Abstract
We address the problem of finding an infinite run with the
optimal cost-time ratio in a one-clock priced timed automaton and pro-
vide an algorithmic solution. Through refinements of the quotient graph
obtained by strong time-abstracting bisimulation partitioning, we con-
struct a graph with time and costs as weights and prove that the min-
imum ratio properties remain unaltered, allowing algorithms for finite
weighted graphs to be applied. The resulting algorithm has an overall
time complexity of O((|Q| · (|T | + |Q|))6 ), |Q| being the number of lo-
cations and |T | the number of transitions of the one-clock priced timed
automaton.
optimal cost-time ratio in a one-clock priced timed automaton and pro-
vide an algorithmic solution. Through refinements of the quotient graph
obtained by strong time-abstracting bisimulation partitioning, we con-
struct a graph with time and costs as weights and prove that the min-
imum ratio properties remain unaltered, allowing algorithms for finite
weighted graphs to be applied. The resulting algorithm has an overall
time complexity of O((|Q| · (|T | + |Q|))6 ), |Q| being the number of lo-
cations and |T | the number of transitions of the one-clock priced timed
automaton.
| Original language | English |
|---|---|
| Publication date | 2011 |
| Number of pages | 9 |
| Publication status | Published - 2011 |
| Event | Annual Doctoral Workshop on Mathematical and Engineering Methods in Computer Science - Lednice, Czech Republic Duration: 14 Oct 2011 → 16 Oct 2011 |
Conference
| Conference | Annual Doctoral Workshop on Mathematical and Engineering Methods in Computer Science |
|---|---|
| Country/Territory | Czech Republic |
| City | Lednice |
| Period | 14/10/2011 → 16/10/2011 |