CN114237054B - 6D interaction control method of aerial robot - Google Patents

6D interaction control method of aerial robot Download PDF

Info

Publication number
CN114237054B
CN114237054B CN202111557112.3A CN202111557112A CN114237054B CN 114237054 B CN114237054 B CN 114237054B CN 202111557112 A CN202111557112 A CN 202111557112A CN 114237054 B CN114237054 B CN 114237054B
Authority
CN
China
Prior art keywords
robot
propeller
aerial robot
control
interaction
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202111557112.3A
Other languages
Chinese (zh)
Other versions
CN114237054A (en
Inventor
雷瑶
宋佳斌
王家鼎
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Fuzhou University
Original Assignee
Fuzhou University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Fuzhou University filed Critical Fuzhou University
Priority to CN202111557112.3A priority Critical patent/CN114237054B/en
Publication of CN114237054A publication Critical patent/CN114237054A/en
Application granted granted Critical
Publication of CN114237054B publication Critical patent/CN114237054B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
    • G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
    • G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
    • G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B64—AIRCRAFT; AVIATION; COSMONAUTICS
    • B64C—AEROPLANES; HELICOPTERS
    • B64C39/00—Aircraft not otherwise provided for
    • B64C39/02—Aircraft not otherwise provided for characterised by special use
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B64—AIRCRAFT; AVIATION; COSMONAUTICS
    • B64U—UNMANNED AERIAL VEHICLES [UAV]; EQUIPMENT THEREFOR
    • B64U2101/00—UAVs specially adapted for particular uses or applications

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Manipulator (AREA)

Abstract

The invention relates to a 6D interactive control method of an aerial robot, which is characterized in that the aerial robot is driven by an inclined propeller to control a complete 6D gesture and apply complete torsion, and a firm end effector is attached; interaction is achieved by admittance control, wherein the outer loop controls the required admittance behavior, i.e. interaction compliance stiffness, damping and mass, while the inverse dynamics based inner loop ensures complete 6D pose tracking, wherein the interaction forces are estimated by an IMU enhanced momentum-based observer. The aerial robot employs a multi-rotor robot with non-collinear fixed tilting propellers NCFTP to independently control translational and angular acceleration without constraint, or to independently control an applied wrench when touching the wrench, to achieve 6D force control. The method is beneficial to improving the reliability of the aerial robot and reducing the weight and the implementation cost of the aerial robot.

Description

6D interaction control method of aerial robot
Technical Field
The invention relates to the field of aerial robots, in particular to a 6D interaction control method of an aerial robot.
Background
For physical interaction, aerial robots are equipped with rigid tools or n-degree of freedom (DoF) robotic arms. In the first case, the tool is firmly fixed to the body. The disadvantage of this solution is that: typical aerial platforms are not driven enough and therefore it is not possible to independently control the 6D (position plus orientation) dynamics of the end effector. In practice, this has shown that internal dynamic stabilization of underactuated multi-rotors cannot be guaranteed in the presence of interactions with fuselage points different from the aircraft centroid (CoM), and is generally neither stable nor practical for practical use.
A second possibility is to connect an n-DoF robotic arm to the aerial platform, aimed at overcoming the under-actuation of the end effector dynamics by exploiting the increase in the number of actuators provided by the arm. In this way, efficient 6D force control on the end effector side is possible. But this solution also has the following drawbacks: i) The mechanical arm greatly reduces effective load and transportation time due to self weight; ii) from a mechanical point of view, the system is much more complex than a single fuselage with rigid tools, and therefore, it is more costly to build and requires more maintenance and repair costs throughout its service life; iii) Lateral forces not provided by the aircraft platform itself, but generated by the dynamic/inertial coupling between the arm and the aerial robot: the dynamic coupling must be properly grasped to obtain the desired gain-6D force control aspect therefrom. This, in turn, requires knowledge of the precise kinetic model and very precise measurements of the system inputs and states (position, orientation, linearity and angular velocity). In fact, these requirements are difficult to achieve under realistic conditions (especially the former). For this reason, for real world verification, it is preferable to employ only kinematic methods, but at the cost of losing the main advantage of introducing a manipulator.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides a 6D interaction control method of an aerial robot.
In order to achieve the above purpose, the invention adopts the following technical scheme: A6D interactive control method of an aerial robot is characterized in that the aerial robot is driven by an inclined propeller to control a complete 6D gesture and apply a complete torsion, and a firm end effector is attached; interaction is achieved by admittance control, wherein the outer loop controls the required admittance behavior, i.e. interaction compliance stiffness, damping and mass, while the inverse dynamics based inner loop ensures complete 6D pose tracking, wherein the interaction forces are estimated by an IMU enhanced momentum-based observer.
Further, the aerial robot employs a multi-rotor robot with non-collinear fixed tilting propellers NCFTP to independently control translational and angular acceleration without constraint, or to independently control an applied wrench when touching the wrench, to achieve 6D force control.
Further, the measurement of the interactive spanner uses a force/torque sensor, wherein spanner measurements of the sensor on the aerial robot wrist are fed into an admittance filter.
Further, the modeling method of the NCFTP six-rotor general model is as follows:
The following coordinate system is established:
Inertial world frame Its axis (unit vector) and origin are denoted by { x W,yW,zW } and O W, respectively;
Frame of bicycle O R-{xW,yW,zW, installed on the robot, aircraft CoM with origin O R and end effector;
End effector frame O E-{xR,yR,zR, has been attached to the robotic end effector and has a start point at interaction point O E;
At the position of AndThe position of O E in the above-mentioned formula is defined by P E∈R3 and PRepresentation of whereinIs a constant time; the rotation matrix is represented by R R epsilon SO (3) and R E epsilon SO (3), respectively representing the framesWhere SO (3) = { a e R 3×3|AAT =i a detA =1 }; relative to inertial systemIs thatIn addition, in the case of the optical fiber,Representing a constant rotation matrix, representingRelative toIs a direction of (2); finally makeAnd AndRelative toRespectively using intra-frame angular velocityAndA representation;
Wherein [ (s ] x ε SO (3) represents the oblique symmetry matrix associated with vector [ s ] R 3; using Newton-Euler form sense, the equation of motion of an aerial robot is expressed as:
wherein m and J epsilon R 3×3 represent the robot mass and its inertial matrix, respectively corresponding to O R and Expressed as g is gravitational acceleration, fεR 3 and τ R∈R3 are force and torque inputs, andAndAn environment on the end effector for external forces and torques applied to the robotic CoM;
in formula (3), translational dynamics are expressed in the frame In (2), while the rotation dynamics are represented in the frameIn (a) and (b);
The inclined hexadecimal is used for NCFTP platforms by the derivation formula (3) without losing the generality of any full-drive high-altitude platform; NCFTP platform based on six rotor structures, equidistant spacing and of screw Equidistant from O R in the x-y plane of (2); complete drive is achieved by means of a rigid adapter, tilting each individual motor-propeller combination; considering the 6 frames to be considered,I=1..6, wherein O Pi coincides with the centre of rotation of the i-th propeller group; relative to Is represented by a rotation matrix:
the ith electric propeller group is relative to Is defined by constant parameters α and β; selecting an alternate alpha symbol for each other propeller in equation (4) to fully actuate the aircraft;
the position of the ith electric propeller group relative to O R is defined as:
Wherein the vectors are listed from the center of rotation to the center vector of motor propeller group O Pi, and from O R to the center of tilt rotation;
The rotating propeller generates thrust and drag in O Pi; both were simulated using positive and negative squares of rotational speed:
Wherein c f and c τ are constant parameters, and the rotation speed of the propeller is determined Correlating with the generated thrust and resistance moment; (-1) i-1 in formula (7) represents the effect on the counter-rotating rotors of all even propellers;
Will be applied to The total force of the upper O R is expressed as:
f 1∈R3×6 contains the geometric and physical characteristics of the aerial robot, i.e. dimensions, inclination angle, thrust coefficient, while u= [ u 1...u6]T ] is a vector that collects 6 square propeller rotational speeds;
under the same conditions, the total torque is combined, and the thrust contribution and drag torque in F B are expressed using formula (6) and formula (7):
wherein F 2∈R3×6 again includes geometric and terrestrial physical properties; by replacing (8) and (9) in the formula (3), obtained are:
this serves as a reliable kinetic model under slow flight conditions.
Further, the control framework of the aerial vehicle is based on an outer loop admittance control and an inner loop full attitude controller; the state of the aerial robot is estimated by UnscentedKalman a filter UKF which fuses the acceleration and angular velocity measurements of the inertial measurement unit IMU with the position and orientation of the attitude sensor.
Further, the inner ring attitude control algorithm aims at finding six required propeller rotational speeds u e R 6, SO that P R and R R track any full-attitude reference trajectory (P R,r(t),RR,rR(t))：R→R3 e SO (3);
By ignoring external forces and torques acting on the structure or end effector of the aerial robot, the dynamic model is rewritten as:
Where g (ω R)∈R6 is the error vector, combined with the gravity effect and the coriolis effect, The robot input jacobian, called J R as the inner ring system; if the input J R is reversible, select:
As control input v is a virtual additional input, thus obtaining:
for exponential convergence of position errors P R,r-PR＝ep to 0, the natural choice is a linear controller:
Wherein the diagonal positive definition gain matrix K p1,Kp2 defines the Hurwitz polynomial;
Developing a directional controller on SO (3);
it is assumed that the R R, AndWherein [. Cndot. ] v represents the inverse mapping from SO (3) to R 3; the attitude tracking error e R∈R3 is given by:
The tracking error of the angular velocity e ω∈R3 is given by:
In order to asymptotically converge the rotation error e R to 0, the following directional controller is selected:
Wherein the diagonal positive definition gain matrix K R1,KR2 again defines the Hurwitz polynomial.
Compared with the prior art, the invention has the following beneficial effects: the 6D interaction control method of the aerial robot improves the reliability of the aerial robot, reduces the weight and the implementation cost of the aerial robot, and is superior to aviation manipulation and physical interaction solutions adopted in the prior art in terms of functionality, reliability, complexity, cost and the like.
Drawings
FIG. 1 shows a NCFTP aerial platform (with end effector fixed inside) with a tilting propeller in an embodiment of the present invention.
Fig. 2 is a vector schematic of an end effector with a rigid connection in an embodiment of the present invention.
FIG. 3 is a block diagram of an inner loop low profile gesture controller in an embodiment of the present invention.
Fig. 4 is a signal block diagram of a control framework in an embodiment of the present invention.
Fig. 5 is an end effector of a hollow center robot in accordance with an embodiment of the present invention.
Fig. 6 is a schematic illustration of sliding with constant force on a hard sloped surface in an embodiment of the present invention.
FIG. 7 is a schematic illustration of a tool-tip used to secure a hole task in a funnel in an embodiment of the invention.
Fig. 8 is a schematic view of a sliding multiple contact platform in an embodiment of the invention.
FIG. 9 is a snapshot diagram highlighting the "slide" task in an embodiment of the present invention.
Detailed Description
The invention will be further described with reference to the accompanying drawings and examples.
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the application. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the present application. As used herein, the singular is also intended to include the plural unless the context clearly indicates otherwise, and furthermore, it is to be understood that the terms "comprises" and/or "comprising" when used in this specification are taken to specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof.
The embodiment provides a 6D interaction control method of an aerial robot, which enables the aerial robot to control a complete 6D gesture (independent of position and direction) and apply a complete torsion (independent of force and torque) by utilizing a tilting propeller drive, and is attached with a firm end effector; interaction is achieved by admittance control, wherein the outer loop controls the required admittance behavior, i.e. interaction compliance stiffness, damping and mass, while the inverse dynamics based inner loop ensures complete 6D pose tracking, wherein the interaction forces are estimated by an IMU enhanced momentum-based observer.
Wherein the aerial robot employs a multi-rotor robot with non-collinear fixed tilting propellers NCFTP to independently control translational and angular acceleration without constraint, or to independently control an applied wrench when touching the wrench, thereby achieving 6D force control.
In this embodiment, the measurement of the interactive wrench uses a force/torque sensor, wherein the wrench measurement of the sensor on the aerial robot wrist is fed into the admittance filter.
A fully driven aircraft equipped with any installed end effector tool is considered an aerial robot. The physical interaction framework presented is generic to any fully automated system capable of tracking full gesture trajectories through end effectors. The general model of the NCFTP six-rotor general model will be further described starting from the general part of the modeling.
Modeling method
The following coordinate system (see fig. 2) is established:
Inertial world frame Its axis (unit vector) and origin are denoted by { x W,yW,zW } and O W, respectively.
Frame of bicycleO R-{xW,yW,zW, an aircraft CoM that has been mounted on the robot, starting at O R and ending at the actuator.
End effector frameO E-{xR,yR,zR }, has been attached to the robotic end effector and has a start point at interaction point O E.
For the general vector v, the notation v * (x=r, E) indicates that the vector v is in a frameIs indicated in (a). If editing of the superscript equivalent line means that the vector is pressed in the inertial frame.
As shown in FIG. 3, inAndThe position of O E in the above-mentioned formula is defined by P E∈R3 and PRepresentation of whereinIs a constant time. The rotation matrix is represented by R R epsilon SO (3) and R E epsilon SO (3), respectively representing the framesWhere SO (3) = { a e R 3×3|AAT =i a detA =1 }. Relative to inertial systemIs thatIn addition, in the case of the optical fiber,Representing a constant rotation matrix, representingRelative toIs a direction of (2). Finally makeAnd AndRelative toRespectively using intra-frame angular velocityAndAnd (3) representing.
Wherein [ (S ] x ε SO (3) represents the oblique symmetry matrix associated with vector [ S ] R 3. Using Newton-Euler form sense, the equation of motion of an aerial robot is expressed as:
wherein m and J epsilon R 3×3 represent the robot mass and its inertial matrix, respectively corresponding to O R and Expressed as g is gravitational acceleration, fεR 3 and τ R∈R3 are force and torque inputs, andAndTo the external forces applied to the robotic CoM and to the environment on the torque end effector.
Notably, as in the field of aerorobotics, in formula (3) the translational dynamics are represented in the frameIn (2), while the rotation dynamics are represented in the frameIs a kind of medium.
The tilted hexadecimal derives equation (3) for NCFTP platforms without losing any commonality of the fully driven aerial platform. NCFTP platform based on six rotor structures, equidistant spacing and of screwEquidistant from O R in the x-y plane of (C). Full drive is achieved by a rigid adapter, tilting each individual motor-propeller combination (see fig. 1 and 2). Considering the 6 frames to be considered,I=1..6, wherein O Pi coincides with the centre of rotation of the i-th propeller group (see fig. 2).Relative toIs represented by a rotation matrix:
the ith electric propeller group is relative to Is defined by constant parameters a and β. The choice of α and β determines the maximum lateral force at the cost of higher lateral force resulting in higher internal force and pre-waste of energy. Selecting an alternate alpha symbol for each other propeller in equation (4) may cause the aircraft to fully operate.
The position of the ith electric propeller group relative to O R is defined as:
Wherein vectors are listed from the center of rotation to the center vector of motor propeller group O Pi, and from O R to the center of tilt rotation (see fig. 2).
It is well known that rotating propellers create thrust and drag in O Pi. Both are well modeled with positive and negative squares of rotational speed:
Most of the symbols used in the present invention are summarized. A value is also shown if it remains unchanged throughout all experiments.
Wherein c f and c τ are constant parameters, and the rotation speed of the propeller is determinedIn relation to the thrust and drag moments generated. (-1) i-1 in formula (7) represents the effect on the counter-rotating rotors of all even propellers.
Will be applied toThe total force of the upper O R is expressed as:
F 1∈R3×6 contains the geometric and physical characteristics of the aerial robot, i.e. dimensions, inclination angle, thrust coefficient, while u= [ u 1...u6]T ] is a vector that collects 6 square propeller rotational speeds.
Under the same conditions, the total torque is combined, and the thrust contribution and drag torque in F B are expressed using formula (6) and formula (7):
Wherein F 2∈R3×6 again includes geometric and terrestrial physical properties. The model and the detailed derivation of F 1 and F 2 and the requirements for full actuation thereof can be found in articles Rajappa (2015) and Michieletto (2017) et al. By replacing (8) and (9) in the formula (3), obtained are:
This was used as a reliable kinetic model under slow flight conditions (< 0.5 m/s). We ignore any aerodynamic effects such as the well known first order effects rotor windage, airframe windage and H forces, as these effects depend linearly on the speed of the vehicle and are therefore negligible for the slow air interactions considered herein.
Controller for controlling a power supply
The control framework of the aerial vehicle is based on an outer loop admittance control and an inner loop full attitude controller (see fig. 4). The state of the aerial robot is estimated by a concentrated Kalman filter (UKF) which fuses the acceleration and angular velocity measurements of the Inertial Measurement Unit (IMU) with the position and direction of the attitude sensor (in our case the motion capture system), using the "perspective point (PnP)" algorithm to easily replace it with an onboard camera. The torque and force of the interaction is estimated by the wrench observer.
Gesture controller
The inner ring attitude control law aims at finding six required propeller rotational speeds u e R 6, SO that P R and R R track any full-attitude reference trajectory (P R,r(t),RR,r(t))：R→R3 e SO (3).
By ignoring external forces and torques acting on the structure or end effector of the aerial robot, the dynamic model is rewritten as:
Where g (ω R)∈R6 is the error vector, combined with the gravity effect and the coriolis effect, J R is referred to as the intra-annular system x robot input jacobian. If the input J R is reversible, select:
As control input v is a virtual additional input, thus obtaining:
static feedback linearization system. The block diagram of fig. 3 shows the architecture of the control scheme.
For exponential convergence of position errors P R,r-PR＝ep to 0, the natural choice is a linear controller:
Wherein the diagonal positive definition gain matrix K p1,Kp2 defines the Hurwitz polynomial.
We now consider direction tracking. A common option for representing direction tracking is to use euler angles. But euler angles are prone to singularity problems. Therefore, the invention directly develops the directional controller on SO (3), thereby avoiding the common problem of local coordinate systems.
Next, assuming that R R,AndWherein [. Cndot. v represents the inverse mapping from SO (3) to R 3. The attitude tracking error e R∈R3 is given by:
The tracking error of the angular velocity e ω∈R3 is given by:
In order to asymptotically converge the rotation error e R to 0, the following directional controller is selected:
Wherein the diagonal positive definition gain matrix K R1,KR2 again defines the Hurwitz polynomial.
Contact wrench estimation
Knowledge of the contact interaction wrench between the tool tip and the environment in order to properly handle the physical interactions of the aerial robot with the external environmentIs indispensable. For this purpose, a force/torque sensor may be mounted on the tool tip of the robot, which may generally provide a reliable measure, but this solution increases both the cost and the weight of the robot. In the field of aerial robots, a more viable solution is to employ a wrench estimator that can provide a sufficiently accurate estimate, expressed as: accurately measure position, velocity and acceleration (if any).
External wrench of robotIt can be seen as the impact of the wrench applied to the tool tip by the environment on the wrench CoM.
The NCFTP stage sensor device used in the experiments can make sufficiently accurate measurements of stage position and velocity (angle and linearity), while only the linear acceleration provided by the IMU can be reasonably used in the wrench observer. Thus, in this context Tomic et al have proposed a mixing method, which has been followed. In more detail, it is an acceleration-based observer proposed by You Kesai mol (Yuksel) et al. To estimate the external interaction force f R on the robot CoM, external torqueObtained by using a momentum-based observer.
Disturbance observer under estimation of contact force requires vehicle acceleration measures, originally proposed by You Kese lux aerial robots, used to estimate contact force
Where L ε R 3×3 is the gain matrix to be designed, andIs an estimate f R. You Kese et al show the error dynamics in the presence of constant or slowly varying external forces.
Where O 3∈R3 is the (3×1) zero vector. Thus, for any positive definite matrix L, the error kinetic index converges to the origin.
Contact torque estimation in order to estimate the interaction torque exerted by the external environment on the tool tip, momentum-based observers have been designed. Referring to the system dynamics (10), the angular momentum q R∈R3 in the frame FR can be calculated as
From (3), the time derivative of (21) can be expressed as
By using (22), estimateCan be regarded as a residual vector
Where t and t 0 are the current time and the initial time, respectively, and K l is a positive fixed gain matrix. By reasonable assumptionThis means that q R(t0) is also zero. By taking the time derivatives of (23) to (22), the following dynamics of the residual vector can be obtained
Equation (24) is a first order low pass dynamic system: it can be readily appreciated that, for any positively defined base matrix K l, when t→infinity,Filter characteristics of observer: a larger gain value may increase the convergence speed, while a smaller value may filter high frequency noise.
The wrench acting on the tool noseAndAfter all are known, the estimated wrench acting on the tool noseThe calculation formula of (2) is
A wrench observer demonstrates one example of precision. In this test, the aerial robot has been firmly connected to an ATI45 force-torque sensor, which itself has been mounted on the test bench, and by letting the aerial robot track the trajectory and using only the inner ring pose controller, the aerial robot can apply the force/moment profile defined by the trajectory on the force/moment sensor. In order to correctly test the limits of the wrench observer, a chirp signal (cosine signal of gradually increasing frequency) which simultaneously surrounds a plurality of axes has been used as a trace.
Interacting spanner compensation
In order to obtain the best results of the admittance filter, a low level tracking with high rigidity is required. This can be achieved by increasing the gain in (14) and (17). The disadvantage of this solution is that noise is also amplified, which may make the low-level system more prone to instability. Furthermore, a true zero tracking error is still not achieved. Instead, to improve convergence, an estimated wrench is usedFeedback to the lower level controller as an additional item in (12). Thus, even contactless flight tracking is improved (see fig. 4) since any steady state error is driven to zero. In more detail, the final control input is given by
If the interaction force and torque are perfectly compensated, the error dynamics are the same as (14) and (17), and therefore the same stability is maintained. Otherwise, under the trivial assumption that the interaction wrench is bounded, the wrench estimation error may also be considered a bounded term. The stability of the proposed controller without interaction has been studied previously. In detail, it has been demonstrated that if feasible for the TiltHex stages, the tracking error converges the exponential error to zero under mild conditions of the initial angular velocity error e ω. Thus, by reproducing the stability theory of the disturbance system without vanishing disturbances, it can be pointed out that in case of a bounded wrench estimation error, the tracking error is eventually bounded. Furthermore, if the interaction wrench is constant, the wrench error will converge to zero, so the tracking error will also disappear and will also converge to zero.
Admittance filter
To obtain a limited force exchange with the environment, compliance behavior may be enforced between the position and orientation of the end effector and the interacting generalized forces. According to position P E,d, direction R E,d, speedAnd accelerationThe planned ideal trajectory of the end effector is assigned, and a corresponding set of reference motion variables may be generated by an admittance filterHas the following dynamic characteristics
Where Deltav E＝vE,d-vE,r is the velocity error and e E is the attitude error given by
(27) Representing the dynamics of the 6 degree of freedom mechanical impedance of inertia M E, damping D E and stiffness K E: these matrices are all positive and apply the appropriate choice in the appropriate way-degrading the system behaviour. Furthermore, in order to guarantee the stability of the whole system, the gain matrix must ensure that the motion controller (inner loop) has a faster dynamic characteristic with respect to the admittance filter. Once the reference trajectory of the end effector is calculated, it should be represented as a CoM reference trajectory for tracking by the inner ring pose controller. Then calculate the reference position and direction of the robot (see FIG. 2) as
Whereas the CoM reference speed and acceleration are obtained by taking the time derivative of (29). In detail, the reference speed is given by:
While the reference acceleration is
The above description is only a preferred embodiment of the present invention, and is not intended to limit the invention in any way, and any person skilled in the art may make modifications or alterations to the disclosed technical content to the equivalent embodiments. However, any simple modification, equivalent variation and variation of the above embodiments according to the technical substance of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims (3)

1. A6D interactive control method of an aerial robot is characterized in that the aerial robot is driven by an inclined propeller to control the complete 6D gesture and apply complete torsion, and a firm end effector is attached; interaction is achieved by admittance control, wherein the outer loop controls the required admittance behavior, i.e. interaction compliance stiffness, damping and mass, while the inverse dynamics based inner loop ensures complete 6D pose tracking, wherein the interaction forces are estimated by an IMU enhanced momentum-based observer;
the aerial robot adopts a multi-rotor robot with a non-collinear fixed inclined propeller, so that translation and angular acceleration are independently controlled under the condition of no constraint, or an applied spanner is independently controlled when the spanner is contacted, and 6D force control is realized;
the modeling method of the six-rotor general model of the non-collinear fixed inclined propeller comprises the following steps:
The following coordinate system is established:
Inertial world frame Its axis and origin are denoted by { x W,yW,zW } and O W, respectively;
Frame of bicycle O R-{xW,yW,zW, installed on the robot, aircraft CoM with origin O R and end effector;
End effector frame O E-{xR,yR,zR, has been attached to the robotic end effector and has a start point at interaction point O E;
At the position of AndThe position of O E in (a) is represented by P E∈R3 and P E R∈R3, respectively; the rotation matrix is represented by R R e SO (3) and R E e SO (3), where SO (3) = { a e R 3×3|AAT =i a detA =1 }; relative to inertial systemIs thatIn addition, R E R ε SO (3) represents a constant rotation matrix, representingRelative toIs a direction of (2); finally, omega R R∈R3 and omega E E∈R3 are carried out,AndRelative toIs represented by intra-frame omega R R and omega E E, respectively;
Wherein [ (s ] x ε SO (3) represents the oblique symmetry matrix associated with vector [ s ] R 3; using Newton-Euler form sense, the equation of motion of an aerial robot is expressed as:
Where m and J ε R 3×3 represent the robot mass and its inertial matrix, g is the gravitational acceleration, f ε R 3 and τ R∈R3 are the force and torque inputs, and f R and τ R R are the external force and torque applied to the robot CoM;
in formula (3), translational dynamics are expressed in the frame In (2), while the rotation dynamics are represented in the frameIn (a) and (b);
The inclined hexadecimal is derived from (3) for a non-collinear fixed inclined propeller platform without losing the generality of any full-drive high-altitude platform; non-collinear fixed inclined propeller platform is based on six-rotor structure, equidistant spacing and of propellers Equidistant from O R in the x-y plane of (2); complete drive is achieved by means of a rigid adapter, tilting each individual motor-propeller combination; considering the 6 frames to be considered,I=1..6, wherein O Pi coincides with the centre of rotation of the i-th propeller group; relative to Is represented by a rotation matrix:
the ith motor propeller group is relative to Is defined by constant parameters α and β; selecting an alternate alpha symbol for each other propeller in equation (4) to fully actuate the aircraft;
The position of the ith motor rotor set relative to O R is defined as:
The rotating propeller generates thrust and drag in O Pi;
Wherein c f and c τ are constant parameters, and the rotation speed of the propeller is determined Correlating with the generated thrust and resistance moment; (-1) i-1 in formula (7) represents the effect on the counter-rotating rotors of all even propellers;
Will be applied to The total force of the upper O R is expressed as:
F 1∈R3×6, and u= [ u 1...u6]T ] is a vector collecting 6 square propeller rotational speeds;
under the same conditions, the total torque is combined, and the thrust contribution and drag torque in F B are expressed using formula (6) and formula (7):
Wherein F 2∈R3×6; by substituting the formulas (8) and (9) into the formula (3), it is obtained:
this serves as a reliable kinetic model under slow flight conditions.
2. The method of 6D interactive control of an aerial robot of claim 1 wherein the measurement of the interactive wrench uses a force/torque sensor, wherein wrench measurements of the sensor on the aerial robot wrist are fed into an admittance filter.
3. The method of 6D interactive control of an aerial robot of claim 1, wherein the control framework of the aerial robot is based on an outer loop admittance control and an inner loop full-pose controller; the state of the aerial robot is estimated by a concentrated Kalman filter UKF which fuses the acceleration and angular velocity measurements of the inertial measurement unit IMU with the position and orientation of the attitude sensor.
CN202111557112.3A 2021-12-18 2021-12-18 6D interaction control method of aerial robot Active CN114237054B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111557112.3A CN114237054B (en) 2021-12-18 2021-12-18 6D interaction control method of aerial robot

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111557112.3A CN114237054B (en) 2021-12-18 2021-12-18 6D interaction control method of aerial robot

Publications (2)

Publication Number Publication Date
CN114237054A CN114237054A (en) 2022-03-25
CN114237054B true CN114237054B (en) 2024-08-09

Family

ID=80758704

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111557112.3A Active CN114237054B (en) 2021-12-18 2021-12-18 6D interaction control method of aerial robot

Country Status (1)

Country Link
CN (1) CN114237054B (en)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CA1311061C (en) * 1989-05-23 1992-12-01 Hidetaka Nose Method of teaching a robot
CN106708078A (en) * 2017-02-21 2017-05-24 西北工业大学 Rapid attitude stabilization method suitable for space robot actuator failure

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101262840B1 (en) * 2011-08-05 2013-05-09 한국과학기술연구원 Human-robot cooperative system and method of parts based on the same
US11305431B2 (en) * 2016-11-10 2022-04-19 Cognibotics Ab System and method for instructing a robot
JP6851837B2 (en) * 2017-01-20 2021-03-31 キヤノン株式会社 Manufacturing methods for control devices, robot systems, control methods, programs, recording media and articles
EP3581341B1 (en) * 2018-06-13 2020-12-23 Siemens Healthcare GmbH Method for operating a robot, data storage having a program code, robot and robot system
CN109240343B (en) * 2018-09-10 2021-11-16 西北工业大学 Rope-tied robot approaching target pose integrated control method
CN109343369B (en) * 2018-11-19 2021-08-10 南京邮电大学 Design method of four-rotor fault-tolerant controller based on non-linear observer
CN110053044B (en) * 2019-03-19 2022-03-22 江苏大学 Model-free self-adaptive smooth sliding mode impedance control method for clamping serial fruits by parallel robot
EP3845346A1 (en) * 2019-12-31 2021-07-07 Fundación Tecnalia Research & Innovation Method, system and computer program product for controlling the teleoperation of a robotic arm
CN111660306B (en) * 2020-05-27 2021-07-20 华中科技大学 Robot variable admittance control method and system based on operator comfort
CN111984024B (en) * 2020-08-25 2022-08-09 福州大学 Disturbance and uncertainty control method based on operation type flying robot
CN112527008B (en) * 2020-12-21 2023-05-02 福州大学 Low-complexity specified performance-based operation type flying robot control method

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CA1311061C (en) * 1989-05-23 1992-12-01 Hidetaka Nose Method of teaching a robot
CN106708078A (en) * 2017-02-21 2017-05-24 西北工业大学 Rapid attitude stabilization method suitable for space robot actuator failure

Also Published As

Publication number Publication date
CN114237054A (en) 2022-03-25

Similar Documents

Publication Publication Date Title
Ryll et al. 6D physical interaction with a fully actuated aerial robot
Choi et al. Nonlinear control of quadrotor for point tracking: Actual implementation and experimental tests
Meng et al. Vibration suppression control of free-floating space robots with flexible appendages for autonomous target capturing
Jimenez-Cano et al. Control of an aerial robot with multi-link arm for assembly tasks
Suarez et al. Lightweight and human-size dual arm aerial manipulator
Lippiello et al. Cartesian impedance control of a UAV with a robotic arm
Caccavale et al. Six-dof impedance control of dual-arm cooperative manipulators
Orsag et al. Stability control in aerial manipulation
Korpela et al. Flight stability in aerial redundant manipulators
CN112558621A (en) Decoupling control-based flying mechanical arm system
Pounds et al. UAV rotorcraft in compliant contact: Stability analysis and simulation
Korpela et al. Towards the realization of mobile manipulating unmanned aerial vehicles (MM-UAV): Peg-in-hole insertion tasks
Wopereis et al. Multimodal aerial locomotion: An approach to active tool handling
Yu et al. An over-actuated multi-rotor aerial vehicle with unconstrained attitude angles and high thrust efficiencies
CN106802570B (en) Method and device for tracking position of unmanned helicopter
CN110641738B (en) Trajectory tracking control method of space five-degree-of-freedom free flying mechanical arm
CN111459188A (en) Multi-rotor nonlinear flight control method based on quaternion
Sun et al. Dynamic modeling and hardware-in-loop simulation for a tail-sitter unmanned aerial vehicle in hovering flight
CN113580127B (en) Multi-rotor aircraft humanoid double-mechanical-arm system and dynamic self-balancing control design method thereof
Stingu et al. Design and implementation of a structured flight controller for a 6DoF quadrotor using quaternions
Korpela et al. Insertion tasks using an aerial manipulator
CN114237054B (en) 6D interaction control method of aerial robot
Gonzalez-Hernandez et al. Altitude control improvement for a Quadrotor UAV using integral action in a sliding-mode controller
Siciliano et al. Six-degree-of-freedom impedance robot control
Anuar et al. Super-twisting sliding mode control with adaptive gain of quadrotor with rigid manipulator

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant