Control method and device for position-based impedance controlled industrial robot
View Patent ↗The present invention relates to a control method and device for position-based impedance controlled industrial robot, and more particularly, a control method and device for position-based impedance controlled industrial robot able to improve contact stabilization with regard to an environment with a variety of stiffness. According to a control method and device for position-based impedance controlled industrial robot in accordance with the present invention, a robust contact stabilization for a position-based impedance controlled industrial robot contacting and interacting with an uncertain actual environment may be guaranteed.
1. A control method for position-based impedance controlled industrial robot in accordance with a force control method for a robot system with a position controller comprising:
(a) setting a first parameter;
(b) detecting a displacement (X=Xd) and a contact force (Fe) of a robot;
(c) calculating a stiffness estimate value of an environment from the displacement (X=Xd) and the contact force (Fe) of the robot;
(d) calculating a second control parameter from the first control parameter and the stiffness estimate value;
(e) judging whether a position variation amount is 0;
(f) calculating an operating force (Fd) of a robot if a position variation amount is 0;
(g) calculating a bounding parameter if a position variation amount is not 0;
(h) calculating a range of bounding parameter values satisfying a condition a robot system is passive using the first parameter and the second parameter, and calculating an operating force (Fd) of a robot according to the calculated range of bounding parameter values;
(i) applying the calculated operating force (Fd) in an impedance controller;
(j) calculating a position correction displacement (Xf) by the impedance controller; and
(k) applying the calculated position correction displacement (Xf) to the position controller.
2. The method according to claim 1 , wherein (a) further comprises,
initializing the first control parameter to a nominal environmental stiffness value;
gradually increasing the first control parameter when a stable contact is maintained while a contact experiment with a high stiffness environment is performed;
gradually decreasing the first control parameter when an unstable contact is generated while a contact experiment with a high stiffness environment is performed; and
setting a maximum value, which guarantees stability, to the first control parameter.
3. The method according to claim 1 , wherein (d) is where the second control parameter is calculated as a smaller value than the first control parameter and the stiffness estimate value.
4. The method according to claim 1 , wherein the operating force (Fd) of a robot is calculated with the following control rule:
F d [n]=F d [n− 1 ]+β[n]Δx d [n].
Here, Fd[n] is an operating force of a robot, β[n] is a bounding parameter, xd[n] is a displacement of a robot.
5. The method according to claim 1 , wherein the bounding parameter is calculated with the following equation:
β
[
n
]
=
F
e
[
n
]
-
F
d
[
n
-
1
]
Δ
x
d
[
n
]
for
Δ
x
d
[
n
]
≠
0.
Here, β[n] is a bounding parameter, Fe[n] is an actual contact force of an environment measured from a force/torque sensor, Fd[n] is an operating force of a robot, xd[n] is a displacement of a robot.
6. The method according to claim 1 , wherein (h) modifies the bounding parameter to the maximum value when the bounding parameter is higher than a set maximum value, modifies the bounding parameter to the minimum value when the bounding parameter is lower than a set minimum value, and maintaining the value of the bounding parameter when the bounding parameter is between the minimum and the maximum value.
7. The method according to claim 6 , wherein the range of the bounding parameter is determined by the following bounding rule:
β max [n ]=min( c 1 ,γ max [n ]),β min [n]=γ min [n]
and,
γ
max
[
n
]
=
c
2
-
F
d
[
n
-
1
]
Δ
x
d
[
n
]
+
c
2
2
+
(
F
d
[
n
-
1
]
Δ
x
d
[
n
]
)
2
γ
min
[
n
]
=
c
2
-
F
d
[
n
-
1
]
Δ
x
d
[
n
]
-
c
2
2
+
(
F
d
[
n
-
1
]
Δ
x
d
[
n
]
)
2
.
Here, βmax[n] is a maximum value of a bounding parameter, βmin[n] is a minimum value of a bounding parameter, Fd[n] is an operating force of a robot, xd[n] is a displacement of a robot, c1 is a first control parameter, c2 is a second control parameter, c1 and c2 are positive values and c1≧c2.
8. The method according to claim 1 , wherein a control method for position-based impedance controlled industrial robot further comprises,
(l) judging whether the value of the bounding parameter exceeds a range satisfying a condition for passivity and is bounded to a maximum value or a minimum value (EBA bounding generated); and
(m) calculating a new reference position (X0) when EBA bounding is generated and applying to the position controller.
9. A control device for position-based impedance controlled industrial robot comprising:
a sample/hold unit for performing input and output with a position-based impedance controlled industrial robot, and interconverting analog signals and digital signals; and
a control unit receiving a physical parameter of the position-based impedance controlled industrial robot from the sample/hold unit, and calculating an operating force, which is to be transmitted to the position-based impedance controlled industrial robot, according to an interaction between the position-based impedance controlled industrial robot and an environment, and transmitting to the position-based impedance controlled industrial robot through the sample/hold unit.
The control unit calculates a bounding parameter for the position-based impedance controlled industrial robot according to the physical parameter, and calculates an operating force of the position-based impedance controlled industrial robot according to a range of a bounding parameter value, in which a robot system satisfies a passivity condition in the environment, and applies to an impedance controller, and applies the calculated position compensation displacement, which is calculated by the impedance controller, to a position controller.
10. The device according to claim 9 , wherein the operating force (Fd) of a robot is calculated with the following control rule:
F d [n]=F d [n− 1 ]+β[n]Δx d [n].
Here, Fd[n] is an operating force of a robot, β[n] is a bounding parameter, xd[n] is a displacement of a robot.
11. The device according to claim 9 , wherein the bounding parameter is calculated with the following equation:
β
[
n
]
=
F
e
[
n
]
-
F
d
[
n
-
1
]
Δ
x
d
[
n
]
for
Δ
x
d
[
n
]
≠
0.
Here, β[n] is a bounding parameter, Fe[n] is an actual contact force of an environment measured from a force/torque sensor, Fd[n] is an operating force of a robot, xd[n] is a displacement of a robot.
12. The device according to claim 9 , wherein the control unit modifies the bounding parameter to the maximum value when the bounding parameter is higher than a set maximum value, modifies the bounding parameter to the minimum value when the bounding parameter is lower than a set minimum value, and maintaining the value of the bounding parameter when the bounding parameter is between the minimum and the maximum value.
13. The device according to claim 6 , wherein the range of the bounding parameter is determined by the following bounding rule:
β max [n ]=min( c 1 ,γ max [n ]),β min [n]=γ min [n],
and,
γ
max
[
n
]
=
c
2
-
F
d
[
n
-
1
]
Δ
x
d
[
n
]
+
c
2
2
+
(
F
d
[
n
-
1
]
Δ
x
d
[
n
]
)
2
γ
min
[
n
]
=
c
2
-
F
d
[
n
-
1
]
Δ
x
d
[
n
]
-
c
2
2
+
(
F
d
[
n
-
1
]
Δ
x
d
[
n
]
)
2
.
Here, βmax[n] is a maximum value of a bounding parameter, βmin[n] is a minimum value of a bounding parameter, Fd[n] is an operating force of a robot, xd[n] is a displacement of a robot, c1 is a first control parameter, c2 is a second control parameter, c1 and c2 are positive values and c1≧c2.
14. The device according to claim 9 , wherein the control unit judges whether the bounding parameter value exceeds a range satisfying a condition for passivity and is bounded to a maximum value or a minimum value (EBA bounding generated), and calculates a new reference position (X0) when EBA bounding is generated and applies to the position controller.