Synthesis of Robust Optimal Real-Time Systems

Abstract

Weighted Timed Games (WTGs for short) are widely used to describe real-time controller synthesis problems, but they rely on an unrealistic perfect measure of time elapse. In order to produce strategies tolerant to timing imprecisions, we consider a notion of robustness, expressed as a parametric semantics, first introduced for timed automata. WTGs are two-player zero-sum games played in a weighted timed automaton in which one of the players, that we call Min, wants to reach a target location while minimising the cumulated weight. The opponent player, in addition to controlling some of the locations, can perturb delays chosen by Min. The robust value problem asks, given some threshold, whether there exists a positive perturbation and a strategy for Min ensuring to reach the target, with an accumulated weight below the threshold, whatever the opponent does.
We provide in this article the first decidability result for this robust value problem. More precisely, we show that we can compute the robust value function, in a parametric way, for the class of divergent WTGs (this class has been introduced previously to obtain decidability of the (classical) value problem in WTGs without bounding the number of clocks). To this end, we show that the robust value is the fixpoint of some operators, as is classically done for value iteration algorithms. We then combine in a very careful way two representations: piecewise affine functions introduced in [Alur et al., 2004] to analyse WTGs, and shrunk Difference Bound Matrices (shrunk DBMs for short) considered in [Sankur et al., 2011] to analyse robustness in timed automata. The crux of our result consists in showing that using this representation, the operator of value iteration can be computed for infinitesimally small perturbations. Last, we also study qualitative decision problems and close an open problem on robust reachability, showing it is EXPTIME-complete for general WTGs.