A quadrotor unmanned aerial vehicle attitude fault-tolerant control method considering unknown wind disturbance

Abstract

The application discloses a quad-rotor unmanned aerial vehicle attitude fault-tolerant control method considering unknown wind disturbance, and relates to the field of unmanned aerial vehicle attitude control design; the method can effectively cope with unknown wind disturbance, actuator partial failure and deviation faults. In the application, unknown wind disturbance and deviation faults are regarded as total disturbance, an improved disturbance compensator is designed by using an adaptive mechanism for the total disturbance, and the size of switching gain is adjusted online according to the control accuracy requirement, so that the waste of controller output energy caused by excessively large switching gain is avoided. For actuator partial failure, a failure factor compensator designed based on an adaptive method enables the unmanned aerial vehicle to timely process faults and recover normal trajectory. An attitude controller is designed by using a non-singular fast terminal sliding mode control technology, so that the robustness and fast response characteristics are ensured, the unmanned aerial vehicle can timely process faults, the fault tolerance of the unmanned aerial vehicle is enhanced, and the flight safety of the unmanned aerial vehicle is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of unmanned aerial vehicle attitude control design, in particular to a quadrotor unmanned aerial vehicle attitude fault-tolerant control method considering unknown wind disturbance. BACKGROUND

[0002] The quadrotor unmanned aerial vehicle has many advantages, of which the most prominent feature is vertical take-off and landing and hovering. With this advantage, the quadrotor unmanned aerial vehicle has a wide range of applications in agricultural investigation, traffic mapping, power grid detection and information measurement. With the wide application of quadrotor unmanned aerial vehicles, people have higher requirements for flight quality. The factors affecting the flight quality of unmanned aerial vehicles mainly include unknown wind disturbance and potential actuator failure. The position of the quadrotor unmanned aerial vehicle in the air needs to be controlled by adjusting the attitude. In order to achieve accurate tracking of the position, the accurate adjustment of the attitude must be achieved first. Therefore, it is of great significance to study a quadrotor unmanned aerial vehicle attitude fault-tolerant control method considering unknown wind disturbance.

[0003] The existing unmanned aerial vehicle attitude control methods include PID control, linear quadratic control and sliding mode control. Among them, the sliding mode control is widely used in the field of flight control due to its strong anti-interference ability. When using the sliding mode control method to design the attitude controller, some conditions must be limited, such as assuming that the system disturbance is bounded and the upper bound is known. However, in the actual flight environment, the upper bound information of the system disturbance is often difficult to obtain, and engineers generally use a larger value to replace the upper bound information. This will make the controller output unnecessary energy. In addition, when using the traditional sliding mode control method to design the attitude controller, it is difficult to achieve fast convergence characteristics. Although the general fast terminal sliding mode control method can improve the convergence characteristics of the system, some design parameters must satisfy the condition of being odd, which limits the adjustment range of the controller. Therefore, a non-singular fast terminal sliding mode control scheme is needed to overcome the problems existing in the sliding mode control, achieve fast convergence characteristics, do not need to know the upper bound information of the disturbance, and design parameters do not need to satisfy the condition of being odd, and consider the possible partial failure and deviation failure of the actuator. SUMMARY

[0004] In order to solve the problems existing in the prior art, the present application provides a quadrotor unmanned aerial vehicle attitude fault-tolerant control method considering unknown wind disturbance, a total disturbance compensator is designed according to an adaptive mechanism. In response to the partial failure of the actuator, the failure factor compensator designed by using the adaptive mechanism can make the unmanned aerial vehicle process the failure faster. An improved non-singular fast terminal sliding mode control technology is adopted to design an attitude controller for the unmanned aerial vehicle, fast convergence characteristics are obtained, and the limitation that the condition parameters of the general fast terminal sliding mode control method are odd is overcome.

[0005] The scheme adopted by the present application is as follows:

[0006] A quadrotor unmanned aerial vehicle attitude fault-tolerant control method considering unknown wind disturbance, comprising the following steps:

[0007] Step one: establish a quadrotor unmanned aerial vehicle attitude model with unknown wind disturbance and unknown actuator failure factors;

[0008] Step two: design a non-singular fast terminal sliding mode surface, including a roll channel sliding mode surface s1, a pitch channel sliding mode surface s2 and a yaw channel sliding mode surface s3;

[0009] Step three: design a failure factor compensator and a total disturbance compensator;

[0010] Step four: design an attitude controller, and verify whether the quadrotor unmanned aerial vehicle attitude model meets the stability requirements;

[0011] Step five: adjust the parameters of the attitude controller, the total disturbance compensator and the failure factor compensator for the quadrotor unmanned aerial vehicle attitude model that meets the stability requirements, so that the performance of the quadrotor unmanned aerial vehicle attitude model reaches the set performance index.

[0012] The quadrotor unmanned aerial vehicle attitude model with unknown wind disturbance and unknown actuator failure factors is shown in formula (1):

[0013]

[0014] wherein, and respectively represent roll angular velocity, pitch angular velocity and yaw angular velocity; and respectively represent roll angular acceleration, pitch angular acceleration and yaw angular acceleration; x , I y and I z are the rotational inertia of the X-axis, Y-axis and Z-axis in the body coordinate system respectively; 0 < ρ1 ≤ 1, 0 < ρ2 ≤ 1 and 0 < ρ3 ≤ 1 are the failure degrees of the roll channel, pitch channel and yaw channel actuators, collectively referred to as failure factors; D1 = δ1 + d1, D2 = δ2 + d2 and D3 = δ3 + d3 are the total disturbances of the roll channel, pitch channel and yaw channel respectively, wherein δ1, δ2 and δ3 are the roll channel deviation fault, pitch channel deviation fault and yaw channel deviation fault respectively, and d1, d2 and d3 are the unknown wind disturbances of the roll channel, pitch channel and yaw channel respectively; |D1| ≤ χ1, |D2| ≤ χ2 and |D3| ≤ χ3, wherein χ1, χ2 and χ3 are unknown constants; and respectively represent the roll channel torque control input, the pitch channel torque control input and the yaw channel torque control input, wherein l is the distance from the propeller center to the body center of gravity, and k is the tension coefficient, For torque coefficient, ω1, ω2, ω3 and ω4 are the rotating speeds of the first rotor, the second rotor, the third rotor and the fourth rotor of the unmanned aerial vehicle respectively.

[0015] The designed non-singular fast terminal sliding mode surface includes a roll channel sliding mode surface s1, a pitch channel sliding mode surface s2 and a yaw channel sliding mode surface s3, and the specific process is as follows:

[0016] The sign function sgn(·) is shown in formula (2):

[0017]

[0018] The roll angle error variable e1, the pitch angle error variable e2 and the yaw angle error variable e3 are shown in formula (3):

[0019]

[0020] Wherein, φ, θ and ψ are the roll angle, the pitch angle and the yaw angle respectively; φ d , θ d and ψ d are the expected roll angle, pitch angle and yaw angle respectively.

[0021] The roll channel sliding mode surface s1, the pitch channel sliding mode surface s2 and the yaw channel sliding mode surface s3 are shown in formula (4):

[0022]

[0023] Wherein, α1, α2 and α3 are normal numbers to be designed, β1, β2 and β3 are normal numbers to be designed, μ1, μ2 and μ3 are normal numbers to be designed, v1, v2 and v3 are normal numbers to be designed, and μ1>v1, μ2>v2, μ3>v3, 2>v1>1, 2>v2>1, 2>v3>1.

[0024] The designed failure factor compensator and total disturbance compensator are as follows:

[0025] The first intermediate variable a1, the second intermediate variable a2, the third intermediate variable a3, the first failure degree relative variable κ1, the second failure degree relative variable κ2 and the third failure degree relative variable κ3 are set, and are shown in formula (5):

[0026]

[0027] Wherein, and are the estimated values of χ1, χ2 and χ3 respectively, l1, l2 and l3 are normal numbers to be designed, and k1, k2 and k3 are normal numbers to be designed.

[0028] The roll channel failure factor compensator Pitch channel failure factor compensator And yaw channel failure factor compensator As shown in formula (6):

[0029]

[0030] Wherein, And Respectively, κ1, κ2 and κ3 are the estimated values of κ1, κ2 and κ3, And Is a normal number to be designed;

[0031] Roll channel total disturbance compensator Pitch channel total disturbance compensator And yaw channel total disturbance compensator As shown in formula (7):

[0032]

[0033] Wherein, I1, i2 and i3 are normal numbers to be designed, λ1, λ2 and λ3 are normal numbers to be designed; when |s i -ι i |≥0, the switching gain With the increase of time, the error variable e i More and more close to the sliding surface; when |s i -ι i |<0, the switching gain begins to decrease, but the error variable e i Will continue to keep in the vicinity of the sliding surface less than i i Of the adjacent area, wherein s i ={s1,s2,s3}, i i ={ι1,ι2,ι3}, e i ={e1,e2,e3}.

[0034] The design attitude controller, specifically:

[0035] Roll channel controller u1, pitch channel controller u2 and yaw channel controller u3, as shown in formula (8):

[0036]

[0037] The verification four rotor unmanned aerial vehicle attitude model stability, specifically:

[0038] Lyapunov candidate function V1, V2 and V3 are selected, as shown in formula (9):

[0039]

[0040] wherein,

[0041] Taking the derivative of V1, V2 and V3, the result is shown as formula (10):

[0042]

[0043] Substituting formula (6), formula (7) and formula (8) into formula (10), the calculation result is shown as formula (11):

[0044]

[0045] The stability threshold condition is that the derivative of the Lyapunov function is less than zero, and specifically is: and

[0046] Beneficial technical effects:

[0047] 1. Compared with the traditional sliding mode control, the controller designed in the application has faster response speed; and compared with the general fast terminal sliding mode control, the controller designed in the application has a wider parameter adjustment range;

[0048] 2. Compared with the general sliding mode control scheme, the controller designed in the application does not need to know the upper bound information of the wind disturbance, and can adjust the size of the switching gain online according to the change of the external disturbance, thereby avoiding the loss of energy to a certain extent;

[0049] 3. When the unmanned aerial vehicle actuator is partially failed or deviates from the fault, the adaptive mechanism is used to compensate for the fault in the method of the application, so that the unmanned aerial vehicle can process the fault faster, enhances the fault tolerance of the unmanned aerial vehicle, and improves the flight safety of the unmanned aerial vehicle; BRIEF DESCRIPTION OF DRAWINGS

[0050] Figure 1 A four-rotor unmanned aerial vehicle attitude fault-tolerant control method considering unknown wind disturbance is provided for the embodiment of the application

[0051] Figure 2 A roll angle tracking curve schematic diagram is provided for the embodiment of the application;

[0052] Figure 3 A pitch angle tracking curve schematic diagram is provided for the embodiment of the application;

[0053] Figure 4 A yaw angle tracking curve schematic diagram is provided for the embodiment of the application; DETAILED DESCRIPTION

[0054] The specific embodiments of the present application are described in further detail below in conjunction with the accompanying drawings and examples.

[0055] In this embodiment, a quadrotor unmanned aerial vehicle attitude fault-tolerant control method considering unknown wind disturbance is provided, a total disturbance compensator is designed according to an adaptive mechanism; in response to partial actuator failure, the failure factor compensator designed by the adaptive mechanism can make the unmanned aerial vehicle process faults faster; an improved non-singular fast terminal sliding mode control technology is used to design an attitude controller for the unmanned aerial vehicle, rapid convergence characteristics are obtained, and the limitation that the condition parameters of the general fast terminal sliding mode control method are odd numbers is overcome, as shown in Figure 1 The specific embodiments of the present application are described in further detail below in conjunction with the accompanying drawings and examples.

[0056] A quadrotor unmanned aerial vehicle attitude fault-tolerant control method considering unknown wind disturbance comprises the following steps:

[0057] Step 1: Establish a quadrotor unmanned aerial vehicle attitude model with unknown wind disturbance and unknown actuator failure factors;

[0058] A quadrotor unmanned aerial vehicle attitude kinematic model with unknown wind disturbance and unknown actuator failure factors is established, as shown in formula (12):

[0059]

[0060] wherein, and respectively represent roll angular velocity, pitch angular velocity and yaw angular velocity; and respectively represent roll angular acceleration, pitch angular acceleration and yaw angular acceleration; x , I y and I z respectively represent the rotational inertia of the X-axis, Y-axis and Z-axis in the body coordinate system; 0 < ρ1≤ 1, 0 < ρ2≤ 1 and 0 < ρ3≤ 1 respectively represent the failure degree of the roll channel, pitch channel and yaw channel actuator, and are collectively referred to as failure factors; D1 = δ1 + d1, D2 = δ2 + d2 and D3 = δ3 + d3 respectively represent the total disturbance of the roll channel, pitch channel and yaw channel, wherein δ1, δ2 and δ3 respectively represent the roll channel deviation fault, pitch channel deviation fault and yaw channel deviation fault, d1, d2 and d3 respectively represent the unknown wind disturbance of the roll channel, pitch channel and yaw channel; |D1|≤ χ1, |D2|≤ χ2, |D3|≤ χ3, wherein χ1, χ2 and χ3 are unknown normal numbers; and respectively represent the roll channel torque control input, pitch channel torque control input and yaw channel torque control input, wherein l is the distance from the propeller center to the body center of gravity, k is the tension coefficient, is the torque coefficient; ω1, ω2, ω3 and ω4 are the rotating speeds of the first rotor, the second rotor, the third rotor and the fourth rotor of the unmanned aerial vehicle respectively.

[0061] In this embodiment, the selected physical parameters of the quad-rotor unmanned aerial vehicle are I x = 0.045 kg·m 2 , I y = 0.045 kg·m 2 , I z = 0.083 kg·m 2 ;

[0062] In this embodiment, the failure faults and the deviation faults occur at the 10th second, the failure factors are ρ1 = 0.5, ρ2 = 0.5, ρ3 = 0.5; the deviation faults are δ1 = -0.002, δ2 = -0.002, δ3 = -0.002;

[0063] In this embodiment, the initial state values are selected as φ(0) = 0, θ(0) = 0, ψ(0) = 0, The wind disturbances are selected as d1 = sin(2t), d2 = sin(2t), d3 = sin(2t);

[0064] Step two: design a non-singular fast terminal sliding mode surface, including a roll channel sliding mode surface s1, a pitch channel sliding mode surface s2 and a yaw channel sliding mode surface s3;

[0065] The sign function sgn(·) is shown in equation (13):

[0066]

[0067] The roll angle error variable e1, the pitch angle error variable e2 and the yaw angle error variable e3 are shown in equation (14):

[0068]

[0069] wherein φ, θ and ψ are the roll angle, the pitch angle and the yaw angle respectively; φ d , θ d and ψ d are the desired roll angle, the desired pitch angle and the desired yaw angle respectively;

[0070] In this embodiment, the desired attitude angles are step signals, as shown in Table 1:

[0071] Table 1

[0072]

[0073] The roll channel sliding mode surface s1, the pitch channel sliding mode surface s2 and the yaw channel sliding mode surface s3 are shown in equation (15):

[0074]

[0075] Where α1, α2 and α3 are positive constants to be designed, β1, β2 and β3 are positive constants to be designed, μ1, μ2 and μ3 are positive constants to be designed, v1, v2 and v3 are positive constants to be designed, satisfying μ1>v1, μ2>v2, μ3>v3, 2>v1>1, 2>v2>1, 2>v3>1;

[0076] In this embodiment, the design parameter values ​​are selected as α1=1, α2=1, α3=1, β1=0.1, β2=0.1, β3=0.1, μ1=10, μ2=10, μ3=10, ν1=1.5, ν2=1.5, ν3=1.5;

[0077] Step 3: Design the failure factor compensator and the total disturbance compensator;

[0078] Let the first intermediate variable a1, the second intermediate variable a2, the third intermediate variable a3, the first relative failure degree variable κ1, the second relative failure degree variable κ2, and the third relative failure degree variable κ3 be set as shown in equation (16):

[0079]

[0080] in, and χ1, χ2, and χ3 are estimated values, l1, l2, and l3 are normal values ​​to be designed, and k1, k2, and k3 are normal values ​​to be designed.

[0081] In this embodiment, the design parameter values ​​are selected as k1=30, k2=30, k3=30, l1=2, l2=2, l3=2;

[0082] Roll path failure factor compensator Pitch Channel Failure Factor Compensator and yaw channel failure factor compensator As shown in equation (17):

[0083]

[0084] in, and The estimated values ​​for κ1, κ2, and κ3 are respectively. and For the positive constants to be designed;

[0085] In this embodiment, the initial value is selected as .

[0086] In this embodiment, the design parameter value is selected as follows:

[0087] Roll channel total disturbance compensator Pitch channel total disturbance compensator Yaw channel total disturbance compensator As shown in equation (18):

[0088]

[0089] wherein, ι1, ι2 and ι3 are normal numbers to be designed, λ1, λ2 and λ3 are normal numbers to be designed; when |s i -ι i |≥0, the switching gain With the increase of time, the error variable e i is closer and closer to the sliding surface; when |s i -ι i |<0, the switching gain begins to decrease, but the error variable e i will continue to remain in the adjacent area with a distance less than ι i from the sliding surface, wherein s i ={s1, s2, s3}, ι i ={ι1, ι2, ι3}, e i ={e1, e2, e3};

[0090] In this embodiment, the parameter values are selected as λ1=100, λ2=100, λ3=100, ι1=0.0005, ι2=0.0005, ι3=0.0005;

[0091] In this embodiment, the initial values are selected as

[0092] Step four: design the attitude controller, and verify whether the attitude model of the quadrotor unmanned aerial vehicle meets the stability requirements;

[0093] The roll channel controller u1, the pitch channel controller u2 and the yaw channel controller u3 are shown in equation (19):

[0094]

[0095] Verify the stability of the attitude model of the quadrotor unmanned aerial vehicle, and the Lyapunov candidate functions V1, V2 and V3 are selected, as shown in equation (20):

[0096]

[0097] wherein,

[0098] Derivations of V1, V2 and V3 are made, and the results are shown as formula (21) :

[0099]

[0100] Substitute formula (17), formula (18) and formula (19) into formula (21), and the calculation results are shown as formula (22) :

[0101]

[0102] The stability threshold condition is that the derivative of the Lyapunov function is less than zero, and specifically is: and

[0103] Step five: adjust the parameters of the attitude controller, the total disturbance compensator and the failure factor compensator for the quad-rotor unmanned aerial vehicle attitude model that meets the stability requirement, so that the performance of the quad-rotor unmanned aerial vehicle attitude model reaches the set performance index.

[0104] To further verify the significant and substantial features of the present application, simulation experiments are carried out.

[0105] As shown in formula (23), the tracking curve of the roll angle is shown in formula (24), the tracking curve of the pitch angle is shown in formula (25), and the tracking curve of the yaw angle is shown in formula (26). Figure 2 Figure 3 Figure 4 As can be seen from the simulation results, the design scheme provided by the present application can still accurately track the expected signal in the case of unknown wind disturbance and actuator failure.​​

Claims

1. A quadrotor unmanned aerial vehicle attitude fault-tolerant control method considering unknown wind disturbance, characterized in that: The method comprises the following steps: Step one: establishing a quadrotor unmanned aerial vehicle attitude model with unknown wind disturbance and unknown actuator failure factors; Step two: design nonsingular fast terminal sliding mode surface, including roll channel sliding mode surface , pitch channel sliding mode surface and yaw channel sliding mode surface ; Step three: designing a failure factor compensator and a total disturbance compensator; Set a first intermediate variable , a second intermediate variable , a third intermediate variable , a first failure degree relative variable , a second failure degree relative variable , a third failure degree relative variable As shown in equation (5): (5) wherein, , and denote roll, pitch and yaw angular velocities, respectively; , and are the moments of inertia of the body coordinate system about axis, axis and axis, respectively; , and are the failure degrees of the roll, pitch and yaw channels actuators, respectively, denoted as failure factors; , and are the estimated values of , and , , and are unknown constants; , and are constants to be designed, , and are constants to be designed; , and are the roll, pitch and yaw angle error variables, respectively; , and are constants to be designed, , and are constants to be designed, , and are constants to be designed, , and are constants to be designed; , and are the desired roll, pitch and yaw angles, respectively. Roll channel failure factor compensator Pitch channel failure factor compensator And yaw channel failure factor compensator As shown in equation (6): (6) wherein, , and are the estimated values of , and , , and are normal numbers to be designed. Roll channel total disturbance compensator Pitch channel total disturbance compensator Yaw channel total disturbance compensator as shown in equation (7): (7) wherein , , , , and are normal numbers to be designed, , and are normal numbers to be designed; when , the switching gain increases with time, the error variable is getting closer and closer to the sliding surface; when is satisfied, the switching gain starts to decrease, but the error variable will continue to stay in the neighborhood of the sliding surface with a distance less than , where , , , ; Step four: designing an attitude controller and verifying whether the quadrotor unmanned aerial vehicle attitude model meets the stability requirement; Step five: adjusting parameters of the attitude controller, the total disturbance compensator and the failure factor compensator for the quadrotor unmanned aerial vehicle attitude model meeting the stability requirement, so that the performance of the quadrotor unmanned aerial vehicle attitude model reaches the set performance index.

2. The attitude fault-tolerant control method for a quadrotor UAV considering unknown wind disturbances according to claim 1, characterized in that: The quadrotor unmanned aerial vehicle attitude model with unknown wind disturbance and unknown actuator failure factors is shown in formula (1): (1) wherein , and represent roll, pitch and yaw angular accelerations, respectively; , and are total disturbances in the roll, pitch and yaw channels, respectively, wherein , and are roll channel off-fault, pitch channel off-fault and yaw channel off-fault, respectively, , and are roll channel unknown wind disturbance, pitch channel unknown wind disturbance and yaw channel unknown wind disturbance, respectively; , , , , and represent roll channel moment control input, pitch channel moment control input and yaw channel moment control input, respectively, wherein is the distance from the center of the propeller to the center of mass of the airframe, is the drag coefficient, is the torque coefficient, , , and are the rotational speeds of the first, second, third and fourth rotors of the UAV, respectively. 3.The method of claim 2, wherein: The design non-singular fast terminal sliding mode surface, including roll channel sliding mode surface , pitch channel sliding mode surface and yaw channel sliding mode surface , the specific process is: sign function as shown in equation (2): (2) roll angle error variable pitch angle error variable and yaw angle error variable as shown in equation (3): (3) wherein , and are roll, pitch and yaw angles, respectively; Roll channel slide surface Pitch channel slide surface Yaw channel slide surface as shown in equation (4): (4) wherein , and satisfy , , , , , .

4. The quadrotor unmanned aerial vehicle attitude fault-tolerant control method considering unknown wind disturbance according to claim 3, characterized in that: The attitude controller is designed, and specifically: Roll channel controller Pitch channel controller And yaw channel controller As shown in equation (8): (8)。 5. The quadrotor unmanned aerial vehicle attitude fault-tolerant control method considering unknown wind disturbance according to claim 4, characterized in that: The stability of the quadrotor unmanned aerial vehicle attitude model is verified, and specifically: Lyapunov candidate function 、 and is chosen as in equation (9): (9) wherein , , , , , ; Taking the derivative of , and , the result is as shown in equation (10): (10) Formula (6), formula (7) and formula (8) are substituted into formula (10), and the calculation result is shown in formula (11): (11) Whether the quadrotor unmanned aerial vehicle attitude model meets the stability requirement is judged through a stable threshold condition.

6. The quadrotor UAV attitude fault-tolerant control method considering unknown wind disturbance according to claim 5, characterized in that: The stable threshold condition is that the derivative of the Lyapunov function is less than zero, specifically: , and .

Images