CN118295450A - A control method for near-ground flight of a rotary-wing UAV - Google Patents

Abstract

The invention discloses a control method for the near-to-ground flight of a rotor unmanned aerial vehicle, which comprises the following steps: carrying out dynamic modeling comprising ground effect on the rotary wing unmanned aerial vehicle to obtain an unmanned aerial vehicle dynamic model, a motor model, an additional thrust model under the ground effect, a ground effect correcting torque model and a ground effect forward resistance model, and carrying out equivalent modeling on the ground effect correcting torque by using an equivalent model method to obtain an under-hanging particle model, and combining the ground effect torque model to obtain an equivalent model of the horizontal correcting torque of the unmanned aerial vehicle; acquiring a calibration value of a parameter in the established model; based on differential flatness characteristics of the unmanned aerial vehicle dynamic model, generating a feedforward reference control instruction according to given unmanned aerial vehicle reference track information and yaw angle information; and combining the feedforward reference control instruction, the actual flight state of the unmanned aerial vehicle and the calibration value of the parameter to generate a final control instruction.

Description

A control method for near-ground flight of a rotary-wing UAV

Technical Field

The invention belongs to the field of control of a rotor-wing unmanned aerial vehicle (UAV) flying close to the ground, and in particular relates to a control method for a rotor-wing unmanned aerial vehicle (UAV) flying close to the ground.

Background technique

Due to the simple mechanical design and high maneuverability, multi-rotor aircraft have been widely used in various scenarios both indoors and outdoors. However, in many applications, close-to-ground flight is inevitable. For example, using a multi-rotor drone with a robotic arm to grab objects close to the ground, planning the trajectory of close-to-ground flight to save energy through ground reverse thrust, automatically landing on the surface of an object, and so on. In these scenarios, when approaching rigid objects, multi-rotor drones will be disturbed by the airflow near the ground or the surface of the object, resulting in oscillations in attitude and difficulty in controlling the position, which has a significant impact on its safety and stability.

This phenomenon is often referred to as ground effect, and researchers have analyzed it extensively. Most work has focused on accurately modeling the "extra lift" created by ground effect. Previous approaches typically fit a function to the relationship between the extra lift and altitude. The model is then used to calculate a feedforward control signal (inversely compensated acceleration) to counteract the disturbance. However, ground effect imposes far more disturbances on multirotor drones than this.

The first type of disturbance is external torque. When a multi-rotor drone flies near a flat ground and its attitude is not horizontal, the rotors closer to the ground will produce greater thrust, thereby generating a torque that causes the attitude of the multi-rotor drone to tend to be horizontal [6]. We call this "horizontal return torque". This may cause the drone's attitude to oscillate or reduce control accuracy.

The second interference is the change in the drag coefficient of the drone's forward movement. For the rotating rotors on the quadcopter, each blade is subject to the lift of the air as well as the drag applied to the rotating plane. This drag is positively correlated with the relative airflow velocity. When hovering, the relative airflow velocity is proportional to the rotor speed. When the drone is flying forward, the relative airflow velocity of the blades moving forward in each rotation cycle is large, while the relative airflow velocity of the blades moving backward is small. This causes the blades to experience different resistance at different phases, and the resultant force of the resistance experienced by the entire rotor is not zero, thus generating what is commonly referred to as rotor drag in the drone field. Since the relative airflow velocity difference between blades at different phases is proportional to the rotor angular velocity, rotor drag is positively correlated with the total thrust. When the drone approaches the ground, the required throttle decreases, so the rotor drag also decreases. The reason why some previous works approximated rotor drag as a constant is that the total thrust is maintained near the hovering throttle during normal flight, which is not applicable when flying close to the ground.

The third disturbance model that is usually not considered is the escape of the lift airflow of the UAV when flying close to the ground at high speed. "X. Kan, J. Thomas, H. Teng, H. G. Tanner, V. Kumar, and K. Karydis, "Analysis of ground effect for small-scale uavs in forward flight," IEEE Robotics and Automation Letters, vol. 4, no. 4, pp. 3860-3867, 2019" analyzes the changes in the additional thrust of the UAV during forward flight near the ground. When the UAV is hovering near the ground, it is blocked by the ground, and the airflow rebounds upward to form a high pressure layer, resulting in additional lift. When the UAV flies forward at high speed, part of the airflow escapes, causing the additional lift to decay.

The above problems make it difficult for the UAV to maintain a tilted posture near the ground to generate lateral acceleration. Therefore, the existing methods are only applicable to simple situations such as takeoff and hovering near the ground, and it is difficult to track a trajectory of flying close to the ground. Therefore, it is necessary to study the UAV dynamics model under the influence of ground effect and improve the stability of UAV control through certain control strategies.

In the process of implementing the present invention, the inventors found that there are at least the following problems in the prior art:

(1) How to deal with “extra push”

The research on ground effect was initially focused on helicopters, and its influencing factors mainly involved the rotor radius and the distance between the blade plane and the ground. Other influencing factors include air density, rotation speed, blade angle of attack, and incoming flow velocity in the direction of the rotation axis. Compared with helicopters, the dynamic model of multi-rotors is relatively simple because helicopters have only one rotor. For multi-rotors, the influence of the air convection area formed by the entire body on the ground effect is greater than the superposition effect of a single rotor. Therefore, some ground effect models designed specifically for multi-rotor UAVs have been proposed one after another. For example, "P. Sanchez-Cuevas, G. Heredia, and A. Ollero, "Characterization of the aerodynamic ground effect and its influence in multirotor control," International Journal of Aerospace Engineering, vol. 2017, 2017." introduced some correction terms in the ground effect model of helicopters to adapt to the characteristics of multi-rotor UAVs. Although these models are already quite accurate, the model parameters are usually large and the identification process is relatively complex.

(2) How to deal with the “horizontal return moment”

In previous studies, a common solution is inner loop tuning. When a multirotor drone approaches the ground and has a certain degree of attitude tilt, it will be affected by the stronger "horizontal return torque" generated by the ground effect. At small angles, the relationship between this torque and the tilt angle can be considered linear. The "horizontal return torque" is similar to adding a fixed weight under the multirotor. Further, it can be equivalent to moving the center of gravity downward and increasing the moment of inertia, which can be handled by adjusting the gains of the attitude and angular velocity control links. Since the intensity of the ground effect varies at different altitudes, a corresponding set of suitable parameters is required for each altitude. However, due to the lack of a torque model, it is relatively difficult to adjust the control parameters.

For example, P. Sanchez-Cuevas, G. Heredia, and A. Ollero, “Characterization of the aerodynamic ground effect and its influence in multirotor control,” International Journal of Aerospace Engineering, vol. 2017, 2017. argue that the “horizontal aligning torque” is not significant enough and only makes the multirotor more stable, so it is not modeled or considered in their control framework. Their research focuses on the torque (called obstacle torque) generated by the additional thrust generated by the ground effect on some rotors when the multirotor flies over an obstacle, thereby pushing the multirotor away from the obstacle. In the experiment, they placed a table under the horizontal flight path of the multirotor, which means that there are actually only two altitudes during the flight: constant altitude and infinite altitude. They used the obstacle torque model as a feedforward to introduce a deviation in the attitude command, and the final output command was the desired attitude of the multirotor. In order to solve the leveling torque problem, the attitude controller parameters of the flight control only need to be adjusted according to the specific altitude of the scene, without adapting to all altitudes.

Another example is G. Shi, X. Shi, M. O’Connell, R. Yu, K. Azizzadenesheli, A. Anandkumar, Y. Yue, and S.-J. Chung, “Neural lander: Stable drone landing control using learned dynamics,” in 2019 International Conference on Robotics and Automation (ICRA). IEEE, 2019, pp. 9784-9790. A network was designed based on the dynamics and state information of a multirotor drone to predict external forces near the ground. They believe that the leveling torque is not large enough to require compensation. However, parameters in their controller framework, such as the angular velocity error gain, can mitigate the oscillations caused by the horizontal torque. Adjusting the parameters to adapt to low-altitude scenarios may reduce control accuracy in high-altitude attitudes, but will not affect landing and low-altitude flight.

(3) Model-free adversarial perturbation methods

The disturbance caused by the ground effect is considered as an unknown disturbance, and attempts to observe and compensate it using a disturbance observer are theoretically feasible. However, in practical applications, the effect is not satisfactory. The core idea of this type of method is to assume that the external disturbance in the next time period is roughly equal to the disturbance observed in the previous time period. This method is effective for some slowly changing external disturbances, such as the gravitational torque caused by the movement of the center of gravity on the XOY plane of the body coordinate system. Because even if the multi-rotor drone has a certain attitude tilt, the lever arm of the gravitational torque remains basically unchanged. However, for other cases, such as the offset of the center of gravity on the Z axis of the body coordinate system or the return torque caused by the ground effect, the external torque is proportional to the tilt angle and changes at a high frequency, which leads to a delay in disturbance estimation, and torque compensation cannot completely eliminate oscillations.

(4) Methods for dealing with thrust attenuation during high-speed flight

At present, the research on thrust attenuation of UAVs during high-speed forward flight near the ground is only at the modeling level, mainly because the mechanism is complex and the model is not yet clear. However, an airflow-induced speed can be calculated based on the blade radius and height above the ground of the UAV (《X.Kan, J.Thomas, H.Teng, H.G.Tanner, V.Kumar, and K.Karydis, "Analysis of ground effect for small-scale uavs in forward flight," IEEE Robotics and Automation Letters, vol.4, no.4, pp.3860-3867, 2019.》). When the thrust is lower than this airflow-induced speed, there will be no significant attenuation. This disturbance can be avoided by controlling the flight speed below this speed.

(5) Methods for dealing with changes in drag coefficient

There have been many studies on the drag formation mechanism and compensation methods of rotary-wing UAVs when flying at high altitudes. However, there is no relevant research on modeling the change of drag coefficient when UAVs fly close to the ground.

Summary of the invention

The purpose of the embodiments of the present application is to provide a control method for a rotorcraft UAV flying close to the ground, so as to solve the above-mentioned problems existing in the related art.

According to a first aspect of an embodiment of the present application, a method for controlling a rotorcraft UAV near-ground flight is provided, comprising:

S1: The rotor UAV is subjected to dynamic modeling including ground effect, and the UAV dynamic model, motor model, additional thrust model under ground effect, ground effect self-aligning torque model, and ground effect forward resistance model are obtained. The ground effect self-aligning torque is equivalent to the hanging mass point model by using the equivalent model method, and the equivalent model of the horizontal self-aligning torque of the UAV is obtained by combining the ground effect torque model;

S2: Obtain the calibration values of the parameters in the model established in step S1;

S3: Based on the differential flatness characteristics of the UAV dynamics model, a feedforward reference control instruction is generated according to given UAV reference trajectory information and yaw angle information;

S4: Generate the final control command by combining the feedforward reference control command, the actual flight status of the UAV and the calibration value of the parameters.

Furthermore, in step S1, the UAV dynamics model is as follows:

ma＝-gz _W +Tz _B +f _G +f _D

Where m is the mass of the aircraft, J is the moment of inertia of the aircraft, g is the acceleration of gravity, T and τ _B are the thrust and torque generated by the rotor, f _G is the additional thrust generated by the ground effect, f _D is the forward resistance, τ _G is the ground effect aligning torque, τ _ext is other unknown torques, z _W = [0,0,1] is the direction of the Z axis of the world coordinate system, and z _B is the direction of the Z axis of the drone body coordinate system in the world coordinate system;

The motor model is the thrust and torque produced by a single motor:

T _i = k _T n _i ²

Where k _T , k _I are the additional thrust coefficient and the anti-torque coefficient respectively, J _R is the rotor moment of inertia;

In the additional thrust model under the influence of ground effect, the total thrust of the motor and the additional thrust caused by the ground effect are:

f _G ＝F _G (h)Tz _B

Where k _T is the thrust coefficient, _ni is the speed of each motor, F _G (h) is a function related to the height above the ground h, and num is the number of rotors;

The model of the self-aligning torque generated by the ground effect is:

Rτ _G ＝ _MG (h)T(z _B ×z _W )

Where _MG (h) is a function related to the height above the ground;

In the ground effect forward resistance model, the forward resistance of the UAV is:

f _D (R,v,h)＝-RD(h)R ^T v

Where D(h) is a function related to the height above the ground;

In the equivalent model of the horizontal self-aligning torque of the UAV, the equivalent moment of inertia of the UAV is:

Further, step S3 includes:

Based on the above-mentioned UAV dynamics model, the additional thrust model under the influence of ground effect and the forward resistance model under the influence of ground effect, the mass-normalized UAV dynamics model is obtained, and the reference thrust and reference attitude are generated in combination with the given UAV reference trajectory information;

Taking the first-order derivative of the acceleration in the mass-normalized UAV dynamics model to generate a reference angular velocity;

The second-order derivative of the acceleration in the mass-normalized UAV dynamics model is calculated to obtain a reference angular acceleration, thereby generating a reference torque.

Furthermore, a reference torque is generated based on the reference angular velocity and the reference angular acceleration:

Where ω _ref is the reference angular velocity, is the reference angular acceleration, and J′(h) is the equivalent moment of inertia of the UAV obtained by the equivalent model of the horizontal self-aligning torque.

Further, step S4 includes:

The expected acceleration is obtained according to the acceleration required by the UAV to offset the forward resistance, the additional lift and the reference acceleration;

According to a one-to-one mapping relationship between the desired attitude and the desired acceleration of the drone, the desired attitude is determined by the desired acceleration, and then the desired angular velocity and the desired angular acceleration are obtained by the attitude error;

Calculate the expected thrust according to the expected acceleration, and calculate the expected torque of the UAV based on the expected angular velocity and the expected angular acceleration, combined with the equivalent model of the horizontal self-aligning torque;

The expected speed is calculated based on the expected thrust and the expected torque, and then the expected throttle of the motor is obtained by combining the reference throttle curve.

Furthermore, without considering the additional torque, the expected torque of the drone is:

Taking the additional torque into account, the expected torque of the drone is:

Where J′(h) is the equivalent moment of inertia of the UAV obtained by the equivalent model of the horizontal self-aligning torque, ω _des is the desired angular velocity, is the expected angular acceleration.

Furthermore, the reference throttle curve _nesc is:

_nesc ( _{tc )} = _c2tc2 ⁺ _{c1tc + c0}

Wherein t _c ∈[0,1] is the throttle of the motor, and {c ₂ ,c ₁ ,c ₀ } are constants calibrated in step S2 .

According to a second aspect of an embodiment of the present application, a control device for a rotorcraft UAV near-ground flight is provided, comprising:

A modeling module is used to perform dynamic modeling of the rotor UAV including ground effect, obtain the UAV dynamic model, motor model, additional thrust model under ground effect, ground effect self-aligning torque model, ground effect forward resistance model, and use the equivalent model method to equate the ground effect self-aligning torque to the lower hanging mass model, and combine the ground effect torque model to obtain the equivalent model of the UAV horizontal self-aligning torque;

A calibration module, used to obtain calibration values of parameters in the model established by the modeling module;

A feedforward reference control instruction generation module, used for generating a feedforward reference control instruction based on the differential flatness characteristics of the UAV dynamics model and according to given UAV reference trajectory information and yaw angle information;

The final control instruction generation module is used to generate the final control instruction by combining the feedforward reference control instruction, the actual flight state of the UAV and the calibration value of the parameters.

According to a third aspect of an embodiment of the present application, there is provided an electronic device, including:

one or more processors;

A memory for storing one or more programs;

When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in the first aspect.

According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, on which computer instructions are stored, and when the instructions are executed by a processor, the steps of the method described in the first aspect are implemented.

The technical solution provided by the embodiments of the present application may have the following beneficial effects:

(1) Accurately model the aligning torque and drag caused by the ground effect when the multi-rotor UAV is flying close to the ground, and form a set of model parameter identification methods;

(2) The UAV dynamics model affected by ground effect is simplified and equivalent so that it can satisfy the differential flatness relationship, thereby generating accurate control feedforward instructions;

(3) By designing a control strategy, the multi-rotor UAV can overcome the negative impact of the ground effect when close to the ground and perform stable hovering and tracking flight trajectory.

It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

Fig. 1 is a flow chart showing a method for controlling a rotorcraft UAV's high-speed flight on the ground according to an exemplary embodiment.

Figure 2 is a schematic diagram of the physical platform and simulation environment used to obtain model parameters.

Fig. 3 Schematic diagram of single motor parameter calibration.

Fig. 4 Schematic diagram of parameter calibration of the ground effect additional thrust model.

Fig. 5 Schematic diagram of parameter calibration of ground effect return torque.

Fig. 6 Schematic diagram of parameter calibration of ground effect forward drag coefficient.

Fig. 7 Schematic diagram of equivalent method for ground effect restoring torque.

FIG8 is a schematic diagram of attitude control error based on this method.

FIG9 is a schematic diagram of a trajectory tracking position control curve based on the present method.

FIG10 is a schematic diagram of the trajectory tracking Euler angle control curve based on the present method.

Fig. 11 is a block diagram of a control device for high-speed flight of a rotary-wing UAV on the ground according to an exemplary embodiment.

Fig. 12 is a schematic diagram showing an electronic device according to an exemplary embodiment.

Detailed ways

Here, exemplary embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application.

The terms used in this application are for the purpose of describing specific embodiments only and are not intended to limit this application. The singular forms of "a", "said" and "the" used in this application and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and/or" used in this article refers to and includes any or all possible combinations of one or more associated listed items.

It should be understood that although the terms first, second, third, etc. may be used in the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present application, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

The present invention is applicable to the planning and control of the flight of rotorcraft UAVs near the ground, and is mostly used on flat indoor ground. Generally speaking, the current rotorcraft UAV products (such as DJI, Daotong, etc.) are applied in outdoor high-altitude (flight altitude above 1m). If the UAV needs to continue high-speed flight at a lower altitude (below 0.5m), the technology described in this patent needs to be used.

FIG1 is a flow chart of a method for controlling a high-speed flight of a rotary-wing UAV on the ground according to an exemplary embodiment. As shown in FIG1 , the method is applied in a terminal and may include the following steps:

S1: Dynamic modeling of a rotary-wing UAV including ground effects;

Specifically, the present application adopts

[x _W ,y _W ,z _W ]＝diag(1,1,1)

Represents the three orthogonal bases of the world coordinate system. The direction of the drone is expressed as

R＝[x _B ,y _B ,z _B ]

{x _B ,y _B ,z _B } is a set of orthogonal bases of the body coordinate system. The position of the multi-rotor UAV's geometric center in the world coordinate system and its derivatives (velocity, acceleration, jerk) are {p, v, a, j} respectively.

Decomposed, z _W = [0,0,1] is the direction of the Z axis of the world coordinate system, z _B is the direction of the Z axis of the drone body coordinate system in the world coordinate system, and the same applies to the other parameters.

The Euler angles of the drone in the world coordinate system are expressed in the order of ZYX rotation as follows: The rotation matrix around the ZYX axis can be expressed as The angular velocity and angular acceleration of the UAV in the body coordinate system are expressed as {ω, β}.

The following description is given using a quad-rotor drone as an example.

(1) UAV dynamics model

The UAV dynamics model is as follows:

ma＝-gz _W +Tz _B +f _G +f _D

Where m is the mass of the aircraft, J is the moment of inertia of the aircraft, g is the acceleration due to gravity, T and τ _B are the thrust and torque generated by all rotors, f _G is the additional thrust generated by the ground effect, f _D is the forward drag, τ _G is the ground effect aligning torque, and τ _ext is the other unknown torque.

For a quadrotor drone, the thrust and torque generated by the rotor can be written as:

Where b is the diagonal wheelbase of the drone. {k _T , k _TX , k _TY } are the thrust coefficient, X-axis torque coefficient, and Y-axis torque coefficient, and the three parameters are close in value. k _I is the motor anti-torque coefficient. _{n i} is the motor speed. M is the mixing matrix, and n ² is a 4x1 matrix synthesized by the squares of the speeds of each motor.

(2) Motor model

The thrust and torque generated by a single motor are:

T _i = k _T n _i ²

Where k _T and k _I are thrust coefficient and anti-torque coefficient respectively, J _R is the rotor moment of inertia, and _ni is the speed of a single motor.

There is a quadratic function mapping relationship between the motor speed and the throttle t _c ∈[0,1] sent by the flight control to the motor controller, which is recorded as the reference throttle curve _nesc :

_nesc ( _{tc )} = _c2tc2 ⁺ _{c1tc + c0}

where {c ₂ ,c ₁ ,c ₀ } are constants.

(3) Additional thrust model under ground effect

The total thrust of the motor is:

Where k _T is the thrust coefficient and _ni is the speed of each motor.

The additional thrust due to ground effect is:

f _G ＝F _G (h)Tz _B

F _G (h) is a function related to the height h above the ground, g ₁ and g ₂ are constants and need to be calibrated.

(4) Ground effect return torque model

The model of the self-aligning torque generated by the ground effect is:

Rτ _G ＝ _MG (h)T(z _B ×z _W )

Where _MG (h) is a function related to height, b is the diagonal wheelbase, and _g3 , _g4 , and _g5 are all constants:

(5) Ground effect forward resistance model

The forward resistance of the drone is:

f _D (R,v,h)＝-RD(h)R ^T v

Where D(h)=diag{ _dx (h), _dy (h),0} is a function related to height.

(6) Equivalent model of horizontal self-aligning moment

For a drone with a mass point rigidly connected below, its dynamic model can be written as:

Where _J1 is the moment of inertia of the body, and _m0 is the mass of the particle. At the same time, if the drone and the ball are regarded as a whole, then the center of the drone will move downward, and its dynamic model can be written as:

Where _J2 is the moment of inertia of the ball and the drone as a whole:

Based on the above-mentioned UAV dynamics model and the additional thrust model under the influence of ground effect, the dynamics model of the UAV affected by the ground effect aligning torque can be written as:

By analogy, the above two situations are equivalent The UAV's moment of inertia can be written as follows, without considering the effect of the aligning torque:

We hope that the function of the equivalent moment of inertia J' of the drone is only related to the height h, but the thrust T in the above formula changes with time. Therefore, assuming that the drone is in a hovering state, the thrust of the drone is only related to the height:

Substituting it into the function of moment of inertia J', we can get

S2: Calibrate the parameters in the model established in step (1) through experiments and simulations;

Figure 2 shows the physical platform and simulation environment used to obtain model parameters, where a) is a platform for a single motor. b) to g) are test platforms that use wooden boards to simulate the ground and collect ground effect interference data experienced by the multi-rotor aircraft. b) to d) The ground height can be adjusted, e) to g) The tilt angle can be adjusted to simulate the pitch angle of the multi-rotor aircraft, h) to n) are schematic diagrams of using CFD simulation to verify the model, h) is an air flow rate scale diagram, and i) to n) are air flow simulation streamlines at different high speeds and tilt angles.

(1) Single motor parameter calibration

Use the single motor calibration platform in a) of Figure 2 to calibrate the parameters of the motor of the multi-rotor drone, and collect the data shown in Figure 3. The parameters of the single motor can be obtained:

Thrust coefficient k _T = 4.0083 × 10 ^-8 N/rpm ²

Reactive torque coefficient k _I = 6.3859×10 ^-10 (N·m)/rpm ²

Rotor moment of inertia J _R = 1.0556 × 10 ^-4 kg/m ²

(2) Mixing matrix parameter calibration

The quadrotor drone is fixed on the test platform in b) to g) of Figure 2 for parameter calibration, and the aircraft is controlled to hover at different heights to obtain:

X-axis torque coefficient k _TX = 4.678×10 ^-8 N/rpm ²

Y-axis torque coefficient k _TY = 3.588×10 ^-8 N/rpm ²

(3) Calibration of ground effect “extra thrust” model parameters

The quadcopter was fixed on the test platform in b) to g) of Figure 2 for parameter calibration, and the aircraft was controlled to hover at different heights. The collected data is shown in Figure 4. It can be seen that the model of this method achieves a similar or even better fitting effect than previous work with fewer parameters. The parameters are:

g ₁ =1.804×10 ^-2

g ₂ =7.339×10 ^-3

(4) Calibration of ground effect “return torque” model parameters

The quadrotor drone is fixed on the test platform in b) to g) of Figure 2 for parameter calibration, and the collected data is shown in Figure 5. The parameters of the ground effect return torque can be obtained:

g ₃ = -3.365×10 ^-1

g ₄ =4.126×10 ^-2

g ₅ =1.768×10 ^-3

(5) Calibration of ground effect “forward drag coefficient” model parameters

After actual flight tests, the collected data is shown in Figure 6. The forward drag coefficient at special points can be calibrated, and other altitudes are filled in by interpolation:

d _x (2.0) = 0.3970 N/(m/s)

d _y (2.0) = 0.3300 N/(m/s)

d _x (0.15) = 0.2367 N/(m/s)

d _y (0.15) = 0.2039 N/(m/s)

S3: generating a feedforward reference control command based on the differential flatness characteristic of the UAV dynamics model;

Figure 7 shows the process of using the equivalent model method to equate the ground effect torque to the hanging mass model. According to the given UAV reference trajectory information (position p _ref , velocity v _ref , acceleration a _ref , jerk j _ref , jerk s _ref , etc.), combined with the yaw angle (z-axis rotation angle) and its derivatives This method can be used to generate reference control instructions (thrust T _ref , attitude ξ _ref , angular velocity ω _ref , angular acceleration Torque τ _Bref , speed n _ref ).

(1) Generate thrust and attitude feedforward reference control commands

Based on the above-mentioned UAV dynamics model, the additional thrust model under the influence of ground effect and the forward resistance model under the ground effect, the UAV dynamics model after mass normalization can be written as:

a＝-gz _W +[1+F _G (h)] _Ta z _B -RD _a (h)R ^T v

Where _Ta = T/m, _Da (h) = diag{ _dax (h), _day (h),0} = D(h)/m is the result of mass normalization of D(h) = diag{ _dx (h), _dy (h),0}. Expand the forward resistance:

a+gz _W +[d _ax (h)x _B ^T v]x _B +[d _ay (h)y _B ^T v]y _B -[1+F _G (h)]T _a z _B =0

Multiplying the above formula by z _B ^T on the left can give the reference output of thrust:

By multiplying the above formula by x _B ^T and y _B ^T on the left, we can get the reference output of the posture R = [x _B , y _B , z _B ]:

x _B ^T α＝0,α＝a+gz _W +d _x (h)v

y _B ^T β＝0,β＝a+gz _W +d _y (h)v

z _B = x _B × y _B

(2) Generate angular velocity feedforward reference control instructions

Since the acceleration of the drone corresponds to the attitude, the first-order derivative of the acceleration corresponds to the angular velocity, and the second-order derivative of the acceleration corresponds to the angular acceleration, the reference angular velocity ω _ref and the reference angular acceleration can be obtained by taking the first-order derivative and the second-order derivative of the acceleration a＝-gz _W +[1+F _G (h)]T _a z _B -RD _a (h)R ^T v

(3) Generate torque feedforward reference control instructions

Reference angular velocity ω _ref and reference angular acceleration After obtaining, the reference moment τ _ref can be obtained by the following method:

where J′(h) is obtained from the equivalent model in S2(6).

S4: Combine the feedforward reference control command and the actual flight status of the UAV to generate the final control command;

According to the reference control instructions and the actual flight status fed back by the sensor, this method can be used to hover and track the trajectory of the multi-rotor UAV. The control effect of angle control is shown in Figure 8. It can be seen that the method based on the proposed method combining the usage model with incremental compensation has the same control effect at high altitude and low altitude. Other methods have a certain degree of error increase at low altitude. The effect of trajectory tracking is shown in Figures 9 and 10. The controller has a good tracking effect on the inner and outer loops of the UAV, that is, the position and angle.

(1) Final acceleration command

The desired acceleration a _des is composed of multiple accelerations:

a _des = a _ref + a _E - a _D - a _G

a _E is the acceleration introduced by position and velocity errors, K ^P and K ^V are gain matrices:

and is the measured position and speed of the UAV, a _D and a _G are the accelerations required to offset the forward resistance and additional lift respectively:

(2) Final attitude, angular velocity, and angular acceleration instructions

Generally speaking, there is a one-to-one mapping relationship between the desired attitude ξ _des and the desired acceleration a _des of the drone:

Therefore, the expected attitude of the UAV can be directly determined by the expected acceleration of the UAV;

The attitude error can be written as:

Thus, the desired angular velocity and angular acceleration instructions are obtained through the PD controller:

ω _des ＝K ^ξ ξ _e +ω _ref

Where ^Kξ and ^Kω are the attitude and angular velocity gain parameter matrices.

(3) Final thrust and torque command

After obtaining the expected acceleration, the expected thrust of the drone is uniquely determined:

Without considering the additional torque, the expected torque of the drone can be written as:

Taking the additional torque into account, the expected torque of the drone can be written as:

in It is the three-axis torque generated by the motor obtained according to the motor speed measured by the sensor. It is the angular acceleration data obtained by the sensor after first-order differentiation and low-pass filtering.

(4) Final motor speed and throttle command

Given the desired thrust and torque, the desired speed can be obtained, where M is the mixing matrix:

According to the inverse function of the reference throttle curve n _esc (t) The desired throttle of the four motors can be obtained through the PI controller:

Corresponding to the aforementioned embodiment of the control method for the near-ground flight of a rotary-wing UAV, the present application also provides an embodiment of a control device for the near-ground flight of a rotary-wing UAV.

FIG11 is a block diagram of a control device for a rotorcraft UAV flying close to the ground according to an exemplary embodiment. Referring to FIG11 , the device may include:

Modeling module 21 is used to perform dynamic modeling of the rotor UAV including ground effect, obtain the UAV dynamic model, motor model, additional thrust model under ground effect, ground effect self-aligning torque model, ground effect forward resistance model, and use the equivalent model method to equate the ground effect self-aligning torque to the lower hanging mass model, and combine the ground effect torque model to obtain the equivalent model of the UAV horizontal self-aligning torque;

A calibration module 22, used to obtain calibration values of parameters in the model established by the modeling module;

A feedforward reference control instruction generating module 23, used for generating a feedforward reference control instruction based on the differential flatness characteristic of the UAV dynamics model and according to the given UAV reference trajectory information and yaw angle information;

The final control instruction generation module 24 is used to generate the final control instruction by combining the feedforward reference control instruction, the actual flight state of the UAV and the calibration value of the parameter.

Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can refer to the partial description of the method embodiments. The device embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present application. A person of ordinary skill in the art can understand and implement it without creative work.

Correspondingly, the present application also provides an electronic device, including: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the control method for the near-ground flight of a rotary-wing UAV as described above. As shown in FIG12, a hardware structure diagram of any device with data processing capability in which a control device for the near-ground flight of a rotary-wing UAV provided in an embodiment of the present invention is located, in addition to the processor, memory and network interface shown in FIG12, any device with data processing capability in which the device in the embodiment is located can also include other hardware according to the actual function of the device with data processing capability, which will not be described in detail.

Accordingly, the present application also provides a computer-readable storage medium on which computer instructions are stored, and when the instructions are executed by the processor, the control method for the near-ground flight of the rotorcraft drone as described above is implemented. The computer-readable storage medium can be an internal storage unit of any device with data processing capabilities described in any of the aforementioned embodiments, such as a hard disk or a memory. The computer-readable storage medium can also be an external storage device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), an SD card, a flash card (Flash Card), etc. equipped on the device. Furthermore, the computer-readable storage medium can also include both an internal storage unit of any device with data processing capabilities and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capabilities, and can also be used to temporarily store data that has been output or is to be output.

Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the contents disclosed herein. The present application is intended to cover any variations, uses or adaptations of the present application, which follow the general principles of the present application and include common knowledge or customary technical means in the art that are not disclosed in the present application.

It should be understood that the present application is not limited to the exact construction that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof.

Claims (10)

1. A method for controlling a rotorcraft UAV near-ground flight, comprising: S1: The rotor UAV is subjected to dynamic modeling including ground effect, and the UAV dynamic model, motor model, additional thrust model under ground effect, ground effect self-aligning torque model, and ground effect forward resistance model are obtained. The ground effect self-aligning torque is equivalent to the hanging mass point model by using the equivalent model method, and the equivalent model of the horizontal self-aligning torque of the UAV is obtained by combining the ground effect torque model; S2: Obtain the calibration values of the parameters in the model established in step S1; S3: Based on the differential flatness characteristics of the UAV dynamics model, a feedforward reference control instruction is generated according to given UAV reference trajectory information and yaw angle information; S4: Generate the final control command by combining the feedforward reference control command, the actual flight status of the UAV and the calibration value of the parameters. 2. The method according to claim 1, characterized in that, in step S1, the UAV dynamics model is as follows: ma＝-gz _W +Tz _B +f _G +f _D Where m is the mass of the aircraft, J is the moment of inertia of the aircraft, g is the acceleration of gravity, T and τ _B are the thrust and torque generated by the rotor, f _G is the additional thrust generated by the ground effect, f _D is the forward resistance, τ _G is the ground effect aligning torque, τ _ext is other unknown torques, z _W = [0,0,1] is the direction of the Z axis of the world coordinate system, and z _B is the direction of the Z axis of the drone body coordinate system in the world coordinate system; The motor model is the thrust and torque produced by a single motor: T _i = k _T n _i ² Where k _T , k _I are thrust coefficient and anti-torque coefficient respectively, J _R is the rotor moment of inertia; In the additional thrust model under the influence of ground effect, the total thrust of the motor and the additional thrust caused by the ground effect are: f _G ＝F _G (h)Tz _B Where k _T is the thrust coefficient, _ni is the speed of each motor, F _G (h) is a function related to the height above the ground h, and num is the number of rotors; The model of the self-aligning torque generated by the ground effect is: Rτ _G ＝ _MG (h)T(z _B ×z _W ) Where _MG (h) is a function related to the height above the ground; In the ground effect forward resistance model, the forward resistance of the UAV is: Where D(h) is a function related to the height above the ground; In the equivalent model of the horizontal self-aligning torque of the UAV, the equivalent moment of inertia of the UAV is: 3. The method according to claim 1, characterized in that step S3 comprises: Based on the above-mentioned UAV dynamics model, the additional thrust model under the influence of ground effect and the forward resistance model under the influence of ground effect, the mass-normalized UAV dynamics model is obtained, and the reference thrust and reference attitude are generated in combination with the given UAV reference trajectory information; Taking the first-order derivative of the acceleration in the mass-normalized UAV dynamics model to generate a reference angular velocity; The second-order derivative of the acceleration in the mass-normalized UAV dynamics model is calculated to obtain a reference angular acceleration, thereby generating a reference torque. 4. The method according to claim 3, characterized in that the reference torque is generated based on the reference angular velocity and the reference angular acceleration: Where ω _ref is the reference angular velocity, is the reference angular acceleration, and J′(h) is the equivalent moment of inertia of the UAV obtained by the equivalent model of the horizontal self-aligning torque. 5. The method according to claim 3, characterized in that step S4 comprises: The expected acceleration is obtained according to the acceleration required by the UAV to offset the forward resistance, the additional lift and the reference acceleration; According to a one-to-one mapping relationship between the desired attitude and the desired acceleration of the drone, the desired attitude is determined by the desired acceleration, and then the desired angular velocity and the desired angular acceleration are obtained by the attitude error; Calculate the expected thrust according to the expected acceleration, and calculate the expected torque of the UAV based on the expected angular velocity and the expected angular acceleration, combined with the equivalent model of the horizontal self-aligning torque; The expected speed is calculated based on the expected thrust and the expected torque, and then the expected throttle of the motor is obtained by combining the reference throttle curve. 6. The method according to claim 5, characterized in that, without considering the additional torque, the expected torque of the UAV is: Taking the additional torque into account, the expected torque of the drone is: Where J′(h) is the equivalent moment of inertia of the UAV obtained by the equivalent model of the horizontal self-aligning torque, ω _des is the desired angular velocity, is the expected angular acceleration. 7. The method according to claim 5, characterized in that the reference throttle curve _nesc is: _nesc ( _{tc )} = _c2tc2 ⁺ _{c1tc + c0} Wherein t _c ∈[0,1] is the throttle of the motor, and {c ₂ ,c ₁ ,c ₀ } are constants calibrated in step S2 . 8. A control device for a rotorcraft UAV flying near the ground, characterized by comprising: A modeling module is used to perform dynamic modeling of the rotor UAV including ground effect, obtain the UAV dynamic model, motor model, additional thrust model under ground effect, ground effect self-aligning torque model, ground effect forward resistance model, and use the equivalent model method to equate the ground effect self-aligning torque to the lower hanging mass model, and combine the ground effect torque model to obtain the equivalent model of the UAV horizontal self-aligning torque; A calibration module, used to obtain calibration values of parameters in the model established by the modeling module; A feedforward reference control instruction generation module, used for generating a feedforward reference control instruction based on the differential flatness characteristics of the UAV dynamics model and according to given UAV reference trajectory information and yaw angle information; The final control instruction generation module is used to generate the final control instruction by combining the feedforward reference control instruction, the actual flight state of the UAV and the calibration value of the parameters. 9. An electronic device, comprising: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 7. 10. A computer-readable storage medium having computer instructions stored thereon, wherein when the instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.