A mars unmanned aerial vehicle and a control method thereof based on a task manifold controller

Abstract

The present application relates to a kind of unmanned aerial vehicle of Mars and its control method based on task manifold controller, the unmanned aerial vehicle KPM-VBLA of Mars includes two magnetic control paddle and two rotor paddle, controllable magnetic field force is generated between the two, the size and direction of the up and down rotor inclination are controlled by the magnetic force between magnetic control paddle and rotor paddle;The method for controlling the above-mentioned unmanned aerial vehicle of Mars based on task manifold includes: S1: define each coordinate system of the unmanned aerial vehicle KPM-VBLA of Mars;S2: establish the six-degree-of-freedom dynamics model of the unmanned aerial vehicle KPM-VBLA of Mars;S3: the control law of task manifold controller is designed, the input control force and moment of the unmanned aerial vehicle KPM-VBLA of Mars are obtained, to realize the control of unmanned aerial vehicle of Mars.In virtual simulation environment, three-dimensional trajectory tracking and path tracking with high-order nonlinear curve characteristics are realized, and the effectiveness of the designed maneuvering performance and control algorithm of unmanned aerial vehicle KPM-VBLA of Mars is verified.

Description

Technical Field

[0001] This invention relates to the integrated application of automatic control, aviation and unmanned aerial vehicle (UAV) design, and particularly to a Mars UAV and a method for controlling the Mars UAV to perform trajectory tracking using a mission manifold controller. Background Technology

[0002] Over decades on Mars, Mars rovers have covered less than 40 miles of relatively flat and open terrain. A Mars drone is a planetary aerial robot capable of flying on the Martian surface and collaborating with Mars rovers to complete surface exploration missions. It fills the gap between the Mars Reconnaissance Orbiter's large-area, low-resolution high-position exploration and the Mars rovers' small-area, high-resolution low-position exploration. Mars drones can enhance rover missions by quickly scouting safe crossing routes or reconnaissance of potential scientific targets, and as a standalone system, they can be used to explore areas that rovers may be unable to reach. Using rotorcraft drones for reconnaissance will increase the scope and breadth of Mars surface exploration missions and provide easier and safer access to caves, craters, ice, and any potentially hazardous locations. Coaxial rotorcraft drones offer advantages such as compact design, high hovering efficiency, relatively high blade Reynolds numbers, and a limited weight and volume. Single-master rotors and multi-master rotors have significant volume constraints and cannot achieve the same tip Reynolds numbers as coaxial rotors within a given volume.

[0003] Since the concept of a Mars drone was proposed in the 1970s, decades of unremitting efforts have culminated in the successful landing of Ingenuity, the first Mars drone, aboard the Perseverance rover, on the Martian surface in March 2021. It has since completed more than ten aerial exploration flights. Images of the Martian surface taken by the Ingenuity drone have significantly improved the efficiency of Mars exploration missions. With advancements in planetary aerial robotics technology, it will become possible for drone swarms to collaboratively explore the surfaces of Mars, Venus, Jupiter, Titan, and other planets in the future.

[0004] The Martian environment differs significantly from Earth's, posing unique and challenging requirements for the control of coaxial rotor UAVs. For example, Martian surface gravity is approximately 38% of Earth's; the thin carbon dioxide atmosphere has a density and speed of sound that are 1.26% and 67% of Earth's, respectively; and the rotor operates at a low Reynolds number. <10000) and relatively high tip Mach number (M> In an environment of 0.2), the resulting aerodynamic performance is very poor. Furthermore, the Martian environment cannot be fully replicated on Earth for Mars drone control testing, but the Mars drone control system must operate smoothly during its first Mars exploration mission. Therefore, detailed modeling, analysis, and simulation are essential, and testing must be conducted in a partially replicated environment.

[0005] Considering mission requirements such as vertical takeoff and landing and maximizing exploration range, the Mars drone should adopt a rotorcraft configuration. On Earth, quadcopter drones have proven to be a successful rotorcraft configuration. However, under Martian surface conditions, issues arise such as the inability to achieve the expected lift. To obtain the maximum possible lift under low Reynolds number and low gravity conditions, the "Ingenuity" Mars drone employs a coaxial dual-rotor helicopter structure. The upper and lower rotors are controlled by a single drive motor for stable speed control and three servo motors for cyclical rotor tilt control. By comprehensively adjusting the upper and lower rotor speeds and rotor tilt, the torque and force that change the drone's attitude and speed are obtained, achieving full-drive control within six degrees of freedom. To achieve stable flight control of the Mars drone under Martian atmospheric conditions, the guidance module generates a reference trajectory based on a pre-designed attitude and waypoints, while the flight control module continuously adjusts the rotor tilt to track the predetermined flight trajectory. Simultaneously, a rotor speed observer feedback loop is added to the rotor speed control loop, effectively addressing the reduced flight control robustness caused by strong coupling between rotor speed and heading control.

[0006] To ensure a sufficiently large stability margin for the flight control system, the guidance module of the "Ingenuity" Mars UAV generates specific reference trajectories that decouple the pitch, roll, and directional channels. Based on the equilibrium points provided by the reference trajectories, the UAV's dynamics and kinematics are linearized. On this basis, the flight control module implements single-input, single-output control for the pitch, roll, and directional channels. Altitude and directional are respectively constructed using PID feedback controllers based on the altitude and directional estimates output by the navigation module, while the pitch and roll channels employ a dual-closed-loop control strategy of PID outer-loop position control + PD inner-loop attitude control. However, under Martian atmospheric conditions, rotor tilt adjustment will alter the rotor flap dynamics and the suck-in effect between the upper and lower rotors, resulting in coupling torques between the pitch / roll and pitch / vertical channels. The single-input, single-output flight control scheme risks amplifying these unmodeled coupling torques, reducing the stability domain of the flight control system. Furthermore, the dynamics of the coaxial dual-rotor UAV are related to airspeed. When the ground speed is less than the airspeed, the controller's PID parameters should be adjusted in real-time based on the measured airspeed. However, due to limitations in the size, weight, and accuracy of the airspeed sensor, the navigation module of the "Ingenuity" Mars drone lacks airspeed measurements. It must rely on a reference trajectory with a large stability margin and design a set of PID parameters to overcome interference from Martian gusts less than 3.5 m / s. This specific trajectory, however, limits the flexibility of the Mars drone in completing its surface exploration mission and cannot meet the needs of future drone swarm collaborative Mars surface exploration missions. Summary of the Invention

[0007] To address the shortcomings of existing technologies and meet the future needs of collaborative planetary surface exploration by UAV swarms, this invention first designs a novel micro-coaxial dual variable tilt rotor UAV, KPM-VBLA, for Mars exploration. It uses a magnetically controlled propeller disk to simulate the spatial geometric stability effect of a flexible gyroscope, ensuring its maneuverability to complete the required exploration tasks. To accomplish different exploration tasks, based on the concept of task manifold design, a unified hybrid scheme for trajectory tracking and path following control of the KPM-VBLA Mars UAV is proposed. This method rationally designs a task manifold including the desired trajectory (or path), velocity, and attitude. The designed controller centrally (without loops) ensures that all state variables of the Mars UAV asymptotically converge to the task manifold, and a Lyapunov function is constructed to prove the global asymptotic stability of the control algorithm. Finally, in virtual simulation environments for Mars crater exploration and Mars periodic slope line observation missions, three-dimensional trajectory tracking and path tracking with high-order nonlinear curve characteristics are achieved using KPM-VBLA as the controlled object, verifying the maneuverability of the designed Mars UAV KPM-VBLA and the effectiveness of the control algorithm. The specific technical solution of this invention is as follows:

[0008] A Mars unmanned aerial vehicle (UAV) KPM-VBLA is characterized by adopting a micro coaxial dual variable tilt rotor structure, specifically including: a fuselage body (6), a battery (2) and a wireless communication module (1), an upper rotor (7), a lower rotor (8), two magnetic propeller disks (12), two rotor disks (10), a control module (5) and a camera (9), a motor (13), a motor shaft (14), and an elastic hinge (15); wherein, an equipment bracket is provided between the upper and lower rotors, and a visual sensor is installed on the outside of the equipment bracket along the front direction of the fuselage body; the control module (5) includes a micro inertial measurement unit and a navigation control computer installed in the fuselage body at the center of the equipment bracket; and two rotor disks. The disk (10) is connected to the motor shaft (14) by an elastic hinge (15). The upper and lower rotors are respectively set on the upper and lower rotor disks (10). Two magnetic control disks (12) are located on the same side as the two rotor disks (10). Electromagnetic materials are embedded at equal intervals on the two magnetic control disks (12). The strength and direction of the magnetic field generated by the electromagnetic materials are controlled by the current. Ferromagnetic materials are embedded at equal intervals on the two rotor disks (10) in correspondence with the ferromagnetic materials of the magnetic control disks (12). Under the action of the gyro precession effect, the magnetic force between the magnetic control disks (12) and the rotor disks (10) will control the size and direction of the tilt angle of the upper and lower rotors, so as to realize the variable control of the tilt angle of the upper and lower rotors.

[0009] Furthermore, it also includes a solar panel (4), a landing gear (3), and a protective cover (11); the solar panel (4) is installed above the upper rotor (7) to charge the battery using solar energy; the protective cover (11) is installed below the lower rotor.

[0010] A control method for the KPM-VBLA Mars unmanned aerial vehicle (UAV) as described above, based on the mission manifold, is characterized by comprising the following steps:

[0011] S1: Define the coordinate systems of the KPM-VBLA Mars UAV;

[0012] S2: Establish a six-degree-of-freedom dynamic model of the Mars unmanned aerial vehicle KPM-VBLA;

[0013] S3: Design a mission manifold controller to obtain the input control force and torque of the Mars UAV KPM-VBLA, so as to achieve control of the Mars UAV.

[0014] Furthermore, step S1 specifically includes the following steps:

[0015] S1-1: Define the ground coordinate system and body coordinate system of the Mars UAV KPM-VBLA: Ground coordinate system O g -X g Y g Z g Origin g For fixed points on the Martian surface, where O g X g The axis points north, O g Y g The axis points east, O g Z g The axis points to the center of Mars; the body coordinate system o b -x b y b z b The origin o b CG for the center of gravity of the Mars drone, where o b x b The axis points to the forward-facing vision sensor, o b z b The axis is perpendicular to o b x b The axis points downwards, o b y b The axis is determined by the right-hand rule and points to the right side of the main fuselage.

[0016] S1-2: The position and attitude kinematic equations in the ground coordinate system described in step S1-1 are obtained by transforming the body coordinate system to the ground coordinate system:

[0017]

[0018] Where Y = [P T ,Θ T ] T Let P = [x, y, z] be the generalized position vector of the Mars drone. T and Θ=]φ,θ,ψ] T These are the position vector and attitude angle vector of the Mars drone in the ground coordinate system, respectively, X = [V T ,ω T ] T Let V be the generalized velocity vector of the Mars drone, where V = [u, v, w]. T and ω=[p,q,r] T These are the velocity vector and rotational angular velocity vector in the body coordinate system, respectively. This is the transformation matrix between the body coordinate system and the ground coordinate system. The transformation matrix for mapping the angular velocity of the fuselage from the body coordinate system to the ground coordinate system;

[0019]

[0020]

[0021] Where, c(·)=cos(·), s(·)=sin(·), t(·)=tan(·);

[0022] S1-3: Define the rotor i-fixed coordinate system of the Mars UAV KPM-VBLA. i -x i y i z i i=1 is the upper rotor, i=2 is the lower rotor, and its origin is o. i Located at the center of rotation of rotor i; o i x i The axis points in the direction of the camera within a plane parallel to the plane of rotor i's rotation; i y i The axis is parallel to the plane of rotation of the rotor and perpendicular to o. i x i Axis, pointing to the left; o i z i The direction of the axis can be determined by the right-hand rule, pointing towards the plane of rotation of rotor i, with upward being positive.

[0023] Furthermore, step S2 specifically includes the following steps:

[0024] S2-1: In the body coordinate system described in step S1-1, the velocity and angular velocity dynamic equations of the KPM-VBLA in the Mars UAV can be obtained based on the Newton-Euler equations as follows:

[0025]

[0026] Among them, F b and M b These represent the net external force and net external torque acting on the center of gravity of the Mars drone in the body coordinate system, {I x ,I y ,I z} for Mars drone orbiting o b x b o b y b and o b z b The moment of inertia of the axis, where m is the mass of the Mars drone;

[0027] S2-2: As can be seen from the structure of the Mars drone KPM-VBLA, the net external force acting on the main body of the fuselage includes the Martian gravity vector F. G The lift vector F generated by the rotor rotation T and stamping resistance vector F R The net external torque acting on the fuselage includes the counter-torque vector M generated by the rotor rotation. Q gyro torque vector M G And rotor lift in the body coordinate system o b x b axis and o b y b The torque vector M generated by the component force on the axis T Therefore, the net external force and net external torque acting on the Mars drone KPM-VBLA can be expressed as:

[0028] F b =F G +F T +F R

[0029] M b =M Q +M G +M T

[0030] S2-3: Based on steps S2-1 and S2-2, the complete six-degree-of-freedom dynamic equation model of the Mars UAV KPM-VBLA can be obtained:

[0031]

[0032] Where M = diag[m,m,m,I] x ,I y ,I z [N] represents the mass matrix of the Mars drone. dTo determine the forces and torques related to the generalized velocity vector X of the Mars drone, N d =[m(vr-wq),m(wp-ur),m(uq-vp),I y -I z )rq,(I z -I x )pr,(I x -I y )pq] T N f These are the forces and torques exerted on the main body of the fuselage by external forces. N u The forces and torques generated by the task flow controller, i.e., the control law of the task flow controller.

[0033] Furthermore, step S3 specifically includes the following steps:

[0034] S3-1: Considering that the flight trajectory or path of the Mars UAV KPM-VBLA can be described by the intersection of a pair of surfaces in three-dimensional space, the expected trajectory of the Mars UAV KPM-VBLA is defined as an implicit function with respect to position P, resulting in:

[0035]

[0036] Where t is time, j A1(t), j A2(t) are three-dimensional row vectors pre-designed based on the flight trajectory, and b(t) are one-dimensional vectors pre-designed based on the flight trajectory. j P = [x j ,y j ,z j ] T Let e ​​be a vector consisting of the j-th powers of the position coordinate components of the Mars drone, where j represents the order describing the desired trajectory or path, and e is the power of the vector. N1 and e N2 For trajectory tracking error components; when N(p,t)=[e N1 e N2 0] T ≠0 3×1 When this occurs, there exists a trajectory tracking error component e. N1 and e N2 A corresponding control law N needs to be designed. u This makes the trajectory tracking error component e N1 →0 and e N2 →0;

[0037] S3-2: Considering the attitude stability during the flight of the Mars UAV KPM-VBLA, the attitude error Φ(P,Θ,t) is defined as the difference between the attitude Θ of the Mars UAV KPM-VBLA and the desired attitude Θ*(P,Θ,t), resulting in:

[0038] Φ(P,Θ,t)=Θ-Θ*(P,Θ,t)

[0039] A corresponding control law N needs to be designed. u This makes the attitude error Φ(P,Θ,t)→0. 3×1 ;

[0040] S3-3: Combining steps S3-1 and S3-2, define the trajectory manifold Ψ. tr =0 6×1 ,get:

[0041]

[0042] Ultimately, the generalized position vector Y of the Mars drone KPM-VBLA needs to converge to the trajectory manifold Ψ. tr =0 6×1 To ensure that the Mars drone KPM-VBLA flies along the desired trajectory with a stable attitude;

[0043] Due to Ψ tr If Ψ is at least first-order differentiable for all independent variables, then tr The total differential with respect to time t can be expressed as:

[0044]

[0045] in,

[0046]

[0047]

[0048]

[0049] S3-4: To reduce energy consumption during flight, the generalized velocity vector X of the Mars UAV KPM-VBLA is constrained when flying along the desired trajectory. The velocity manifold is defined as follows:

[0050]

[0051] in, V is the velocity squared error vector. c The desired flight velocity vector; ultimately, the generalized velocity vector X of the Mars drone KPM-VBLA needs to converge to the velocity manifold Ψ. v =0 6×1This allows the Mars drone KPM-VBLA to achieve the desired velocity vector V. c Maintain a stable spatial geometry during flight;

[0052] S3-5: To ensure that the Mars drone KPM-VBLA flies along the desired attitude and velocity of the exploration mission, the trajectory manifold Ψ is utilized. tr =0 and velocity manifold Ψ v =0, synthesizing task manifold Ψ=0, yields:

[0053] Ψ=Ψ tr +AΨ v =0

[0054] Where A is a 6th-order diagonal weighted coefficient matrix, the matrix A is designed such that if and only if Ψ tr →0 and Ψ v →0, then Ψ→0, therefore, the trajectory tracking error component e in step 3-1 can also be made... N1 →0 and e N2 →0 and the attitude error Φ(P,Θ,t) in step 3-2 →0 3×1 ;

[0055] S3-6: To ensure that the generalized position vector Y, generalized velocity vector X, and their first derivatives of the Mars UAV KPM-VBLA converge to the mission manifold Ψ = 0, a 6th-order constant coefficient matrix T is designed to satisfy the following formula:

[0056]

[0057] S3-7: Based on step S2-3, the KPM-VBLA dynamic equations of the Mars UAV are obtained. Combining steps S3-5 and S3-6, the nonlinear mission manifold control law N is obtained. u for:

[0058]

[0059] Where, K0=J1J ∑PX +J Θ J ∑ΘX K1 = TAK 11 +K 12 K 11 =[0 6×3 J1J ∑PΘ +J Θ J ∑ΘΘ ], K 12 =(T+A)J s +TAГ s , J1 = J P +J V JV =

[0060] The nonlinear task manifold control law N u It can be guaranteed that all state variables asymptotically converge to the task manifold Ψ = 0, that is... Based on the aforementioned nonlinear task manifold control law N u This allows us to obtain the forces and torques generated by the task manifold controller.

[0061] Furthermore, to prove the nonlinear task manifold control law N in step S3-7... u Global asymptotic stability, select

[0062] Lyapunov function Taking the derivative with respect to time, we get:

[0063]

[0064] Among them, to make Only matrix T in steps S3-6 is needed -1 Minimum eigenvalue T min A value greater than 0 is sufficient. Considering that T is a diagonal matrix, it is only necessary for all diagonal elements of T to be positive.

[0065] The simulation method for trajectory tracking and path following control of the KPM-VBL Mars UAV using the aforementioned control method involves constructing a 3D virtual scene of the Santa Maria Martian crater and the periodic slope lines of Mars based on Unity 3D in a virtual simulation environment, and then applying the task manifold control law N designed in step S3 above. u The simulation of controlling the KPM-VBLA Mars drone to realize the Mars crater exploration mission and the Mars periodic slope line observation mission.

[0066] Furthermore, for the Santa Maria Mars Crater exploration mission, the desired trajectory is designed to descend along the crater wall into the crater, then fly around the crater wall in a spiral, and finally fly out of the crater and fly in a straight line to the desired destination. The trajectory tracking control is performed using the mission manifold controller in step 3 and the existing sliding mode controller.

[0067] Furthermore, for the Mars periodic slope line observation mission, the desired trajectory is designed as a high-order curve that conforms to the slope line stripe characteristics, and the path following control is performed by the mission manifold controller in step 3.

[0068] The beneficial effects of this invention compared to the prior art are as follows:

[0069] 1. This invention addresses the exploration needs of Martian unmanned aerial vehicles (UAVs) for nonlinear geometric features such as Martian craters and periodic slope lines. Considering the numerous challenges UAVs face when performing missions in the Martian environment, an innovative micro-coaxial dual variable tilt rotor UAV, KPM-VBLA, was designed. Unlike traditional underactuated quadcopter UAVs, this UAV uses magnetically controlled propeller disks to decompose lift, achieving full-drive control of attitude and velocity. It can fly in a geometrically stable state, making it more suitable for Martian exploration missions primarily focused on visual information acquisition. A six-degree-of-freedom nonlinear dynamic model of the KPM-VBLA Martian UAV was established using Newton-Euler vector mechanics.

[0070] 2. Inspired by the cooperative synthesis control method, this invention proposes a mission manifold controller to accomplish nonlinear geometric terrain probing missions. The mission manifold is a weighted synthesis of the trajectory manifold and the velocity manifold, where the trajectory manifold represents the desired trajectory and attitude error as a high-order nonlinear implicit function. A velocity manifold is designed to constrain the velocity of the Mars drone during flight. The convergence of all state variables in the closed-loop system is guaranteed by designing the weighting coefficient matrix and cooperative gain matrix of the mission manifold controller. The stability analysis of this controller is proven by the constructed Lyapunov function.

[0071] 3. In the 3D virtual simulation environment of two missions—exploration of Martian craters and observation of Martian periodic slope lines—the Mars UAV KPM-VBLA achieved rapid tracking of exploration trajectories similar to high-order nonlinear geometric terrain with relatively low energy consumption, verifying the effectiveness of the mission manifold control algorithm. Attached Figure Description

[0072] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the embodiments will be briefly described below. Referring to the accompanying drawings will provide a clearer understanding of the features and advantages of the present invention. The drawings are illustrative and should not be construed as limiting the present invention in any way. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort. Wherein:

[0073] Figure 1 This is a structural diagram of the KPM-VBLA micro coaxial dual variable tilt rotor Mars unmanned aerial vehicle of the present invention;

[0074] Figure 2 This is a cross-sectional view of the KPM-VBLA magnetocontrol propeller disk structure of the Mars unmanned aerial vehicle of the present invention;

[0075] Figure 3 This is a schematic diagram illustrating the coordinate system definitions of the KPM-VBLA Mars UAV of the present invention;

[0076] Figure 4This is an image of the Santa Maria Martian crater taken by a high-resolution imaging scientific experiment according to an embodiment of the present invention.

[0077] Figure 5 (a) is a 3D virtual model of the Santa Maria Martian Crater according to an embodiment of the present invention;

[0078] Figure 5 (b) is a trajectory diagram of the Mars drone in the Santa Maria crater of Mars according to an embodiment of the present invention;

[0079] Figure 6 This is an airspeed tracking curve of the KPM-VBLA Mars drone in an embodiment of the present invention for a mission to explore Martian craters.

[0080] Figure 7 (a) and (b) are the speed and attitude response diagrams of the KPM-VBLA Mars UAV in the embodiment of the present invention for exploring Martian craters.

[0081] Figure 8 This is a position error curve of the KPM-VBLA Mars drone used in this invention for exploring Martian craters.

[0082] Figure 9 (a) and (b) are respectively the feedback control force and torque curves of the KPM-VBLA Mars drone in the embodiment of the present invention for exploring Martian craters.

[0083] Figure 10 This is a high-resolution imaging scientific experiment of this invention, showing a periodic slope line map of Mars.

[0084] Figure 11 (a) is a 3D virtual model of the periodic slope line of Mars according to an embodiment of the present invention;

[0085] Figure 11 (b) is a trajectory diagram of the Mars drone in a 3D virtual model of the periodic slope line of Mars according to an embodiment of the present invention;

[0086] Figure 12 This is an airspeed tracking curve of the KPM-VBLA Mars drone, an embodiment of the present invention, during a Mars periodic slope line observation mission.

[0087] Figure 13 (a) and (b) are the speed and attitude response diagrams of the Mars UAV KPM-VBLA performing a Mars periodic slope line observation mission according to an embodiment of the present invention.

[0088] Figure 14 This is a position error curve diagram of the Mars periodic slope line observation mission performed by the KPM-VBLA Mars unmanned aerial vehicle according to an embodiment of the present invention;

[0089] Figure 15 (a) and (b) are the feedback control force and torque curves of the Mars UAV KPM-VBLA performing periodic slope line observation mission on Mars according to an embodiment of the present invention.

[0090] Explanation of reference numerals in the attached figures:

[0091] 1-Wireless communication module, 2-Battery, 3-Landing bracket, 4-Solar panel, 5-Control module, 6-Fuselage body, 7-Upper rotor, 8-Lower rotor, 9-Camera, 10-Rotor disk, 11-Protective cover, 12-Magnetic control rotor disk, 13-Motor, 14-Motor shaft, 15-Flexible hinge. Detailed Implementation

[0092] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments of the present invention and the features thereof can be combined with each other. Many specific details are set forth in the following description to provide a thorough understanding of the present invention; however, the present invention may also be implemented in other ways different from those described herein. Therefore, the scope of protection of the present invention is not limited to the specific embodiments disclosed below.

[0093] This invention proposes a nonlinear mission manifold controller based on the trajectory (path) tracking problem of a Mars unmanned aerial vehicle (UAV), enabling the UAV to achieve stable flight with reduced energy consumption while completing the trajectory (path) tracking task. First, a trajectory manifold composed of the desired trajectory and attitude with nonlinear geometric characteristics is defined for the Mars UAV. To reduce the energy consumption of the control system, a velocity manifold is defined to describe the generalized velocity constraints of the Mars UAV. Based on this, a six-dimensional mission manifold is synthesized using weighted averages. Then, the feedback control force and control torque of the Mars UAV's KPM-VBLA, i.e., the mission manifold control law N, are designed using the mission manifold. u This approach allows the generalized position and generalized velocity to converge to the trajectory manifold and velocity manifold, respectively. The stability of this nonlinear mission manifold controller is analyzed. This scheme avoids the decomposition of "internal" and "external" feedback loops in the controller, eliminating the need to process 12 state variables (6 generalized position and 6 generalized velocity) in a cascaded manner. Instead, it directly utilizes the six-dimensional mission manifold to obtain the control force and control torque of the Mars UAV KPM-VBLA, ensuring stability of velocity and angular velocity while bringing the trajectory tracking error to zero.

[0094] Specifically, such as Figure 1The diagram shows the structure of the KPM-VBLA, a micro coaxial dual variable tilt rotor Mars drone designed according to this invention. The KPM-VBLA Mars drone adopts a micro coaxial dual variable tilt rotor structure, specifically including: a solar panel 4, a landing gear 3, a fuselage body 6, a battery 2 and a wireless communication module 1, an upper rotor 7, a lower rotor 8, two magnetic propeller disks 12, two rotor disks 10, a control module 5, a camera 9, a protective cover 11, a motor 13, and a motor shaft 14. The camera 9 is mounted on the control module 5. A visual sensor is installed on the outer side of the equipment bracket at the center of the upper and lower rotors along the front direction of the fuselage body. A micro inertial measurement unit and a navigation control computer are installed in the fuselage body at the center of the equipment bracket as the control module. The small solar panel 4 above the upper rotor uses solar energy to charge the battery. A protective cover 11 is installed under the lower rotor to block Martian sand particles stirred up by the rotor rotation during takeoff and landing, reducing potential damage to the rotor shaft caused by Martian sand particles. The rotor disks 10 and the motor shaft 14 are connected by elastic hinges 15. like Figure 2 The magnetic control rotor disks 12 shown are located coaxially with the upper and lower rotor disks 10, which are fixed to the upper and lower rotors, respectively. As long as they are coaxial and adjacent, the magnetic control rotor disks 12 can be located either above or below the rotor disks 10. Electromagnetic materials are embedded at equal intervals on the magnetic control rotor disks 12, and the strength and direction of the magnetic field generated by these materials are controlled by an electric current. On the rotor disks 10, which are fixed to the motor shaft 14 on the fuselage body 6, ferromagnetic materials are embedded at equal intervals corresponding to the electromagnetic materials of the magnetic control rotor disks 12. Under the gyro precession effect, since the rotor disks and motor shafts are connected by an elastic hinge, the tilt magnitude and direction of the rotor disks can be controlled by controlling the magnetic field on the magnetic control rotor disks. In other words, the magnetic force between the magnetic control rotor disks 12 and the rotor disks 10 controls the magnitude and direction of the tilt angle of the upper and lower rotors, achieving variable control of the tilt angle of the upper and lower rotors.

[0095] A Mars unmanned aerial vehicle and its control method based on a mission manifold controller include the following steps:

[0096] S1: Define the coordinate systems of the KPM-VBLA micro coaxial dual variable tilt rotor Mars UAV. The coordinate system definitions are as follows: Figure 3 As shown, o i -x i y i z i Let i be the fixed coordinate system of the rotor, o b (CG)-x b y b z b For the body coordinate system, O g -X g Y g Z gThis is a ground coordinate system.

[0097] Preferably, the specific steps of step S1 are as follows:

[0098] S1-1: Define the ground coordinate system and body coordinate system of the Mars UAV KPM-VBLA: Ground coordinate system O g -X g Y g Z g Origin g For fixed points on the Martian surface, where O g X g The axis points north, O g Y g The axis points east, O g Z g The axis points to the center of Mars; the body coordinate system o b -x b y b z b The origin o b CG for the center of gravity of the Mars drone, where o b x b The axis points to the forward-facing vision sensor, o b z b The axis is perpendicular to o b x b The axis points downwards, o b y b The axis is determined by the right-hand rule and points to the right side of the main fuselage.

[0099] S1-2: The position and attitude kinematic equations in the ground coordinate system described in step S1-1 are obtained by transforming the body coordinate system to the ground coordinate system:

[0100]

[0101] Where Y = [P T ,Θ T ] T Let P = [x, y, z] be the generalized position vector of the Mars drone. T and Θ=[φ,θ,ψ] T These are the position vector and attitude angle vector of the Mars drone in the ground coordinate system, respectively, X = [V T ,ω T ] T Let V be the generalized velocity vector of the Mars drone, where V = [u, v, w]. T and ω=[p,q,r] T These are the velocity vector and rotational angular velocity vector in the body coordinate system, respectively. This is the transformation matrix between the body coordinate system and the ground coordinate system. The transformation matrix for mapping the angular velocity of the fuselage from the body coordinate system to the ground coordinate system;

[0102]

[0103]

[0104] Where, c(·)=cos(·), s(·)=sin(·), t(·)=tan(·);

[0105] S1-3: Define the rotor i-fixed coordinate system of the Mars UAV KPM-VBLA. i -x i y i z i i=1 is the upper rotor, i=2 is the lower rotor, and its origin is o. i Located at the center of rotation of rotor i; o i x i The axis points in the direction of the camera within a plane parallel to the plane of rotor i's rotation; i y i The axis is parallel to the plane of rotation of the rotor and perpendicular to o. i x i Axis, pointing to the left; o i z i The direction of the axis can be determined by the right-hand rule, pointing towards the plane of rotation of rotor i, with upward being positive.

[0106] S2: Establish a six-degree-of-freedom dynamic model of the Mars unmanned aerial vehicle KPM-VBLA.

[0107] Preferably, S2-1: In the body coordinate system described in step S1-1, the velocity and angular velocity dynamic equations of the KPM-VBLA in the Mars UAV can be obtained based on the Newton-Euler equations as follows:

[0108]

[0109] Among them, F b and M b These represent the net external force and net external torque acting on the center of gravity of the Mars drone in the body coordinate system, {I x ,I y ,I z} for Mars drone orbiting o b x b o b y b and o b z b The moment of inertia of the axis, where m is the mass of the Mars drone;

[0110] S2-2: As can be seen from the structure of the Mars drone KPM-VBLA, the net external force acting on the main body of the fuselage includes the Martian gravity vector F. G The lift vector F generated by the rotation of the upper and lower rotors T and stamping resistance vector F R The net external torque acting on the fuselage includes the counter-torque vector M generated by the rotation of the upper and lower rotors. Q gyro torque vector M G The lift of the upper and lower rotors in the body coordinate system b x b axis and o b y b The torque vector M generated by the component force on the axis T The force and torque vectors related to the upper and lower rotors mentioned above can be obtained in the rotor i-fixed coordinate system and then transformed to the body coordinate system; therefore, the net external force and net external torque acting on the Mars UAV KPM-VBLA can be expressed as:

[0111] F b =F G +F T +F R

[0112] M b =M Q +M G +M T

[0113] S2-3: Based on steps S2-1 and S2-2, the complete six-degree-of-freedom dynamic equation model of the Mars UAV KPM-VBLA can be obtained:

[0114]

[0115] Where M = diag[m,m,m,I] x ,I y ,I z [N] represents the mass matrix of the Mars drone. d To determine the forces and torques related to the generalized velocity vector X of the Mars drone, N d =[m(vr-wq),m(wp-ur),m(uq-vp),I y -I z )rq,(I z -I x )pr,(I x -I y )pq] T N f These are the forces and torques exerted on the main body of the fuselage by external forces. N uThe forces and torques generated by the task flow controller, i.e., the control law of the task flow controller.

[0116] S3: Design a nonlinear task manifold control law N for a task manifold controller. u That is, the input control force and torque of the Mars UAV KPM-VBLA described in step S2 are obtained to achieve control of the Mars UAV.

[0117] Preferably, the specific steps of step S3 are as follows:

[0118] S3-1: Considering that the flight trajectory or path of the Mars UAV KPM-VBLA can be described by the intersection of a pair of surfaces in three-dimensional space, the expected trajectory of the Mars UAV KPM-VBLA is defined as an implicit function with respect to position P, resulting in:

[0119]

[0120] Where t is time, j A1(t), j A2(t) are three-dimensional row vectors pre-designed based on the flight trajectory, and b(t) are one-dimensional vectors pre-designed based on the flight trajectory. The flight trajectory can change with time or remain unchanged. j P = [x j ,y j ,z j ] T Let e ​​be a vector consisting of the j-th powers of the position coordinate components of the Mars drone, where j represents the order describing the desired trajectory or path, and e is the power of the vector. N1 and e N2 For trajectory tracking error components; when N(P,t)=[e N1 e N2 0] T ≠0 3×1 When this occurs, there exists a trajectory tracking error component e. N1 and e N2 The corresponding control law N needs to be designed in the following steps. u This makes the trajectory tracking error component e N1 →0 and e N2 →0;

[0121] S3-2: Considering the attitude stability during the flight of the Mars UAV KPM-VBLA, the attitude error Φ(P,Θ,t) is defined as the difference between the attitude Θ of the Mars UAV KPM-VBLA and the desired attitude Θ*(P,Θ,t), resulting in:

[0122] Φ(P,Θ,t)=Θ-Θ*(P,Θ,t)

[0123] The corresponding control law N needs to be designed in the following steps. u This makes the attitude error Φ(P,Θ,t)→0. 3×1 ;

[0124] S3-3: Combining steps S3-1 and S3-2, define the trajectory manifold Ψ. tr =0 6×1 ,get:

[0125]

[0126] Ultimately, the generalized position vector Y of the Mars drone KPM-VBLA needs to converge to the trajectory manifold Ψ. tr =0 6×1 To ensure that the Mars drone KPM-VBLA flies along the desired trajectory with a stable attitude;

[0127] Due to Ψ tr If Ψ is at least first-order differentiable for all independent variables, then tr The total differential with respect to time t can be expressed as:

[0128]

[0129] in,

[0130]

[0131]

[0132]

[0133] S3-4: To reduce energy consumption during flight, the generalized velocity vector X of the Mars UAV KPM-VBLA is constrained when flying along the desired trajectory. The velocity manifold is defined as follows:

[0134]

[0135] in, V is the velocity squared error vector. c The desired flight velocity vector; ultimately, the generalized velocity vector X of the Mars drone KPM-VBLA needs to converge to the velocity manifold Ψ. v =0 6×1 This allows the Mars drone KPM-VBLA to achieve the desired velocity vector V. c Maintain a stable spatial geometry during flight;

[0136] S3-5: To ensure that the Mars drone KPM-VBLA flies along the desired attitude and velocity of the exploration mission, the trajectory manifold Ψ is utilized. tr =0 and velocity manifold Ψ v =0, synthesizing task manifold Ψ=0, yields:

[0137] Ψ=Ψ tr +AΨ v =0

[0138] Where A is a 6th-order diagonal weighted coefficient matrix, the matrix A is designed such that if and only if Ψ tr →0 and Ψ v →0, then Ψ→0, therefore, the trajectory tracking error component e in step 3-1 can also be made... N1 →0 and e N2 →0 and the attitude error Φ(P,Θ,t) in step 3-2 →0 3×1 It can be seen that the synthesis of the task manifold Ψ=0 allows the controller to handle only 6-dimensional problems, thus unifying the flight control tasks of stabilizing attitude and velocity with the trajectory tracking tasks.

[0139] S3-6: To ensure that the generalized position vector Y, generalized velocity vector X, and their first derivatives of the Mars UAV KPM-VBLA converge to the mission manifold Ψ = 0, a 6th-order constant coefficient matrix T, also known as the cooperative gain matrix, is designed to satisfy the following formula:

[0140]

[0141] The convergence of all state variables in the closed-loop system is guaranteed by designing the weighting coefficient matrix and the cooperative gain matrix of the task manifold controller.

[0142] S3-7: Based on step S2-3, the KPM-VBLA dynamic equations of the Mars UAV are obtained. Combining steps S3-5 and S3-6, the nonlinear mission manifold control law N is obtained. u for:

[0143]

[0144] Where, K0=J1J ∑PX +J Θ J ∑ΘX K1 = TAK 11 +K 12 K 11 =[0 6×3 J1J ∑PΘ +J Θ J ∑ΘΘ ], K 12 =(T+A)J s +TAГ s , J1 = J P +J V ,

[0145] The nonlinear task manifold control law N u It can be guaranteed that all state variables asymptotically converge to the task manifold Ψ = 0, that is... Based on the aforementioned nonlinear task manifold control law N u The matrix can be decomposed to obtain the forces and torques generated by the task manifold controller.

[0146] To prove step S3-7, the nonlinear task manifold control law N u For global asymptotic stability, the Lyapunov function is chosen.

[0147] Taking the derivative with respect to time, we get:

[0148]

[0149] Among them, to make Only matrix T in steps S3-6 is needed -1 Minimum eigenvalue T min A value greater than 0 is sufficient. Considering that T is a diagonal matrix, it is only necessary for all diagonal elements of T to be positive.

[0150] Based on this, this invention utilizes a 3D virtual scene constructed using Unity 3D, depicting the Santa Maria crater and Martian periodic slope lines, to conduct control simulations of the KPM-VBLA Mars UAV performing crater exploration and periodic slope line observation missions. This further verifies the control effect and accuracy of the mission manifold controller, as well as the controlled performance of the designed micro-coaxial dual variable tilt rotor Mars UAV KPM-VBLA. The specific method is as follows: For the Santa Maria crater exploration mission, the desired trajectory is designed to descend along the crater wall into the crater, then fly in a spiral around the crater wall, and finally fly out of the crater and then in a straight line to the desired endpoint. Trajectory tracking control is performed using the mission manifold controller proposed in this invention and a sliding mode controller, which currently offers superior trajectory tracking control. For the Martian periodic slope line observation mission, the desired trajectory is designed as a high-order curve conforming to the slope line stripe characteristics, and path following control is performed using the mission manifold controller proposed in this invention.

[0151] The following specific embodiment illustrates the control effect and accuracy verification of the mission manifold controller of the present invention in controlling the KPM-VBLA Mars UAV to perform a Mars exploration mission. Figure 5 and Figure 11 This is a 3D virtual model of the Santa Maria Martian crater and the periodic slope lines of Mars constructed according to the present invention.

[0152] Considering the need for comprehensive and detailed imaging of the Martian crater interior, the KPM-VBLA Mars UAV mission is designed to fly to above the crater rim, descend diagonally to near the crater bottom, then spiral upwards along the crater wall to the top for scanning images of the crater wall, and finally fly diagonally back to the crater edge. The descent depth is set at 6m, the spiral axis is located at the center of the crater, the radius of the spiral's base and top surfaces are both 40m, and the vertical ascent speed is 0.01m / s. The desired velocity V is set. c =1 m / s, initial position P0 = [-8, -40, 0] T m, initial linear velocity v0 = [0.1, 0.2, 0] T m / s, initial attitude Θ0=[0,0,0] T The rad value is used in all other states, and is 0 in all other states. The designed exploration trajectory is a concatenation of three continuous curves, each described by one of the following three manifolds.

[0153]

[0154]

[0155]

[0156] Where R = 40m, t1 = 10s is the time for the Mars rover to switch from the oblique path to the spiral trajectory, and t2 = (10 + 200π)s is the time for the Mars drone KPM-VBLA to switch from the spiral trajectory to the straight path.

[0157] like Figure 8 As shown, under both the mission manifold controller and the sliding mode controller, the steady-state error of trajectory tracking within the 700s trajectory tracking mission time does not exceed 0.3m, indicating that the Mars UAV can accurately track the quadratic mission trajectory under both controllers to complete the Martian crater exploration mission. During the first 10s of the oblique descent, the sliding mode controller has a position error overshoot of 0.09m at startup, which is 4 times that of the mission manifold controller, and a steady-state error of 0.013m. At the 10s, the KPM-VBLA ends the oblique descent and begins a spiral ascent. At this time, the mission manifold controller has an initial position error of 0.275m, which is higher than the 0.1m of the sliding mode controller. However, the convergence time is 2s, which is lower than the 5s of the sliding mode controller, and the steady-state error is 0.0011m, slightly less than the 0.0013m of the sliding mode controller. At 638s, the KPM-VBLA ends the spiral ascent and begins to fly obliquely to the edge of the crater. At this point, the task manifold controller and sliding mode controller transition smoothly, and the position error of the task manifold controller returns to zero. This is because the inclined line is exactly the tangent to the spiral, and the two trajectories connect smoothly. Figure 8It can be seen that the sliding mode controller exhibits significant speed overshoot at the start of the KPM-VBLA descent and at the 10-second KPM-VBLA trajectory switch, with overshoot rates as high as 28% and 10%, respectively, while the task manifold controller shows no speed overshoot. (Comparison) Figure 7 (a) From the perspective of velocity components, the velocity changes in the x and z directions are drastic, which explains why the trajectory error of the mission manifold controller is higher than that of the sliding mode controller when the KPM-VBLA performs trajectory switching at 10 seconds. The mission manifold controller performs well in velocity tracking because it considers velocity constraints in the synthesis of the mission manifold. In terms of convergence time, the mission manifold controller is comparable to, or even lower than, the sliding mode controller. This is because the mission manifold controller unifies the traditional "inner" and "outer" loop command tracking errors during mission manifold synthesis, thus accelerating the response speed of the Mars UAV to control commands. Figure 7 (b) The three-axis attitude remains at 0, indicating that the designed Mars UAV KPM-VBLA has spatial geometric stability capability.

[0158] In the extreme environment of Mars, the limited power of drones needs to be allocated to functions such as flight control, imaging, communication, and survival heating. Therefore, it is necessary to optimize the flight control module to minimize power consumption while maintaining control accuracy. Figure 9 As shown in (a), the maximum feedback control force in the three axes under the mission manifold controller does not exceed 1.5N. Specifically, the feedback control force in the x-axis direction is 0.3N at 10s, generating x-axis acceleration for tracking along the helical trajectory. The feedback control force in the z-axis direction is -1.4N at 10s, indicating that the Mars UAV generates acceleration in the opposite direction to the z-axis at 10s for deceleration and descent. Figure 8 The attitude angle of the Mars drone remained at 0 throughout. Figure 9 (b) The three-axis feedback control torque of the Mars UAV remains at 0 throughout the trajectory tracking mission, indicating that the attitude of the Mars UAV remains stable throughout the mission, ensuring stable camera images during imaging for data acquisition. While stable trajectory tracking can be achieved under the sliding mode controller, the feedback control force along the x-axis is 0.5 N and the feedback control force along the z-axis is -1.7 N when the trajectory is switched at the 10-second mark. These figures are 1.67 times and 1.21 times those under the mission manifold controller, respectively, placing higher demands on energy supply.

[0159] Simulation results demonstrate that the designed Mars UAV KPM-VBLA possesses satisfactory flight performance and spatial geometric stability, enabling it to successfully complete the Santa Maria crater exploration mission. This also proves the effectiveness of the proposed mission manifold controller design method. Furthermore, considering the unique challenges of trajectory tracking accuracy and energy conservation in the Martian environment, the proposed mission manifold controller offers certain advantages for the trajectory tracking task of the KPM-VBLA Mars UAV during Mars exploration missions.

[0160] While advanced nonlinear controllers such as sliding mode controllers can solve the control problem of quadratic task trajectories, they cannot meet the control requirements of higher-order task trajectories. Let the desired velocity V... c =1m / s, initial position P0=

[0161] [-35,-120,-5] T m, initial linear velocity v0 = [0.1, -0.2, 0] T m / s, initial attitude Θ0=[0,0,0] T rad, other states are 0. Considering the... Figure 14 The mission to observe the periodic slope lines of Mars along their unique shape requires fitting a high-order curve that conforms to the shape characteristics of the Martian periodic slope lines as the desired path for the observation mission, described by the following manifold.

[0162]

[0163] Depend on Figure 14 It can be seen that the initial point of the Mars UAV KPM-VBLA was not on the desired path. After about 10 seconds of flight, it converged to the desired path, and during the subsequent path tracking process, the tracking error always approached 0, demonstrating the good performance of the Mars UAV KPM-VBLA in tracking the 4th order curve shown in the manifold above under the mission manifold controller. Figure 12 The data shows that the KPM-VBLA Mars drone converged to the desired speed of 1 m / s in 3.5 seconds, and maintained good speed stability until 400 seconds later, although the velocity components in the x and y axes... Figure 13 (a) shows the variation. For example... Figure 13 (b) The attitude remained in a spatially geometrically stable state throughout the 400s flight, despite... Figure 11 The flight path in (a) has three turning points: A, B, and C. This demonstrates the effect of considering kinematic constraints in the velocity manifold and the ability of the designed Mars UAV KPM-VBLA to maintain geometric stability while maneuvering.

[0164] Figure 15(a) and (b) show the feedback control force and torque obtained from the control law of the mission manifold. At the initial moment, there are force changes of 0.1N and 0.2N in the x and y axes, respectively. This is to allow the Mars drone to quickly converge to the desired path. Then, at 140s, 175s, and 250s, force changes of no more than 0.2N occur, corresponding to… Figure 11 (a) The forces required at the three turning points A, B, and C are used to change the velocity direction of the Mars drone. The control force in the z-direction remains at -1N because the Mars drone is initially at the desired altitude of -5m and maintains this altitude throughout the flight; changes in the x and y directions do not cause additional motion in the z-direction. The constant zero torque across all three axes corresponds to the geometrically stable state of the Mars drone's KPM-VBLA. The changes in feedback force and torque further demonstrate the low power consumption requirements of the mission manifold controller.

[0165] Simulation results demonstrate that the designed Mars UAV KPM-VBLA retains spatial geometric stability even when the velocity direction changes, and does not generate coupling between axes, making it capable of handling high-order path tracking tasks. Furthermore, the effectiveness of the Mars UAV KPM-VBLA in maneuvering along high-order desired paths under the mission manifold controller is verified; that is, the mission manifold controller ensures that path tracking errors and velocity tracking errors converge rapidly to zero.

[0166] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A control method for a Mars unmanned aerial vehicle (UAV) based on a mission manifold, wherein the Mars UAV employs a micro-coaxial dual variable tilt rotor structure, comprising: The fuselage consists of a main body (6), a battery (2) and a wireless communication module (1), an upper rotor (7), a lower rotor (8), two magnetic propeller disks (12), two rotor disks (10), a control module (5) and a camera (9), a motor (13), a motor shaft (14), and a flexible hinge (15). An equipment bracket is provided between the upper and lower rotors. A visual sensor is installed on the outside of the equipment bracket along the front direction of the main body. The control module (5) includes a micro-inertial measurement unit and a navigation control computer installed in the main body at the center of the equipment bracket. The two rotor disks (10) are connected to the motor shaft (14) by flexible hinges (15). The lower rotor is respectively mounted on the upper and lower rotor disks (10); two magnetically controlled rotor disks (12) are respectively located on one side coaxial with the two rotor disks (10), and electromagnetic materials are embedded at equal intervals on the two magnetically controlled rotor disks (12), and the magnetic field strength and direction generated by the electromagnetic materials are controlled by current; on the two rotor disks (10), ferromagnetic materials are embedded at equal intervals corresponding to the electromagnetic materials of the magnetically controlled rotor disks (12); under the action of gyro precession effect, the magnetic force between the magnetically controlled rotor disks (12) and the rotor disks (10) will control the magnitude and direction of the tilt angle of the upper and lower rotors, realizing variable control of the tilt angle of the upper and lower rotors, characterized by including the following steps: S1: Define the coordinate systems of the Mars UAV; S2: Establish a six-degree-of-freedom dynamic model of the Mars drone; S3: Design a nonlinear task manifold control law N for a task manifold controller. u The input control force and torque of the Mars drone are obtained to achieve control of the Mars drone; Step S1 specifically includes the following steps: S1-1: Define the ground coordinate system and body coordinate system of the Mars UAV: ​​Ground coordinate system O g -X g Y g Z g Origin g For fixed points on the Martian surface, where O g X g The axis points north, O g Y g The axis points east, O g Z g The axis points to the center of Mars; the body coordinate system o b -x b y b z b The origin o b CG for the center of gravity of the Mars drone, where o b x b The axis points to the forward-facing vision sensor, o b z b The axis is perpendicular to o b x b The axis points downwards, o b y b The axis is determined by the right-hand rule and points to the right side of the main fuselage. S1-2: The position and attitude kinematic equations in the ground coordinate system described in step S1-1 are obtained by transforming the body coordinate system to the ground coordinate system: Where Y = [P T ,Θ T ] T Let P = [x, y, z] be the generalized position vector of the Mars drone. T and Θ=[φ,θ,ψ] T These are the position vector and attitude angle vector of the Mars drone in the ground coordinate system, respectively, X = [V T ,ω T ] T Let V be the generalized velocity vector of the Mars drone, where V = [u, v, w]. T and ω=[p,q,r] T These are the velocity vector and rotational angular velocity vector in the body coordinate system, respectively. This is the transformation matrix between the body coordinate system and the ground coordinate system. The transformation matrix for mapping the angular velocity of the fuselage from the body coordinate system to the ground coordinate system; Where, c(·)=cos(·), s(·)=sin(·), t(·)=tan(·); S1-3: Define the fixed coordinate system of the rotor i of the Mars UAV. i -x i y i z i i=1 is the upper rotor, i=2 is the lower rotor, and its origin is o. i Located at the center of rotation of rotor i; o i x i The axis points in the direction of the camera within a plane parallel to the plane of rotor i's rotation; i y i The axis is parallel to the plane of rotation of the rotor and perpendicular to o. i x i Axis, pointing to the left; o i z i The direction of the axis can be determined by the right-hand rule, pointing towards the plane of rotation of rotor i, with upward being positive; Step S2 specifically includes the following steps: S2-1: In the body coordinate system described in step S1-1, the velocity and angular velocity dynamic equations of the Mars UAV are obtained based on the Newton-Euler equations as follows: Among them, F b and M b These represent the net external force and net external torque acting on the center of gravity of the Mars drone in the body coordinate system, {I x ,I y ,I z } for Mars drone orbiting o b x b o b y b and o b z b The moment of inertia of the axis, where m is the mass of the Mars drone; S2-2: As can be seen from the structure of the Mars drone, the net external force acting on the main body of the fuselage includes the Martian gravity vector F. G The lift vector F generated by the rotation of the upper and lower rotors T and stamping resistance vector F R The net external torque acting on the fuselage includes the counter-torque vector M generated by the rotation of the upper and lower rotors. Q gyro torque vector M G The lift of the upper and lower rotors in the body coordinate system b x b axis and o b y b The torque vector M generated by the component force on the axis T Therefore, the net external force and net external torque acting on the Mars drone are expressed as: F b =F G +F T +F R M b =M Q +M G +M T S2-3: Based on steps S2-1 and S2-2, the complete six-degree-of-freedom dynamic equation model of the Mars UAV is obtained in the body coordinate system: Where M = diag[m,m,m,I] x ,I y, I z [N] represents the mass matrix of the Mars drone. d To determine the forces and torques related to the generalized velocity vector X of the Mars drone, N d =[m(vr-wq),m(wp-ur),m(uq-vp),(I y -I z )rq,(I z -I x )pr,(I x -I y )pq] T N f These are the forces and torques exerted on the main body of the fuselage by external forces. N u The forces and torques generated by the task flow controller, i.e., the control law of the task flow controller. Step S3 specifically includes the following steps: S3-1: Considering that the flight trajectory or path of the Mars UAV can be described by the intersection of a pair of surfaces in three-dimensional space, the expected trajectory of the Mars UAV is defined as an implicit function with respect to position P, resulting in: Where t is time, j A1(t), j A2(t) are three-dimensional row vectors pre-designed based on the flight trajectory, and b(t) are one-dimensional vectors pre-designed based on the flight trajectory. j P = [x j ,y j ,z j ] T Let e ​​be a vector consisting of the j-th powers of the position coordinate components of the Mars drone, where j represents the order describing the desired trajectory or path, and e is the power of the vector. N1 and e N2 For trajectory tracking error components; when N(P,t)=[e N1 e N2 0] T ≠0 3×1 When this occurs, there exists a trajectory tracking error component e. N1 and e N2 The corresponding control law N needs to be designed. u This makes the trajectory tracking error component e N1 →0 and e N2 →0; S3-2: Considering the attitude stability during the flight of the Mars UAV, the attitude error Φ(P,Θ,t) is defined as the difference between the attitude Θ of the Mars UAV and the desired attitude Θ. * The difference between (P,Θ,t) yields: Φ(P,Θ,t)=Θ-Θ * (P,Θ,t) The corresponding control law N needs to be designed. u This makes the attitude error Φ(P,Θ,t)→0. 3×1 ; S3-3: Combining steps S3-1 and S3-2, define the trajectory manifold Ψ. tr =0 6×1 ,get: Ultimately, the generalized position vector Y of the Mars drone needs to converge to the trajectory manifold Ψ. tr =0 6×1 This is to ensure that the Mars drone flies in a stable attitude along the desired trajectory; Due to Ψ tr If Ψ is at least first-order differentiable for all independent variables, then tr The total differential with respect to time t is expressed as: in, e={N,Φ},f={P,Θ}; S3-4: To reduce energy consumption during flight, the generalized velocity vector X of the Mars drone is constrained when flying along the desired trajectory. The velocity manifold is defined as follows: in, V is the velocity squared error vector. c Let X be the desired flight velocity vector; ultimately, the generalized velocity vector X of the Mars drone needs to converge to the velocity manifold Ψ. v =0 6×1 This allows the Mars drone to move at the desired velocity vector V. c Maintain a stable spatial geometry during flight; S3-5: To ensure that the Mars drone flies along the desired attitude and velocity of the exploration mission, the trajectory manifold Ψ is utilized. tr =0 and velocity manifold Ψ v =0, synthesizing task manifold Ψ=0, yields: Ψ=Ψ tr +AΨ v =0 Where A is a 6th-order diagonal weighted coefficient matrix, the matrix A is designed such that if and only if Ψ tr →0 and Ψ v →0, then Ψ→0, therefore, the trajectory tracking error component e in step S3-1 can also be made... N1 →0 and e N2 →0 and the attitude error Φ(P,Θ,t) in step S3-2 →0 3×1 ; S3-6: To ensure that the generalized position vector Y, the generalized velocity vector X, and their first derivatives of the Mars UAV converge to the mission manifold Ψ = 0, a 6th-order constant coefficient matrix T is designed to satisfy the following formula: S3-7: Based on step S2-3, the dynamic equations of the Mars drone are obtained. Combining steps S3-5 and S3-6, the nonlinear mission manifold control law N is obtained. u for: Where, K0=J1J ∑PX +J Θ J ΣΘX K1 = TAK 11 +K 12 K 11 =[0 6×3 J1J ΣPΘ +J Θ J ΣΘΘ ], K 12 =(T+A)J s +TAГ s , J1 = J P +J V , m = {P, Θ}, n = {X, Θ}; The nonlinear task manifold control law N u It can be guaranteed that all state variables asymptotically converge to the task manifold Ψ = 0, that is... Based on the aforementioned nonlinear task manifold control law N u This allows us to obtain the forces and torques generated by the task manifold controller.

2. The control method according to claim 1, characterized in that, To prove step S3-7, the nonlinear task manifold control law N u For global asymptotic stability, the Lyapunov function is chosen. Taking the derivative with respect to time, we get: Among them, to make Only matrix T in steps S3-6 is needed -1 Minimum eigenvalue T min A value greater than 0 is sufficient. Considering that T is a diagonal matrix, it is only necessary for all diagonal elements of T to be positive.

3. The control method according to claim 1, characterized in that, The Mars drone also includes a solar panel (4), a landing gear (3), and a protective cover (11); the solar panel (4) is installed above the upper rotor (7) to charge the battery using solar energy; and the protective cover (11) is installed below the lower rotor.

4. A simulation method for trajectory tracking and path following control of the Mars unmanned aerial vehicle using the control method described in any one of claims 1-3, characterized in that, In a virtual simulation environment, a 3D virtual scene of the Santa Maria Martian crater and the periodic slope lines of Mars is constructed based on Unity 3D, and the task manifold control law N designed in step S3 above is used. u Control simulation for controlling Mars drones to achieve Mars crater exploration missions and Mars periodic slope line observation missions.

5. The simulation method according to claim 4, characterized in that, For the Santa Maria Mars Crater exploration mission, the desired trajectory is to descend along the crater wall into the crater, then fly around the crater wall in a spiral, and finally fly out of the crater and fly in a straight line to the desired destination. The trajectory tracking control is performed by the mission manifold controller in step S3 and the existing sliding mode controller.

6. The simulation method according to claim 5, characterized in that, For the Mars periodic slope line observation mission, the desired trajectory is designed as a high-order curve that conforms to the slope line stripe characteristics, and the path following control is performed by the mission manifold controller in step S3.

Images