Artifact for TACAS 2019 paper: Minimal-Time Synthesis for Parametric Timed Automata

This artifact contains the means to reproduce the experimental results from the
paper. In the paper we propose algorithms that synthesize a/all parameter
valuation(s) in a PTA, such that the time to reach the target location is
minimized. Additionally, we also propose algorithms that synthesize a/all
parameter valuation(s) that minimize a specific parameter.

We consider the following algorithms in our experiments:
- Synth: the classical synthesis algorithm (without time minimization)
- MTSynth: our minimal-time synthesis algorithm
- MPSynth: our minimal-parameter synthesis algorithm (which is used to compute
the minimal time)
- MTSynth-noRed: Our MTSynth algorithm without using state-space reductions
- MTReach: our minimal-time reachability algorithm (that returns at least one
parameter valuation that reaches the target location in minimal time)
- MPReach: our minimal-parameter reachability algorithm (see MTReach)

For the experiments, we collected 68 PTA models and properties from the
IMITATOR benchmarks library, and ran each experiment (algorithm+model
combination).

The most interesting result is that the minimal-time synthesis version is in
general faster in computing results when compared to the classical synthesis
algorithm. We can also observe that MTSynth is faster than MPSynth (in MTSynth
we use a shortest-path search, and in MPSynth we do not). When comparing
reachability with synthesis, we see that the results are fairly similar, with
only a few instances in which only reachability is computable in the allowed
time.