LAUSR

Abstract We propose a method that simultaneously identifies a sparse transfer matrix and disturbance for a multi-zone building’s dynamics from input-output measurements. An l1-regularized least-squares optimization problem is solved to obtain a sparse solution, so that only dominant interactions among zones are retained in the model. The disturbance is assumed to be piecewise-constant: the assumption aids identification and is motivated by the nature of occupancy that determines the disturbance. Application of our method on data from a simulation model shows promising results.