Flight envelope protection method and system for actuator asymmetric amplitude saturation

By using a six-degree-of-freedom control-guided attitude model and safety-priority weighted quadratic programming, the problems of asymmetric amplitude saturation and underactuated residual internal dynamics of hypersonic vehicle actuators were solved, achieving reliable protection of the flight envelope and improved control stability.

CN122776870BActive Publication Date: 2026-11-17SHANDONG JIAOTONG UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202611239409.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-08-17
Publication Date
2026-11-17
Estimated Expiration
2046-08-17

AI Technical Summary

Technical Problem

Existing attitude control methods for hypersonic vehicles fail to effectively address the insufficient boundary recovery capability and underactuated residual dynamic divergence caused by actuator asymmetric amplitude saturation, resulting in control command oscillations and flight safety hazards.

Method used

A six-degree-of-freedom control and guidance attitude model is adopted. The input constraint compatible safety set is screened by combining the control barrier function and the higher-order control barrier function. The flight envelope constraint, actuator constraint and internal dynamic control Lyapunov function are processed by safety-priority weighted quadratic programming to generate feasible control surface commands.

Benefits of technology

It effectively suppresses control command oscillations, constrains the system state within the feasible safety domain, improves the safety and control stability of near-boundary flight processes, and prevents internal dynamic divergence.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122776870B_ABST
    Figure CN122776870B_ABST
Patent Text Reader

Abstract

This invention proposes a flight envelope protection method and system for actuator asymmetric amplitude saturation, relating to the field of hypersonic vehicle control technology. It includes: establishing a six-degree-of-freedom control and guidance attitude model, selecting angle of attack, sideslip angle, and roll angle as command outputs, and using pitch rate as a characterization of the residual internal dynamics generated by the two-input, three-output structure; constructing a geometric flight envelope based on the upper and lower bounds of the angle of attack and sideslip angle, and combining asymmetric amplitude constraints of the ailerons and elevators, obtaining an input constraint-compatible safety set through feasibility screening using control barrier functions and higher-order control barrier functions; establishing a safety-priority weighted quadratic programming problem, using flight envelope constraints and actuator constraints as hard constraints, the internal dynamic control Lyapunov function as a semi-hard constraint, and attitude tracking as a soft constraint; solving the quadratic programming problem to achieve near-boundary flight envelope protection. This invention achieves reliable near-boundary protection of the flight envelope.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of hypersonic vehicle control technology, and in particular to a flight envelope protection method and system for actuator asymmetric amplitude saturation. Background Technology

[0002] Hypersonic vehicles exhibit strong coupling, strong nonlinearity, and rapidly changing aerodynamic parameters during near-space flight. Angle of attack and sideslip angle directly affect aerodynamic loads, inlet operating conditions, and flight safety; they are both controlled variables for attitude tracking and critical variables within the flight envelope that require strict constraints.

[0003] Existing attitude control schemes primarily optimize tracking accuracy, convergence characteristics, and anti-interference capabilities, generally using the geometric flight envelope as a safety criterion while neglecting the impact of actuator amplitude limitations and asymmetric deflection permissions. The state within the geometric envelope may not necessarily be able to recover to a safe region. When using control barrier functions for envelope protection, failure to verify the feasibility of constraints under asymmetric control limits can easily lead to problems such as out-of-bounds errors and control command oscillations. Furthermore, the underactuated attitude architecture of the aircraft possesses residual internal dynamics; constraining only the external attitude output cannot suppress the continuous divergence of internal states such as angular rate, making it difficult to guarantee flight safety across the entire time domain. Summary of the Invention

[0004] To address the aforementioned issues, this invention proposes a flight envelope protection method and system for actuator asymmetric amplitude saturation, achieving reliable protection of the flight envelope near its boundary and mitigating the risk of exceeding limits even when the instantaneous state is within the geometric envelope.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a flight envelope protection method for actuator asymmetric amplitude saturation, comprising: A six-degree-of-freedom control and guidance attitude model is established, with angle of attack, sideslip angle and roll angle selected as command outputs, and aileron and elevator deflection selected as inputs. Pitch rate is used as the residual internal dynamics generated by the two-input three-output underactuated structure. Based on the upper and lower bounds of the angle of attack and sideslip angle, the geometric flight envelope is constructed. Combined with the asymmetric amplitude constraints of the aileron and elevator, the input constraint compatible safety set is obtained by feasibility screening through the control barrier function and higher-order control barrier function. A safety-priority weighted quadratic programming problem is established, with flight envelope constraints and actuator constraints as hard constraints, and internal dynamic control Lyapunov functions as semi-hard constraints, and attitude tracking as soft constraints. Solve the quadratic programming problem and output the calculated aileron and elevator commands to the actuator to achieve near-boundary flight envelope protection.

[0006] Secondly, the present invention provides a flight envelope protection system for actuator asymmetric amplitude saturation, comprising: The model building module is configured to establish a six-degree-of-freedom control and guidance attitude model, select angle of attack, sideslip angle and roll angle as command outputs, select aileron and elevator deflection as inputs, and use pitch rate as a characterization of the residual internal dynamics generated by the two-input three-output underactuated structure. The safety set construction module is configured to construct the geometric flight envelope based on the upper and lower bounds of the angle of attack and sideslip angle, and obtain the input constraint compatible safety set by combining the asymmetric amplitude constraints of the aileron and elevator through the feasibility screening of the control barrier function and higher-order control barrier function. The quadratic programming module is configured to establish a safety-priority weighted quadratic programming problem, using flight envelope constraints and actuator constraints as hard constraints, and internal dynamic control Lyapunov functions as semi-hard constraints, and attitude tracking as soft constraints. The solution module is configured to solve the quadratic programming problem and output the aileron and elevator commands obtained from the solution to the actuator to achieve near-boundary flight envelope protection.

[0007] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the flight envelope protection method for actuator asymmetric amplitude saturation as described in the first aspect.

[0008] Fourthly, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the flight envelope protection method for actuator asymmetric amplitude saturation described in the first aspect.

[0009] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention characterizes the underactuated attitude dynamics of an aircraft based on a six-degree-of-freedom attitude model. Angle of attack, sideslip angle, and roll angle are used as output commands, and the residual pitch rate is used to describe the internal dynamics of the system. A safety set adapted to asymmetric amplitude constraints of the ailerons and elevators is obtained by filtering using a control barrier function. Multiple constraints are configured hierarchically through a safety-priority weighted quadratic programming, with flight envelope and control surface constraints set as hard constraints, internal dynamic stability conditions as semi-hard constraints, and attitude tracking targets as soft constraints, to solve for feasible control surface commands. This strategy no longer relies solely on geometric boundaries to determine safety; it can coordinate safety constraints and tracking requirements in flight scenarios near the envelope boundary, effectively suppressing control command oscillations, preventing dynamic divergence within constraints, and ensuring that the system state remains continuously within the feasible safety domain, thus improving the safety and control stability of near-boundary flight processes.

[0010] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute a limitation thereof.

[0011] Figure 1 A main flowchart of a flight envelope protection method for actuator asymmetric amplitude saturation provided in an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the relationship between the input constraint-compatible safety set and the geometric flight envelope provided in an embodiment of the present invention. Figure 3 This is a comparison chart of the angle-of-attack response curves of the embodiments of the present invention, the nominal control method, and the geometric envelope control barrier function method; Figure 4 for Figure 3 A magnified view of the angle-of-attack response curve in the 20-25s interval; Figure 5 This is a comparison chart of the sideslip angle response curves of the embodiments of the present invention, the nominal control method, and the geometric envelope control barrier function method; Figure 6 This is a comparison chart of the roll angle response curves of the embodiments of the present invention, the nominal control method, and the geometric envelope control barrier function method; Figure 7 This is a comparison chart of the actuator aileron input response curves of the embodiments of the present invention, the nominal control method, and the geometric envelope control barrier function method; Figure 8 for Figure 7 A magnified view of the aileron input response curve in the 5-6s interval; Figure 9 This is a comparison chart of the actuator elevator input response curves of the embodiments of the present invention, the nominal control method, and the geometric envelope control barrier function method; Figure 10 This is a comparison chart of the safety margin response curves of the upper boundary of the flight envelope angle of attack of the embodiments of the present invention, the nominal control method, and the geometric envelope control barrier function method; Figure 11 for Figure 10 A magnified view of the 20-25s interval of the safety margin response curve at the upper boundary of the angle of attack; Figure 12 This is a comparison chart of the safety margin response curves of the lower boundary of the flight envelope angle of attack of the embodiments of the present invention, the nominal control method, and the geometric envelope control barrier function method; Figure 13This is a comparison chart of the safety margin response curves of the upper boundary of the flight envelope sideslip angle of the embodiments of the present invention, the nominal control method, and the geometric envelope control barrier function method; Figure 14 This is a comparison chart of the safety margin response curves of the lower boundary of the flight envelope sideslip angle for embodiments of the present invention, the nominal control method, and the geometric envelope control barrier function method; Figure 15 A comparison chart of dynamic response curves within the residual pitch rate provided in the embodiments of the present invention; Figure 16 for Figure 15 A magnified view of the dynamic response curve within the residual pitch rate range of 9.8~10.2s; Figure 17 A comparison chart of the internal dynamic Lyapunov function response curves provided in the embodiments of the present invention; Figure 18 A comparison diagram of angle-of-attack tracking error response during residual internal dynamic adjustment and attitude tracking recovery processes provided in an embodiment of the present invention; Figure 19 A comparison diagram of roll angle tracking error response during residual internal dynamic adjustment and attitude tracking recovery process provided in an embodiment of the present invention; Figure 20 A comparison diagram of the response of the internal dynamic constraint relaxation variable in the residual internal dynamic adjustment and attitude tracking recovery process provided in an embodiment of the present invention; Figure 21 A comparison diagram of the attitude tracking constraint relaxation variable response during the residual internal dynamic adjustment and attitude tracking recovery process provided in an embodiment of the present invention. Detailed Implementation

[0012] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0013] Hypersonic vehicles exhibit strong coupling, strong nonlinearity, and rapid changes in aerodynamic parameters during near-space flight. Angle of attack and sideslip angle determine the vehicle's aerodynamic loads and simultaneously affect the airflow capture state of the inlet. Exceeding these limits can easily lead to inlet instability and excessive thermal loads, directly impacting flight safety. The flight envelope defines the safe operating range of the vehicle; therefore, angle of attack and sideslip angle are not only attitude tracking variables but also critical variables within the flight envelope that must be strictly limited to a safe range.

[0014] Existing attitude control methods for hypersonic vehicles primarily focus on tracking accuracy, convergence speed, disturbance rejection performance, or fault tolerance. Even if these methods can improve attitude tracking response under nominal conditions, flight safety is generally guaranteed by the assumption that the vehicle remains within its geometric flight envelope. However, vehicle control surfaces have maximum deflection angle limitations, meaning actuator amplitudes are restricted. Furthermore, the maximum deflection angles on the upper and lower sides of some control surface configurations are inconsistent, resulting in asymmetrical deflection permissions—a hardware constraint. Due to this constraint, flight states near the geometric envelope boundaries may lack feasible control surface commands to suppress the continuous decay of safety margins. In other words, merely remaining within the geometric envelope does not guarantee that the vehicle possesses the boundary recovery capability to actively return to a safe region.

[0015] Control barrier functions and quadratic programming safety filtering provide feasible tools for flight envelope protection. If constraints are constructed solely based on the geometric flight envelope without verifying the existence of feasible control solutions by combining asymmetric control surface amplitude constraints, a mismatch will occur between theoretical constraints and the actual capabilities of the actuators. This will lead to control problems such as minor state out-of-bounds errors within the sampling interval, frequent switching of control surface commands, and lack of feasible solutions in quadratic programming.

[0016] Furthermore, the aircraft's attitude system is an underactuated structure, relying on two physical actuator inputs to control the three attitude outputs: angle of attack, sideslip angle, and roll angle. The external output channels cannot fully represent all the information of the six-dimensional attitude dynamics, and the system has residual internal dynamics that cannot be directly controlled by external outputs. If explicit constraints are not applied to the internal dynamics, even if the external tracking performance of the angle of attack, sideslip angle, and roll angle is up to standard, internal states such as pitch rate may continue to accumulate and diverge, ultimately compromising the feasibility of full-state motion.

[0017] To address the problems of existing methods that equate the geometric envelope with the recoverable safety envelope, neglect the feasibility of recovering from actuator asymmetric amplitude saturation at the lower boundary, and fail to explicitly adjust the residual internal dynamics in underactuated systems, this invention proposes a flight envelope protection method, system, medium, and device for actuator asymmetric amplitude saturation. It constructs a six-DOF attitude model of the aircraft, defines the pitch rate as the residual internal dynamic for a two-input, three-output underactuated system, integrates the geometric flight envelope and control surface amplitude constraints, determines the input-compatible safety set based on the control barrier function, and constructs a safety-priority weighted quadratic programming approach. Multiple constraints are set hierarchically to solve the control surface commands online, achieving flight envelope protection under near-boundary conditions.

[0018] Example 1 like Figure 1 As shown, this embodiment discloses a flight envelope protection method for actuator asymmetric amplitude saturation, including the following steps: S1: Establish a six-degree-of-freedom control and guidance attitude model, select angle of attack, sideslip angle and roll angle as command outputs, select aileron and elevator deflection as inputs, and use pitch rate as a characterization of the residual internal dynamics generated by the two-input three-output underactuated structure. S2: Construct the geometric flight envelope based on the upper and lower bounds of the angle of attack and sideslip angle, and combine the asymmetric amplitude constraints of the aileron and elevator to obtain the input constraint compatible safety set through feasibility screening of the control barrier function and higher-order control barrier function; S3: Establish a safety-priority weighted quadratic programming problem, with flight envelope constraints and actuator constraints as hard constraints, and internal dynamic control Lyapunov functions as semi-hard constraints, and attitude tracking as soft constraints. S4: Solve the quadratic programming problem and output the calculated aileron and elevator commands to the actuator to achieve near-boundary flight envelope protection.

[0019] Next, combined Figure 1 This embodiment provides a detailed description of a flight envelope protection method for actuator asymmetric amplitude saturation.

[0020] I. Six-DOF Controlled Guiding Attitude Model Under the assumption of rigid body dynamics, a six-degree-of-freedom control and guidance attitude model for a hypersonic vehicle is established. The state vector and input vector are defined as follows: , .in, , and These represent the angle of attack, sideslip angle, and roll angle, respectively. , and These represent roll rate, yaw rate, and pitch rate, respectively. and These represent the aileron deflection and elevator deflection, respectively.

[0021] The six-degree-of-freedom control and guidance attitude model can be written as: (1) in, to , , , , , , to These are the model coefficients determined by the aircraft's mass, velocity, moment of inertia, aerodynamic derivative, and aerodynamic moment derivative.

[0022] The model retains the key coupling relationships between angle of attack, sideslip angle, roll angle, and three-axis angular rates; it also retains the basic input-output main effect channel, namely the elevator. Directly affects the angle of attack and pitch rate channels, ailerons This directly affects the roll and yaw rate channels. The key coupling relationship refers to the kinematic coupling between attitude angles and three-axis angular rates, as well as the dynamic cross-coupling between the roll, yaw, and pitch angular motion channels. Specifically, the angle of attack change is affected by both the pitch rate and roll-slip motion, the sideslip angle change is affected by both the yaw rate and roll-angle of attack motion, and the roll angle change is directly related to the roll rate; at the same time, the three-axis angular rate dynamics includes cross-coupling terms formed by the sideslip angle, angle of attack, and the product of different axis angular rates. Therefore, a change in any attitude channel can affect other attitude channels through the above coupling terms, and the three attitude motions cannot be treated completely independently.

[0023] II. Hybrid Relative Rank Structure and Residual Internal Dynamic Coordinates The selected instruction output is The relative order represents the equivalent number of integrators between the control input u and the output y in a linear / nonlinear system. Mathematically, it is defined as the number of derivatives r taken continuously with respect to the output y until u first appears explicitly in the derivative expression. For the six-degree-of-freedom control-guided attitude model shown in equation (1), due to the angle of attack... Find the first derivative Its expression directly contains Therefore, angle of attack It has a first-order relative order with respect to the physical actuator input; due to the side slip angle and roll angle When finding the first derivative, it does not appear explicitly in the expression. Therefore, sideslip angle and roll angle It has a second relative order with respect to the physical actuator input.

[0024] Therefore, it satisfies And the system's overall relative order 6 represents the dimension of the system state vector. Wherein, , , These represent the relative orders corresponding to the angle of attack, sideslip angle, and roll angle, respectively.

[0025] Since the sum of relative orders is less than the state dimension, in addition to the 5-dimensional external dynamics that can be decomposed into the state space, the system also has a residual internal dynamic degree of freedom not included in the external output channel. This degree of freedom cannot be directly constrained by output feedback and its stability needs to be handled separately. In this embodiment, the pitch rate is selected as the residual internal dynamic coordinate, i.e. .

[0026] Define the external coordinates as From equation (1), the original state variable reconstruction relation can be obtained as follows: , , , , ,as well as .

[0027] The original state variable reconstruction relationship is obtained directly from the definition of external coordinates and equation (1). Based on the input-output relative order, the external coordinates are defined as follows: And select the residual internal dynamic coordinates Therefore, there is a direct result. , , , .

[0028] Furthermore, according to formula (1) ,as well as ,get .

[0029] According to formula (1) ,as well as The results are as follows: .

[0030] Therefore, all six original state variables can be derived from... The only recovery. Furthermore, by achievable The pitch rate equation of equation (1) Replacing each of the original state variables with the above reconstruction relation yields: (2) Since the residual internal dynamics are not directly observable and controllable in the external dynamic subspace after input-output linearization, their boundedness cannot be indirectly guaranteed by the external output tracking error. Furthermore, if the residual degrees of freedom diverge, it will directly threaten the structural strength of the aircraft and the robustness of the control system. Therefore, to quantify the amplitude of the residual internal dynamics, an internal dynamic control Lyapunov function is defined. .

[0031] III. Geometric Flight Envelope and Input Constraint Compatible Safety Set The flight envelope represents the closed geometry of an aircraft's permissible flight range and limitations. Its shape is determined by multiple factors, including engine performance, aerodynamic characteristics, and structural strength. In this embodiment, the angle of attack and sideslip angle are used as flight envelope protection variables, and the geometric flight envelope is defined as follows: (3) in, For the work area, This is the safety margin function.

[0032] When the protection boundary is given by the upper and lower limits of the angle of attack and sideslip angle, the safety margin function is: , , , .

[0033] in, , , , These are the safety margin functions for the upper boundary of the angle of attack, the lower boundary of the angle of attack, the upper boundary of the sideslip angle, and the lower boundary of the sideslip angle, respectively. For scheduling variables, , , and These are the upper and lower boundaries of the angle of attack and sideslip angle, respectively.

[0034] Directly satisfy Simply indicating that the state is within the geometric boundary does not guarantee that there exists an actuator with feasible control input, ensuring the system trajectory remains constrained within the safe set. If a feasible execution input is lacking, the system state will evolve out of the safe region, leading to a safety violation. Therefore, this embodiment unifies the retained control barrier function constraints and higher-order control barrier function constraints into a single form. .

[0035] in, The input coefficient matrix is ​​used to control the barrier function constraints. For the corresponding constraint boundary vectors, both are determined by the current flight state. The aircraft dynamics model, as well as the constructed control barrier functions and higher-order control barrier functions, are determined. Specifically, after expanding the CBF / HOCBF constraints corresponding to each flight envelope boundary, the control input is included. The terms of 'A' are moved to the left side of the inequality, and the remaining terms are moved to the right side, thus uniformly represented as 'A'. .

[0036] In actual flight, due to limitations imposed by aerodynamic hinge torque and hydraulic power, the deflection limits of control surfaces in the positive and negative directions are often unequal. For example, the upward deflection of the elevator (generating a nose-down moment) may be limited by a mechanical stop of -20°, while the downward deflection (generating a nose-up moment) is limited by actuator power of only +15°. This asymmetry between positive and negative limits is known as asymmetric amplitude saturation. Considering the asymmetric amplitude saturation of the ailerons and elevator, the allowable input set of the actuator is defined as follows: (4) Wherein, at least one actuator channel satisfies This asymmetry results in different recovery capabilities for the upper and lower boundaries. Whether boundary recovery is feasible needs to be determined by considering both the current direction of motion and the remaining control surface permissions.

[0037] Define the set of hard-safe feasible inputs as: (5) Further define the input constraint compatible safety set as: (6) thus, It is the geometric flight envelope A subset of states that can be recovered and kept safe by at least one actuator with permissible input. For example... Figure 2 As shown.

[0038] Figure 2 This indicates that the input constraints are compatible with the safety set. It is the geometric flight envelope A subset of. This indicates that the state is outside the geometric flight envelope. The geometric flight envelope is defined by a red dashed line. Enclosed by the blue dashed line, the input constraint-compatible safety set is defined by the blue dashed line. The blue area represents a state that still has boundary recovery capability under actuator amplitude constraints, while the yellow area represents a state that, although within the geometric flight envelope, lacks feasible control inputs that satisfy the constraints. This illustrates that merely satisfying the geometric envelope constraints does not guarantee state recovery; both the current motion state and the available control permissions of the actuator must also be considered.

[0039] For a safety margin function of relative order one The derivative of the safety margin function already explicitly contains rudder deflection; therefore, a control barrier function constraint is applied. .

[0040] For a safety margin function of relative order two Since the first derivative of the safety margin function does not contain rudder deflection, the approach velocity cannot be directly changed through rudder deflection. It is necessary to first construct a combined variable containing velocity information so that the rudder deflection is reflected in the second derivative. Therefore, auxiliary variables are defined first. Then apply higher-order control barrier function constraints. .in, , and All are extended classes function, , , , These are the Lie derivatives of the corresponding functions along the vector fields f and g, respectively.

[0041] In one implementation, the geometric flight envelope is subjected to input feasibility screening, specifically for the upper boundary of the angle of attack. ,when When the input constraint remains approximately constant within a single control update cycle, the following input constraint compatibility determination form can be obtained: (7) in, This represents the barrier gain coefficient corresponding to the angle-of-attack safety margin.

[0042] Equation (7) shows that near the upper limit of the angle of attack, the positive pitch rate increases the demand for elevator restoration authority. Even if the angle of attack has not yet reached the upper limit, it may not belong to the input constraint compatible safety set due to insufficient remaining elevator authority.

[0043] IV. Nominal Command Generation and Physical Control Surface Assignment To maintain attitude tracking performance, this embodiment generates nominal control commands in the three-dimensional virtual output channel and assigns them to the input of the two-dimensional physical actuator.

[0044] The virtual output channel is , These are the status-related items for the virtual output channel.

[0045] In one implementation, Possible options: (8) Let the virtual nominal control command be The nominal physical control surface input is obtained from the following weighted allocation problem:

[0046] when When the matrix is ​​positive definite, the analytical form of the above expression is:

[0047] The above A baseline command that is only performance-oriented cannot directly guarantee safety, so it is necessary to enter a safety-priority secondary programming process for constraint filtering. For virtual nominal control commands, The input performance matrix is ​​the input to the virtual output channel of the physical actuator. and It is a positive definite weight matrix.

[0048] Define the lifting tracking error vector And define the tracking control Lyapunov function. ,in , It is a positive definite symmetric weight matrix.

[0049] V. Safety-Priority Weighted Secondary Programming The decision variable of the safety-priority weighted quadratic programming is the physical actuator input. and two slack variables and The objective function is: (9) in, The actuator input is from the previous sampling time. and The matrix is ​​a positive definite weight matrix. The first term makes the filtered input close to the nominal input, the second term suppresses input changes at adjacent control update times, and the third and fourth terms penalize intra-dynamic relaxation and tracking relaxation, respectively. By setting... This prioritizes internal dynamic relaxation over tracking relaxation. , These represent the internal dynamic relaxation penalty coefficient and the tracking relaxation penalty coefficient, respectively.

[0050] The constraints of the quadratic programming are: (10) in, This indicates the asymmetric amplitude constraints for the ailerons and elevator. For attitude tracking control, Lyapunov function, and The parameters are positive. The first set of constraints are hard safety constraints for the flight envelope, the second set are hard constraints for the asymmetric actuator amplitude, the third set are semi-hard constraints for residual internal dynamics, and the fourth set are soft constraints for attitude tracking. The flight envelope and actuator constraints do not have slack variables, and therefore are strictly satisfied when quadratic programming is feasible.

[0051] At each control update moment, the flight control computer maintains the current state. Input from the previous moment Solving the quadratic programming problem for the parameters yields the following results: Ultimately, it will Output to the aileron and elevator actuators.

[0052] As one implementation method, a flight altitude of 31.5479 km and a flight speed of 5890.3 m / s can be selected, with asymmetric amplitude constraints on the ailerons and elevator. Angle of attack envelope is The sideslip angle envelope is The control update cycle is 0.001 s. This implementation is used to verify the input constraint compatible safety set construction, near-boundary flight envelope protection, and residual dynamic adjustment capabilities of the method in this embodiment.

[0053] like Figure 2 The corresponding input constraint-compatible safety set is constructed to illustrate the relationship between the geometric flight envelope and the input constraint-compatible safety set; for example... Figures 3-14 Corresponding to near-boundary flight envelope protection, it is used to display attitude response, actuator input, and flight envelope safety margin; at the same time, a magnified view is set to take into account the overall change trend over the entire time period and the subtle features of the curves in key sections, which facilitates a complete and clear comparison of the response differences of different control methods; such as Figures 15-21 Corresponding to residual internal dynamic adjustment, this is used to demonstrate the changes in residual pitch rate, internal dynamic Lyapunov function, tracking error, and relaxation variables. Among these, flight altitude, flight speed, actuator asymmetric amplitude constraints, and the envelopes of angle of attack and sideslip angle are common parameters in the above simulations; the control update cycle... Mainly used for Figures 3-21 The closed-loop dynamic simulation is shown.

[0054] Figures 3-6 By comparing the attitude responses of nominal control, Ref.-Geo-CBF, and the proposed method, it can be seen that the proposed method, while maintaining the sideslip angle in compliance with envelope constraints and the roll angle in command tracking, actively preserves a safety margin when the angle of attack approaches its upper boundary, avoiding touching or exceeding the upper boundary of the angle of attack due to over-tracking commands. Here, Ref. represents the reference command, and Bounds represents the state constraint boundaries.

[0055] Figures 7-9 By comparing the aileron and elevator control inputs of the three methods, it can be seen that the method in this embodiment can generate a smoother control input within the asymmetric control surface amplitude constraint range. In contrast, the elevator of the geometric flight envelope control method exhibits frequent large-scale switching, indicating that the method in this embodiment can improve the smoothness of the control input while satisfying actuator constraints.

[0056] Figures 10-14 By comparing the safety margins corresponding to the boundaries of the angle of attack and sideslip angle, it can be seen that the method of this embodiment can keep each safety margin non-negative, and the upper boundary of the angle of attack still retains a positive safety margin in the near-boundary state. In contrast, the nominal control and geometric flight envelope control have smaller safety margins or exceed the boundaries near the upper boundary of the angle of attack, thus indicating that the method of this embodiment has better near-boundary flight envelope protection capability.

[0057] Figures 15-17By comparing the control results with and without residual internal dynamic adjustment, it can be seen that without considering residual internal dynamic adjustment, the pitch rate and its Lyapunov function increase rapidly, leading to premature termination of the simulation. With the method of this embodiment, the residual pitch rate remains bounded, and the corresponding Lyapunov function remains small, indicating that the introduced residual internal dynamic adjustment can effectively suppress internal state divergence. Here, Proposed QP represents the quadratic programming solution in this embodiment.

[0058] Figures 18-21 By comparing the tracking errors and relaxation variables of the two methods, it can be seen that the method without residual internal dynamic adjustment terminates prematurely due to the divergence of internal state; the method in this embodiment can keep the system running continuously and gradually converge the tracking errors of angle of attack and roll angle after switching, while the relaxation variable remains small and approaches zero, indicating that attitude tracking performance can be restored on the basis of meeting safety constraints and internal dynamic adjustment requirements.

[0059] Specifically, Figure 3 In the first method, the nominal control curve closely follows the target reference curve throughout the entire process. In the second method, the geometric flight envelope control (Ref.-Geo-CBF) rapidly increases the angle of attack during the initial phase, quickly converging to the target reference curve and maintaining stability. The proposed method exhibits a more gradual rate of angle of attack ascent, slowly approaching the target reference curve before stably following it. None of the three control methods exceeded the constraint boundaries. Compared to the other two methods, the proposed method shows a smoother angle of attack change, a more stable dynamic process, and a shock-free control response.

[0060] Figure 4 In the figure, a magnified view of the steady-state phase of the angle of attack reveals that the angle of attack of the geometric flight envelope control (Ref.-Geo-CBF) fluctuates periodically around the target reference value, with peak values ​​slightly higher than the upper boundary and trough values ​​lower than the target reference value. This figure further illustrates local fluctuations that are not easily observed in the original global figure.

[0061] Figure 5 In summary, the sideslip angles of all three control methods remained within a small range. The nominal control curve closely followed the target reference curve (Ref.) throughout the entire process, with a high degree of overlap. The nominal control curve and the proposed method remained essentially near zero, while the geometric flight envelope control (Ref.-Geo-CBF) remained slightly above zero. None of the three methods exceeded the sideslip angle constraint boundaries, and the overall sideslip angle remained stable.

[0062] Figure 6In summary, all three control methods achieve the target reference value for roll angle. Nominal control and Ref.-Geo-CBF control increase the roll angle more rapidly, while the proposed method increases it more gradually, subsequently converging to the target reference value. Comparatively, the proposed method exhibits a smoother roll angle change.

[0063] Figure 7 In summary, the aileron control values ​​of all three control methods are within the allowable range. The aileron control values ​​of the nominal control and the proposed method are relatively small overall, and the geometric flight envelope control (Ref.-Geo-CBF) is also mainly distributed near zero on a global scale. None of the control values ​​exceed the aileron constraint boundaries.

[0064] Figure 8 China: To Figure 7 A magnified view of the aileron control values ​​reveals that nominal control exhibits significant periodic fluctuations; the geometric flight envelope control (Ref.-Geo-CBF) has smaller control values ​​with slight fluctuations; and the proposed method shows continuous and gradual changes in control values. This figure more clearly illustrates the differences in the three methods under small control values.

[0065] Figure 9 In the middle section, the elevator control quantity of the geometric flight envelope control (Ref.-Geo-CBF) exhibits continuous, large, and rapid changes, repeatedly approaching its upper limit; the dense red area represents the intermittent saturation of the Ref.-Geo-CBF curve due to saturation. The elevator control quantities of the nominal control and the proposed method remain within a smaller range, with relatively stable changes; the curves near zero correspond to the baseline control input without additional constraint intervention. Because the large oscillations of the Ref.-Geo-CBF widen the range of the vertical axis, the small changes of the Nominal and Proposed methods near zero are visually compressed, appearing flat on the horizontal axis. However, in reality, after the initial rapid convergence, the Nominal remains within a trim range very close to zero throughout the subsequent flight, with a gentle trend and no high-frequency oscillations. In contrast, the proposed method avoids large and frequent changes in the elevator control quantity.

[0066] Figure 10 In the middle: For the upper angle of attack constraint function During operation, the nominal control drops below zero; the geometric flight envelope control (Ref.-Geo-CBF) rapidly decreases to near zero and exhibits continuous small fluctuations; the proposed method remains positive throughout and gradually stabilizes. This demonstrates that the proposed method can maintain a certain angle-of-attack safety margin during this process.

[0067] Figure 11 In Chinese: Geometric flight envelope control (Ref.-Geo-CBF) When magnified locally during the steady-state phase, it can be seen that the value fluctuates periodically around zero and crosses the zero boundary multiple times. This figure further illustrates the local constraint changes that are not easily observed in the global graph.

[0068] Figure 12 In the middle: For the lower angle of attack constraint function All three control methods remained positive throughout the process. Nominal control and Ref.-Geo-CBF reached stable values ​​relatively quickly, while the proposed method gradually stabilized after a gentle rise from a lower initial value. None of the three methods touched the lower angle-of-attack constraint boundary.

[0069] Figure 13 In the middle: For the upper side slip angle constraint function All three control methods consistently maintain a value above zero. The nominal control curve and the curve of the proposed method highly overlap; the curve of the geometric flight envelope control (Ref.-Geo-CBF) remains basically stable with the curve of the proposed method. The proposed method has a slightly larger positive margin, indicating that the sideslip angle is always within the allowable range.

[0070] Figure 14 In the middle: For the lower side slip angle constraint function All three control methods consistently maintained above zero. The nominal control curve and the curve of the proposed method highly overlapped; both the geometric flight envelope control (Ref.-Geo-CBF) and the proposed method remained stable, with the proposed method exhibiting a slightly larger positive margin, indicating that the lower sideslip constraint boundary was not reached throughout the process.

[0071] Figure 15 Internal state of geometric flight envelope control (Ref.-Geo-CBF) It increases rapidly after about 3 seconds and terminates prematurely due to numerical divergence; the method of this embodiment (Proposed QP) The value remains small throughout the entire control process, and remains bounded and gradually stabilizes after a 10-second switch. In contrast, the method in this embodiment can effectively suppress the divergence of internal states and ensure the continuous operation of the control process.

[0072] Figure 16 China: To Figure 15 The method in this embodiment is near the 10-second switching time. By zooming in on a specific area, it can be seen that before switching... The system remained largely stable; after the switch, there was a brief period of slight fluctuation, which then quickly weakened and gradually stabilized. This indicates that although the switching process produced transient changes, it did not cause sustained oscillations or divergence in the internal state.

[0073] Figure 17 In Chinese: Geometry Flight Envelope Control (Ref.-Geo-CBF) It rapidly increases to a large value after approximately 3 seconds, and terminates prematurely with the control process; the method of this embodiment (Proposed QP) Throughout the process, the values ​​remain within a small range close to zero. In contrast, the method in this embodiment maintains the stability of internal state-related quantities, preventing them from continuously increasing.

[0074] Figure 18 In the first embodiment, the angle-of-attack error of the geometric flight envelope control (Ref.-Geo-CBF) increases rapidly after approximately 3 seconds and terminates prematurely. The proposed QP method maintains a small error before switching, exhibits a brief, small change after 10 seconds of switching, and then rapidly decreases to near zero, with the target reference curve (Ref.) consistently remaining at the zero-error baseline. This demonstrates that the proposed method can recover angle-of-attack tracking and maintain stability after switching.

[0075] Figure 19 In the first embodiment, the roll angle error of the Geo-Flight Envelope Control (Ref.-Geo-CBF) gradually decreases during the initial stage of operation, but the control process terminates prematurely at approximately 3 seconds. In the second embodiment, the roll angle error of the proposed QP method decreases steadily from an initial large negative value and eventually converges to near zero, with the target reference curve (Ref.) consistently remaining at the zero-error baseline. This demonstrates that the proposed method can continuously achieve roll angle tracking with stable error changes.

[0076] Figure 20 In Chinese: The method of this embodiment (Proposed QP) Overall, the effect remains on a very small scale, with brief peaks at the initial moment and the 10-second switchover point, followed by rapid decay to near zero. The target reference curve (Ref.) represents the desired zero-value equilibrium state, and its curve consistently lies near zero on both the horizontal and vertical axes. This indicates that the impact of the switchover is short-lived. It can quickly recover to a smaller range.

[0077] Figure 21 In Chinese: The method of this embodiment (Proposed QP) In the initial stage, the value rapidly decreases and approaches zero. A small transient peak occurs at the 10-second switching point, followed by a rapid decay back to near zero. The target reference curve (Ref.) represents the desired zero-value equilibrium state, and its curve remains near the zero value on both the horizontal and vertical axes. This demonstrates that the method in this embodiment can achieve this balance both before and after the switching process. The fluctuations remained small, and the switching process did not cause sustained volatility.

[0078] Compared to quadratic programming based solely on the control barrier function of the geometric flight envelope, this embodiment first filters out states in the geometric envelope that lack actuator recoverability, and then performs online safety filtering, which reduces the risk of near-boundary sampling gap overruns and repeated control surface switching. Compared to schemes without internal dynamic control Lyapunov function constraints, this embodiment suppresses the growth of residual pitch rate through semi-hard internal dynamic constraints, which helps maintain the feasibility of quadratic programming across the entire time domain.

[0079] This specific embodiment characterizes a two-input, three-output underactuated structure and its residual internal dynamics based on a six-degree-of-freedom attitude model. Combined with asymmetric constraints on the control surfaces, a true safe set compatible with the inputs is obtained through screening using a control barrier function. A hierarchical weighted quadratic programming model is constructed to handle envelope hard constraints, internal dynamic semi-hard constraints, and attitude tracking soft constraints in a hierarchical manner. By distinguishing between the geometric envelope and the actuator feasible safe region, the feasibility of the barrier constraints is verified, internal dynamic divergence is suppressed, near-boundary flight over-limit risks are addressed, and the envelope protection theory for underactuated aircraft is improved.

[0080] The above embodiments are merely preferred embodiments of the present invention. Without departing from the basic idea of ​​the present invention, the method can be extended to hypersonic vehicle safety control scenarios that include actuator rate limits, aerodynamic parameter uncertainties, or sampling data invariance margins.

[0081] Example 2 This embodiment provides a flight envelope protection system for actuator asymmetric amplitude saturation, including: The model building module is configured to establish a six-degree-of-freedom control and guidance attitude model, select angle of attack, sideslip angle and roll angle as command outputs, select aileron and elevator deflection as inputs, and use pitch rate as a characterization of the residual internal dynamics generated by the two-input three-output underactuated structure. The safety set construction module is configured to construct the geometric flight envelope based on the upper and lower bounds of the angle of attack and sideslip angle, and obtain the input constraint compatible safety set by combining the asymmetric amplitude constraints of the aileron and elevator through the feasibility screening of the control barrier function and higher-order control barrier function. The quadratic programming module is configured to establish a safety-priority weighted quadratic programming problem, using flight envelope constraints and actuator constraints as hard constraints, and internal dynamic control Lyapunov functions as semi-hard constraints, and attitude tracking as soft constraints. The solution module is configured to solve the quadratic programming problem and output the aileron and elevator commands obtained from the solution to the actuator to achieve near-boundary flight envelope protection.

[0082] Example 3 This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in the flight envelope protection method for actuator asymmetric amplitude saturation as described in Embodiment 1 above.

[0083] Example 4 This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in the flight envelope protection method for actuator asymmetric amplitude saturation as described in Embodiment 1 above.

[0084] The steps or modules involved in Embodiments 2 to 4 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0085] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A flight envelope protection method for actuator asymmetric amplitude saturation, characterized in that, include: A six-degree-of-freedom control and guidance attitude model for an aircraft is established, with angle of attack, sideslip angle, and roll angle selected as command outputs, and aileron and elevator deflection selected as inputs. Pitch rate is used as the residual internal dynamics generated by the two-input three-output underactuated structure. The six-degree-of-freedom control and guidance attitude model is used to characterize the underactuated attitude dynamics of the aircraft. The aircraft attitude system is an underactuated structure, and the angle of attack, sideslip angle, and roll angle are controlled by two physical actuator inputs. The geometric flight envelope is constructed based on the upper and lower bounds of the angle of attack and sideslip angle. Combined with the asymmetric amplitude constraints of the aileron and elevator, an input constraint-compatible safety set is obtained through feasibility screening using control barrier functions and higher-order control barrier functions. The specific process of generating the input constraint-compatible safety set includes: considering the asymmetric amplitude saturation of the aileron and elevator, defining the allowable input set of the actuator based on the actual physical deflection limits of the aileron and elevator. ; wherein, is an input vector, and denote aileron deflection and elevator deflection, respectively; In the current state, select the rudder deflection commands from the set of permissible inputs that ensure the system state does not exceed the safety boundary, thus forming a hard-safe feasible input set: ; wherein, is an input coefficient matrix of the control barrier function constraint, is a corresponding constraint boundary vector, both determined by the current flight state , the aircraft dynamics model, and the constructed control barrier function and higher order control barrier function. Then, the state points in the geometric flight envelope that have a non-empty intersection with the hard-safe feasible input set are defined as the input constraint compatible safety set, expressed as: ; in, For geometric flight envelope; A safety-priority weighted quadratic programming problem is established, with flight envelope constraints and actuator constraints as hard constraints, and the Lyapunov function of internal dynamic control as a semi-hard constraint, and attitude tracking as a soft constraint; the quadratic programming constraints are: ; in, For the physical actuator input, the first set of constraints is the flight envelope hard safety constraint, and the second set of constraints... The first group represents the asymmetric amplitude constraints for the ailerons and elevator, which are hard constraints for asymmetric actuator amplitudes. The hard constraints formed by the flight envelope constraints and actuator constraints do not have slack variables. The third group of constraints are residual internal dynamic semi-hard constraints. The first set of constraints is the Lyapunov function for internal dynamic control; the fourth set of constraints is the soft constraint for attitude tracking. For attitude tracking control, Lyapunov function, and It is a positive parameter. and There are two slack variables; Solve the quadratic programming problem and output the calculated aileron and elevator commands to the actuator to achieve near-boundary flight envelope protection.

2. The flight envelope protection method for actuator asymmetric amplitude saturation as described in claim 1, characterized in that, In the six-degree-of-freedom control and guidance attitude model, the state vector includes the angle of attack, sideslip angle, roll angle, and three-axis angular rate, and the control vector includes the aileron deflection and elevator deflection; among them, the elevator directly affects the angle of attack and pitch rate, and the aileron directly affects the roll and yaw rate.

3. The flight envelope protection method for actuator asymmetric amplitude saturation as described in claim 1, characterized in that, The process of obtaining an input constraint-compatible safety set through feasibility screening using control barrier functions and higher-order control barrier functions specifically includes: For the first-order safety margin function, a control barrier function constraint is directly applied to limit the upper limit of the speed at which the state approaches the safety boundary; For the second-order safety margin function, first construct auxiliary variables containing position and velocity information, and then apply higher-order control barrier function constraints to indirectly limit the upper limit of acceleration when the state approaches the safety boundary; The first-order safety margin function is the safety margin function corresponding to the angle of attack, and the second-order safety margin function is the safety margin function corresponding to the sideslip angle and the roll angle.

4. The flight envelope protection method for actuator asymmetric amplitude saturation as described in claim 1, characterized in that, It also includes generating nominal control commands in the three-dimensional virtual output channel to maintain attitude tracking performance, and then allocating the nominal control commands to the two-dimensional physical actuator input through the input performance matrix, and using weighted optimization to make the allocation result approximate the virtual commands based on least squares.

5. The flight envelope protection method for actuator asymmetric amplitude saturation as described in claim 4, characterized in that, The aforementioned safety-priority weighted quadratic programming problem uses flight envelope constraints and actuator constraints as hard constraints, and the Lyapunov function of internal dynamic control as a semi-hard constraint, and attitude tracking as a soft constraint. Specifically, it includes: The objective function of the quadratic programming problem consists of a first term that makes the filtered input close to the nominal control command, a second term that suppresses the rate of change of the input in adjacent control cycles, and a third and a fourth term that penalize the internal dynamic relaxation and the tracking relaxation, respectively. The weights are adjusted so that the penalty for the internal dynamic relaxation has a higher priority than that for the tracking relaxation. In the constraints, the flight envelope constraint and actuator amplitude constraint do not contain slack variables, the internal dynamic Lyapunov function constraint is configured with slack variables, and the attitude tracking constraint is configured with slack variables.

6. The flight envelope protection method for actuator asymmetric amplitude saturation as described in claim 1, characterized in that, Solving the quadratic programming problem and outputting the calculated aileron and elevator commands to the actuators to achieve near-boundary flight envelope protection specifically includes: at each control update time, collecting the current flight state and the actuator input at the previous time, using these as parameters to solve the quadratic programming problem online, and outputting the calculated optimal aileron deflection and elevator deflection to the corresponding actuators.

7. A flight envelope protection system for actuator asymmetric amplitude saturation, implementing the flight envelope protection method for actuator asymmetric amplitude saturation as described in claim 1, characterized in that, include: The model building module is configured to establish a six-degree-of-freedom control and guidance attitude model of the aircraft, selecting angle of attack, sideslip angle, and roll angle as command outputs, and aileron and elevator deflection as inputs, and using pitch rate as the residual internal dynamics generated by the two-input three-output underactuated structure; wherein, the six-degree-of-freedom control and guidance attitude model is used to characterize the underactuated attitude dynamics of the aircraft, the aircraft attitude system is an underactuated structure, and the angle of attack, sideslip angle, and roll angle are controlled by two physical actuator inputs; The safety set construction module is configured to construct a geometric flight envelope based on the upper and lower bounds of the angle of attack and sideslip angle, and combine the asymmetric amplitude constraints of the aileron and elevator to obtain an input constraint compatible safety set through feasibility screening using control barrier functions and higher-order control barrier functions. The generation process of the input constraint compatible safety set includes: defining the allowable input set of the actuator based on the actual physical deflection limits of the aileron and elevator; in the current state, selecting the rudder deflection commands from the allowable input set that can prevent the system state from exceeding the safety boundary to form a hard safety feasible input set; and then defining the state points in the geometric flight envelope that have a non-empty intersection with the hard safety feasible input set as the input constraint compatible safety set. The quadratic programming module is configured to establish a safety-priority weighted quadratic programming problem, using flight envelope constraints and actuator constraints as hard constraints, and internal dynamic control Lyapunov functions as semi-hard constraints, and attitude tracking as soft constraints. The solution module is configured to solve the quadratic programming problem and output the aileron and elevator commands obtained from the solution to the actuator to achieve near-boundary flight envelope protection.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the flight envelope protection method for actuator-oriented asymmetric amplitude saturation as described in any one of claims 1-6.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the flight envelope protection method for actuator-oriented asymmetric amplitude saturation as described in any one of claims 1-6.

Citation Information

Patent Citations

  • Hypersonic aircraft trajectory tracking control method for performance recovery

    CN110515392A

  • Hypersonic flight vehicle ascending stage attitude fault-tolerant control method considering damage

    CN117519277A