We present a new release of PRISM-games, a tool for verification and strategy synthesis for stochastic games. PRISM-games 2.0 significantly extends its functionality by supporting, for the first time: (ⅰ) long-run average (mean-payoff) and ratio reward objectives, e.g., to express energy consumption per time unit; (ⅱ) strategy synthesis and Pareto set computation for multi-objective properties; and (ⅲ) compositional strategy synthesis, where strategies for a stochastic game modelled as a composition of subsystems are synthesised from strategies for individual components using assume-guarantee contracts on component interfaces. We demonstrate the usefulness of the new tool on four case studies from autonomous transport and energy management.
展开▼