Fault-tolerant control methods, devices and aircraft for aircraft

By employing a cascaded control method of position outer loop and attitude inner loop and a constrained secondary thrust matrix reconstruction in an eVTOL aircraft, the problems of timeliness and accuracy of power compensation response were solved, achieving precise thrust compensation and attitude stabilization after rotor failure, and reducing computational complexity and cost.

CN123061584APending Publication Date: 2026-11-10YONGJIANG LAB
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610994011.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-03
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

In the existing technology, eVTOL aircraft have difficulty ensuring both timeliness and accuracy of response during power compensation, which may lead to over- or under-compensation of power, resulting in overall failure.

Method used

A cascaded control method based on position outer loop and attitude inner loop is adopted to reconstruct the thrust matrix of available rotors. The target thrust matrix of available rotors is determined by constrained secondary thrust matrix reconstruction. The thrust matrix is ​​then optimized by combining the standard weighted L2 objective and the solver to achieve smooth adjustment of the thrust matrix.

Benefits of technology

It achieves precise compensation for thrust loss and attitude deviation after rotor failure, improves the response speed and accuracy of power compensation, reduces the chip computing power requirements of the algorithm, and reduces costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123061584A_ABST
    Figure CN123061584A_ABST
Patent Text Reader

Abstract

This application relates to a fault-tolerant control method, device, and aircraft for an aircraft, comprising: identifying a faulty rotor; reconstructing the thrust matrix of available rotors based on a cascaded control method of position outer loop and attitude inner loop; performing a constrained secondary thrust matrix reconstruction based on the thrust matrix reconstruction to determine the target thrust matrix of the available rotors; and smoothly adjusting the motor output of each available rotor based on the target thrust matrix to enter a fault-tolerant flight mode, thereby ensuring timely response of power compensation while maintaining accuracy.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of aircraft technology, and in particular to an aircraft fault-tolerant control method, device and aircraft. Background Technology

[0002] With the development of UAM (urban air mobility), the strategic demand for low-altitude aircraft is showing a significant upward trend. eVTOL (electric vertical take-off and landing aircraft), as the core technology carrier of future urban transportation, usually adopts a distributed electric propulsion system configuration, that is, six or more motors and their controllers as power sources. When one or more of the motors and controllers fail, the other motors and controllers need to provide sufficient power compensation instantaneously to maintain the flight attitude of the eVTOL.

[0003] However, if the power compensation algorithm is too complex and the response speed is too slow, other motors and controllers will not be able to provide sufficient power compensation in time. If the power compensation algorithm is too rigid and not flexible enough, it is easy to have large calculation deviations. For example, over-compensation may cause the instantaneous thrust to exceed the motor limit, which may lead to the overall failure of eVTOL.

[0004] Ensuring both timely and accurate power compensation has become a pressing technical challenge. Summary of the Invention

[0005] Based on this, embodiments of this application provide a fault-tolerant control method and system for aircraft to solve the technical problem that the timeliness and accuracy of power compensation response cannot be simultaneously achieved.

[0006] In a first aspect, embodiments of this application provide an aircraft fault-tolerant control method, including:

[0007] Identify the faulty rotor;

[0008] Based on the cascaded control method of position outer loop and attitude inner loop, the thrust matrix of available rotors is reconstructed.

[0009] Based on the reconstructed thrust matrix, a constrained secondary thrust matrix reconstruction is performed to determine the target thrust matrix of the available rotor.

[0010] Based on the target thrust matrix, the motor output of each available rotor is smoothly adjusted to enter a fault-tolerant flight mode.

[0011] In one embodiment, a constrained secondary thrust matrix reconstruction is performed based on the initial thrust matrix reconstruction to determine the target thrust matrix for the usable rotor, including:

[0012] Based on the inequality constraints of yaw degradation and combined with the standard weighted L2 target, the target thrust matrix is ​​determined.

[0013] In one embodiment, the target thrust matrix is ​​determined based on the inequality constraints of yaw degradation and in conjunction with the standard weighted L2 target, including:

[0014] Using the OSQP solver for quadratic programming, and employing strictly convex quadratic programming, we find the unique global optimal solution for the standard weighted L2 norm objective.

[0015] The unique global optimal solution is used as the target thrust matrix.

[0016] In one embodiment, a constrained secondary thrust matrix reconstruction is performed based on the initial thrust matrix reconstruction to determine the target thrust matrix for the usable rotor, including:

[0017] Based on the equality constraint of constant lift, and combined with the standard weighted L2 target, the target thrust matrix is ​​determined.

[0018] In one embodiment, determining the faulty rotor includes:

[0019] Determine whether the rotor is faulty by comparing at least two of the following parameters: throttle command, propeller speed, motor current, and aircraft attitude residual.

[0020] In one embodiment, based on a cascaded control method of the outer position loop and the inner attitude loop, the thrust matrix of the available rotor is reconstructed, including:

[0021] The outer position loop determines the total lift and desired attitude of the aircraft based on its position and velocity, and outputs the total lift and desired attitude as control targets to the inner attitude loop.

[0022] The attitude inner loop is based on the desired attitude. The three-axis torque corresponding to the desired attitude is analyzed by a proportional-integral-derivative PID controller and used as the reconstructed thrust matrix.

[0023] In one embodiment, the method further includes:

[0024] If the motor of the faulty rotor is a dual three-phase motor or a multi-phase motor, set the thrust of the faulty rotor to at least half of the thrust before the fault, and treat the faulty rotor as a usable rotor.

[0025] In one embodiment, the method further includes:

[0026] If the motor of the faulty rotor is a single-phase AC motor or a standard three-phase AC motor, set the thrust of the faulty rotor to zero and downgrade the faulty rotor to an unusable rotor.

[0027] Thrust matrix reconstruction for available rotors includes:

[0028] Degrading the thrust matrix involves removing unusable rotor channels.

[0029] Secondly, embodiments of this application also provide an aircraft fault-tolerant control device, including: a processor, which executes executable code to implement the above-described method.

[0030] Thirdly, embodiments of this application also provide an aircraft, including: multiple rotors, a memory, and a controller;

[0031] The memory stores executable code, and the controller executes the executable code in the memory to achieve the following:

[0032] Identify the faulty rotor from multiple rotors;

[0033] Based on the cascaded control method of position outer loop and attitude inner loop, the thrust matrix of available rotors is reconstructed.

[0034] Determining the target thrust matrix of available rotors based on constrained quadratic thrust matrix reconstruction;

[0035] Based on the target thrust matrix, the motor output of each available rotor is smoothly adjusted to enter a fault-tolerant flight mode.

[0036] This application provides a fault-tolerant control method for aircraft. After a rotor failure, the method reconstructs the thrust matrix based on position and attitude, and then reconstructs it based on a constrained secondary thrust matrix to determine the target thrust matrix of the usable rotor. Therefore, it can more accurately compensate for the thrust loss and attitude deviation torque after the rotor failure. Furthermore, the constrained secondary thrust matrix reconstruction is fast, has low requirements for chip computing power, facilitates algorithm portability, and effectively reduces costs. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of this application or the conventional technology, the drawings used in the description of the embodiments or the conventional technology will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 A schematic flowchart of the aircraft fault-tolerant control method of the first embodiment provided in this application;

[0039] Figure 2 A schematic flowchart of the aircraft fault-tolerant control method of the second embodiment provided in this application;

[0040] Figure 3 This is a schematic diagram of the coordinate system for a six-rotor aircraft provided in an embodiment of this application.

[0041] Figure 4 This is a schematic diagram of the cascaded control method based on the outer position loop and the inner attitude loop provided in the embodiments of this application. Detailed Implementation

[0042] To facilitate understanding of this application, it will be described in more detail below. However, it should be understood that this application can be implemented in many different forms and is not limited to the embodiments or examples described herein. Rather, these embodiments or examples are provided to provide a more thorough and complete understanding of the disclosure of this application.

[0043] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein in the specification of this application is for the purpose of describing particular implementations or embodiments only and is not intended to be limiting of this application. The optional scope of the term "and / or" as used herein includes any one of two or more of the related listed items, as well as any and all combinations of the related listed items, including any two related listed items, any more related listed items, or a combination of all related listed items.

[0044] Regarding the common malfunctions of aircraft rotors, see [link / reference]. Figure 1 As shown, the first embodiment of this application provides a fault-tolerant control method for an aircraft, including:

[0045] S101: Identify the faulty rotor.

[0046] S102: Based on the cascaded control method of position outer loop and attitude inner loop, the thrust matrix of the available rotor is reconstructed.

[0047] S103: Based on the thrust matrix reconstruction, perform a constrained secondary thrust matrix reconstruction to determine the target thrust matrix of the available rotor.

[0048] In S102, the position outer loop and attitude inner loop can be control circuits inside the aircraft or software modules in the aircraft's controller.

[0049] For example, the outer position loop can be used to: receive the desired position, calculate the required acceleration of the aircraft based on the difference between the current position and the desired position, as well as the difference between the current velocity and the target velocity; the outer position loop can perform attitude calculation together with the current yaw angle and other information and the acceleration, and output the attitude calculation results to the inner attitude loop as the control target, which is responsible for ensuring the convergence of the flight trajectory.

[0050] The current yaw angle can be provided by the aircraft's built-in attitude sensor, which can include, but is not limited to, at least one of the following sensors: gyroscope, inertial accelerometer, GPS, etc. For example, the yaw angle can be obtained through a gyroscope.

[0051] The attitude inner loop can be used to: receive attitude calculation results from the position outer loop, quickly calculate the desired torque and desired lift corresponding to the attitude calculation results using a PID (Proportional-Integral-Derivative) controller, and reconstruct the initial thrust matrix based on the desired torque and desired lift. The attitude inner loop can have two PID controllers: one as an angle controller to calculate the desired torque, and the other as an angular velocity controller to calculate the desired lift.

[0052] The above-mentioned reconstruction of the secondary thrust matrix can be achieved, but is not limited to, using the constrained quadratic programming (QP) method.

[0053] In an optional embodiment of this application, the above control method can be compatible with both dual three-phase motors and multi-phase motors, that is, the motor of the faulty rotor can be a dual three-phase motor or a multi-phase motor. If the motor of the faulty rotor is a dual three-phase motor or a multi-phase motor, the thrust of the faulty rotor is set to at least half of the thrust before the fault, and the faulty rotor is treated as a usable rotor.

[0054] If the motor of the faulty rotor is a single-phase AC motor or a standard three-phase AC motor, set the thrust of the faulty rotor to zero and downgrade the faulty rotor to an unusable rotor.

[0055] During the process of thrust matrix reconstruction or secondary thrust matrix reconstruction, channels of unusable rotors can be deleted from the thrust matrix to achieve thrust matrix downgrading, thereby reducing the amount of computation.

[0056] The constraints in the above-mentioned constrained secondary thrust matrix reconstruction can be inequality constraints for yaw degradation and / or equality constraints for lift invariance. This secondary thrust matrix reconstruction can be based on the standard weighted L2 norm objective.

[0057] See Figure 2 As shown, a second embodiment of this application provides a fault-tolerant control method for an aircraft, comprising:

[0058] S201: Based on at least one of the following parameters, such as throttle command, actual rotor speed, motor current, and attitude residual, perform fault detection to determine whether a rotor fault exists.

[0059] S202: If a rotor malfunction exists, lock the malfunctioning rotor.

[0060] For example, the drive PWM of the faulty rotor can be directly idled to 0 and not participate in channel allocation.

[0061] S203: Control the remaining rotor matrix to reconstruct and reduce the dimensionality of the allocation matrix.

[0062] S204: Adjust the attitude loop gain via the angle controller and / or angular velocity controller.

[0063] For example, both angle controllers and speed controllers can use PID controllers. By using PID parameters, the proportional gain, integral gain, and derivative gain can be adjusted sequentially to achieve angle control and angular velocity control.

[0064] The proportional gain adjustment can be done by gradually increasing the P value from the current value, making small adjustments at each step, until the drone's attitude shows slight oscillations, then lowering the P value to 50% of the oscillation value to avoid instability.

[0065] Integral gain tuning can be done by: after determining a stable P value through proportional gain tuning, gradually increasing the I value until the attitude steady-state deviation is quickly eliminated. The upper limit of the I value is controlled to prevent integral saturation and avoid control delay.

[0066] Differential gain adjustment can be done by gradually increasing the D value after the P and I values ​​are properly adjusted, until there is no overshoot or reciprocating oscillation after the attitude control is activated, thus avoiding high-frequency jitter.

[0067] S205: The remaining rotor thrust is calculated using the constrained quadratic programming (QP) method.

[0068] S206: Based on motor saturation limiting and combined with low-pass filtering anti-oscillation processing, the thrust smooth increment is determined.

[0069] S207: Entering fault-tolerant flight mode.

[0070] Entering fault-tolerant flight mode can include at least one of the following operations: fault protection settings, sensor fault tolerance mode activation, and power fault tolerance adaptation.

[0071] Fault protection settings can include: setting fault-tolerant actions for scenarios such as loss of connection and low battery in the flight control ground station.

[0072] For example, the COM_OF_LOSS_T parameter can be configured to set the trigger delay after the offline link is interrupted, and the COM_OBL_RC_ACT parameter can be used to select fault-tolerant actions such as return to home / fixed point / landing after the connection is lost.

[0073] The sensor fault tolerance mode can be enabled by: enabling GNSS failure dead reckoning mode, configuring the NAV_TRAFF_AVOID parameter to enable traffic avoidance fault tolerance, and relying on backup sensors to maintain stable flight for a short period of time when GPS signal is lost.

[0074] The power fault tolerance adaptation settings can be as follows: For available rotors, set a millisecond-level motor fault residual judgment threshold. If any motor is confirmed to have failed according to the residual judgment threshold, automatically lock the fault channel corresponding to the motor, reconstruct the power distribution matrix of the remaining motors, and maintain attitude stability.

[0075] During the process of reconstructing the dimension reduction allocation matrix by controlling the remaining rotor matrix, the thrust vector of the failed rotor channel can be set to zero.

[0076] For example, the thrust vector of a failed rotor channel can be set to zero by reducing the dimensionality of the distribution matrix.

[0077] Assume the normal thrust distribution matrix is ​​as follows Formula (1)

[0078] Represents the force helical vector

[0079] in Indicates torque, This represents the three-axis torque, where u represents the thrust command for n rotors. , where n represents the total number of all rotors, and A represents the standard control allocation matrix.

[0080] If the kth rotor fails, then its thrust T is reduced. k Set it to zero and split it into the failure channel, as follows:

[0081] Formula (2)

[0082] in, The thrust vector representing the healthy rotor channel is a decision variable in the constrained quadratic programming (QP) calculation. This represents the thrust vector of the failed rotor channel.

[0083] In the constrained quadratic programming (QP) of the optional embodiment of this application, the desired total thrust can be calculated based on the current aircraft weight and acceleration, and combined with the current aircraft attitude, the desired total thrust can be decomposed into desired triaxial torques, and the desired total thrust and desired triaxial torques can be combined into a desired force spiral. .

[0084] In establishing a standard weighted L2 norm objective based on the expected force spiral, the goal is to minimize the deviation of the rotor thrust from its nominal value, i.e., to minimize the deviation of the aircraft's Z-axis, thereby avoiding sudden drops caused by abrupt thrust changes. The standard weighted L2 norm objective can be expressed, but is not limited to, by the following formula:

[0085] Formula (3)

[0086] Where J represents f(u) h Regarding u h The function, Indicates u h As a decision optimization variable, to minimize J This represents the thrust vector available in the rotor channel. This represents the lift distribution matrix for the available rotor channels. Represents the spiral of expectation force. Indicates the force screw tracking error. This indicates the reference value of the nominal thrust of the healthy rotor before the failure. This represents the force screw tracking error, and Q represents the weight. The solution process can be as follows:

[0087] Step 1, let the optimization variable x = Expanding formula (3) yields formula (4):

[0088] Formula (4)

[0089] in, This represents the thrust vector available in the rotor channel. This represents the lift distribution matrix for the available rotor channels, where Q represents the weights. Represents the spiral of expectation force. This represents the thrust smoothing matrix, whose input controls (typically including total thrust, roll moment, pitch moment, and yaw moment). This indicates the reference value of the nominal thrust of the healthy rotor before the failure.

[0090] Step 2, rearrange formula (4) to obtain formula (5) with a constant term:

[0091] Formula (5)

[0092] After discarding the constant term, we get:

[0093] Formula (6)

[0094] By comparing the formula (6) that discards constant terms with the standard form of quadratic programming. We can obtain:

[0095] Formula (7)

[0096] Where Q represents the weight, g represents an n-dimensional column vector, which is the coefficient vector of the linear term of the objective function and defines the linear weight of the optimization variable x; H represents an n-order real symmetric Hessian matrix, which is the coefficient matrix of the quadratic term of the objective function and determines the convexity of the quadratic programming: if H is positive semi-definite, it is a convex quadratic programming; if H is positive definite, it is a strictly convex quadratic programming, in which case there exists a globally unique optimal solution.

[0097] Step 3: Combining the upper and lower thrust limits of each available rotor, add inequality constraints to formula (7): Its matrix form is .

[0098] Because the aircraft's degrees of freedom decrease after some rotors fail, the weights of each available rotor must be set accordingly. When this happens, you can force priority tracking. , lower Weights are used to implement yaw degradation. The weights are set to... This allows us to prioritize lift, which in turn ensures the roll angle. Pitch angle It automatically reduces yaw rate. It can be used to quantify the penalty for the deviation of the current thrust from the steady-state thrust at the previous moment, thereby suppressing thrust jumps. For example, the larger the R term, the stronger the penalty, and the thrust tends to be closer to the historical value; the smaller the R term, the smaller the thrust fluctuations are allowed. Taking R = I (identity matrix) applies a uniform constraint to all rotors. Without the R term, the above formula only cares about the torque tracking term (weight Q). Solving QP will result in optimal torque, but the thrust will jump significantly, thus affecting the stability of the aircraft.

[0099] Step 4: Using H as positive definite, strictly convex quadratic programming, find the globally unique optimal solution to obtain x.

[0100] The solver used in the above solution process can be, but is not limited to, QSQP (Operator Splitting Quadratic Programming), ECOS (Embedded Cone Optimizer), qpOASES (Quadratic Programming Online Active Set Evolution Strategy), etc.

[0101] In an optional embodiment of this application, a lighter OSQP (Operator Splitting Quadratic Program) can be used, with a single-step execution time of 0.2~1ms on the MCU / DSP, which meets the flight control control cycle of 50~200HZ.

[0102] In step four, equality constraints can also be added to force the lift before and after the fault. To maintain consistency, the objective function can now be optimized as follows:

[0103] Formula (8)

[0104] The weight matrix is ​​adjusted to a third-order matrix. Take submatrix A τ =A h (2:4), then we get:

[0105]

[0106] in Indicates u h As a decision optimization variable, to minimize J This represents the thrust vector available in the rotor channel. This represents the lift distribution matrix for the available rotor channels. This represents the three-axis moment matrix of the available rotor channels, where Q represents the weights. Represents the spiral of expectation force. This represents the thrust smoothing matrix, whose input controls (typically including total thrust, roll moment, pitch moment, and yaw moment). This indicates the reference value of the nominal thrust of the healthy rotor before the failure. Indicates roll angle The weight, Indicates pitch angle The weight, This indicates the weight of the yaw angle.

[0107] The upper and lower thrust limits of each usable rotor depend on the output capacity limits of the motor and controller and the propeller limits of that healthy rotor.

[0108] In the above embodiments, for dual three-phase motors or multi-phase motors (4-phase, 5-phase, 12-phase), assuming the thrust of the failed rotor is not zeroed, and the power drops to half its original value after the failure of the k-th rotor section, the calculation can continue only by modifying the thrust constraint. .

[0109] This application provides a hexagonal eVTOL with a regular hexagonal rotor distribution, the rotor distribution of which is as follows: Figure 3As shown, in the body coordinate system (x, y, z) of the six-rotor eVTOL, x represents the forward direction of the nose, y represents the rightward direction of the fuselage, and z represents the vertical upward direction. The azimuth angles of each rotor relative to the x-axis are as follows: In the dimensionless case, assume the arm length Thrust coefficient Torque coefficient Its standard allocation matrix is:

[0110]

[0111] Assuming rotor 1 fails, its thrust is set to zero, and the first column of the standard allocation matrix is ​​deleted, resulting in the allocation matrix for the usable rotors:

[0112] For example, in hovering conditions (a dimensionless example; in practical applications, the unit of lift can be Newtons), the total lift of the six-rotor system offsets the weight of the six-rotor eVTOL:

[0113]

[0114] The total lift is 6, which offsets the weight of the six-rotor eVTOL.

[0115] Assuming the hexacoach eVTOL is in a perfectly horizontal attitude, that is:

[0116]

[0117] After a malfunction, yaw control is downgraded, and the yaw angle is no longer forcibly controlled.

[0118]

[0119] The expectation force spiral is:

[0120]

[0121] Assuming the thrust of a single rotor ranges from 0.5 to 2, the weighting matrix is ​​set as follows:

[0122]

[0123] Among them, lift > roll angle / pitch angle > yaw angle.

[0124]

[0125] R represents the thrust smoothing matrix, whose input controls (lift, roll moment, pitch moment, and yaw moment).

[0126] Assuming the nominal thrust of a healthy rotor, i.e., the thrust per path during normal hovering, is 1:

[0127]

[0128] After calling the QSOP solver, the optimal thrust for the available rotor channels is obtained as follows:

[0129]

[0130] Assuming the six-rotor eVTOL uses a dual three-phase motor, and the faulty rotor fails with a thrust range of 0 to 0.5 N, while other parameters remain unchanged, the thrust of the faulty rotor can be maximized to 0.5 N. Then, the QSOP solver is called to allocate five usable rotors, yielding the optimal thrust for each of the six rotor channels:

[0131]

[0132] After determining the optimal thrust for each rotor channel, the motor corresponding to each rotor channel can be saturated and limited. The thrust can be smoothly increased based on low-pass filtering to prevent oscillation and avoid causing fuselage vibration.

[0133] For example, a first-order low-pass filter algorithm can be used to smoothly increase thrust:

[0134] Let the filter coefficients a∈(0,1), T(n) represent the currently calculated thrust value, T(n-1) represent the historical thrust value, and T represent the output value:

[0135] T = a×T(n) + (1-a)×T(n-1) Formula (9)

[0136] With the sampling frequency remaining constant, the smaller 'a' is, the lower the cutoff frequency, the smoother the filtering result, and the more smoothly the thrust increases.

[0137] This application also provides an aircraft fault-tolerant control device, including: a processor, which executes executable code to implement the above control method.

[0138] This application also provides an aircraft, including: multiple rotors, a memory, and a controller;

[0139] The memory stores executable code, and the controller executes the executable code in the memory to achieve the following:

[0140] Identify the faulty rotor from multiple rotors;

[0141] Based on the cascaded control method of position outer loop and attitude inner loop, the thrust matrix of available rotors is reconstructed.

[0142] Determining the target thrust matrix of available rotors based on constrained quadratic thrust matrix reconstruction;

[0143] Based on the target thrust matrix, the motor output of each available rotor is smoothly adjusted to enter a fault-tolerant flight mode.

[0144] See Figure 4 As shown, the cascaded control method based on the outer position loop and the inner attitude loop can include:

[0145] Each available rotor's motor drive circuit may include an outer position loop and an inner attitude loop.

[0146] The aircraft's sensors can send the detected position information as the current position to the outer position loop.

[0147] The position outer loop can include a position controller and a velocity controller, wherein the position controller is used to determine the desired position and desired velocity. The position outer loop can calculate the total lift required by the aircraft and the desired attitude based on the difference between the current position and the desired position, and the difference between the current velocity and the target velocity.

[0148] For example, the desired position can be obtained from the input command. After the outer position loop receives the desired position, it outputs the desired velocity through the position controller (e.g., proportional controller P), and then enters the velocity controller (e.g., proportional-integral controller PI) to output the desired acceleration. Then, combined with the desired acceleration and yaw angle, the desired attitude is calculated through the attitude solution matrix. The outer position loop sends the desired attitude to the inner attitude loop.

[0149] The aircraft's sensors send current attitude information to the attitude inner loop, which then outputs the desired torque and lift through the angle controller and angular velocity controller. This smoothly adjusts the motor's output torque and lift to match the desired torque and lift. The angle controller and angular velocity controller can be, but are not limited to, PID controllers.

[0150] Fault-tolerant flight mode refers to a working mode in which the flight control system of an aircraft can maintain safe flight or smooth degraded landing through redundant design, fault diagnosis and control reconfiguration in the event of component failure, actuator failure or sensor abnormality.

[0151] Once in fault-tolerant flight mode, the aircraft can continue to monitor rotor health, report fault location and severity, and trigger a control reconfiguration strategy if a new rotor fault is added.

[0152] In the embodiments of this application, the control reconstruction strategy may include, but is not limited to, at least one of the following: reconstruction of the dimension reduction allocation matrix, thrust matrix reconstruction, and constrained quadratic thrust matrix reconstruction.

[0153] If the FDI module determines that a new faulty rotor has been added and initiates the control reconfiguration strategy, then the reconfiguration of the dimension reduction allocation matrix and the reconfiguration of the constrained secondary thrust matrix will be performed when the rotor motor of the aircraft is a single motor.

[0154] If the rotor motor of the aircraft is a multiphase motor, and the FDI module determines that a new faulty rotor has been added and initiates the control reconfiguration strategy, then the thrust of the newly added faulty rotor will be reduced according to the condition of its faulty motor, and thrust matrix reconfiguration and constrained secondary thrust matrix reconfiguration will be performed.

[0155] For example, if the motor of the newly added faulty rotor is a dual three-phase motor or a multi-phase motor, the thrust of the faulty rotor is set to at least half of the thrust before the fault, and the faulty rotor is treated as a usable rotor, and the thrust matrix is ​​reconfigured.

[0156] For example, if the motor of the newly added faulty rotor is a five-phase permanent magnet synchronous motor, the thrust of the faulty rotor is set to at least four-fifths of the thrust before the fault, and the faulty rotor is treated as a usable rotor, and thrust matrix reconstruction and constrained secondary thrust matrix reconstruction are performed.

[0157] For example, if the motor of the newly added faulty rotor is a nine-phase symmetrical winding motor, the thrust of the faulty rotor is set to at least 82% of the thrust before the fault, and the faulty rotor is treated as a usable rotor, and thrust matrix reconstruction and constrained secondary thrust matrix reconstruction are performed.

[0158] Treating a faulty rotor as a usable rotor and performing thrust matrix reconstruction can be achieved by sequentially adjusting the proportional gain, integral gain, and derivative gain to realize angle control and angular velocity control.

[0159] Constrained reconstruction of the secondary thrust matrix may include minimizing the deviation of rotor thrust from the nominal value during the process of establishing a standard weighted L2 norm target based on the desired force spiral.

[0160] This application provides a fault-tolerant control method for aircraft. After a rotor failure, the method reconstructs the thrust matrix based on position and attitude, and then reconstructs it based on a constrained secondary thrust matrix to determine the target thrust matrix of the usable rotor. Therefore, it can more accurately compensate for the thrust loss and attitude deviation torque after the rotor failure. Furthermore, the constrained secondary thrust matrix reconstruction is fast, has low requirements for chip computing power, facilitates algorithm portability, and effectively reduces costs.

[0161] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0162] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application.

Claims

1. A fault-tolerant control method for aircraft, characterized in that, include: Identify the faulty rotor; Based on the cascaded control method of position outer loop and attitude inner loop, the thrust matrix of available rotors is reconstructed. Based on the aforementioned thrust matrix reconstruction, a constrained secondary thrust matrix reconstruction is performed to determine the target thrust matrix of the available rotor. Based on the target thrust matrix, the motor output of each available rotor is smoothly adjusted to enter a fault-tolerant flight mode.

2. The control method according to claim 1, characterized in that, Based on the aforementioned thrust matrix reconstruction, a constrained secondary thrust matrix reconstruction is performed to determine the target thrust matrix of the available rotor, including: Based on the inequality constraints of yaw degradation and combined with the standard weighted L2 target, the target thrust matrix is ​​determined.

3. The control method according to claim 2, characterized in that, Based on the inequality constraints of yaw degradation and combined with the standard weighted L2 target, the target thrust matrix is ​​determined, including: Using the OSQP solver for quadratic programming, and employing strictly convex quadratic programming, we find the unique global optimal solution for the standard weighted L2 norm objective. The unique global optimal solution is used as the target thrust matrix.

4. The control method according to claim 1, characterized in that, Based on the aforementioned thrust matrix reconstruction, a constrained secondary thrust matrix reconstruction is performed to determine the target thrust matrix of the available rotor, including: Based on the equality constraint of constant lift, and combined with the standard weighted L2 target, the target thrust matrix is ​​determined.

5. The control method according to claim 1, characterized in that, The determination of the faulty rotor includes: Determine whether the rotor is faulty by comparing at least two of the following parameters: throttle command, propeller speed, motor current, and aircraft attitude residual.

6. The control method according to claim 1, characterized in that, The cascaded control method based on the outer position loop and the inner attitude loop reconstructs the thrust matrix of the available rotors, including: The outer position loop determines the total lift and desired attitude required by the aircraft based on the aircraft's position and velocity, and outputs the total lift and desired attitude as control targets to the inner attitude loop. The attitude inner loop, based on the desired attitude, uses a proportional-integral-derivative PID controller to analyze the three-axis torque corresponding to the desired attitude, which is then used as the reconstructed thrust matrix.

7. The control method according to claim 1, characterized in that, The method further includes: If the motor of the faulty rotor is a dual three-phase motor or a multi-phase motor, the thrust of the faulty rotor is set to at least half of the thrust before the fault, and the faulty rotor is treated as a usable rotor.

8. The control method according to claim 1, characterized in that, The method further includes: If the motor of the faulty rotor is a single-phase AC motor or a standard three-phase AC motor, the thrust of the faulty rotor is set to zero, and the faulty rotor is downgraded to an unusable rotor. The thrust matrix reconstruction of the available rotors includes: The thrust matrix is ​​downgraded by removing the channels of the unusable rotors.

9. A fault-tolerant control device for aircraft, characterized in that, include: processor, The processor executes executable code to implement the method of any one of claims 1 to 8.

10. An aircraft, characterized in that, include: Multiple rotors, memory, and controller; The memory stores executable code, and the controller executes the executable code in the memory to achieve the following: Identify the faulty rotor from the plurality of rotors; Based on the cascaded control method of position outer loop and attitude inner loop, the thrust matrix of available rotors is reconstructed. The target thrust matrix of the available rotor is determined based on the constrained quadratic thrust matrix reconstruction. Based on the target thrust matrix, the motor output of each available rotor is smoothly adjusted to enter a fault-tolerant flight mode.