Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Introduction
Why do we need it?
Former research
This paper in comparison
Multi-vehicle coordination.
However
Former research
The cooperative control task under consideration is average
consensus
We learned about in class (consensus protocol for singleintegrator agents) and can further be read on in Olfati-Saber
and Murray (2004).
We saw in our homework (consensus protocol for doubleintegrator agents) and can further be read on in Ren and
Atkins (2007).
In practice those control laws have to be implemented on
digital platforms
The traditional method is time-scheduled periodic sampling
measurements are taken periodically according to a constant
sampling period and the controllers are updated
synchronously. Implementation is investigated in:
Xie, Liu, Wang and Jia (2009); Ren and Cao (2008)
Both Mazo and Tabuada (2011) and Wang and Lemmon (2011)
suggest that event-based scheduling reduces the number of
transmissions required in networked control systems.
Preliminaries
Problem statement
Single-integrator agents:
x i t ui t , i V , where xi(t) R is the state and
ui(t) R its control input.
Double-integrator agents:
0
xi t
0
1
0
xi t ui t , i V , where
0
1
Single-integrator agents
Construction
Theorem
Remarks
Construction
Construction cont.
Construction cont.
Theorem
f i t , ei t ei t c0 c1e t
Remarks
Remarks cont.
Delayed communication
Construction
Theorem
Remarks
Construction
Theorem
f i t , ei t ei t c0 c1e t
Remarks
Simulations
Simulation network
Single-Integrator agents
Double-Integrator agents
Simulation network
Single-integrator agents
Double-integrator agents
Double-integrator agents
Construction
Theorem
Construction
Theorem