Fuzzy behavior integration and action fusion for robotic excavation

This paper discusses control behavior integration and bucket action fusion for excavation control of a robotic front-end-loader type machine. To utilize the experience and expertise from skilled human operators, a fuzzy-logic based control approach is developed. A hierarchical excavation control arc...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on industrial electronics (1982) 1996-06, Vol.43 (3), p.395-402
Main Authors: Shi, X., Lever, P.J.A., Wang, F.-Y.
Format: Article
Language:English
Subjects:
Applied sciences
Automatic control
Computer science
control theory
systems
Control system synthesis
Control theory. Systems
Exact sciences and technology
Fuzzy control
Geology
Humans
Laboratories
Robot control
Robot kinematics
Robot sensing systems
Robotics
Robotics and automation
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper discusses control behavior integration and bucket action fusion for excavation control of a robotic front-end-loader type machine. To utilize the experience and expertise from skilled human operators, a fuzzy-logic based control approach is developed. A hierarchical excavation control architecture decomposes excavation goals to tasks, then tasks to behaviors, and finally behaviors to actions. The excavation actions are primitive and can be executed directly by an excavation machine. Finite state machines are used to specify the coordination and integration of behaviors for task execution and actions for behavior implementation. A simple strategy for action fusion is proposed based on fuzzy logic reasoning and the COA defuzzification method. Finally, laboratory experiments are conducted using a PUMA 560 robot arm and a Zebra force/torque sensor in a simulated rock excavation environment. Experimental results indicate that the proposed approach in this paper has led to more efficient task execution than previous approaches.
ISSN:0278-0046
1557-9948
DOI:10.1109/41.499812