The fast and user-friendly generation of controller models is a crucial prerequisite for the widespread implementation and scalability of Model Predictive Control (MPC) for buildings. Apart from this, these models can be employed for evaluating other more and more relevant characteristics of buildings, for example, energy-saving potentials in user behavior or the building’s flexibility potential to reduce or shift loads. In this work, the ALICE2Modelica toolchain is presented, which enables the automated and scalable generation of building models for use in Modelica-based building control and simulation. The toolchain is based on the developed mini-language ALICE, which constitutes a straightforward approach to describe geometrical information of the envelope of rooms. Based on ALICE files, floor plans in SVG format, and parametrizable Modelica room templates, the toolchain automatically generates Modelica room and building models including Heating, Ventilation and Air Conditioning (HVAC) systems. Therefore, it is suitable for the modeling of large-scale multi-zone buildings. In a case study, based on an existing parameter estimation module and data from building operation, selected parameters of the toolchain-generated Modelica models are calibrated and user behavior is evaluated for an office building with respect to energy efficiency and energy-saving potentials.

2024 Open Source Modelling and Simulation of Energy Systems, OSMSES2024, Vienna, Austria, 3 Sep 2024 - 4 Sep 2024; 1-7 (2024).