A fault-tolerant obstacle avoidance control method for fixed-wing UAV

By combining path planning, disturbance observer and non-singular terminal sliding mode control, the problems of obstacle avoidance and fault tolerance of fixed-wing drones in complex environments are solved, and stable flight under unknown external disturbances and thrust failures are achieved. Track tracking.

CN117872751BActive Publication Date: 2024-11-12NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202311864438.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-12-29
Publication Date
2024-11-12
Estimated Expiration
2043-12-29

AI Technical Summary

Technical Problem

The prior art is difficult to simultaneously realize obstacle avoidance and fault tolerance control of fixed-wing drones in complex external environments, especially in the case of unknown external disturbances and thrust failures, resulting in flight instability and trajectory tracking difficulties.

Method used

Combining path planning, disturbance observer and non-singular terminal sliding mode control, a dynamic model and actuator failure model are established, non-linear disturbance observer and improved non-singular terminal sliding mode method are designed, and artificial potential field method and escape force field Solve obstacle avoidance and thrust loss problems, and achieve fault-tolerant control.

Benefits of technology

It effectively avoids obstacle collisions and thrust failures, ensures stable flight of the drone in complex environments and precise tracking of expected trajectories, and improves the safety and reliability of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117872751B_ABST
    Figure CN117872751B_ABST
Patent Text Reader

Abstract

The present invention discloses a fixed-wing unmanned aerial vehicle fault-tolerant obstacle avoidance control method, and relates to the technical field of unmanned aerial vehicles. The method first improves the traditional artificial potential field method, solves the unreachable target and local minimum problems existing in the traditional artificial potential field method, and then converts the dynamic model of the fixed-wing unmanned aerial vehicle into an affine form, considers the situation of thrust failure, and establishes a fixed-wing unmanned aerial vehicle actuator fault model. Afterwards, the lumped uncertainty in the fault model is estimated by designing a disturbance observer. Finally, a fault-tolerant controller is obtained by deduction and design based on an improved non-singular terminal sliding mode method. The designed improved non-singular terminal sliding mode fault-tolerant controller is used to accurately track the planned path to achieve the purpose of obstacle avoidance and fault tolerance. The present invention is used for obstacle avoidance of fixed-wing unmanned aerial vehicles and fault-tolerant control when thrust fails under unknown external disturbances.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to an obstacle avoidance method for a fixed-wing unmanned aerial vehicle and an improved non-singular terminal sliding mode fault-tolerant control method in the presence of unknown external disturbances and thrust failures, and belongs to the field of aircraft fault-tolerant obstacle avoidance control. Technical Background

[0002] In recent years, drones have been widely used in many fields due to their high speed and ability to carry a variety of payloads. Especially in the military field, drones have become an indispensable force. However, as the intensity of battles increases, the environment becomes more and more complex. The requirements for drones in terms of safety, reliability, and stability are also getting higher and higher. The complex and changing external environment, unknown external interference, and failure of drone actuators will not only have an adverse effect on flight, but may even lead to some catastrophic consequences; avoiding collisions between drones and surrounding objects is also an aspect worth considering. Therefore, in order to improve the ability of fixed-wing drone systems to perform various tasks and ensure that the system has high safety and reliability, it is necessary to study the obstacle avoidance and fault-tolerant control of drones.

[0003] At present, few research results consider comprehensive flight safety issues including obstacle avoidance and actuator fault tolerance at the same time. Moreover, when such flight safety requirements are combined with the original target tracking flight mission, the difficulty of controller design will be further increased. The present invention combines path planning, disturbance observer and non-singular terminal sliding mode control to propose a solution to this complex problem. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a path planning obstacle avoidance algorithm for fixed-wing UAVs and a non-singular terminal sliding mode fault-tolerant control method for thrust loss, so as to ensure that the fixed-wing UAV has obstacle avoidance capability and can fly stably and track the desired trajectory in the event of actuator failure.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] The present invention is a fault-tolerant obstacle avoidance control method for a fixed-wing UAV, and the control method comprises the following processes:

[0007] S1 establishes the dynamic model of fixed-wing UAV and sets the control input;

[0008] S2 establishes the actuator fault model;

[0009] S3 sets the obstacle as a sphere, and obtains the collision range between the drone and the obstacle according to the obstacle radius and the wing length of the drone; then combines the coordinates of the obstacle and the influence radius to obtain the repulsive field in the artificial potential field method; introduces the escape force F e To solve the local minimum problem, we can get the minimum distance between obstacles in the direction of the UAV.

[0010] S4 designs a nonlinear disturbance observer to estimate the lumped uncertainty of obstacles, and uses the estimated value and combines it with the improved non-singular terminal sliding mode method to deduce and design a fault-tolerant controller;

[0011] S5 returns to the fixed-wing UAV fault model based on the obtained control input, performs fault-tolerant control on the fixed-wing UAV under thrust loss conditions and tracks the desired trajectory.

[0012] Furthermore, the established fixed-wing UAV dynamics model is specifically as follows:

[0013]

[0014] Among them, x, y, z represent the three-axis position coordinates, V represents the airspeed, χ represents the heading angle, γ represents the track angle, φ represents the tilt angle, T represents the thrust, D represents the drag, L represents the lift, m represents the mass of the drone, and g represents the acceleration due to gravity;

[0015] The control input quantity is specifically set as follows:

[0016] Set control input u s =[u s1 ,u s2 ,u s3 ] T =[T,L sinφ,L cosφ] T , set P f =[x,y,z] T is the outer ring position vector, we get:

[0017]

[0018] in,

[0019]

[0020] Furthermore, the establishment of the actuator fault model is specifically as follows:

[0021] Consider thrust failure, external disturbances and parameter inaccuracies:

[0022] u(t)=(1-ρ(t))u s (t)+δ(t) (4)

[0023] d=[d x ,d y ,d z ] T (5)

[0024] Δf=[Δf x ,Δf y ,Δf z ] T (6)

[0025] Among them, ρ(t)=diag{ρ1,0,0}, 0≤ρ1<1 represents the thrust loss ratio, δ(t)=[δ p ,0,0] T represents the bias fault, δ p is the magnitude of the bias fault;

[0026] Substituting (4) to (6) into (2), we obtain:

[0027]

[0028] Among them, F(P f ,t)=-Gρu s (t)+Gδ(t)+Δf+d is the uncertain term.

[0029] Furthermore, the S3 is specifically:

[0030] Treat all obstacles of the drone as spheres, and set the obstacle radius to r ob , the wing length of the drone is r uav , then the collision range between the drone and the obstacle is r col =r ob +r uav ; Then set the coordinates of the obstacle to P o =[x o ,y o ] T , the influence radius is d0, and the repulsive field in the traditional artificial potential field method is defined as follows:

[0031]

[0032] Among them, k rep is the repulsion coefficient, d(P,P o )=||PP o || is the distance between the current position of the drone and the obstacle;

[0033] Set the repulsive force on the drone to the negative gradient of the repulsive field:

[0034]

[0035] The direction of the repulsive force is along the straight line between the drone and the obstacle, pointing toward the obstacle;

[0036] The unreachable target problem is solved by improving the repulsive field function. The improved repulsive field function is:

[0037]

[0038] The repulsive force on the improved drone is:

[0039]

[0040] in,

[0041]

[0042] By introducing the escape force F e To solve the local minimum problem, the escape force applied is defined as follows:

[0043]

[0044] Among them, F d =||F att ||-||F rep ||;

[0045] The angle used to apply the escape force is defined as follows:

[0046]

[0047] Among them, r min is the minimum distance between n obstacles perpendicular to the direction of the drone's travel at this time.

[0048] Furthermore, the S4 is specifically:

[0049] When the total uncertainty in the fault model is unknown, a nonlinear disturbance observer is designed to predict the uncertainty F(P f ,t) make an estimate, then quote the estimated value and combine it with the improved non-singular terminal sliding mode method to deduce and design the fault-tolerant controller;

[0050] Design auxiliary variables Where υ is the auxiliary subsystem, which is defined as follows:

[0051]

[0052] Among them, α1, α2, α3, α4>0, 0<λ1<1, λ2>1 are the observer parameters to be designed;

[0053] The estimated value of the lumped uncertainty F is:

[0054]

[0055] Assume that the expected path given by trajectory planning is Defining Tracking Error Then the improved non-singular terminal sliding surface is defined as follows:

[0056]

[0057] in, ξ>0, 0<τ<1, p and q are odd numbers and satisfy 1<q<p<2q is the parameter to be designed, P f (0) is the initial state of the system;

[0058] The fault-tolerant controller based on the improved non-singular terminal sliding mode is designed as:

[0059]

[0060] Among them, k1>0, k2>0 are the controller parameters to be designed.

[0061] The beneficial effects of the present invention are:

[0062] (1) The present invention improves the traditional artificial potential field method, solves the problem of unreachable target and local minimum value existing in the traditional artificial potential field method, and can help fixed-wing UAVs to avoid obstacles.

[0063] (2) The present invention discloses a fault-tolerant control method for fixed-wing UAV based on improved non-singular terminal sliding mode, which can maintain stable flight of the fixed-wing and accurately track the desired trajectory when the UAV is subject to unknown external disturbances and actuator failures. The designed disturbance observer can achieve fast and accurate estimation of unknown external disturbances and faults.

[0064] (3) It has great practical significance and application prospects in the fault-tolerant obstacle avoidance control of fixed-wing UAVs. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 This is a flow chart of a fault-tolerant obstacle avoidance control method for a fixed-wing UAV;

[0066] Figure 2 This is the error diagram when a thrust failure occurs in a fixed-wing UAV without obstacle avoidance;

[0067] Figure 3 This is a control signal diagram when a thrust failure occurs in a fixed-wing UAV without obstacle avoidance;

[0068] Figure 4 This is the observer estimation value diagram when a thrust failure occurs in a fixed-wing UAV without obstacle avoidance;

[0069] Figure 5 This is the observer estimation error diagram when a thrust failure occurs in a fixed-wing UAV without obstacle avoidance;

[0070] Figure 6 Path planning and trajectory tracking diagram for fixed-wing UAVs;

[0071] Figure 7 is the error diagram during trajectory tracking;

[0072] Figure 8 This is the control signal diagram during trajectory tracking. DETAILED DESCRIPTION

[0073] The control method of the present invention is further explained in conjunction with the attached diagrams.

[0074] (a) Establishment of fixed-wing UAV dynamics model:

[0075]

[0076] Among them, x, y, z represent the three-axis position coordinates, V represents the airspeed, χ represents the heading angle, γ represents the track angle, φ represents the tilt angle, T represents the thrust, D represents the drag, and L represents the lift. m represents the mass of the drone, and g represents the acceleration due to gravity.

[0077] The mathematical expression of resistance D is as follows:

[0078]

[0079] Where ρ is the air density, C D0 is the zero-lift drag coefficient, V w is the gust, S is the wing area, k d is the lifting resistance, k n is the load factor effectiveness.

[0080] The mathematical model of gust is as follows:

[0081]

[0082] in, For normal wind shear, V m is the average wind speed. w For gust turbulence, the mean is 0 and the standard deviation is 0.9V m is a Gaussian random variable.

[0083] (b) Take the control input u s =[u s1 ,u s2 ,u s3 ] T =[T,L sinφ,L cosφ] T, take P f =[x,y,z] T is the outer ring position vector, we get:

[0084]

[0085] in,

[0086]

[0087] (c) Establishing actuator fault model:

[0088] Consider thrust failure, external disturbance, parameter inaccuracy:

[0089] u(t)=(1-ρ(t))u s (t)+δ(t) (6)

[0090] d=[d x ,d y ,d z ] T (7)

[0091] Δf=[Δf x ,Δf y ,Δf z ] T (8)

[0092] Among them, ρ(t)=diag{ρ1,0,0}, 0≤ρ1<1 represents the thrust loss ratio, δ(t)=[δ p ,0,0] T represents the bias fault, δ p is the magnitude of the bias fault.

[0093] Substituting (6) to (8) into (4), we obtain:

[0094]

[0095] Among them, F(P f ,t)=-Gρu s (t)+Gδ(t)+Δf+d is the lumped uncertainty term.

[0096] (d) Improvement of the traditional artificial potential field method

[0097] In most typical mission scenarios, the altitude of fixed-wing drones remains basically unchanged during flight, so this invention only considers obstacle avoidance problems within a two-dimensional plane. Take the current position coordinates of the drone as P = [x, y] T , the target point position coordinates are P tar =[x tar ,y tar ]T Before the drone reaches the target point, it will be affected by the attraction, and the farther away from the target point, the greater the attraction. Therefore, the definition of the gravitational field in the traditional artificial potential field method is as follows:

[0098]

[0099] Among them, k att is the gravitational coefficient, d(P,P tar )=||PP tar || is the distance between the current position of the drone and the target point.

[0100] The gravitational force on the drone is the negative gradient of the gravitational field:

[0101]

[0102] The direction of gravity is along the straight line between the drone and the target point, pointing to the target point.

[0103] Similarly, the repulsive field generated by the obstacle is also related to the distance, but the repulsive field has an effective range. The closer the distance is within the range, the greater the repulsive force is, and the repulsive force outside the range is zero. In the present invention, all obstacles are treated as spheres, and the radius of the obstacle is r ob , the wing length of the drone is r uav , then the collision range between the drone and the obstacle is r col =r ob +r uav . Assume that the coordinates of the obstacle are P o =[x o ,y o ] T , the influence radius is d0, and the definition of the repulsive field in the traditional artificial potential field method is as follows:

[0104]

[0105] Among them, k rep is the repulsion coefficient, d(P,P o )=||PP o || is the distance between the current position of the drone and the obstacle.

[0106] The repulsive force on the drone is the negative gradient of the repulsive field:

[0107]

[0108] The direction of the repulsive force is along the straight line between the drone and the obstacle, pointing toward the obstacle.

[0109] The unreachable target problem is solved by improving the repulsive field function. The improved repulsive field function is:

[0110]

[0111] The repulsive force on the improved drone is:

[0112]

[0113] in,

[0114]

[0115] By introducing the escape force F e To solve the problem of local minima, the escape force that can be applied is defined as follows:

[0116]

[0117] Among them, F d =||F att ||-||F rep ||.

[0118] The angle at which the escape force can be applied is defined as follows:

[0119]

[0120] Among them, r min is the minimum distance between n obstacles perpendicular to the direction of the drone's travel at this time.

[0121] (e) When the lumped uncertainty in the fault model is unknown, a nonlinear disturbance observer is designed to calculate the lumped uncertainty F(P f ,t) make an estimate, then quote the estimated value and combine it with the improved non-singular terminal sliding mode method to deduce and design the fault-tolerant controller;

[0122] Design auxiliary variables Where υ is the auxiliary subsystem, which is defined as follows:

[0123]

[0124] Among them, α1, α2, α3, α4>0, 0<λ1<1, λ2>1 are the observer parameters to be designed.

[0125] The estimated value of the lumped uncertainty F is:

[0126]

[0127] Let the expected path given by trajectory planning be Defining Tracking Error Then the improved non-singular terminal sliding surface is defined as follows:

[0128]

[0129] in, ξ>0, 0<τ<1, p and q are odd numbers and satisfy 1<q<p<2q is the parameter to be designed, P f (0) is the initial state of the system.

[0130] The fault-tolerant controller based on the improved non-singular terminal sliding mode is designed as:

[0131]

[0132] Among them, k1>0, k2>0 are the controller parameters to be designed.

[0133] (f) According to the obtained control input u s , returning to the fixed-wing UAV fault model, the fixed-wing UAV under thrust loss condition is fault-tolerantly controlled and the trajectory is tracked.

[0134] The effectiveness of the present invention is verified by performing simulation below:

[0135] The model parameters of the fixed-wing UAV are m = 10 kg, g = 9.81 kg / m 2 ,ρ=1.225kg / m 3 , S=1.37m 2 , C do =0.02, k d =0.1, k n =1, V m =4m / s. The relevant parameters of the observer are α1=1.5, α2=0.72, α3=0.6, α4=0.1, λ1=0.6, λ2=7 / 5. The relevant parameters of the improved non-singular terminal sliding mode are ξ=0.5,p=7,q=5. The relevant parameters of the controller are k1=diag{7,5,20},k2=diag{2,2,2}. The external disturbance is d=[-1.6,-1.8,-1] T , parameter uncertainty is Δf = [-1.2, 2.6, 1.6] T The thrust fault occurs at t = 10s, the fault size is ρ1 = 0.4, δ p = -5. The starting point of the drone is set to P0 = [0m, 0m] T , the target point is set to P tar =[1000m,0m] T The obstacle coordinate settings are shown in the following table:

[0136] Table 1

[0137] Obstacle coordinates

[0138]

[0139]

[0140] The simulation results show that the improved artificial potential field method solves the unreachable target and local minimum problems of the traditional artificial potential field. The proposed fault-tolerant algorithm can achieve fault-tolerant control under thrust failure and track the expected trajectory. The designed observer can accurately estimate the lumped uncertainty. Figure 2 is the tracking error of the system when there is no obstacle avoidance. It can be clearly seen that the tracking error will change suddenly after the fault is injected, but it can quickly recover and reach stability under fault-tolerant control. Figure 3 This is the control signal of the system when there is no obstacle avoidance. It can be seen that a thrust failure occurs at the 10th second, and then quickly reaches a new stable value under fault-tolerant control. Figure 4 and Figure 5 are the estimated values ​​and fault errors of the observer respectively. It can be seen from the figure that the observer can quickly estimate external disturbances and faults. Figure 6 Planning the trajectory for UAV tracking. It can be seen from the figure that the improved artificial potential field method can help the UAV avoid obstacles and solve the problems of unreachable targets and local minima. Figure 7 and Figure 8 are the errors and control signals in the trajectory tracking process. It can be seen from the figure that the designed improved non-singular terminal sliding mode fault-tolerant controller can accurately track the planned trajectory.

[0141] In summary, with respect to the obstacle avoidance and fault tolerance issues under thrust failure of fixed-wing UAVs, the method of the present invention can effectively implement fault-tolerant obstacle avoidance control.

[0142] The above embodiments are only for illustrating the technical idea of ​​the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the present invention.

Claims

1. A fault-tolerant obstacle avoidance control method for a fixed-wing UAV, characterized in that: The control method comprises the following steps: S1 establishes the dynamic model of fixed-wing UAV and sets the control input; S2 establishes the actuator fault model; S3 sets the obstacle as a sphere, and obtains the collision range between the drone and the obstacle based on the obstacle radius and the wing length of the drone; then combines the coordinates of the obstacle and the influence radius to obtain the repulsive field in the artificial potential field method; By introducing the escape force F e To solve the local minimum problem, we can get the minimum distance between obstacles in the direction of the UAV’s movement. S4 designs a nonlinear disturbance observer to estimate the lumped uncertainty of obstacles, and cites the estimated value and combines it with the improved non-singular terminal sliding mode method to deduce and design a fault-tolerant controller; S4 is specifically: When the total uncertainty in the fault model is unknown, a nonlinear disturbance observer is designed to predict the uncertainty F(P f ,t) make an estimate, then quote the estimated value and combine it with the improved non-singular terminal sliding mode method to deduce and design the fault-tolerant controller; Design auxiliary variables Where υ is the auxiliary subsystem, which is defined as follows: Among them, α1, α2, α3, α4>0, 0<λ1<1, λ2>1 are the observer parameters to be designed; The estimated value of the lumped uncertainty F is: Assume that the expected path given by trajectory planning is Defining Tracking Error Then the improved non-singular terminal sliding surface is defined as follows: Among them, 0 < τ < 1, p and q are odd numbers and satisfy 1 < q < p < 2q, which are parameters to be designed, P f (0) is the initial state of the system; P f is the outer loop position vector; The fault-tolerant controller based on the improved non-singular terminal sliding mode is designed as: Among them, k1>0, k2>0 are the controller parameters to be designed; S5 returns to the fixed-wing UAV fault model according to the obtained control input, performs fault-tolerant control on the fixed-wing UAV under thrust loss conditions and tracks the desired trajectory.

2. The method for controlling a fixed-wing UAV with fault tolerance and obstacle avoidance according to claim 1, characterized in that: The specific steps of establishing the fixed-wing UAV dynamics model are as follows: Among them, x, y, z represent the three-axis position coordinates, V represents the airspeed, χ represents the heading angle, and γ represents the track angle; represents the tilt angle, T is the thrust, D is the drag, L is the lift, m is the mass of the drone, and g is the acceleration due to gravity; The setting control input is specifically: Set control input u s =[u s1 ,u s2 ,u s3 ] T =[T,Lsin ,Lcos ] T , set P f =[x,y,z] T is the outer ring position vector, we get: in, 3. The method for controlling a fixed-wing UAV with fault tolerance and obstacle avoidance according to claim 1, characterized in that: The establishment of the actuator fault model is specifically as follows: Consider thrust failure, external disturbances and parameter inaccuracies: u(t)=(1-ρ(t))u s (t)+δ(t) (8) d=[d x ,d y ,d z ] T (9) Δf=[Δf x ,Δf y ,Δf z ] T (10) Among them, ρ(t)=diag{ρ1,0,0}, 0≤ρ1<1 represents the thrust loss ratio, δ(t)=[δ p ,0,0] T represents the bias fault, δ p is the magnitude of the bias fault; Substituting (8) to (10) into (6), we obtain: Among them, F(P f ,t)=-Gρu s (t)+Gδ(t)+Δf+d is the uncertain term.

4. The method for controlling a fixed-wing UAV with fault tolerance and obstacle avoidance according to claim 1, characterized in that: S3 is specifically as follows: all obstacles of the drone are treated as spheres, and the radius of the obstacle is r ob , the wing length of the drone is r uav , then the collision range between the drone and the obstacle is r col =r ob +r uav ; Then set the coordinates of the obstacle to P o =[x o ,y o ] T , the influence radius is d0, and the repulsive field in the traditional artificial potential field method is defined as follows: Among them, k rep is the repulsion coefficient, d(P,P o )=||PP o || is the distance between the current position of the drone and the obstacle; Set the repulsive force on the drone to the negative gradient of the repulsive field: The direction of the repulsive force is along the straight line between the drone and the obstacle, pointing toward the obstacle; The unreachable target problem is solved by improving the repulsive field function. The improved repulsive field function is: The repulsive force on the improved drone is: in, d(P,P tar ) is the distance between the UAV and the target point, and the escape force F is introduced. e To solve the local minimum problem, the escape force applied is defined as follows: Among them, F d =||F att ||-||F rep ||;F att is the gravitational force on the drone; The angle used to apply the escape force is defined as follows: Among them, r min is the minimum distance between n obstacles perpendicular to the direction of the UAV’s travel at this time.

Citation Information

Patent Citations

  • Design method of four-rotor fault-tolerant controller based on nonlinear observer

    CN109343369A

  • Sliding-mode fault-tolerant control method for constrained unmanned aerial vehicle system under actuator fault

    CN115373363A

  • Fast terminal sliding mode-based fault-tolerant cooperative control method for bee colony unmanned aerial vehicles

    CN115903908A

  • Rotor unmanned aerial vehicle obstacle avoidance sliding mode fault-tolerant control method

    CN117055593A