Stochastic Simulation and Inference using Modelica

Abstract

The physical modelling and simulation of systems with inherent uncertainty still poses significant issues when using Modelica and its tools. At present, both language and tools are fundamentally deterministic and offer limited support for handling uncertainty; this limits the scope of using Modelica in certain domains, e.g. feedback control systems. We propose a frame- work for incorporating uncertainty in Modelica sim- ulation and analysis tasks. We do this by coupling a Modelica model with exogenous stochastic models. Finally, we apply this approach to the domain of build- ing modelling