|
Overview Topics Approach
_____________________________________________________________________________________
Overview
In recent years, there has been a growing
interest in understanding on the one hand, how various animal aggregations
such as fish schools, bird flocks, deer herds, etc. coordinate their
collective motions to perform useful tasks and on the other, how groups of
mobile autonomous agents such as AUV schools, UAV flocks, etc., might be
instructed to cooperate in a similar manner.
Cooperative Control is concerned with engineered systems
that can be characterized as: .collection of interconnected decision-making
components (systems) with limited processing capabilities, locally sensed
information and limited inter-component communications, all seeking to
achieve a collective (global) objective.
Multi-agent coordinated motion
planning problems arise in various contexts, such as Air Traffic Management,
robotics, and Unmanned Aerial Vehicle. In most cases, certain separation
constraints between the agents have to be guaranteed due to physical,
safety, or efficiency reasons.
Formation control in a five agents joint manoeuvre
Cooperative Control
is
‘multi-agent control’,
‘distributed control’,
‘networked control’,
‘swarms’, or
‘coordinated control’
_____________________________________________________________________________________
Topics
·
Multi-agent
hybrid systems (MASH)
·
Stochastic reachability analysis for MASH
·
Communication
and control co-design
·
Optimal coordinated motion planning
·
Hierarchical
control of multi-agent systems
_____________________________________________________________________________________
Approach
The key challenge in cooperative
control for multi-agent systems could be how to design the local
interaction rules and coordination principles among agents so as to achieve
certain desired global behaviors.
In our approach, we consider that an
agent is described by a hybrid dynamical system whose position is given by
its location in three dimensional space. Then, we
consider a collection of N agents that are performing a shared task, where
the task depends on the relationship between the locations of the
individual vehicles.
The vehicles are able to
communicate with each in carrying out the task, with the individual
vehicles able to communicate with some subset of the other vehicles.
The communication network of
(stochastic) hybrid systems supports the information structure for the
cooperative controller that defines “who knows what and when they know it.”
This network may be defined via a process algebra communication rules that
describe how the communication takes place.
|
