Automated translation of non-HA hybrid models to implementation while preserving verified behaviors

Develop automated model-translation techniques that extend beyond hybrid automata to other hybrid dynamical system formalisms—such as mixed logical dynamical systems, switched systems, piecewise affine models, and hybrid input/output automata—so they can be automatically converted into implementation artifacts (e.g., Simulink/Stateflow diagrams) while preserving verified behaviors such as trajectory equivalence.

Background

The paper discusses verification methods and highlights practical challenges in translating verified hybrid models into deployable tools. Prior work (Bak 2019) enables automatic conversion of hybrid automata into trajectory-equivalent Simulink/Stateflow diagrams, thus preserving verified behaviors during implementation.

However, that solution is limited to hybrid automata, leaving an open problem of extending automated translation to other hybrid modeling frameworks used in power systems, where ensuring behavioral fidelity during implementation is critical for safety and reliability.