A quadrotor unmanned aerial vehicle fault-tolerant control method and system considering random disturbance

Abstract

The application provides a quad-rotor unmanned aerial vehicle fault-tolerant control method and system considering random disturbance, and belongs to the field of unmanned aerial vehicle flight control. The method comprises the following steps: a random control system model is established by fusing wiener process random disturbance, actuator failure and input saturation, a six-degree-of-freedom control problem is decoupled into position and attitude tracking sub-problems and the expected attitude angle is inversely solved; a mixed constraint strategy fusing a delay state conversion function and a nonlinear state-dependent mapping is adopted to apply constraints to state variables; based on an adaptive backstepping control framework, a controller is designed for six-degree-of-freedom sub-channels of the quad-rotor unmanned aerial vehicle, so that the final input control signals of the degree-of-freedom sub-channels are obtained; a comprehensive Lyapunov function covering six degrees of freedom of the unmanned aerial vehicle is constructed, and it is proved that all signals of the closed-loop system are ultimately uniformly bounded. The application can synchronously cope with multiple complex factors, improve flight stability and reliability of the unmanned aerial vehicle in a complex environment, and is suitable for quad-rotor unmanned aerial vehicle control in multiple scenes.

Description

Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) flight control technology, and particularly relates to a fault-tolerant control method and system for quadrotor UAVs that takes into account random disturbances. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Quadrotor drones, with their vertical takeoff and landing capabilities, maneuverability, and low cost, have been widely used in disaster relief, agricultural plant protection, logistics distribution, and industrial inspection, becoming an important tool for operations in complex scenarios. However, the system disturbances and inherent malfunction risks they face during actual flight pose serious challenges to control stability and reliability, and existing technologies still have many limitations.

[0004] From the perspective of system disturbances, UAVs are susceptible to factors such as motor operating noise, rotor airflow disturbances, and structural vibrations during flight. These disturbances often exhibit continuous random characteristics, making them difficult to accurately characterize using deterministic models. Traditional control systems are mostly designed based on linear or deterministic nonlinear models under ideal operating conditions, which cannot effectively adapt to the dynamic changes of random disturbances. This leads to decreased attitude tracking accuracy, sluggish dynamic response, and even flight attitude instability in complex environments.

[0005] Regarding system-specific faults, actuators, as core components for UAV attitude and position control, are prone to partial failure or bias faults under long-term high-load operation, directly weakening control efficiency. Simultaneously, due to physical limitations such as motor power and mechanical structure, control inputs inevitably face saturation constraints; excessive manipulation can easily lead to actuator overload damage, further deteriorating control performance. Existing fault-tolerant control methods mostly address actuator faults or external disturbances individually, lacking the ability to coordinate multiple factors such as "random disturbance-actuator fault-input saturation," often resulting in control strategy failure under coupled influences.

[0006] In addition, most existing state constraint control methods impose strict restrictions on the initial state of the system, requiring that the initial attitude, position and other state variables must be strictly within the preset constraint boundaries, otherwise the controller will not be able to start normally; and some advanced control strategies introduce a large number of parameters to be estimated online in order to cope with uncertainties, which leads to a surge in computational complexity, making it difficult to adapt to the limited hardware computing power of small and medium-sized UAVs, which greatly limits their promotion and application in engineering practice. Summary of the Invention

[0007] To overcome the shortcomings of the prior art, the present invention provides a fault-tolerant control method and system for quadrotor UAVs that considers random disturbances, aiming to solve the problem of degraded control performance of quadrotor UAVs caused by uncertain disturbances and faults.

[0008] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:

[0009] The first aspect of this invention provides a fault-tolerant control method for a quadcopter unmanned aerial vehicle that takes into account random disturbances;

[0010] A fault-tolerant control method for a quadcopter unmanned aerial vehicle (UAV) considering random disturbances includes:

[0011] A pose model of a quadrotor UAV, modified by an inherent stochastic process, is established, and actuator faults and saturation constraints are integrated into the same framework to obtain a stochastic control system model of the quadrotor UAV.

[0012] Based on the stochastic control system model of the quadcopter UAV, the desired roll angle and pitch angle are obtained by decoupling the control variables of the UAV's six degrees of freedom system.

[0013] The error state equation of the system is constructed. To meet the full-state constraint requirements of the UAV, a hybrid constraint strategy that integrates delayed state transition function and nonlinear state dependency mapping is adopted to apply constraints to the state variables in different time periods.

[0014] Based on the adaptive backstepping control framework, controllers are designed for the six degrees of freedom sub-channels of the quadrotor UAV to obtain the final input control signals for the corresponding degrees of freedom sub-channels.

[0015] We construct a comprehensive Lyapunov function covering all six degrees of freedom of the UAV, and using stochastic system theory, prove that all signals in the closed-loop system are eventually uniformly bounded.

[0016] As a further technical solution, a quadrotor UAV pose model modified by an inherent stochastic process is established, and actuator faults and saturation constraints are integrated into the same framework to obtain a stochastic control system model for the quadrotor UAV, including:

[0017] A stochastic differential equation model is obtained by modifying the classical quadrotor UAV dynamics model with stochastic perturbation based on Wiener process.

[0018] The actuator fault model and saturation constraints are integrated into the stochastic differential equation model, and the actual control input is obtained by combining them. The stochastic control system model of the quadcopter UAV is then reconstructed.

[0019] The stochastic control system model of the quadcopter UAV is as follows:

[0020]

[0021] in, And respectively represent the roll angle, pitch angle, yaw angle, and the UAV's position in the inertial frame. Direction, that is: , , , , , ; express Sub-channel pose dimension state quantity. express Sub-channel velocity dimension state quantity, express The random perturbation intensity function of the sub-channel pose dimension. express The random perturbation intensity function of the sub-channel velocity dimension. express Sub-channel control input, express The system state nonlinear function and unknown uncertainties are input in the sub-channel. express Unknown control coefficients of sub-channels caused by faults express The subchannel is affected by unknown bias and external disturbances caused by faults. express The output signal of the sub-channel.

[0022] As a further technical solution, the desired roll angle and pitch angle are:

[0023]

[0024]

[0025] in, For the desired trajectory of the roll angle, For the desired trajectory at the pitch angle, The desired trajectory is the yaw angle. After reconstruction Virtual control input for the position sub-channel.

[0026] As a further technical solution, the delayed state transition function for:

[0027]

[0028] in, For the system's state dimension, The set value for the constraint start time.

[0029] The nonlinear state dependency mapping is:

[0030]

[0031]

[0032] in, , They are respectively The subchannel pose and velocity variables after state mapping. for Sub-channel pose tracking error, express Sub-channel velocity dimension state quantity, They are set respectively and Constrain upper and lower bounds.

[0033] As a further technical solution, based on the adaptive backstepping control framework, controllers are designed for each of the six degrees of freedom sub-channels of the quadcopter UAV, obtaining the final input control signals for the corresponding degree of freedom sub-channels, including:

[0034] definition Sub-channel error variables, and Lyapunov functions designed using the backstepping method;

[0035] Calculate the infinitesimal differential operator based on Itoh's stochastic differential formula;

[0036] The system's unknowns and random disturbances are approximated using a radial basis function neural network, and the disturbances, errors, and upper bounds of the neural network weights are uniformly estimated using a lumped parameter adaptive law.

[0037] The Nussbaum gain technique is applied to address the problem of unknown control gain caused by actuator failure and saturation, thus obtaining the final input control signal of the system.

[0038] As a further technical solution, the aforementioned The sub-channel error variable is defined as:

[0039]

[0040]

[0041] in, Designed for subsequent steps Virtual control rate of sub-channels .

[0042] The The sub-channel adaptive law and the final input control signal are:

[0043]

[0044]

[0045]

[0046]

[0047] in, The selected Nussbaum function, For the design of positive constants, for For variables The first-order partial derivative, , express The maximum values ​​of the subchannel neural network weight norm, error norm, and external disturbance. Indicates to The estimate.

[0048] As a further technical solution, the comprehensive Lyapunov function covering all six degrees of freedom of the UAV is as follows:

[0049]

[0050] in, for The Lyapunov function constructed by the sub-channels, They are respectively Sub-channel Error variables defined by pose and velocity dimensions They are respectively The sub-channel pose and velocity dimensions are adaptively estimated.

[0051] A second aspect of the present invention provides a fault-tolerant control system for a quadcopter unmanned aerial vehicle that takes into account random disturbances.

[0052] A fault-tolerant control system for a quadcopter unmanned aerial vehicle (UAV) considering random disturbances, comprising:

[0053] The stochastic control system model building module is configured to: establish a quadrotor UAV pose model modified by an inherent stochastic process, and integrate actuator faults and saturation constraints into the same framework to obtain a stochastic control system model of the quadrotor UAV.

[0054] The roll and pitch angle calculation module is configured to: based on the stochastic control system model of the quadcopter UAV, obtain the desired roll and pitch angles by decoupling the control variables of the UAV's six degrees of freedom system;

[0055] The state variable constraint addition module is configured to: construct the error state equation of the system, and, in response to the full state constraint requirements of the UAV, adopt a hybrid constraint strategy that integrates delayed state transition function and nonlinear state dependency mapping to apply constraints to state variables in different time periods.

[0056] The controller design module is configured to: design controllers for the six degrees of freedom sub-channels of the quadcopter UAV based on the adaptive backstepping control framework, and obtain the final input control signals for the corresponding degrees of freedom sub-channels;

[0057] The stability proof module is configured to: construct a comprehensive Lyapunov function covering all six degrees of freedom of the UAV, and, in conjunction with stochastic system theory, prove that all signals in the closed-loop system are eventually uniformly bounded.

[0058] A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon that, when executed by a processor, implements the steps of a fault-tolerant control method for a quadcopter unmanned aerial vehicle considering random disturbances as described in the first aspect of the present invention.

[0059] A fourth aspect of the present invention provides an electronic device including a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of a fault-tolerant control method for a quadcopter unmanned aerial vehicle considering random disturbances as described in the first aspect of the present invention.

[0060] The above one or more technical solutions have the following beneficial effects:

[0061] (1) This invention breaks through the limitations of traditional methods that only deal with disturbances or faults. For the first time, it integrates random disturbances (based on Wiener process characterization), actuator faults (partial failures and bias faults) and input saturation constraints into the same control framework. By reconstructing the system model and designing a robust controller, it achieves stable control under multiple adverse factors. It can effectively adapt to harsh and complex flight scenarios such as military reconnaissance and disaster relief, and significantly reduce flight risks caused by environmental and system faults.

[0062] (2) This invention employs a hybrid constraint strategy combining delayed state transition and nonlinear state-dependent mapping. State constraints are not applied initially, reserving a state adjustment window for the controller and avoiding controller failure caused by initial state out-of-bounds errors in traditional methods. Furthermore, the constraint boundaries can be dynamically adjusted based on the desired trajectory and real-time state, without strictly limiting initial conditions, thus better meeting the flexible requirements of UAV start-up, shutdown, and operation in practical engineering. By utilizing a radial basis function neural network to approximate the system's unknowns and random disturbances, and combining this with a lumped parameter adaptive law to uniformly estimate disturbances, errors, and the upper bound of network weights, there is no need to estimate each parameter individually, significantly reducing the number of variables to be calculated online. Simultaneously, it avoids complex multi-model or high-dimensional adaptive law designs, effectively reducing the computational burden and adapting to the limited hardware computing power of small and medium-sized UAVs, meeting real-time control requirements.

[0063] (3) By introducing Nussbaum gain technology, the unknown control gain problem caused by actuator failure and input saturation is accurately handled. Combining the six-degree-of-freedom integrated Lyapunov function and stochastic system theory, it is theoretically proven that all signals of the closed-loop system are eventually uniformly bounded, ensuring the accuracy of attitude and position tracking. In practical applications, it can effectively suppress the influence of random disturbances and maintain stable flight even when the actuator fails, significantly improving the reliability and robustness of UAV control.

[0064] 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

[0065] 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 an improper limitation of the invention.

[0066] Figure 1 This is a flowchart of the method in the first embodiment.

[0067] Figure 2 This is a schematic diagram of the three-dimensional flight trajectory of the quadcopter UAV in the first embodiment.

[0068] Figure 3 This is a schematic diagram of the actual running trajectory and reference trajectory of the XYZ axes in the first embodiment.

[0069] Figure 4 This is a schematic diagram of the actual and reference operating trajectories of the yaw angle, roll angle, and pitch angle in the first embodiment.

[0070] Figure 5 This is a schematic diagram of the system control input for the first embodiment.

[0071] Figure 6This is a system structure diagram of the second embodiment. Detailed Implementation

[0072] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, 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 invention pertains.

[0073] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0074] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0075] Example 1

[0076] This embodiment discloses a fault-tolerant control method for quadrotor UAVs considering random disturbances. First, a multi-factor integrated control model is constructed to decouple the control problem and solve the desired attitude angle. Then, the initial state boundary crossing problem is solved through a hybrid constraint strategy. Based on the adaptive backstepping method, a control signal is generated by combining radial basis neural networks, lumped parameter adaptive laws, and Nussbaum gain techniques. Finally, the boundedness of the system signal is proved by Lyapunov functions, which can improve the flight stability and reliability of UAVs in complex environments and is suitable for multiple application scenarios.

[0077] Specifically, such as Figure 1 As shown, a fault-tolerant control method for a quadcopter unmanned aerial vehicle (UAV) considering random disturbances includes:

[0078] Step S1: Establish a quadrotor UAV pose model modified by an inherent stochastic process, and integrate actuator faults and saturation constraints into the same framework to obtain a stochastic control system model of the quadrotor UAV.

[0079] Step S11: Using a stochastic perturbation based on the Wiener process, the classic quadcopter UAV dynamics model is modified, and the following stochastic differential equation model is established:

[0080] (1)

[0081] (2)

[0082] (3)

[0083] (4)

[0084] (5)

[0085] (6)

[0086] in, = , , , , , These represent the roll angle, pitch angle, yaw angle, and the three-dimensional coordinates of the UAV's center of mass in the inertial frame, respectively. = , , , , , These represent the corresponding velocity components, These represent the total lift, roll moment, pitch moment, and yaw moment of the system, respectively. Represent the moment of inertia of each wing about the fuselage and along the direction of the fuselage, respectively. Moment of inertia in the direction, Indicates angular velocity, These represent the corresponding aerodynamic damping coefficients; and These represent the unknown and uncertain terms in the modeling process, respectively. This indicates the interference caused to the drone system by external environmental factors such as wind. This represents the intensity function of random perturbation. Indicates the quality of the drone. Represents gravitational acceleration, Wiener process It can be considered as a continuous random fluctuation caused by rotors, airframe, motors, circuits, and sensors.

[0087] Step S12: Further incorporate the input fault model and saturation constraints into the above stochastic differential equation model. The actuator fault model takes the following form:

[0088] (7)

[0089] in, For control input after actuator failure, For actual control input The unknown time-varying fault coefficient characterizes partial failure of the actuator. Or completely normal , The bias fault is unknown.

[0090] The saturation constraint takes the following form:

[0091] (8)

[0092] in, To account for the system input after saturation constraints, , The maximum and minimum values ​​are saturated inputs.

[0093] In this embodiment, the actual saturation characteristics are approximated by a smooth hyperbolic tangent function, and the actual input is represented as:

[0094] (9)

[0095] in, The smoothing function is... The approximate error is expressed in the following specific expressions: , .

[0096] Furthermore, according to the mean value theorem, linearizing the smoothing function near the zero point yields a saturation model of the following form:

[0097] (10)

[0098] in, for The derivative at a certain operating point is a time-varying gain.

[0099] In this way, the hard saturation nonlinearity of the actuator is transformed into the sum of a linear term with time-varying gain and a bounded disturbance term, laying the foundation for designing a robust controller with anti-saturation capability.

[0100] Combining the above actuator fault and saturation constraints, the actual control input acting on the system can be obtained as follows:

[0101] (11)

[0102] As can be seen from Equation (11), after introducing this fault model, the controller design faces two problems: first, the control gain changes in an unknown way, making it difficult to evaluate the control efficiency; second, the additional unknown bias will bring continuous error signals.

[0103] Step S13: To facilitate controller design, and considering the complex coupling relationship between the control input and system attitude dynamics in the above steps, the control input is redefined as follows:

[0104]

[0105] in: These represent the virtual control inputs for the six sub-channels. , ,

[0106] .

[0107] The random system of the quadcopter UAV is reconstructed to obtain the following form:

[0108] (12)

[0109] in, And respectively represent the roll angle, pitch angle, yaw angle, and the UAV's position in the inertial frame. Direction, that is: , , , , , ; express Sub-channel pose dimension state quantity. express Sub-channel velocity dimension state quantity, express The random perturbation intensity function of the sub-channel pose dimension. express The random perturbation intensity function of the sub-channel velocity dimension. express Sub-channel control input, express The system state nonlinear function and unknown uncertainties are input in the sub-channel. express Unknown control coefficients of sub-channels caused by faults express The subchannel is affected by unknown bias and external disturbances caused by faults. express The output signal of the sub-channel. Specifically:

[0110] ; ; ; = ;

[0111] = , , , , , ; =

[0112] .

[0113] The stochastic system model established in this step describes continuous random fluctuations through Wiener processes and incorporates them into the system dynamics model, laying a more realistic theoretical foundation for the subsequent design of robust controllers that can actively suppress such disturbances.

[0114] Step S2: Based on the stochastic control system model of the quadcopter UAV, the desired roll angle and pitch angle are obtained by decoupling the control variables of the UAV's six degrees of freedom system.

[0115] The complex six-degree-of-freedom control problem is decomposed into a relatively independent six-degree-of-freedom sub-channel tracking control problem. Based on the redefinition of the control input in step S1, the desired roll angle and pitch angle are solved inversely, as follows:

[0116] (13)

[0117] (14)

[0118] in, For the desired trajectory of the roll angle, For the desired trajectory at the pitch angle, The desired trajectory is the yaw angle. After reconstruction Virtual control input for the position sub-channel.

[0119] Step S3: Construct the error state equation of the system. To meet the full-state constraint requirements of the UAV, a hybrid constraint strategy that integrates delayed state transition function and nonlinear state dependency mapping is adopted to apply constraints to the state variables in different time periods.

[0120] The system tracking error is set as follows:

[0121] (15)

[0122] in, for Sub-channel pose tracking error, for The pose state of the sub-channel. for The desired trajectory of the sub-channel.

[0123] Set the following preset boundaries:

[0124] (16)

[0125] in, The upper and lower bounds of the constraints are set.

[0126] Traditional state constraint methods typically require the initial system state to be strictly within the constraint boundaries; otherwise, once the constraints are violated, the controller will malfunction. To address this issue, this embodiment employs a delayed constraint mechanism, where the initial state is within the constraint boundaries during the initial time period. Within this period, no state constraints are applied, providing the controller with an adjustment period to guide the state into the feasible region; from the preset time... To initiate strict full-state constraints, the above objective is achieved by combining a delayed state transition function with a nonlinear state dependency mapping.

[0127] The delayed state transition function used in this mechanism The specific format is as follows:

[0128] (17)

[0129] in, For the system's state dimension, The set value for the constraint start time.

[0130] Based on the above constraints and objectives, the following nonlinear state-dependent mapping is designed. Its mathematical properties strictly guarantee the equivalence of the boundedness of the mapped variables and the original variables satisfying the constraints.

[0131] (18)

[0132] (19)

[0133] in, , They are respectively The subchannel pose and velocity are variables after state mapping.

[0134] Step S4: Based on the adaptive backstepping control framework, design a controller for the stochastic control system model of the quadcopter UAV to obtain the final input control signal of the system.

[0135] Based on the backstepping control framework, a controller is designed for the reconstructed stochastic system model. The following section discusses the design of the controller for the first... We will design a controller using a single degree of freedom as an example.

[0136] Sub-channel error variables and The definition is as follows:

[0137] (20)

[0138] (twenty one)

[0139] in, Designed for subsequent steps Virtual control rate of sub-channels .

[0140] Differentiate equation (18):

[0141] (twenty two)

[0142] in, The derivative of the delayed state transition function. for Regarding time The first-order partial derivative, for For variables The first-order partial derivative, for For variables The second-order partial derivative is in the following form:

[0143]

[0144]

[0145]

[0146] Let the Lyapunov function be:

[0147]

[0148] By applying Itoh's formula, we can obtain the infinitesimal operator of the Lyapunov function. for:

[0149] (twenty three)

[0150] Using Young's inequality, the following scaling is performed:

[0151] (twenty four)

[0152] The introduction of random terms introduces numerous uncertainties into the system, significantly increasing the difficulty of analytical solutions. Therefore, in this embodiment, a radial basis function neural network with a Gaussian function is used to approximate these uncertainties. Combining formulas (23) and (24), the term to be approximated is selected as:

[0153] (25)

[0154] in, The input to the neural network is specifically: .

[0155] The output representation of the neural network is selected as follows:

[0156] (27)

[0157] in, For the ideal weight vector, The number of nodes in the neural network. To minimize the approximation error, For the activation function vector, It is a Gaussian function.

[0158] Combining equations (23)-(27), we can obtain:

[0159] (28)

[0160] Perform the following scaling transformation:

[0161] (29)

[0162] in, and , This represents the maximum value of the ideal weight norm and the minimum approximation error norm of the neural network. Indicates to The estimate, This represents the estimation error. After the above scaling, it is not necessary to estimate the neural network weights and errors separately, thereby reducing unknown variables and lowering the complexity of the system controller design.

[0163] Using inequalities on equation (29) Perform the following scaling:

[0164] (30)

[0165] in, For the design of positive constants, It is a constant.

[0166] From the error definition (21), we can obtain:

[0167] (31)

[0168] After the above steps, equation (30) can be expressed as:

[0169] (32)

[0170] Design adaptive update rate and virtual control rate as follows:

[0171] (33)

[0172] (34)

[0173] in, , For the design of positive constants.

[0174] Substituting the adaptive update rate and the virtual control rate into equation (32):

[0175] (35)

[0176] Using Yang's inequality for scaling:

[0177] (36)

[0178] Equation (35) can be written in the following form:

[0179] (37)

[0180] in, For positive numbers, the expression is: .

[0181] Differentiate equation (19):

[0182] (38)

[0183] in, for Regarding time The first-order partial derivative, for For variables The first-order partial derivative, for For variables The first-order partial derivative, for For variables The second-order mixed partial derivative is in the following form:

[0184]

[0185]

[0186]

[0187]

[0188] Differentiate equation (22) using equations (39) and (40):

[0189] (39)

[0190] in, , Indicates to Infinitesimal operators, Representing variables The intensity of random disturbances, specifically in the form of: , .

[0191] Define the Lyapunov function: The infinitesimal operator of the Lyapunov function is calculated using Iton's formula:

[0192] (40)

[0193] Using Young's inequality, the following scaling is performed:

[0194] (41)

[0195] In this embodiment, a radial basis function neural network is used to approximate the model uncertainties and unknowns caused by random perturbations of the quadcopter UAV. The unknown function to be approximated is selected as follows:

[0196] (42)

[0197] in, The input to the neural network is specifically: .

[0198] The output representation of the neural network is selected as follows:

[0199] (43)

[0200] in, For the ideal weight vector, The number of nodes in the neural network. To minimize the approximation error, For the activation function vector, It is a Gaussian function.

[0201] Substituting the neural network into equation (40):

[0202] (44)

[0203] Similarly, to reduce the complexity of controller design, the following inequality is used to perform lumped scaling on the ideal neural network weight norm, minimum approximation error norm, and external disturbances:

[0204] (45)

[0205] in, and , This represents the maximum value of the neural network weight norm, error norm, and external disturbance. Indicates to The estimate, This indicates the estimation error.

[0206] Using inequalities Scaling equation (46):

[0207] (46)

[0208] in, For the design of positive constants.

[0209] Combining equations (44) and (46), we get:

[0210] (47)

[0211] Design adaptive estimation update rate as follows:

[0212] (48)

[0213] in, For the design of positive constants.

[0214] Because actuator failure and input saturation constraints are considered in this embodiment, the system control gain is unknown and time-varying, making it difficult for traditional feedback control to guarantee stability. Therefore, the Nussbaum gain technique is introduced to address this issue, and its design is as follows:

[0215] (49)

[0216] (50)

[0217] (51)

[0218] in, The selected Nussbaum function, For the design of positive constants.

[0219] Substituting equations (48)-(51) into equation (47) yields:

[0220] (52)

[0221] Continuing to use Yang's inequality for scaling, we can obtain:

[0222] (53)

[0223] The final expression (52) is expressed as

[0224] (54)

[0225] in, , For positive numbers, the expression is: .

[0226] Step S5: Construct a comprehensive Lyapunov function covering all six degrees of freedom of the UAV, and prove, in conjunction with stochastic system theory, that all signals in the closed-loop system are eventually uniformly bounded.

[0227] The Lyapunov function with six degrees of freedom is designed as follows:

[0228] (55)

[0229] Based on the content of step S4, the infinitesimal operator of formula (54) calculated using Itō's formula satisfies:

[0230] (56)

[0231] in, .

[0232] Further equation (56) can be written as:

[0233] (57)

[0234] in, .

[0235] Taking the expectation of equation (57) yields:

[0236] (58)

[0237] in, .

[0238] Combining equations (55) and (58), we can see that: signal and All are bounded, thus obtaining signals. and It is bounded, and the virtual control signal Is with and Since it is a linear function, the virtual control signal is also bounded. (Combined formula) It can be known and It is also bounded. Therefore, according to the contents of equations (18) and (19) in step S3, we can obtain that... , When bounded, it can be guaranteed that the state signal always remains within the constraint range.

[0239] Furthermore, based on the method provided in the above embodiments, numerical simulation verification was performed on the quadcopter UAV using the above parameters, and the results are as follows. Figures 2 to 5 As shown. By Figure 2 The three-dimensional flight trajectory of the drone shown is as follows: Figure 3 A comparison of the actual trajectories of each axis (XYZ) with the reference trajectory shows that the quadcopter UAV can effectively track the reference trajectory and exhibits good tracking performance. Figure 4 The actual and reference trajectories for yaw, roll, and pitch angles are shown. It can be seen that even when the yaw angle is not initially within the set constraints, the method of this invention can still ensure its tracking effect. Furthermore, from... Figure 5 As shown in the system control inputs (total lift, yaw moment, roll moment, and pitch moment), all control inputs satisfy the saturation constraint conditions throughout the entire process.

[0240] Simulation experiments demonstrate that the proposed method can still ensure stable tracking control of the attitude and position system of a quadrotor UAV under complex conditions, including simultaneous model uncertainties, random external disturbances, actuator failures, and input saturation constraints. This method not only achieves safe operation under multi-state constraints but also improves the robustness and reliability of the system under abnormal conditions through effective fault tolerance and saturation compensation mechanisms, providing a feasible technical solution for high-performance flight control of quadrotor UAVs in real-world complex environments.

[0241] Example 2

[0242] This embodiment discloses a fault-tolerant control system for a quadcopter unmanned aerial vehicle that takes into account random disturbances;

[0243] like Figure 6 As shown, a fault-tolerant control system for a quadcopter unmanned aerial vehicle (UAV) considering random disturbances includes:

[0244] The stochastic control system model building module is configured to: establish a quadrotor UAV pose model modified by an inherent stochastic process, and integrate actuator faults and saturation constraints into the same framework to obtain a stochastic control system model of the quadrotor UAV.

[0245] The roll and pitch angle calculation module is configured to: based on the stochastic control system model of the quadcopter UAV, obtain the desired roll and pitch angles by decoupling the control variables of the UAV's six degrees of freedom system;

[0246] The state variable constraint addition module is configured to: construct the error state equation of the system, and, in response to the full state constraint requirements of the UAV, adopt a hybrid constraint strategy that integrates delayed state transition function and nonlinear state dependency mapping to apply constraints to state variables in different time periods.

[0247] The controller design module is configured to: design controllers for the six degrees of freedom sub-channels of the quadcopter UAV based on the adaptive backstepping control framework, and obtain the final input control signals for the corresponding degrees of freedom sub-channels;

[0248] The stability proof module is configured to: construct a comprehensive Lyapunov function covering all six degrees of freedom of the UAV, and, in conjunction with stochastic system theory, prove that all signals in the closed-loop system are eventually uniformly bounded.

[0249] Example 3

[0250] The purpose of this embodiment is to provide a computer-readable storage medium.

[0251] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in a fault-tolerant control method for a quadcopter unmanned aerial vehicle considering random disturbances as described in Example 1.

[0252] Example 4

[0253] The purpose of this embodiment is to provide an electronic device.

[0254] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in a fault-tolerant control method for a quadcopter unmanned aerial vehicle considering random disturbances as described in Embodiment 1.

[0255] The steps and methods involved in the apparatuses of Embodiments 2, 3, and 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.

[0256] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0257] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A fault-tolerant control method for a quadcopter unmanned aerial vehicle considering random disturbances, characterized in that, include: A pose model of a quadrotor UAV, modified by an inherent stochastic process, is established, and actuator faults and saturation constraints are integrated into the same framework to obtain a stochastic control system model of the quadrotor UAV. Based on the stochastic control system model of the quadcopter UAV, the desired roll angle and pitch angle are obtained by decoupling the control variables of the UAV's six degrees of freedom system. The system's error state equation is constructed. To address the full-state constraint requirements of the UAV, a hybrid constraint strategy combining a delayed state transition function and a nonlinear state dependency mapping is adopted to apply constraints to the state variables in different time periods. The delayed state transition function... for: in, For the system's state dimension, The set value for the constraint start time; The nonlinear state dependency mapping is: in, , They are respectively The variables of subchannel pose and velocity after state mapping are: Sub-channel pose tracking error, express Sub-channel pose dimension state quantity. express Sub-channel velocity dimension state quantity, for The expected trajectory of the sub-channel; The upper and lower bounds of the constraints are set; Based on the adaptive backstepping control framework, controllers are designed for each of the six degrees of freedom sub-channels of a quadrotor UAV, obtaining the final input control signals for the corresponding degree of freedom sub-channels, including: definition Sub-channel error variables, and Lyapunov functions designed using the backstepping method; the... The sub-channel error variable is defined as: in, , for Sub-channel error variables; For virtual control rate, ; The The sub-channel adaptive law and the final input control signal are: in, Indicates an adaptive update rate; The selected Nussbaum function, For the design of positive constants, for For variables The first-order partial derivative, express The maximum values ​​of the subchannel neural network weight norm, error norm, and external disturbance. Indicates to The estimate; Calculate the infinitesimal differential operator based on Itoh's stochastic differential formula; The system's unknowns and random disturbances are approximated using a radial basis function neural network, and the disturbances, errors, and upper bounds of the neural network weights are uniformly estimated using a lumped parameter adaptive law. The Nussbaum gain technique was applied to address the problem of unknown control gain caused by actuator failure and saturation, resulting in... The final input control signal for the sub-channel; We construct a comprehensive Lyapunov function covering all six degrees of freedom of the UAV, and using stochastic system theory, prove that all signals in the closed-loop system are eventually uniformly bounded.

2. The fault-tolerant control method for a quadcopter UAV considering random disturbances as described in claim 1, characterized in that, A pose model of a quadrotor UAV, modified by an inherent stochastic process, is established, and actuator faults and saturation constraints are integrated into the same framework to obtain a stochastic control system model of the quadrotor UAV, including: A stochastic differential equation model is obtained by modifying the classical quadrotor UAV dynamics model with stochastic perturbation based on Wiener process. The actuator fault model and saturation constraints are integrated into the stochastic differential equation model, and the actual control input is obtained by combining them. The stochastic control system model of the quadcopter UAV is then reconstructed. The stochastic control system model of the quadcopter UAV is as follows: in: And respectively represent the roll angle, pitch angle, yaw angle, and the UAV's position in the inertial frame. Direction, that is: , , , , , ; express Sub-channel pose dimension state quantity. express Sub-channel velocity dimension state quantity, express The random perturbation intensity function of the sub-channel pose dimension. express The random perturbation intensity function of the sub-channel velocity dimension. express Sub-channel control input, express The system state nonlinear function and unknown uncertainties are input in the sub-channel. express Unknown control coefficients of sub-channels caused by faults express The subchannel is affected by unknown bias and external disturbances caused by faults. express The output signal of the sub-channel The Wiener process is considered as a continuous random fluctuation caused by the rotor, airframe, motor, circuitry, and sensors.

3. The fault-tolerant control method for a quadcopter UAV considering random disturbances as described in claim 1, characterized in that, The desired roll and pitch angles are: in, For the expected trajectory of the roll angle, For the desired trajectory at the pitch angle, For the desired trajectory at the yaw angle, After reconstruction Virtual control input for the position sub-channel.

4. The fault-tolerant control method for a quadcopter UAV considering random disturbances as described in claim 1, characterized in that, The comprehensive Lyapunov function covering all six degrees of freedom of the UAV is as follows: in, for The Lyapunov function constructed by the sub-channels, They are respectively Error variables defined by the pose and velocity dimensions of the sub-channels. They are respectively The sub-channel pose and velocity dimensions are adaptively estimated.

5. A fault-tolerant control system for a quadcopter unmanned aerial vehicle (UAV) considering random disturbances, employing the fault-tolerant control method for a quadcopter UAV considering random disturbances as described in any one of claims 1-4, characterized in that, include: The stochastic control system model building module is configured to: establish a quadrotor UAV pose model modified by an inherent stochastic process, and integrate actuator faults and saturation constraints into the same framework to obtain a stochastic control system model of the quadrotor UAV. The roll and pitch angle calculation module is configured to: based on the stochastic control system model of the quadcopter UAV, obtain the desired roll and pitch angles by decoupling the control variables of the UAV's six degrees of freedom system; The state variable constraint addition module is configured to: construct the error state equation of the system, and, in response to the full state constraint requirements of the UAV, adopt a hybrid constraint strategy that integrates delayed state transition function and nonlinear state dependency mapping to apply constraints to state variables in different time periods. The controller design module is configured to: design controllers for the six degrees of freedom sub-channels of the quadcopter UAV based on the adaptive backstepping control framework, and obtain the final input control signals for the corresponding degrees of freedom sub-channels; The stability proof module is configured to: construct a comprehensive Lyapunov function covering all six degrees of freedom of the UAV, and, in conjunction with stochastic system theory, prove that all signals in the closed-loop system are eventually uniformly bounded.

6. A computer-readable storage medium having a program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the fault-tolerant control method for a quadcopter unmanned aerial vehicle that takes into account random disturbances as described in any one of claims 1-4.

7. An electronic device comprising a memory, a processor, and a 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 fault-tolerant control method for a quadcopter unmanned aerial vehicle that considers random disturbances as described in any one of claims 1-4.

Images