Method and system for constructing target test flight task-oriented simulation model
By using a task-driven approach, the model architecture and parameter set of the target flight test mission are determined, and sensitivity analysis and reinforcement learning optimization are performed. This solves the problem of insufficient accuracy of existing flight test simulation models in high-risk flight test missions, and realizes the construction of high-precision simulation models to support simulation verification of specific flight test subjects.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- COMMERCIAL AIRCRAFT CORP OF CHINA LTD
- Filing Date
- 2025-12-31
- Publication Date
- 2026-04-10
AI Technical Summary
Existing flight test simulation models suffer from insufficient specificity and limited simulation accuracy when facing high-risk, highly dynamic, and specific target flight test missions. In particular, they are unable to meet the accuracy requirements in high-risk subjects such as determining the minimum takeoff speed and the minimum ground control speed.
By adopting a "task-driven, precise optimization" approach, we determine the model architecture and parameter set of the target flight test mission, conduct sensitivity analysis to screen key parameters, and use reinforcement learning algorithms to optimize the initial simulation model to build a high-precision simulation model.
A high-precision simulation model for specific target flight test missions was constructed, which can provide reliable simulation data support for flight test plan formulation, risk prediction and result analysis, and improve the pertinence and optimization efficiency of the simulation model.
Smart Images
Figure CN121835408A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of flight test engineering, and in particular to a method and system for constructing a simulation model for a target-oriented flight test task. BACKGROUND
[0002] With the deepening of the digital transformation of the aviation industry, the industry has clearly proposed the strategic requirement of building a model-based digital flight test engineering and realizing model-based air-ground integrated simulation verification. A high-fidelity flight test simulation model is the core foundation and key prerequisite for achieving the above goals. During the flight test, a large number of flight test subjects are usually pre-rehearsed and risk assessed by using flight test simulators and other devices to provide important reference for actual flight test, and the flight test simulation model plays a crucial role. However, the existing flight test simulation model is mainly suitable for ordinary flight test subjects and is usually designed for universality. When facing high-risk and high-dynamic specific target flight test tasks, there are often problems of insufficient pertinence and limited simulation accuracy.
[0003] Therefore, there is an urgent need to construct a method of flight test simulation model that can accurately simulate specific target flight test tasks. SUMMARY
[0004] The following presents a simplified summary of one or more aspects in order to provide a basic understanding of such aspects. This summary is not an extensive overview of all contemplated aspects, and is intended to neither identify key or critical elements of all aspects nor delineate the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed description that is presented later.
[0005] In order to solve one or more of the above existing problems, the present disclosure proposes a method for constructing a simulation model for a target-oriented flight test task. This method changes the inefficient mode of "blanket optimization" of all model parameters in the traditional way, and adopts an efficient mode of "task-driven and precise optimization". The simulation model constructed by this method also has high confidence for specific target flight test tasks, and can provide reliable simulation data support for flight test scheme development, risk prediction and result analysis.
[0006] One aspect of the present disclosure provides a method for constructing a simulation model for a target flight test mission. The method can include determining a model architecture and a parameter set required for constructing the simulation model for the target flight test mission based on the target flight test mission; constructing an initial simulation model for the target flight test mission based on the model architecture and the parameter set, wherein the initial simulation model for the target flight test mission includes a regular sub-model and a key sub-model; performing a sensitivity analysis on parameters in the parameter set associated with the key sub-model to filter out one or more sensitive parameters that have an impact on the accuracy of the simulation model for the target flight test mission exceeding a pre-set threshold; and optimizing the initial simulation model for the target flight test mission based on the one or more sensitive parameters using a parameter optimization algorithm to obtain an optimized simulation model for the target flight test mission.
[0007] In an example, the method can further include verifying the optimized simulation model for the target flight test mission to confirm that the optimized simulation model for the target flight test mission meets a pre-set accuracy requirement.
[0008] In an example, determining the model architecture and the parameter set based on the target flight test mission can further include establishing a target flight test scenario for the target flight test mission, wherein establishing the target flight test scenario includes specifying a flight test object and a flight test strategy, conducting a criticality analysis, forming a flight test point matrix, developing a flight test plan, and confirming a flight test risk; and determining the model architecture and the parameter set related to the target flight test mission based on the target flight test scenario, wherein the key sub-model is determined based on the target flight test mission.
[0009] In an example, the parameter optimization algorithm can include a reinforcement learning algorithm. In some cases, optimizing the initial simulation model for the target flight test mission using the reinforcement learning algorithm can further include setting up an interactive environment for an agent in the reinforcement learning algorithm. In some specific cases, setting up the interactive environment for the agent in the reinforcement learning algorithm can include modifying variables required by the target flight test mission, determining an adjustment action for the one or more sensitive parameters; and determining a reward function based on a plurality of expected objectives to be met by the initial simulation model for the target flight test mission; setting hyperparameters; and optimizing the initial simulation model for the target flight test mission using the reinforcement learning algorithm based on the reward function and the hyperparameters.
[0010] In an example, the parameters in the parameter set associated with the key sub-model can be pre-filtered or prioritized before performing the sensitivity analysis on the parameters.
[0011] In an example, the target flight test mission can be a flight test mission related to ground motion.
[0012] In an example, the conventional sub-models included in the initial simulation model for the targeted flight test mission can include at least one of the following: a body aerodynamics sub-model, a dynamics sub-model, a power system model, an engine model, and / or a power system sub-model, etc. In some cases, the body aerodynamics sub-model, the dynamics sub-model, the engine model, and the power system sub-model employ validated models.
[0013] In an example, the focus sub-models included in the initial simulation model for the targeted flight test mission can include at least one of the following: a landing gear sub-model, and / or a tail skid sub-model, etc.
[0014] In an example, the one or more sensitivity parameters can include at least one of the following: a tire brake time friction coefficient, a tire lateral friction coefficient, a tire rolling friction coefficient, a bumper stiffness coefficient, a bumper damping characteristic, and / or an anti-skid brake model PID parameter, etc.
[0015] In an example, performing the sensitivity analysis on the parameters in the parameter set associated with the focus sub-models can further include: determining an observation; obtaining a check value of the parameters in the parameter set associated with the focus sub-models on the observation; and performing a local sensitivity analysis on each of the parameters based on the observation to determine a parameter with the highest sensitivity as the first sensitivity parameter.
[0016] In an example, the method can further include: based on the determination of the first sensitivity parameter, performing a local sensitivity analysis on each of the other parameters in the parameter set based on the observation again to determine a parameter with the highest sensitivity as a second sensitivity parameter, wherein the local sensitivity analysis employs a regression-based local sensitivity analysis method.
[0017] In an example, using the parameter optimization algorithm to optimize the initial simulation model for the targeted flight test mission based on the one or more sensitivity parameters can further include: using the parameter optimization algorithm to optimize the initial simulation model for the targeted flight test mission based on the first sensitivity parameter and the second sensitivity parameter.
[0018] One aspect of the present disclosure provides a system for constructing a simulation model for a target flight test mission. The system can include a determination module configured to determine a model architecture employed for constructing the simulation model for the target flight test mission and a parameter set required for constructing the simulation model for the target flight test mission based on the target flight test mission; a construction module configured to construct an initial simulation model for the target flight test mission based on the model architecture and the parameter set, wherein the initial simulation model for the target flight test mission includes a regular sub-model and a key sub-model, and wherein the key sub-model is determined based on the target flight test mission; an analysis module configured to perform a sensitivity analysis on parameters in the parameter set associated with the key sub-model to filter out one or more sensitive parameters that have an impact on accuracy of the simulation model for the target flight test mission exceeding a pre-set threshold; and an optimization module configured to optimize the initial simulation model for the target flight test mission based on the one or more sensitive parameters using a parameter optimization algorithm to obtain an optimized simulation model for the target flight test mission.
[0019] One aspect of the present disclosure provides a non-transitory computer-readable storage medium having instructions stored thereon. The instructions can be executed by a processor to perform the method as described above.
[0020] The simulation model construction method for a target flight test mission proposed by the present disclosure determines a model architecture and a parameter set based on a specific target flight test mission, ensuring that the model is constructed with pertinence and purpose. The simulation model construction method for a target flight test mission also realizes rational allocation of modeling resources by distinguishing between a “regular sub-model” and a “key sub-model”, avoiding inefficient modification of non-target model parts. The simulation model construction method for a target flight test mission then introduces sensitivity analysis to identify “sensitive parameters” that have a relatively large impact on simulation accuracy for the current target flight test mission from the parameter set, thereby focusing optimization computing resources on key model parts. The simulation model construction method for a target flight test mission finally performs directional optimization for the filtered key parameters, significantly improving optimization efficiency and convergence speed.
[0021] The present disclosure is provided to introduce some concepts in a simplified form, which will be further described in the following detailed description. The present disclosure does not intend to identify key or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter. Other aspects, features, and / or advantages of the embodiments will be apparent from the description contained herein and from the claims. BRIEF DESCRIPTION OF DRAWINGS
[0022] For a more complete understanding of the foregoing features of the present disclosure, reference is made to the detailed description taken in connection with the accompanying drawings in which various aspects are represented by various figures. It is to be noted, however, that the figures are presented for the purpose of illustration and description and are not intended to limit the scope of the disclosure in any way. In the figures: Figure 1 A flow chart of a method for constructing a simulation model for a target-oriented flight test mission is shown according to an embodiment of the present disclosure.
[0023] Figure 2 A curve showing the variation of the take-off roll distance with the rolling friction coefficient pull-off according to an embodiment of the present disclosure is shown.
[0024] Figure 3 A curve showing the variation of the front landing gear tire compression with the rolling friction coefficient pull-off according to an embodiment of the present disclosure is shown.
[0025] Figure 4 A curve showing the variation of the left side main landing gear tire compression with the rolling friction coefficient pull-off according to an embodiment of the present disclosure is shown.
[0026] Figure 5 A curve showing the variation of the right side main landing gear tire compression with the rolling friction coefficient pull-off according to an embodiment of the present disclosure is shown.
[0027] Figure 6 A curve showing the variation of the front landing gear strut compression with the rolling friction coefficient pull-off according to an embodiment of the present disclosure is shown.
[0028] Figure 7 A curve showing the variation of the left side main landing gear strut compression with the rolling friction coefficient pull-off according to an embodiment of the present disclosure is shown.
[0029] Figure 8 A curve showing the variation of the right side main landing gear strut compression with the rolling friction coefficient pull-off according to an embodiment of the present disclosure is shown.
[0030] Figure 9 A curve showing the variation of the take-off roll distance with the tire lateral friction coefficient pull-off according to an embodiment of the present disclosure is shown.
[0031] Figure 10 A curve showing the variation of the front landing gear tire compression with the tire lateral friction coefficient pull-off according to an embodiment of the present disclosure is shown.
[0032] Figure 11 A curve showing the variation of the main landing gear tire compression with the tire lateral friction coefficient pull-off according to an embodiment of the present disclosure is shown.
[0033] Figure 12A plot of front landing gear strut compression versus tire lateral friction coefficient pull-off is shown according to an embodiment of the disclosure.
[0034] Figure 13 A plot of main landing gear strut compression versus tire lateral friction coefficient pull-off is shown according to an embodiment of the disclosure.
[0035] Figure 14 A plot of front landing gear shock absorber compression stiffness data is shown according to an embodiment of the disclosure.
[0036] Figure 15 A plot of takeoff distance versus strut compression stiffness pull-off is shown according to an embodiment of the disclosure.
[0037] Figure 16 A plot of front landing gear tire compression versus strut compression stiffness pull-off is shown according to an embodiment of the disclosure.
[0038] Figure 17 A plot of main landing gear tire compression versus strut compression stiffness pull-off is shown according to an embodiment of the disclosure.
[0039] Figure 18 A plot of front landing gear strut compression versus strut compression stiffness pull-off is shown according to an embodiment of the disclosure.
[0040] Figure 19 A plot of left main landing gear strut compression versus strut compression stiffness pull-off is shown according to an embodiment of the disclosure.
[0041] Figure 20 A plot of right main landing gear strut compression versus strut compression stiffness pull-off is shown according to an embodiment of the disclosure.
[0042] Figures 21-22 A plot of front landing gear shock absorber damping characteristics - positive travel and reverse travel is shown according to an embodiment of the disclosure.
[0043] Figure 23 A plot of takeoff distance versus strut damping coefficient pull-off is shown according to an embodiment of the disclosure.
[0044] Figure 24 A plot of front landing gear tire compression versus strut damping coefficient pull-off is shown according to an embodiment of the disclosure.
[0045] Figure 25 A plot of main landing gear tire compression versus strut damping coefficient pull-off is shown according to an embodiment of the disclosure.
[0046] Figure 26A plot showing the change in front landing gear strut compression with strut damping coefficient pull-off according to an embodiment of the present disclosure is shown.
[0047] Figure 27 A plot showing the change in main landing gear strut compression with strut damping coefficient pull-off according to an embodiment of the present disclosure is shown.
[0048] Figure 28 A schematic diagram of an agent action according to an embodiment of the present disclosure is shown.
[0049] Figure 29 A state comparison schematic diagram before and after optimizing a model using a reinforcement learning correction algorithm according to an embodiment of the present disclosure is shown.
[0050] Figure 30 A block diagram of a system for constructing a simulation model for a target-oriented flight test task according to an embodiment of the present disclosure is shown.
[0051] Figure 31 A block diagram of a device including a system for constructing a simulation model for a target-oriented flight test task according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION
[0052] The detailed description set forth below is intended as a description of various configurations and is not intended to represent the only configurations in which the concepts described herein can be practiced. The detailed description includes specific details for the purpose of providing a thorough understanding of the various concepts. However, it will be apparent to those skilled in the art that these concepts can be practiced without these specific details.
[0053] It will be appreciated by persons of the art that the scope of the present disclosure is intended to cover all aspects of the present disclosure, whether implemented independently of, or in combination with, any other aspect of the present disclosure. For example, aspects of the present disclosure can be used in combination with any other aspect of the present disclosure. The scope of the present disclosure is intended to cover any and all such aspects. Additionally, the scope of the present disclosure is intended to cover any and all such aspects, whether implemented independently of, or in combination with, any other aspect of the present disclosure. For example, the scope of the present disclosure includes any and all aspects of the disclosed aspects, alone or in combination.
[0054] While specific aspects are described in this specification, the aspects are not intended to be limiting, and changes can be made to the aspects described without departing from the scope of the present disclosure. While the aspects of the present disclosure have been described with reference to one or more particular embodiments, those skilled in the art will understand that various other modifications can be made thereto and will appreciate that the various configurations described herein are not the only ones that satisfy the requirements, purposes, constraints and / or objectives of the present disclosure. Changes can be made to the disclosure in light of the above and the disclosure is further unlimited by the disclosure recited here, but can be practiced or carried out in various ways. The disclosure is not limited to the details shown and described herein but can be implemented with variations made to the details without departing from the scope of the disclosure.
[0055] During the flight test period, a large number of flight test subjects are usually verified by using devices such as flight test simulators to provide important references for actual flight tests, and a flight test simulation model plays a crucial role. However, the existing flight test simulation model is mainly applicable to general flight test subjects (for example, air test subjects), and is usually designed for universality. The optimization goal of such a method is to improve the average simulation accuracy and comprehensive confidence of the model under a wide range of working conditions, so that the corrected model can better meet the simulation verification needs of conventional flight test subjects. When facing specific target flight test tasks (for example, flight test tasks related to the ground) with high risks and high dynamics, there are often problems of insufficient pertinence and limited simulation accuracy.
[0056] For example, the current landing gear model mainly performs parameter correction for conventional flight test subjects (such as normal takeoff, taxiing, landing, etc.) to improve accuracy, but the corrected model is applicable to the simulation verification of conventional subjects and cannot meet the simulation verification accuracy of specific flight test subjects (for example, high-risk subjects such as determining the minimum unstick speed (VMU) and the minimum ground control speed (VMCG)). For example, in the flight test subject for determining the VMU, it is required to dynamically adjust the control according to the test point state so that the aircraft can reach and maintain the maximum unstick pitch attitude at a predetermined speed. This puts higher requirements on the response accuracy of the simulation model in the longitudinal dynamics (such as pitch angle, pitch angle rate, etc.). That is, the simulation based on the existing universal model is difficult to achieve the accuracy level required by such high-risk flight test subjects in the transient response and steady-state fidelity of the longitudinal dynamics of the aircraft (such as pitch angle, pitch angle rate, pitch moment, normal overload, etc.).
[0057] To solve one or more of the above existing problems, the present disclosure proposes a method for constructing a simulation model for a target flight test task, which aims to construct a corresponding dynamic simulation model for a specific target flight test subject (such as determining the VMU, VMCG, etc.) and to make corresponding directional correction to the simulation model. The method of the present disclosure breaks through the bottleneck of the existing “universal” modeling method, can construct a high-precision simulation model and realize efficient directional correction of the simulation model, so that the obtained simulation model can directly serve specific high-risk flight test tasks to support pre-flight verification and in-flight monitoring of specific flight test subjects.
[0058] It is worth noting that the embodiments of the present disclosure are described below by taking an aircraft as an example, but those skilled in the art will appreciate that the embodiments according to the present disclosure can also be applied to other types of aircraft without departing from the scope of the present disclosure. The present disclosure is further described below in conjunction with specific embodiments and drawings, but should not be limited in the protection scope of the present disclosure.
[0059] Figure 1A flowchart illustrating a method 100 for constructing a simulation model for a target-oriented flight test task according to an embodiment of the present disclosure is shown. In one example, the method 100 for constructing a simulation model for a target-oriented flight test task can be performed by the system 200 as shown in FIG. 1. In this embodiment, the method 100 for constructing a simulation model for a target-oriented flight test task can include: Figure 2 At 102, a model architecture employed for constructing a simulation model for a target-oriented flight test task and a parameter set required for constructing the simulation model for the target-oriented flight test task can be determined based on the target-oriented flight test task. In one example, the target-oriented flight test task can be a flight test task related to ground motion. In some cases, the flight test task related to ground motion can include, but is not limited to, a flight test task for determining a VMU, a flight test task for determining a VMCG, and the like.
[0060] At 104, an initial simulation model for the target-oriented flight test task can be constructed based on the model architecture and the parameter set, wherein the initial simulation model for the target-oriented flight test task includes a regular sub-model and a key sub-model.
[0061] In one example, determining the model architecture and the parameter set based on the target-oriented flight test task can further include establishing a target flight test scenario of the target-oriented flight test task. In some cases, establishing the target flight test scenario includes clarifying a flight test object and a flight test strategy, conducting a criticality analysis, forming a flight test point matrix, making a flight test plan, and confirming a flight test risk, and determining the model architecture and the parameter set related to the target-oriented flight test task based on the target flight test scenario, wherein the key sub-model is determined based on the target-oriented flight test task. That is, establishing the target flight test scenario can clarify a modeling idea of the simulation model and required parameters.
[0062] In one optional example, the regular sub-model included in the initial simulation model for the target-oriented flight test task can include at least one of a body aerodynamic sub-model, a dynamics sub-model, an engine model, and / or a power system sub-model, and the like. In some cases, the body aerodynamic sub-model, the dynamics sub-model, the engine model, and the power system sub-model employ a validated model.
[0063] In one optional example, the key sub-model included in the initial simulation model for the target-oriented flight test task can include at least one of a landing gear sub-model, and / or a tail skid sub-model, and the like.
[0064] At 106, a sensitivity analysis can be performed on parameters in the parameter set associated with the key sub-model to screen out one or more sensitive parameters that have an impact on the accuracy of the simulation model for the target-oriented flight test task exceeding a preset threshold.
[0065] In an optional example, the one or more sensitivity parameters can include at least one of: a tire braking friction coefficient, a tire lateral friction coefficient, a tire rolling friction coefficient, a shock absorber stiffness coefficient, a shock absorber damping characteristic, and / or an anti-skid braking model PID parameter, etc.
[0066] In an optional example, the parameters associated with the focus sub-model in the parameter set can be pre-screened or prioritized prior to performing the sensitivity analysis on the parameters.
[0067] In an optional example, performing the sensitivity analysis on the parameters associated with the focus sub-model in the parameter set can further include: determining an observation; obtaining a test value of the parameters associated with the focus sub-model in the parameter set on the observation; and performing a local sensitivity analysis on each of the parameters based on the observation to determine a parameter with a highest sensitivity as the first sensitivity parameter.
[0068] Generally, different aircraft parameters can be selected as the observation and the state variable according to the characteristics of the target flight test mission. For example, a flight test mission for determining the VMU generally focuses on the longitudinal response of the aircraft. Thus, the height, speed, pitch angle, pitch angle velocity, heading displacement, etc. of the aircraft can be selected as the observation, and the landing gear longitudinal related parameters (such as the landing gear shock strut stiffness, damping coefficient, longitudinal static and dynamic friction coefficients, etc.) can be pre-screened for the sensitivity analysis. For another example, a flight test mission for determining the VMCG generally focuses on the lateral response of the aircraft. Thus, the height, speed, pitch angle, yaw angle, yaw angle velocity, roll angle, lateral displacement, etc. of the aircraft can be selected as the observation, and the landing gear lateral related parameters (such as the landing gear shock strut stiffness, damping coefficient, lateral static and dynamic friction coefficients, etc.) can be pre-screened for the sensitivity analysis.
[0069] In an optional example, the method can further include: based on the determination of the first sensitivity parameter, performing a local sensitivity analysis on each of the other parameters in the parameter set based on the observation again to determine a parameter with a highest sensitivity as a second sensitivity parameter, wherein the local sensitivity analysis employs a regression-based local sensitivity analysis method.
[0070] In an optional example, using the parameter optimization algorithm to optimize the initial simulation model for the target flight test mission based on the one or more sensitivity parameters can further include: using the parameter optimization algorithm to optimize the initial simulation model for the target flight test mission based on the first sensitivity parameter and the second sensitivity parameter.
[0071] At 108, the initial simulation model for the target flight test task can be optimized to obtain an optimized simulation model for the target flight test task based on the one or more sensitivity parameters using a parameter optimization algorithm. In one example, the parameter optimization algorithm can include a reinforcement learning algorithm, and optimizing the initial simulation model for the target flight test task using the reinforcement learning algorithm can further include: building an interactive environment for an agent in the reinforcement learning algorithm, including: modifying variables of the target flight test task requirements, determining adjustment actions for the one or more sensitivity parameters; and determining a reward function based on a plurality of expected targets to be met by the initial simulation model for the target flight test task; setting hyperparameters; and optimizing the initial simulation model for the target flight test task using the reinforcement learning algorithm based on the reward function and the hyperparameters.
[0072] In one embodiment, optionally, the method can further include: verifying the optimized simulation model for the target flight test task to confirm that the optimized simulation model for the target flight test task meets a preset accuracy requirement.
[0073] The present disclosure proposes a method for constructing a simulation model for a target flight test task. The method changes the inefficient mode of traditionally performing full parameter traversal optimization on all model parameters, and instead adopts an efficient mode of "task-driven, accurate optimization". The simulation model constructed via the method also has high confidence for a specific target flight test task, and can provide reliable simulation data support for flight test scheme formulation, risk prediction, and result analysis.
[0074] The above method will be described below in conjunction with specific embodiments, but should not limit the protection scope of the present disclosure.
[0075] Example embodiments
[0076] When the target flight test task is a flight test task for determining a VMU, the model architecture adopted to construct the simulation model for the flight test task for determining the VMU and the required parameter set can be determined.
[0077] Subsequently, an initial simulation model for determining the flight test mission of the VMU can be constructed based on the model architecture and the parameter set, where the initial simulation model for determining the flight test mission of the VMU can include regular sub-models and key sub-models. In one example, determining the model architecture and the parameter set based on the target flight test mission being the flight test mission for determining the VMU can further include establishing a flight test scenario for determining the VMU. In some cases, establishing the flight test scenario for determining the VMU can include specifying the flight test object and the flight test strategy, conducting a criticality analysis, forming a flight test point matrix, developing a flight test plan, and confirming the flight test risk. In one example, the model architecture to be employed for constructing the initial simulation model and the required parameter set can be determined based on the established flight test scenario. Since the flight test scenario for determining the VMU is a scenario oriented to the ground flight test of the aircraft, the focus of modeling and correction is on the force and response between the aircraft and the ground.
[0078] In one example, the regular sub-models included in the initial simulation model can include a body aerodynamic sub-model (e.g., an aircraft aerodynamic sub-model), a dynamics sub-model, a power system model, and / or an engine sub-model, etc. In some cases, the regular sub-models can employ mature models that have been verified.
[0079] In one example, the key sub-models included in the initial simulation model can be determined as sub-models related to the ground. For example, the key sub-models can include a landing gear sub-model, and / or a tail skid sub-model, etc. In some cases, the key sub-models can be modeled individually according to the parameters determined by the flight test scenario.
[0080] In one example, the parameter set required for constructing the simulation model for determining the flight test mission of the VMU can include a tire friction coefficient, a tail skid system friction coefficient, an aircraft pitch angle, landing gear shock strut related parameters (e.g., compression curve, damping curve), etc.
[0081] Since there are numerous modeling parameters involved in the landing gear sub-model, and / or the tail skid sub-model, etc., and not all parameters have a significant impact on the accuracy of the model. If all parameters are called to correct the model, the workload is large, the efficiency is low, and the effect cannot be significantly improved. Therefore, a sensitivity analysis is performed before calling the parameters to correct the model, and parameters that have a greater impact on the accuracy of the model are screened out for model correction. Therefore, the parameters associated with the key sub-models in the parameter set can be subjected to a sensitivity analysis to screen out one or more sensitive parameters that have an impact on the accuracy of the simulation model for determining the flight test mission of the VMU exceeding a preset threshold.
[0082] In one example, the parameters associated with the focus sub-model in the parameter set can be pre-screened or prioritized before the sensitivity analysis. For example, the parameters associated with the landing gear sub-model, the ski sub-model, and / or the like can be pre-screened or prioritized before the sensitivity analysis. In some cases, the pre-screening or prioritization can be based at least in part on the target flight test mission or based on modular analysis. Subsequently, the pre-screened parameters or the prioritized parameters can be subjected to the sensitivity analysis. As one example, since the flight test scenario for determining the VMU is a scenario for the airplane ground flight test, the working condition of the landing gear of the airplane is complex, the priority of the parameters associated with the landing gear sub-model can be determined as the highest, and the sensitivity analysis of the parameters associated with the landing gear sub-model is prioritized. In some cases, the landing gear can include a bumper strut, a wheel, a tire, a brake system, and / or the like. As one example, the following key parameters associated with the landing gear are pre-screened for the sensitivity analysis: tire brake friction coefficient, tire lateral friction coefficient, tire rolling friction coefficient, bumper stiffness coefficient, bumper damping characteristics (e.g., positive and negative stroke damping curves), and / or anti-skid brake model PID parameters, and / or the like.
[0083] In one example, the sensitivity analysis of the parameters associated with the focus sub-model in the parameter set can further include determining an observation, obtaining a test value of the parameters associated with the focus sub-model in the parameter set on the observation, and performing a local sensitivity analysis on each of the parameters based on the observation to determine a parameter with the highest sensitivity as a first sensitive parameter. For example, the observation can be determined as a takeoff distance, a landing distance, a landing gear bumper strut compression amount, and / or the like. Through the one-dimensional and two-dimensional regression, the test values (e.g., partial F test values) of the above-mentioned key parameters on the observations can be obtained, and the sensitivity of each observation on different parameters (e.g., the sensitivity of the F value, the greater the F value, the more sensitive) can be obtained.
[0084] In one optional example, the method can further include, based on the determination of the first sensitive parameter, performing a local sensitivity analysis on other parameters in the parameter set based on the observation again to determine a parameter with the highest sensitivity as a second sensitive parameter, wherein the local sensitivity analysis employs a regression-based local sensitivity analysis method.
[0085] Based on the takeoff process, one-dimensional sensitivity analysis can be performed on the pre-screened parameters. For example, one of the parameters can be pulled by ±20%, the changes of each simulation variable are observed, and a regression analysis and a partial F test of the parameter are performed on a number of observations to test whether the parameter has a significant impact on the model. The greater the F value, the more significant the impact.
[0086] Since the brake is not used in the normal takeoff process, the sensitivity of the brake friction coefficient is not analyzed.
[0087] Firstly, the rolling friction coefficient is pulled off, and the main wheel pull-off value can be: From left to right, in turn: pull off -20%, -15%, -10%, -5%, original data, +5%, +10%, +15%, +20%. Through the cycle simulation, the change of each simulation variable with the parameter pull-off is obtained. Figure 2 The take-off distance is shown with the change curve of the rolling friction coefficient pull-off. Figure 3 The front landing gear tire compression amount is shown with the change curve of the rolling friction coefficient pull-off. Figure 4 The left main landing gear tire compression amount is shown with the change curve of the rolling friction coefficient pull-off. Figure 5 The right main landing gear tire compression amount is shown with the change curve of the rolling friction coefficient pull-off. Figure 6 The front landing gear strut compression amount is shown with the change curve of the rolling friction coefficient pull-off. Figure 7 The left main landing gear strut compression amount is shown with the change curve of the rolling friction coefficient pull-off. Figure 8 The right main landing gear strut compression amount is shown with the change curve of the rolling friction coefficient pull-off. It can be seen that the take-off distance is very sensitive to the tire rolling friction coefficient, and increases with the increase of the rolling friction coefficient, while the landing gear related variables are not sensitive to the rolling friction coefficient.
[0088] Subsequently, the tire lateral friction coefficient is pulled off, and in turn, -20%, -15%, -10%, -5%, 0%, +5%, +10%, +15%, +20% are pulled off. Through the cycle simulation, the change of each simulation variable with the parameter pull-off is obtained. Figure 9 The take-off distance is shown with the change curve of the tire lateral friction coefficient pull-off. Since there is no lateral movement in the normal take-off process under no wind condition, the take-off distance does not change with the change of the tire lateral friction coefficient. Figure 10 The front landing gear tire compression amount is shown with the change curve of the tire lateral friction coefficient pull-off. Figure 11 The main landing gear tire compression amount is shown with the change curve of the tire lateral friction coefficient pull-off. Figure 12 The front landing gear strut compression amount is shown with the change curve of the tire lateral friction coefficient pull-off. Figure 13 The main landing gear strut compression amount is shown with the change curve of the tire lateral friction coefficient pull-off.
[0089] For example, in the case of temperature 25℃, the buffer compression stiffness (i.e. compression curve two-dimensional list) is pulled off, and the buffer compression stiffness is as follows. Figure 14A front landing gear shock absorber compression stiffness data curve is shown. The curve is pulled in increments of -20%, -15%, -10%, -5%, 0%, +5%, +10%, +15%, +20% overall. Through a cycle simulation, the changes in each simulation variable with parameter pull are obtained, as follows.
[0090] Figure 15 A takeoff distance versus strut compression stiffness pull curve is shown. As shown, strut compression stiffness has an effect on takeoff distance, which is not significant, and takeoff distance decreases as strut compression stiffness increases. Figure 16 A front landing gear tire compression versus strut compression stiffness pull curve is shown. Figure 17 A main landing gear tire compression versus strut compression stiffness pull curve is shown. Figure 18 A front landing gear strut compression versus strut compression stiffness pull curve is shown. Figure 19 A left main landing gear strut compression versus strut compression stiffness pull curve is shown. Figure 20 A right main landing gear strut compression versus strut compression stiffness pull curve is shown. As shown, the shock strut compression is highly sensitive to compression stiffness, and other variables are less sensitive to compression stiffness.
[0091] The shock absorber damping characteristics (i.e., a two-dimensional list of positive and negative travel damping coefficients) are pulled in increments of -20%, -15%, -10%, -5%, 0%, +5%, +10%, +15%, +20% overall. Through a cycle simulation, the changes in each simulation variable with parameter pull are obtained, as follows. Figure 21 A front landing gear shock absorber damping characteristics (positive travel) is shown, and Figure 22 A front landing gear shock absorber damping characteristics (negative travel) is shown. The curve is pulled in increments of -20%, -15%, -10%, -5%, 0%, +5%, +10%, +15%, +20% overall. Through a cycle simulation, the changes in each simulation variable with parameter pull are obtained, as follows.
[0092] Figure 23 A takeoff distance versus strut damping coefficient pull curve is shown. As shown, shock strut damping coefficient has an effect on takeoff distance, which is not significant, and takeoff distance increases as strut compression stiffness increases. Figure 24 A front landing gear tire compression versus strut damping coefficient pull curve is shown. Figure 25 A main landing gear tire compression versus strut damping coefficient pull curve is shown. Figure 26 A front landing gear strut compression versus strut damping coefficient pull curve is shown. Figure 27 A main landing gear strut compression versus strut damping coefficient pull curve is shown. When shock strut damping coefficient changes, it should affect the slope of the compression amount decrease process, but in Figure 27The change curve shown can be seen that the change is not obvious, considering that the damping coefficient is too large, the 20% pull bias has little effect. Subsequently, F test quantification analysis is performed to observe the correlation between the strut compression amount and the strut damping coefficient.
[0093] The takeoff distance is taken as the observation quantity, and the takeoff distance is the absolute value of the x-axis position of the aircraft at the start of the simulation minus the x-axis position when the flight height reaches 15 m above the airport: .
[0094] Then select the difference between the buffer compression amount and the test flight data, take the absolute value of each sampling step, and then take the average value. Take the average value as the observation quantity, denoted as .
[0095] Based on the above observation quantity, the partial F test is performed on each parameter, and the F value list is shown as follows.
[0096] Table 1 Partial F test table
[0097] As shown in Table 1 above, the takeoff distance is most sensitive to the tire rolling friction coefficient, and the buffer strut compression amount is most sensitive to the compression stiffness, and these two parameters can be considered to be added to the model correction.
[0098] As an example, in this embodiment, a one-parameter sensitivity analysis and a two-parameter sensitivity analysis are disclosed. Those skilled in the art will appreciate that the sensitivity analysis can be extended to three or more parameters without departing from the scope of the present disclosure.
[0099] In one example, optimizing the initial simulation model using a parameter optimization algorithm based on one or more sensitivity parameters can further include optimizing the initial simulation model using a parameter optimization algorithm based on a first sensitivity parameter and a second sensitivity parameter. For example, after determining the tire rolling friction coefficient and the buffer compression stiffness as the sensitivity parameters, the initial simulation model can be optimized using a parameter optimization algorithm based on the tire rolling friction coefficient and the buffer compression stiffness.
[0100] In one example, the parameter optimization algorithm can include a reinforcement learning algorithm, and using the reinforcement learning algorithm to optimize the initial simulation model for the target flight test task can further include: building an interaction environment for an agent in the reinforcement learning algorithm, including: correcting variables required by the target flight test task, determining adjustment actions for one or more sensitivity parameters; and determining a reward function based on a plurality of expected targets to be met by the initial simulation model for the target flight test task; setting hyperparameters; and using the reinforcement learning algorithm to optimize the initial simulation model for the target flight test task based on the reward function and the hyperparameters. For example, according to the principle of aircraft motion response and the flight test scene established in the foregoing, an interaction module of the reinforcement learning agent is established, i.e., the state, action, and reward function are established.
[0101] State: Correct the relevant variables of the indicators required by the target flight test task, such as acceleration distance, acceleration time, airspeed, altitude, pitch angle, and / or pitch angle velocity, etc. Action: Adjustment action for the parameter to be corrected, for example, the parameter to be corrected can include parameters a and b, and there are five actions, as shown in Figure 28 . Figure 28 A schematic diagram of the agent action is shown.
[0102] Since the model needs to meet multiple expected targets, the reward function of each target is calculated according to the percentage error, and the reward function is larger when the result is closer to the flight test data. Assuming that the target requirement of the parameter a is s% error, the actual error q %, then the reward function Q is: .
[0103] In one embodiment, optionally, the method can further include verifying the optimized simulation model for the target flight test task to confirm that the optimized simulation model for the target flight test task meets the preset accuracy requirement. For example, after the interaction module is established, a deep Q network (DQN) reinforcement learning correction algorithm can be selected and hyperparameters can be set to carry out model correction.
[0104] For example, the optimized simulation model can be tested and verified.
[0105] As an optional example, ground flight test data and simulation results are compared and verified. As an example, ground minimum control speed flight test, normal takeoff, interrupted takeoff, deceleration stopping distance, and other flight test data can be used for verification. When the accuracy does not meet the requirement, the parameter optimization algorithm is repeatedly used to iteratively optimize the initial simulation model until the accuracy requirement is met. Figure 29 A state comparison schematic diagram before and after the model is optimized using the DQN reinforcement learning correction algorithm is shown, where the state space dimension is n and the action is discretized into 5 levels.
[0106] As another optional example, a digital virtual flight module can also be established for result verification, taking full-power takeoff as an example: ■Digital virtual flight task 1) Before takeoff, the aircraft is stationary on the runway, the longitudinal axis of the aircraft body and the front wheel are aligned with the center line of the runway. Step on the brake, push the throttle lever to the takeoff power / thrust position after starting the engine, and then release the brake; 2) The aircraft accelerates from the takeoff starting point to V R ; 3) After the speed reaches the nose wheel lifting speed V R , the pilot gradually applies the pull rod force, the aircraft nose wheel is lifted and enters the two-wheel acceleration taxiing; continue to accelerate to the takeoff speed V LOF , the aircraft takes off and starts climbing; 4) When the aircraft is higher than the takeoff plane by 35 ft, the flight speed reaches V 2 , and during this period, the horizontal distance covered by the aircraft is 1.15 times the takeoff distance under full-power working condition.
[0107] ■Task digital model
[0108] Based on the above control procedure, the digital virtual flight instructions of the full-power takeoff task can be written as:
[0109] In the above formula, is the takeoff thrust throttle; is the pull-up pitch angle command after the aircraft reaches the nose wheel lifting speed V R , generally 10°~15°, and the pitch angle rate range is 2.5°~3° / s; the track angle command =0, indicating that the pilot controls the aircraft to take off parallel to the runway. The above virtual flight is carried out, and the correctness of the model response is analyzed to verify its function.
[0110] In the method, the corresponding agent interaction model and reward function need to be established in combination with the principles of aircraft dynamics and reinforcement learning algorithm, and the parameters are adjusted according to the real flight scene of the aircraft to limit the model correction process from diverging and falling into local optimum.
[0111] According to the simulation purposes of ground modeling and correction model, a corresponding simulation program is designed for testing and verification, preferably, real flight test data is used for comparison and verification to ensure that the model accuracy meets the target requirements. Alternatively, a digital virtual flight module can also be established for verification.
[0112] The simulation model construction method for target flight test task proposed by the present disclosure determines the model architecture and parameter set based on a specific target flight test task, ensuring that the model is constructed with pertinence and purpose. The simulation model construction method for target flight test task also realizes the rational allocation of modeling resources by distinguishing between "regular sub-models" and "key sub-models", avoiding inefficient modification of non-target model parts. The simulation model construction method for target flight test task then introduces sensitivity analysis to identify "sensitivity parameters" from the parameter set that have a relatively large impact on the simulation accuracy of the current target flight test task, thereby focusing optimization computing resources on key model parts. The simulation model construction method for target flight test task finally performs directional optimization on the selected key parameters, significantly improving optimization efficiency and convergence speed.
[0113] Figure 30 A block diagram of a system 3000 for constructing a simulation model for a target flight test task according to one embodiment of the present disclosure is shown. In this embodiment, the system 3000 can include a determination module 3005, a construction module 3010, an analysis module 3015, and an optimization module 3020. In one example, the determination module 3005 can determine a model architecture employed to construct a simulation model for a target flight test task and a parameter set required to construct the simulation model for the target flight test task based on the target flight test task. The construction module 3010 can construct an initial simulation model for the target flight test task based on the model architecture and the parameter set, wherein the initial simulation model for the target flight test task includes regular sub-models and key sub-models, and wherein the key sub-models are determined based on the target flight test task. The analysis module 3015 can perform sensitivity analysis on parameters in the parameter set associated with the key sub-models to screen one or more sensitivity parameters that have an impact on the accuracy of the simulation model for the target flight test task exceeding a preset threshold. The optimization module 3020 can use a parameter optimization algorithm to optimize the initial simulation model for the target flight test task based on the one or more sensitivity parameters to obtain an optimized simulation model for the target flight test task.
[0114] Figure 31 A block diagram of a device 3100 including a system 500 for constructing a simulation model for a target flight test task according to one embodiment of the present disclosure is shown. The device shows a general hardware environment in which the present disclosure can be applied according to exemplary embodiments of the present disclosure.
[0115] Reference will now be made to Figure 31A device 3100 is described, which is an exemplary embodiment of a hardware device that can be applied to aspects of the present disclosure. The device 3100 can be any machine that is configured to perform processing and / or computations, can be, but is not limited to, a workstation, a server, a desktop computer, a laptop computer, a tablet computer, or any combination thereof. The above systems can be implemented in whole or in part by the device 3100 or a similar device or system.
[0116] The device 3100 can include elements that can connect or communicate with the bus 3102, possibly via one or more interfaces. For example, the device 3100 can include the bus 3102, one or more input devices 3105, one or more output devices 3110, one or more processors 3115, and one or more memories 3120, among others.
[0117] The processor 3115 can be any type of processors, and can include but not limited to general purpose processors and / or special purpose processors (e.g., special purpose chips), intelligent hardware devices (e.g., general purpose processors, DSPs, CPUs, microcontrollers, ASICs, FPGAs, programmable logic devices, discrete gate or transistor logic components, discrete hardware components, or any combination thereof). In some cases, the processor 3115 can be configured to operate a memory array using a memory controller. In other cases, a memory controller (not shown) can be integrated into the processor 3115. The processor 3115 can be configured to execute computer-readable instructions stored in the memory to perform various functions described herein.
[0118] The memory 3120 can be any storage device that can implement data storage. The memory 3120 can include but not limited to a disk drive, an optical storage device, a solid-state memory, a floppy disk, a flexible disk, a hard disk, a magnetic tape, or any other magnetic medium, an optical disk, or any other optical medium, a ROM (read only memory), a RAM (random access memory), a cache, and / or any other memory chip or cartridge, and / or any other medium from which a computer can read data, instructions, and / or code. The memory 3120 can store computer-executable software 3125 including computer-readable instructions that, when executed, cause the processor to perform various functions described herein. The memory 3120 can have various data / instructions / code for implementing various functions described herein.
[0119] Software 3125 can be stored in memory 3120 and can include, without limitation, an operating system, one or more applications, drivers, and / or other data and code. The instructions to perform the various functions described herein can be included in one or more of the applications, and the various elements of the device 3100 described above can be implemented by the processor 3115 reading and executing the instructions of one or more of the applications. In some cases, the software 3125 can not be directly executable by the processor but can, for example when compiled and executed, cause a computer to perform the various functions described herein.
[0120] Input device 3105 can be any type of device that can input information to the computing device.
[0121] Output device 3110 can be any type of output device that can output information. In one case, output device 3110 can be any type of image output device that can display information.
[0122] Those skilled in the art can clearly understand from the above-mentioned embodiments that the present disclosure can be implemented by software with necessary hardware or by hardware, firmware, etc. Based on such understanding, the embodiments of the present disclosure can be partially implemented in software. The computer software can be stored in a readable storage medium such as a floppy disk, a hard disk, an optical disk, or a flash memory of a computer. The computer software includes a series of instructions to make a computer (e.g., a personal computer, a service station, or a network terminal) execute a method or a part thereof according to various embodiments of the present disclosure.
[0123] The word "exemplary" is used herein to mean "serving as an example, instance, or illustration." Any aspect described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other aspects.
[0124] The previous description is provided to enable any person skilled in the art to practice the various aspects described herein. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other aspects. Thus, the claims are not intended to be limited to the aspects shown herein, but is to be accorded the full scope consistent with the language claims, wherein reference to an element in the singular is not intended to mean "one and only one" unless specifically so stated, but rather "one or more." Unless specifically stated otherwise, the term "some" refers to one or more. All structural and functional equivalents to the elements of the various aspects described throughout this disclosure that enable a person skilled in the art to practice the disclosure, are expressly incorporated in potential as equivalents of this disclosure.
Claims
1. A method for constructing a simulation model for a target-oriented flight test mission, comprising: The model architecture used to construct the simulation model for the target flight test mission and the parameter set required for constructing the simulation model for the target flight test mission are determined based on the target flight test mission. An initial simulation model for a target-oriented flight test mission is constructed based on the model architecture and the parameter set, wherein the initial simulation model for a target-oriented flight test mission includes a conventional sub-model and a key sub-model. Sensitivity analysis is performed on the parameters in the parameter set that are associated with the key sub-model to screen out one or more sensitive parameters that have an impact on the accuracy of the simulation model for the target-oriented flight test mission that exceeds a preset threshold. as well as Based on one or more of the aforementioned sensitivity parameters, a parameter optimization algorithm is used to optimize the initial simulation model of the target-oriented flight test mission to obtain an optimized simulation model of the target-oriented flight test mission.
2. The method of claim 1, further comprising: The optimized simulation model of the target-oriented flight test mission was verified to confirm that the optimized simulation model of the target-oriented flight test mission meets the preset accuracy requirements.
3. The method as described in claim 1, characterized in that, Determining the model architecture and parameter set based on the target flight test mission further includes: Establishing the target flight test scenario for the target flight test mission, wherein establishing the target flight test scenario includes: identifying the flight test object and strategy, conducting criticality analysis, forming a flight test point matrix, formulating a flight test plan, and confirming flight test risks; and The model architecture and parameter set related to the target flight test mission are determined based on the target flight test scenario, wherein the key sub-models are determined based on the target flight test mission.
4. The method as described in claim 1, characterized in that, The parameter optimization algorithm includes a reinforcement learning algorithm, and using the reinforcement learning algorithm to optimize the initial simulation model of the target-oriented flight test mission further includes: Building an interactive environment for the agent in the reinforcement learning algorithm includes: modifying variables required by the target test flight mission, determining adjustment actions for one or more sensitivity parameters; and determining a reward function based on multiple expected goals to be met by the initial simulation model of the target test flight mission. Setting hyperparameters; and The initial simulation model of the target-oriented flight test mission is optimized using the reinforcement learning algorithm based on the reward function and the hyperparameters.
5. The method as described in claim 1, characterized in that, The parameters are pre-screened or prioritized before performing sensitivity analysis on the parameters in the parameter set that are associated with the key sub-model.
6. The method as described in claim 1, characterized in that, The target flight test mission is a flight test mission related to ground motion, wherein: The initial simulation model for the target-oriented flight test mission includes at least one of the following conventional sub-models: a body aerodynamic sub-model, a dynamics sub-model, a propulsion system model, an engine model, and / or a propulsion system sub-model, wherein the body aerodynamic sub-model, the dynamics sub-model, the engine model, and the propulsion system sub-model are validated models; and / or The initial simulation model for the target-oriented flight test mission includes key sub-models comprising at least one of the following: a landing gear model, and / or a tail skid model; and / or The one or more sensitivity parameters include at least one of the following: tire friction coefficient during braking, tire lateral friction coefficient, tire rolling friction coefficient, shock absorber stiffness coefficient, shock absorber damping characteristics, and / or anti-skid braking model PID parameters.
7. The method as described in claim 1, characterized in that, Sensitivity analysis of the parameters in the parameter set associated with the key sub-model further includes: Determine the observations; Obtain the test values of the parameters in the parameter set associated with the key sub-model against the observations; and Based on the observations, a local sensitivity analysis is performed on each of the parameters to determine the parameter with the highest sensitivity as the first sensitivity parameter.
8. The method of claim 7, further comprising: Based on the determination of the first sensitivity parameter, the local sensitivity analysis is performed again on each of the other parameters based on the observations, so as to determine the parameter with the highest sensitivity as the second sensitivity parameter, wherein the local sensitivity analysis adopts the regression-based local sensitivity analysis method.
9. The method as described in claim 8, characterized in that, Optimizing the initial simulation model of the target-oriented flight test mission using a parameter optimization algorithm based on the one or more sensitivity parameters further includes: The initial simulation model of the target-oriented flight test mission is optimized using the parameter optimization algorithm based on the first sensitivity parameter and the second sensitivity parameter.
10. A system for constructing a simulation model for a target-oriented flight test mission, comprising: The determining module is used to determine, based on the target flight test mission, the model architecture used to construct the simulation model for the target flight test mission and the parameter set required to construct the simulation model for the target flight test mission; A construction module is used to construct an initial simulation model for a target-oriented flight test mission based on the model architecture and the parameter set. The initial simulation model for the target-oriented flight test mission includes a regular sub-model and a key sub-model, and the key sub-model is determined based on the target flight test mission. An analysis module is used to perform sensitivity analysis on the parameters in the parameter set that are associated with the key sub-model in order to screen out one or more sensitive parameters that have an impact on the accuracy of the simulation model of the target-oriented flight test mission that exceeds a preset threshold. as well as An optimization module is used to optimize the initial simulation model of the target-oriented flight test mission based on one or more sensitivity parameters using a parameter optimization algorithm to obtain an optimized simulation model of the target-oriented flight test mission.
11. A non-transient computer-readable storage medium having instructions stored thereon, the instructions being executable by a processor to perform the method as claimed in any one of claims 1-9.