Using Machine-Learning Algorithms to improve CO2-based Demand-controlled Natural Ventilation
Keywords:Single-sided natural ventilation, CO2 demand control, regression algorithms, system identification
Although natural ventilation is applicable to most buildings, architects today struggle to integrate it as an alternative to mechanical ventilation systems. This paper presents the application of regression algorithms and identification methods to single-sided natural ventilation with 1 opening (SS1). The result of this work provides a predictive control-oriented model of CO2 demand-controlled natural ventilation with SS1 configuration.