Integrating Formal Verification and Simulation of Hybrid Systems

An increasing number of today's systems can be characterised as cyber-physical, or hybrid systems that combine the concurrent continuous environment and discrete computational logic. In order to develop such systems as safe and reliable one needs to be able to model and verify them from the early stages of the development process. Current modelling technologies allow us to specify the abstractions of these systems in terms of the procedural or declarative modelling languages and visual notations, and to simulate their behaviour over a period of time for analysis. Other means of modelling are formal methods, which define systems in terms of logics and enable rigorous analysis of system properties. While the first class of technologies provides a natural notation for describing physical processes, but lacks the formal proof, the second class relies on mathematical abstractions to rationalise and automate the complex task of formal verification. The benefits of both technologies can be significantly enhanced by a collaborative methodology. Due to the complexity of the considered systems and the formal proof process it is critical that such a methodology is based on a top-down development process that fully supports abstraction and refinement. We develop this idea into a tool extension for the state of the art Rodin platform for system-level formal modelling and analysis in the Event-B language. The developed tool enables integration of the physical simulation with refinement-based formal verification in Event-B, thus enhancing the capabilities of Rodin with the simulation-based validation that supports refinement. The tool utilises the Functional Mock-up Interface (FMI) standard for industrial-grade model exchange and co-simulation and is based on a co-simulation principle between the discrete models in Event-B and continuous physical models of FMI. It provides a graphical environment for model import, composition and co-simulation, and implements a generic simulation algorithm for discrete-continuous co-simulation