Name
Abstract | ||
|---|---|---|
ABSTRACTWe consider the problem of synthesizing the most robust controllers using the Abstraction-Based Controller Design (ABCD). First, we perform a finite-state abstraction of the continuous dynamic system. We then synthesize a most robust control strategy in the finite space by formulating it as a two-player game. Finally, we refine the strategy to a controller for the original problem. To preserve robustness, we consider the specifications for the controllers to be expressed in Robust Linear Temporal Logic (rLTL), which allows the reasoning about how robust the specification is. However, the current algorithms for rLTL synthesis do not compute optimally robust controllers. It only considers the worst-case analysis for reactive synthesis. Hence, we develop two new notions of adaptive strategies. One is Weakly Adaptive strategy, which, in response to the opponent's bad choices, adaptively changes the degree of satisfaction we want to achieve to ensure the optimality w.r.t. the current stage. The second one is Strongly adaptive strategy, which is weakly adaptive that also maximizes the chances of the opponent making a bad choice. We show that the computability problem for both the strategies is not harder than the classical one and can be solved in doubly-exponential time. |
Year | DOI | Venue |
|---|---|---|
2021 | 10.1145/3447928.3457210 | Cyber-physical Systems |
DocType | Citations | PageRank |
Conference | 0 | 0.34 |
References | Authors | |
0 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Satya Prakash Nayak | 1 | 0 | 0.34 |
| Daniel Neider | 2 | 153 | 21.97 |
| Martin Zimmermann 0002 | 3 | 35 | 10.88 |