Theoretical Computer Science
ICALP'03 Proceedings of the 30th international conference on Automata, languages and programming
Timed control with observation based and stuttering invariant strategies
ATVA'07 Proceedings of the 5th international conference on Automated technology for verification and analysis
Efficient on-the-fly algorithms for partially observable timed games
FORMATS'07 Proceedings of the 5th international conference on Formal modeling and analysis of timed systems
A lattice theory for solving games of imperfect information
HSCC'06 Proceedings of the 9th international conference on Hybrid Systems: computation and control
Algorithms for omega-regular games with imperfect information
CSL'06 Proceedings of the 20th international conference on Computer Science Logic
Template-Based controller synthesis for timed systems
TACAS'12 Proceedings of the 18th international conference on Tools and Algorithms for the Construction and Analysis of Systems
Solving games via three-valued abstraction refinement
CONCUR'07 Proceedings of the 18th international conference on Concurrency Theory
Hi-index | 0.00 |
We present a novel approach to the safety controller synthesis problem with partial observability for real-time systems. This in general undecidable problem can be reduced to a decidable one by fixing the granularity of the controller: finite sets of clocks and constants in the guards. Current state-of-the-art methods are limited to brute-force enumeration of possible granularities or manual choice of a finite set of observations that a controller can track. We address this limitation by proposing a counterexample-guided method to successively refine a set of observations until a sufficiently precise abstraction is obtained. The size of the abstract games and strategies generated by our approach depends on the number of observation predicates and not on the size of the constants in the plant. Our experiments demonstrate that this results in better performance than the approach based on fixed granularity when fine granularity is necessary.