05-06-2012, 03:31 PM
Biologically Inspired Behaviour Design for Autonomous
Robotic Fish
[attachment=159]
Abstract:
Behaviour-based approach plays a key role for mobile robots to operate safely in unknown or dynamically
changing environments. We have developed a hybrid control architecture for our autonomous robotic fish that consists of
three layers: cognitive, behaviour and swim pattern.
Introduction
The biologically inspired or behaviour based ap-
proach has been widely adopted in the design of robotic
systems, including design of mechanical structure,
robot locomotion, navigation algorithm and multi-
robot cooperation. A number of biologically inspired
robot systems have been built over the last twenty
years, such as the robot phonotaxis by Horchler[1] and
biomimetic “robo-lobsters” by Grasso[2]. It has advan-
tages of easy design, quick response, model-free, and
robustness. It enables a mobile robot to operate in dy-
namic or unpredictable environments and can quickly
response to the changes in the real world.
Control architecture and behaviour
layer
Our robotic fish is controlled under a hybrid con-
trol architecture[19] which is presented in Fig. 1. It con-
sists of three layers: Cognitive Layer, Behaviour Layer
and Swim Pattern Layer. The centre of the architec-
ture is the behaviour layer where most instant decisions
are made. The swim pattern layer is at the bottom
of our three-layer architecture, which consists of all
swim patterns for basic control purpose. Five swim
patterns are designed here, namely Cruise-Straight,
Ascend/Descend, Coast, Sharp-Left/Right-Turn and
Cruise-in-Left/Right-Turn. Each swim pattern is a se-
ries of joint motion functions which control the tail
joints to realize an associative movement in order to
propel or maneuver the robotic fish. In the behaviour
layer, states and actions are directly related to the
robot physical sensors and swim patterns.
Design of robotic fish behaviours
The behaviours for the robotic fish are designed
based on the real fish behaviours rather than man-
made machines such as the submarine. Each behaviour
we designed is inspired by real fish. The following be-
haviours are implemented.
Conclusion
In this paper, we mainly discuss the design of the
behaviour layer in the hybrid control architecture of
our robotic fish, including the design of individual be-
haviours. All individual robotic fish behaviours are in-
spired by the real fish behaviours and implemented by
FLCs because FLC has advantage on non-linear con-
trol and real-time implementation.
Robotic Fish
[attachment=159]
Abstract:
Behaviour-based approach plays a key role for mobile robots to operate safely in unknown or dynamically
changing environments. We have developed a hybrid control architecture for our autonomous robotic fish that consists of
three layers: cognitive, behaviour and swim pattern.
Introduction
The biologically inspired or behaviour based ap-
proach has been widely adopted in the design of robotic
systems, including design of mechanical structure,
robot locomotion, navigation algorithm and multi-
robot cooperation. A number of biologically inspired
robot systems have been built over the last twenty
years, such as the robot phonotaxis by Horchler[1] and
biomimetic “robo-lobsters” by Grasso[2]. It has advan-
tages of easy design, quick response, model-free, and
robustness. It enables a mobile robot to operate in dy-
namic or unpredictable environments and can quickly
response to the changes in the real world.
Control architecture and behaviour
layer
Our robotic fish is controlled under a hybrid con-
trol architecture[19] which is presented in Fig. 1. It con-
sists of three layers: Cognitive Layer, Behaviour Layer
and Swim Pattern Layer. The centre of the architec-
ture is the behaviour layer where most instant decisions
are made. The swim pattern layer is at the bottom
of our three-layer architecture, which consists of all
swim patterns for basic control purpose. Five swim
patterns are designed here, namely Cruise-Straight,
Ascend/Descend, Coast, Sharp-Left/Right-Turn and
Cruise-in-Left/Right-Turn. Each swim pattern is a se-
ries of joint motion functions which control the tail
joints to realize an associative movement in order to
propel or maneuver the robotic fish. In the behaviour
layer, states and actions are directly related to the
robot physical sensors and swim patterns.
Design of robotic fish behaviours
The behaviours for the robotic fish are designed
based on the real fish behaviours rather than man-
made machines such as the submarine. Each behaviour
we designed is inspired by real fish. The following be-
haviours are implemented.
Conclusion
In this paper, we mainly discuss the design of the
behaviour layer in the hybrid control architecture of
our robotic fish, including the design of individual be-
haviours. All individual robotic fish behaviours are in-
spired by the real fish behaviours and implemented by
FLCs because FLC has advantage on non-linear con-
trol and real-time implementation.