Upload
others
View
7
Download
0
Embed Size (px)
Citation preview
Effect of Stabilization Control for Cooperation between Tele-Robot Systems
with Force Feedback by Using Master-Slave Relation
Kazuya K𝐚𝐚𝐚𝐚𝐚𝐚𝐚𝐚𝐚𝐚𝐚𝐚𝐚𝐚†, Yutaka Ishibashi†, Pingguo Hung‡, and Yuichiro Tateiwa†
† Nagoya Institute of Technology, ‡Seijoh University
2020 IEEE Conference on Computer Applications (ICCA)February 27 - 28, 2020, Novotel Hotel, Yangon, Myanmar
Outline Background Problems Purpose Tele-robot systems with force feedback Adaptive Δ-causality control Stabilization control Experiment method Experiment results Conclusion and Future work
Background Remote robot systems with force feedback have actively been
researched. It is possible to transmit the information about the shape,
weight, and softness of a remote object by using haptic interface devices.
The efficiency and accuracy of work can largely be improved by using the remote robot system with force feedback.
However, when we transmit haptic information over the
Internet, which does not guarantee QoS (Quality of
Service)
Network delay, delay jitter, and
packet loss・QoE (Quality of Experience) degrades・Instability phenomena occur.
QoS control and stabilization control
Previous work *1 Applied the adaptive Δ-causality control to the cooperative work of
the two tele-robot systems with force feedback by using master-slave relation .
Investigated the effect of the adaptive ∆-causality control by experiment.
The influence of network delay can greatly be alleviated by the control
The reaction force is set to a value by multiplying 0.5 to the force detected by the robot's force sensor.
We can use the stabilization control with filters.The effect of the stabilization control has not been clarified quantitatively in the systems using the master-slave relation.
Problems*1 K. Kanaishi, et al., in Proc. ICTCE, Nov. 2019.
Purpose
We perform the stabilization control with filters for the tele-robot systems using the master-slave relation with force feedback in which the adaptive ∆-causality control is exerted.
Clarify the effect of the stabilization control.
Investigate the effect of the adaptive ∆-causality control under the stabilization control.
This work
Remote robot systems with force feedback
Carrying Object TogetherMove a wooden stick together by the two industrial robot arms
while watching video. In order to move the robot arms in almost the same way always,
we push and drop the uppermost block of the piled building blocks by moving the robot arms together with the force feedback devices.Toggle clamp handof slave robot arm Wooden stick
Toggle clamp handof master robot arm
Building blocks
Calculation method of position
Position of robot
𝑺𝑺𝑡𝑡 = �𝑴𝑴𝑡𝑡−1 + 𝑽𝑽𝑡𝑡−1 ( 𝑽𝑽𝑡𝑡−1 ≤ 𝑉𝑉max)
𝑴𝑴𝑡𝑡−1 + 𝑉𝑉max𝑽𝑽𝑡𝑡−1𝑽𝑽𝑡𝑡−1
(otherwise)
𝑺𝑺𝒕𝒕 : Position vector of industrial robot at time 𝒕𝒕( 𝒕𝒕 > 𝟏𝟏)𝑴𝑴𝒕𝒕: Position vector of haptic interface device at time 𝒕𝒕𝑽𝑽𝒕𝒕 : Velocity vector of industrial robot𝑽𝑽𝐦𝐦𝐚𝐚𝐦𝐦 : Maximum velocity of industrial robot (5mm / ms)
Calculation method of force
Force outputted at master terminal
𝑭𝑭𝑡𝑡(m) = 𝐾𝐾scale𝑭𝑭𝑡𝑡−1
(s)
𝑭𝑭𝒕𝒕(𝐦𝐦) : Reaction force outputted at the master terminal at time
𝑡𝑡(>0)
𝑭𝑭𝒕𝒕(𝐚𝐚) : Force received at the master terminal from the slave
terminal at time 𝑡𝑡 (> 0)
𝑲𝑲𝐚𝐚𝐬𝐬𝐚𝐚𝐬𝐬𝐬𝐬 : Force scale which changes 𝑭𝑭𝒕𝒕−𝟏𝟏(𝐚𝐚)
𝑲𝑲𝐚𝐚𝐬𝐬𝐚𝐚𝐬𝐬𝐬𝐬 = 𝟏𝟏.𝟎𝟎
Demo video (Carrying object together)
Slave robot Master robot
Master - slave relation
Output positioninformation
Follow master robot.
Output positioninformation
Robot arm’s motion is delayed.
Send position information
Large network delay
Large force may be applied to an object.
Send position information
Network delay : Δ ms
Wait for Δ ms without outputOutput position information
Output positioninformation
Robot arm’s motion is not
delayed.
Adaptive Δ-causality control (1/2)
No large force may be applied to an object.
Δ is set to the smoothed network delay 𝑫𝑫𝒕𝒕.
�𝑫𝑫𝟎𝟎 = 𝒅𝒅𝟎𝟎𝑫𝑫𝒕𝒕 = 𝜶𝜶𝑫𝑫𝒕𝒕−𝟏𝟏 + 𝟏𝟏 − 𝜶𝜶 𝒅𝒅𝒕𝒕 (𝒕𝒕 ≥ 𝟏𝟏)
Smoothing coefficient 𝜶𝜶 = 𝟎𝟎.𝟗𝟗𝟗𝟗𝟗𝟗*2
𝒅𝒅𝒕𝒕 : Network delay at time t
Information received after generation time + Δ is discarded as old and useless information.
Adaptive Δ-causality control (2/2)*2 T. Abe et al., Journal of IEICE, vol. 91, no. 2, Feb. 2008.
Stabilization control with filters
The stabilization control with filters uses the wave filter in combination with the phase control filter.
The control can make the remote robot system with force feedback stable against any network delay*3.
*3 P. Huang et al., IEICE, CQ2017-79, Nov. 2017.
Experiment method
We generated a constant delay (called the additional delay) for each packet transmitted between the two systems by a network emulator (NIST Net).
We measured the force sensed by the force sensor and the robot arm position .
Experiment results (1/2)
No stabilization control Stabilization control
Additional delay : 200 ms
Position and Force in the x-axis of each robot under QoS control
Posi
tion
[mm
]
Posi
tion
[mm
]
Forc
e [N
]
Forc
e [N
]
Elapsed time [sec.] Elapsed time [sec.]
Elapsed time [sec.] Elapsed time [sec.]
Master robotSlave robot
Master robot
Master robot
Slave robot
Slave robot
Slave robot
Master robot
Experiment results (2/2)Position and Force in the x-axis of each robot under stabilization control
No QoS control QoS control
Additional delay : 200 ms
Posi
tion
[mm
]
Posi
tion
[mm
]
Forc
e [N
]
Forc
e [N
]
Elapsed time [sec.] Elapsed time [sec.]
Elapsed time [sec.] Elapsed time [sec.]
Master robot
Master robot
Master robot
Master robotSlave robot
Slave robot Slave robot
Slave robot
Conclusion
The instability phenomena can greatly be suppressed by the stabilization control in the systems.
The adaptive ∆-causality control is effective under the stabilization control.
We applied the stabilization control with filters and the adaptive ∆-causality control for cooperative work between the tele-robot systems with force feedback by using a master-slave relation.
We investigated the effects of the stabilization control and the adaptive ∆-causality control.
Future work
Apply the stabilization control with filters and the adaptive ∆-causality control to the systems with an equal relationship and investigate their effects.
Switch the master-slave relationship dynamically according to the network delay in the systems.
Effect of Stabilization Control for Cooperation between Tele-Robot Systems with Force Feedback �by Using Master-Slave RelationOutlineBackgroundProblemsPurposeRemote robot systems with force feedbackCarrying Object TogetherCalculation method of position Calculation method of forceDemo video (Carrying object together)Master - slave relationスライド番号 12スライド番号 13Stabilization control with filtersExperiment method Experiment results (1/2)Experiment results (2/2)ConclusionFuture workスライド番号 20Experiment results