ReTrofiT: A Software to Solve Optimization and Identification Problems Applied to Building Energy Management

Abstract : Energy management systems in buildings greatly contribute to the improvement of overall energy efficiency. Monitoring systems can lead to significant reductions of the global energy use by increasing occupants' awareness of the consumptions or by enabling the implementation of more efficient regulation strategies. Model predictive control consists in computing optimal heating or cooling strategies by taking into account the future evolution of the state of the building under forecast weather or use conditions. Demand response strategies in smart grids consist in adjusting energy demand at the end-user level to reduce the overall demand thus resulting in end-user customer bill savings, increase of electricity market stability and of electricity supply reliability. Further, today the building construction practices evolve towards a more performance-based approach in which the concern becomes the performance of the final building rather than the means employed to construct it. All the aforementionned applications rely on the ability to accurately predict a building's behavior using a calibrated model. In building energy applications, uncertainties in input data of modelling tools often lead to important discrepancies between the model predictions and the real performance. The desired model response can be obtained if the internal parameters of the model are calibrated using on-site measurements and model identification methods. The paper presents the software ReTrofiT that was specifically designed to treat this kind of problems. ReTrofiT is first of all a dynamic building simulation code with multizone-type assumptions. It integrates a set of tools and algorithms to set up and solve minimization problems as well as to compute sensitivities. All these operations can be done with a negligible computation by means of the adjoint model that is intrinsically implemented.