JP3305050B2 - Space robot controller - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a space robot controller, and more particularly to a space robot having a robot body, a manipulator extended from the robot body, and a robot body control actuator for controlling the robot body. Related to a space robot controller for controlling a robot.

[0002]

2. Description of the Related Art In future space activities, assembling and maintaining large structures in orbit, supplying materials to space stations and unmanned platforms, conducting various space experiments such as materials science and life science, Various operations such as recovery and repair of satellites and unmanned exploration as a precursor to manned exploration of the moon and Mars are required. Some of these tasks are appropriate to be performed by astronauts,
In consideration of rational utilization of astronaut's ability, necessary cost, and the like, it is desired that a robot perform a part of on-orbit work.

[0003] As a space robot, JE installed in the Japanese Experiment Module of the space station FREEDOM
In addition to a large manipulator installed at the space station like MRMS, it has a propulsion device such as a thruster and an attitude control device such as a flywheel and has a function to move to the target location and performs various tasks with the mounted manipulator A free-flying space robot capable of performing the following has been proposed. Technical test hygiene type VII
It is a technology test satellite scheduled to be launched in 1997, and is a free-flying space robot.

[0004] Since a free-flying space robot has a moving function, it is expected to perform various tasks such as carrying work between the space station and an unmanned platform, and maintenance and inspection work around the space station.

However, in a free-flying space robot,
When the onboard manipulator is moved, the robot is suspended in the weightless space,
A problem arises in that the posture of the robot main body is changed by the force and the torque (hereinafter, referred to as force / torque). If the target trajectory of the manipulator is generated so that the tip of the manipulator grasps the target based on the values of the position and posture of the robot body before the work, if the posture fluctuation of the robot body occurs due to the recoil accompanying the manipulator movement The tip of the manipulator is positioned at a position different from the target, so that the target operation cannot be performed.
Various researches have been conducted to solve the problem of the posture change of the robot body accompanying the movement of the manipulator.

[0006] For example, Japanese journal of robot science, VOL.
7, no. In the “Object capture control experiment using a free-floating space tele-robot model” described in No. 6, if the propulsion device / attitude control device of the space robot itself is not used, the momentum
Utilizing that angular momentum is preserved, a generalized Jacobian matrix that specifies the relationship between the joint speed of the robot body and the manipulator and the speed at the manipulator tip is derived, and decomposition speed control is performed using the generalized Jacobian matrix. By doing this, it was shown that the manipulator fleet could be positioned at the target position even if the posture of the robot body fluctuated due to the recoil of the manipulator. However, when the space robot main body is varied, problems occur other than the positioning of the manipulator tip. For example, a communication antenna mounted on the robot body needs to be pointed with high precision. If the posture of the robot body greatly changes due to the recoil caused by the movement of the manipulator, communication may become impossible. Therefore, it is necessary to control the robot body with high precision so that the position and posture of the robot body do not change even if the manipulator moves.

When the recoil due to manipulator motion is compensated by the propulsion device and attitude control device of the robot body, it is important that the force / torque of the reaction does not exceed the allowable maximum output of the propulsion device and attitude control device of the robot body. is there.
In particular, when the flywheel is used as an attitude control device, the maximum output torque is not so large, so that it is impossible to prevent the reaction torque of the manipulator from exceeding the allowable maximum output.

In "Cooperative control of arm system and attitude system in space robot" described in the preliminary proceedings of the 10th Annual Meeting of the Robotics Society of Japan, the recoil due to manipulator motion in a space robot is caused by the propulsion device and attitude control device of the robot body. A control method that does not exceed the allowable value has been proposed. The recoil is calculated by performing inverse dynamics calculation from the target trajectory of the manipulator generated by remote operation of the operator, etc., and if it exceeds the allowable value, the motion speed of the manipulator is reduced so that the reaction does not exceed the allowable value. I have to. Since the inverse dynamics calculation requires a large amount of calculation, it is performed by a high-speed computer on the ground, and a trajectory with a reduced speed is generated and then transferred to the space robot as a target trajectory on the orbit.

However, this method requires a computer equipped with an expensive high-speed real-time OS on the ground. This is because the inverse dynamics requires a large amount of calculation, and it is necessary to correct a target trajectory generated by remote control of an operator in real time. Although it is necessary to accurately identify the dynamic characteristics of the space robot in advance, it is not always easy to perform accurate identification on a powered ground.

[0010]

As described above, in a free-flying space robot in which the posture variation of the robot body is problematic due to the recoil caused by the movement of the manipulator, the allowable capacity of the actuator such as the flywheel of the robot body is considered. Trajectory generation method of a manipulator existed, but it required expensive high-speed computers to calculate the recoil by inverse dynamics calculation, and it was necessary to accurately identify the dynamic characteristics of the space robot in advance There was a point.

SUMMARY OF THE INVENTION An object of the present invention is to provide a space robot control device which does not need to identify the dynamic characteristics of a space robot and does not require an expensive high-speed computer, taking into account the allowable capacity of the actuator for controlling the robot body. It is to be.

[0012]

In order to achieve the above object, a space robot controller according to the present invention comprises a robot body, a manipulator extended from the robot body, and a robot for controlling the robot body. A space robot control device for controlling a space robot including a main body control actuator, wherein a virtual value to be applied to a tip of the manipulator required for a current value of a tip position of the manipulator to follow its target value. A work coordinate control device that calculates a target force and torque, and a value of the force and torque received by the robot body due to the calculated virtual target force and torque does not exceed an allowable maximum value of the robot body control actuator. A force and torque correction device for correcting the target force and torque, A joint torque converter for calculating a target joint torque to be applied to a joint of the manipulator required to realize the target force and torque corrected by the writing and torque correcting device; and A joint torque control unit, comprising: a tip position calculation unit that calculates and updates the current value of the tip position of the manipulator from a value of a joint angle of the joint when the target joint torque is realized, I do.

It is preferable that the joint torque control unit is realized as a torque control system that feeds back the value of a torque sensor attached to each joint of the manipulator.

[0014]

The coordinates of the virtual target force and torque (expressed as force / torque) to be applied to the tip of the manipulator so as to follow the target trajectory of the manipulator generated by remote operation of the operator or the like. Calculated by the device.

Based on the target force / torque value calculated by the work coordinate control device, a virtual force / torque value received by the robot body as a reaction is calculated. If the virtual force / torque value received by the robot body exceeds the maximum allowable output of the robot body control actuator that controls the propulsion and attitude of the robot body, it exceeds the maximum allowable output of the robot body control actuator. Is corrected by a force / torque corrector so that there is no This correction decomposes the target force / torque value into a target trajectory follow-up component which is a component in the target trajectory direction and an error compensation component other than the target trajectory follow-up component, and the robot receives the target trajectory follow-up component. This is performed by preferentially reducing the value of the virtual force / torque to be less than or equal to the maximum allowable output of the robot body control actuator.

The value of the target joint torque to be applied to the joint of the manipulator required for realizing the target force / torque value corrected by the force / torque correction device is calculated by the joint torque conversion device. The calculated target joint torque at each joint is realized by the joint control unit, and the current value of the manipulator tip position is calculated and updated from the joint angle value of each joint when the target joint torque is realized by the tip position calculation unit. I do. The updated current value follows the target value more closely. This prevents the force / torque value of the reaction to the robot body due to the movement of the manipulator from exceeding the maximum allowable output of the actuator for controlling the robot body, and the position of the distal end of the manipulator generated by remote control by the operator. Movement can be realized. In addition, since calculations requiring a large amount of calculation such as inverse dynamics calculations are not required, an expensive high-speed computer can be omitted.

The joint torque control unit is realized as a torque control system that feeds back the value of a torque sensor attached to each joint of the manipulator. Even if present, the value of the target joint torque can be accurately realized, and high-speed followability of the tip of the manipulator can be realized.

[0018]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a space robot control device for controlling the space robot 8.

As shown in FIG. 4, the space robot 8 includes a bot main body 10, a manipulator 11 extended from the robot main body 10, and a robot main body control actuator 12 for controlling the robot main body 10. I have.

The space robot controller is a working coordinate controller for calculating a virtual target force / torque to be applied to the tip of the manipulator 11 for the current value of the tip position 11a of the manipulator 11 to follow the target value. 1
The virtual force / torque received by the robot body 10 by the calculated target force / torque does not exceed the maximum allowable value of the robot body control actuator 12.
A force / torque correction device 2 for correcting the calculated target force / torque, and a target to be applied to the joint 11b of the manipulator 11 required to realize the target force / torque corrected by the force / torque correction device 2 A joint torque converter 6 for calculating joint torque, a joint torque control unit 7 for achieving a target joint torque, and a joint position of the manipulator 11 based on a joint angle value of the joint 11b when the target joint torque is achieved. A tip position calculator 9 that calculates and updates the current value is provided.

As shown in FIG. 2, the force / torque correcting device 2
Is a target force / torque decomposing unit 3 that decomposes a target trajectory tracking component, which is a component in a trajectory direction that targets a target force / torque value, into an error compensation component other than the target trajectory tracking component; A robot main body load force / torque calculation unit 4 for calculating a virtual force / torque value received by the robot, and an arm tip target force / torque correction unit 5 for correcting the force / torque at the tip position 11a of the manipulator 11. I have.

When the target tip position of the manipulator tip is given as an input by remote control or the like by an operator, the work coordinate control device 1 outputs the value of the target tip position of the manipulator tip and the current manipulator output from the tip position calculation unit 9. The target force / torque value at the distal end position 11a of the manipulator 11 is calculated from the value of the distal end position 11a.

The tip position calculator 9 calculates the tip position 11a of the manipulator 11 from the joint sensor signal attached to the space robot 8.

The work coordinate control device 1 is configured as a PID control system that does not require a model to be controlled, and calculates a target force / torque such that the manipulator tip position 11a follows a target value.

The vector of the target value at the tip position 11a of the manipulator 11 is Xd, and the target force at the tip position 11a is F.
Assuming that d is the target torque and Md is the current position vector of the distal end position of the manipulator 11, X is a composite vector Fd 'of target force / torque to be applied to the distal end position 11a of the manipulator 11 as follows. .

[0026]

(Equation 1) Δx = xd−x Kp: proportional gain matrix Kd: differential gain matrix Ki: integral gain matrix The value of the target force / torque at the tip position 11a of the manipulator 11 is the target force / torque in the target force / torque correction device 2. It is input to the decomposition unit 3. The value of the target position at the tip of the manipulator 11 is also input to the target force / torque decomposing unit 3, and the target force / torque value at the tip of the manipulator is converted into a target trajectory tracking component and an error compensation component from the target trajectory as follows. Decompose. The error compensating component from the target trajectory is a component of the target force / torque vector at the tip of the manipulator in the normal direction of the target trajectory, and is a component that works to reduce the displacement of the tip of the manipulator from the target trajectory.

On the other hand, the target trajectory tracking component is a component that works to reduce an error along the trajectory among errors between the target position and the current manipulator tip position. The decomposition calculation is performed as follows.

As shown in FIG. 3, (fxt, fyt) is a target trajectory tracking component, and (fxv, fyv) is an error compensation component from the target trajectory.

[0029]

(Equation 2) Updating of the value of Xtd is performed at each cycle of the robot control system by passing through the position X of the tip of the manipulator.

[0030]

(Equation 3) The intersection between the plane perpendicular to the target and the target trajectory is calculated, and a new Xtd
Is performed. Vtd is a time differential vector of Xtd, and is a tangential vector of the target trajectory. Similar to the decomposition of the target force, the decomposition of the target torque is performed by calculating the tangent component in the posture vector space from the values of the target posture vector at the manipulator tip and the value of the manipulator tip posture vector. Note that the symbol "/" indicates the calculation of the inner product.

The trajectory tracking component and the trajectory error compensating component of the target force / torque output from the target force / torque decomposing unit 3 are input to the robot body load force / torque calculating unit 4. The value of the reaction force / torque applied to the robot body 10 when the target force / torque, which is the output of (1), is generated at the tip of the manipulator. Robot body 10
The calculation of the reaction force / torque applied to is calculated assuming a state of static balance as follows.

Fdv = Fv Mbv = Fv × x + Mv (3) Fbt = Ft Mbt = Ft × x + Mb (Fbv, Fbv) is the reaction force / torque of the robot body due to the trajectory error compensation component of the target force / torque at the tip.
(bt, Mbt) is the reaction force / torque of the robot body due to the trajectory following component of the target force / torque at the tip. Further, x is a position vector (only position, not including posture) of the tip of the manipulator 11, and is a value in the coordinate system of the robot body 10. The symbol x indicates an outer product. Therefore, the force / torque calculated by the equation (3) is the force and torque applied to the origin of the coordinate system of the robot body 10 assuming a state of static balance.

The output of the robot body load force / torque calculation unit 4 is input to the arm tip target force / torque correction unit 5,
It is calculated how much the trajectory following component and the trajectory error compensation component of the force / torque value of the target tip should be reduced.

In general, in the space robot 8, in the robot body control actuator 12, the thrust in the translation direction is generated by the thruster, and the torque in the rotation (posture) direction is generated by the flywheel or the CMG. Therefore, the maximum allowable output of the actuator 12 for controlling the robot body exists individually for the force and the torque. The allowable maximum value of force is Fmax, and the allowable maximum value of torque is Mmax.
Then, the value of the force / torque at the target tip is corrected as follows. rm = | Mbv | / | Mmax |, rf = | Fbv
| / | Fmax | a) | Mbv | ≧ | Mmax | (| rm |> |
rf |) Fd = Fv · | Mmax | / | Mbv | Md = Mv · | Mmax | / | Mbv | b) | Fbv | ≧ | Fmax | (| rf |> |
rm |) Fd = Fv · | Fmax | / | Fbv | Md = Mv · | Fmax | / | Fbv | c) | Mbv + Mbt | ≦ | Mmax | and | F
bd + Fbt | ≦ | Fmax | Fd = Fv + Ft Md = Mv + Mt d) Except for c), | Mbv | <| Mmax | and |
If Fbv | <| Fmax | 1) Find a positive α1 satisfying | Mbv + α1 · Mbt | = | Mmax |.

| Mvv + α1 · Mbt | = | Mmax | ↓ | Mbt | ² α1 ² +2 (Mbv · Mbt) α1 + |
Mbv | ² − | Mmax | ² = 0 Calculate α1 as a solution of the above quadratic equation.

2) | Fbv + α2 · Fbt | = | F
max | is found.

3) α1 and α2 are compared, and the smaller value is set as α.

4) Fd = α · Ft + Fv Md = α · Mt + Mt The force / torque value of the manipulator tip corrected as described above is input to the tip force / torque → joint torque converter 6, and the tip is calculated as follows. Each joint torque that realizes the force / torque is calculated.

Td = J ^T · Fd ′ J: Jacobian matrix The above is a conversion formula assuming a static balance state. By modifying the target force / torque value at the tip of the manipulator as described above, the reaction force / torque received by the robot body due to the movement of the manipulator can be prevented from exceeding the allowable maximum force / torque of the actuator of the robot body. it can.

The value of each joint target torque, which is the output of the joint torque converter 6, is input to a current-controlled joint torque controller 7. The joint torque control unit 7 realizes each joint target torque, which is an input value, by controlling the motor current for driving each joint 11b. Therefore, the tip position of the robot body 8 moves so as to follow the target tip position.

As described above, by leaving the error compensation component from the trajectory of the manipulator 11 preferentially, even if the movement along the target trajectory may be delayed with respect to the target trajectory, Deviation can be minimized. Generally, if the movement of the manipulator deviates from the target trajectory, a collision with the surrounding environment may occur, which is not desirable. It is naturally desirable that the movement along the target trajectory also follows the target trajectory, but collisions with the surrounding environment can be avoided as long as they are on the target trajectory. In addition, it is thought that the minimum required functions can be provided because the work time is increased but the target position is finally reached.

If a torque sensor is attached to each joint of the manipulator and a torque control system is configured to realize the target joint torque which is the output of the joint torque converter 6, even if the friction of the speed reducer is large, It is possible to accurately realize the value of the target torque of the joint, and it is possible to realize more accurate target location followability.

[0043]

As described above, according to the configuration of the present invention, the force / torque received by the robot body as a reaction due to the movement of the manipulator is limited in real time so as not to exceed the maximum allowable value of the actuator for controlling the robot body. Becomes possible. As a result, even if the manipulator receives a reaction due to the movement of the manipulator, it is possible to prevent the position and posture of the robot body from changing. In addition, since calculations requiring a large amount of calculation such as inverse dynamics calculations are not required, an expensive high-speed computer can be omitted.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a schematic configuration of an embodiment of the present invention.

FIG. 2 is a block diagram illustrating a configuration of a force / torque correction device.

FIG. 3 is a diagram showing a target trajectory error compensation component of a target trajectory follow-up component of a target force / torque at the tip of the manipulator.

FIG. 4 is a plan view showing a schematic configuration of a space robot.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 work coordinate control device 2 force / torque correction device 3 target force / torque decomposition unit 4 robot body load force / torque calculation unit 5 arm tip target force / torque correction unit 6 joint torque conversion device 7 joint torque control unit 8 space robot 9 Tip position calculation unit 10 Robot body 11 Manipulator 11a Tip position of manipulator 11b Joint 12 Robot body control actuator

Continuation of front page (56) References JP-A-4-129688 (JP, A) JP-A-4-3318104 (JP, A) JP-A-5-158540 (JP, A) JP-A-5-178295 (JP) JP-A-5-238493 (JP, A) JP-A-6-198585 (JP, A) JP-A-3-178788 (JP, A) JP-A-1-116704 (JP, A) (58) Field surveyed (Int.Cl. ⁷ , DB name) B25J 3/00-3/04 B25J 9/10-9/22 B25J 13/00-13/08 B25J 19/02-19/06

Claims (1)

(57) [Claims] 1. A space robot control device for controlling a space robot including a robot body, a manipulator extended from the robot body, and a robot body control actuator for controlling the robot body. A working coordinate control device for calculating a virtual target force and torque to be applied to the tip of the manipulator and required for the current value of the tip position of the manipulator to follow the target value; and the calculated virtual target. A force and torque correcting device for correcting the target force and torque so that the value of the force and torque received by the robot body by the force and torque does not exceed the maximum allowable value of the actuator for controlling the robot body; To achieve the target force and torque corrected by the A joint torque converter for calculating a target joint torque to be applied to a joint of the manipulator required for: a joint torque control unit for realizing the target joint torque; and a joint of the joint when the target joint torque is realized. A space robot control device comprising: a tip position calculation unit that calculates and updates the current value of the tip position of the manipulator from an angle value.