Model reduction for indoor-air behavior in control design for energy-efficient buildings

We present two methods for generating reduced-order models for the indoor-air environment in control design for energy efficient buildings. In one method, projection-based H₂-optimal interpolatory model reduction techniques are employed on a high-dimensional state-space model resulting from finite element discretization of convection-diffusion equations. In the second approach, we obtain reduced-models directly from input and output measurements using a data-driven model reduction approach. In both cases, we produce reduced-order models that accurately represent the input-output behavior of the full-order system.