subplot(3,1,3) plot(flycut{i}.y.time,flycut{i}.y.signals.values); hold on plot(simout{i}.tout,simout{i}.y.data,'r'); xlabel('time(s)') legend({'flight parameter data y','driving result y'}); end The above detailed embodiments are a description of the present invention. It should not be considered that the specific embodiments of the present invention are limited to these descriptions. For those skilled in the art, several simple deductions and substitutions can be made without departing from the concept of the present invention, and all of these should be considered to fall within the protection scope of the present invention.
Claims
1. A flight parameter data-driven reproduction method, characterized in that, The method includes: Modify the six-degree-of-freedom flight simulation program; Feature actions are manually extracted from historical flight data records, and representative data segments are extracted and saved as action 1 template, action 2 template, ... action n template; Select an action template, match the recorded flight parameter data with the action template using identification methods, then extract and segment the matching data segments into data segment 1, data segment 2, ... data segment n; Manually configure driver options SWGDAS1 and SWGDAS2 according to the driver intent; The modified six-degree-of-freedom flight simulation program is called repeatedly, and the simulation is driven according to the drive options SWGDAS1 and SWGDAS2. The simulation results are recorded in the output result file as output result 1, output result 2, ..., output result n. Compare data segment 1 and output result 1, data segment 2 and output result 2, ..., data segment n and output result n in sequence to create a graph, and then output it to a document.
2. The method according to claim 1, characterized in that, The modification of the six-degree-of-freedom flight simulation program includes: Branches are set at the control stick control commands, throttle stick control commands, sensor acquisition signals, and control surface positions, respectively, and flight parameter input interfaces are added to allow subsequent use of flight parameters recorded from test flights for driving.
3. The method according to claim 2, characterized in that, After manually extracting characteristic actions from historical flight data parameters and saving representative data segments as action 1 template, action 2 template, ... action n template, the process also includes: Save the extracted Action 1 template, Action 2 template, ... Action n template to the template library for easy reuse later.
4. The method according to claim 3, characterized in that, The selection of the action template involves matching the recorded flight parameter data with the action template using identification methods, including: Select an action template from the template library, and match the recorded flight parameter data with the action template using identification methods.
5. The method according to claim 4, characterized in that, The step of selecting an action template from the template library and matching the recorded flight parameter data with the action template using identification methods includes: Select an action template from the template library, and use interpretation rules, image recognition, and fast Fourier transform methods to obtain feature information based on the type of the identified object. Then match the recorded flight parameter data with the action template.
6. The method according to claim 5, characterized in that, The meaning of the drive option SWGDAS1 is as follows: SWGDAS1=1 indicates that the drive is operated solely using pilot stick commands; SWGDAS1=2 indicates that the pilot's stick control commands and sensor-acquired signals are used simultaneously for driving; SWGDAS1=3 indicates that the drive is directly driven by the control surface position.
7. The method according to claim 6, characterized in that, The meaning of the drive option SWGDAS2 is as follows: SWGDAS2=0 indicates that the pilot does not use stick control commands or control surface positions for propulsion. SWGDAS2=0 indicates that the pilot's throttle control commands are not used for propulsion; SWGDAS2=1 indicates that the drive is performed using the pilot's throttle control commands.
8. The method according to claim 1, characterized in that, Data segment 1, data segment 2, ... data segment n are used to drive the modified simulation program.