An hour-ahead predictive control strategy for maximizing natural ventilation in passive buildings based on weather forecasting

Due to the increasing demands of energy conservation and emission reduction, passive buildings with extra-low energy consumption and near zero energy consumption have attracted more attentions as a promising sustain-able building type. Due to passive techniques such as air-tightness, passive buildings have a special demand of mechanical fresh air systems after closing windows. So it is indispensable to explore elaborate control strategies for maximizing natural ventilation, for the purpose of postponing the use of fossil energy. By taking meteoro-logical forecast parameters as the input, an hour-head model-based predictive control (MPC) strategy of natural ventilation was proposed in this paper. Specifically, simulated multi-dimensional datasets obtained by setting different control actions under the same disturbances were obtained as the database for data-driven prediction models, which take weather forecast parameters as the input and labels of thermal comfort levels and air volumes as the output. It has been verified that although both with extended working hours and more stable comfort levels, the MPC strategy with priority to thermal comfort is superior to that with priority to fresh air volume. Furthermore, when compared with the traditional control strategy, ventilation hours via this MPC strategy could be increased by 56.3% from 366 h to 572 h. With this MPC model, the self-adaptive adjustment of window openness could be achieved by taking hour-ahead weather forecasting parameters as the input, thus to facilitate the climatic self-adaptive intelligent operation of passive buildings in future.

Automated summary

This paper proposes an hour-ahead model-based predictive control (MPC) strategy for natural ventilation in passive buildings. By using weather forecast parameters as input, the strategy maximizes the control of natural ventilation to delay the use of fossil energy. Simulation datasets were used to build data-driven prediction models, with weather forecast parameters as inputs and thermal comfort levels and air volumes as outputs. The results show that the MPC strategy with priority to thermal comfort is superior to that with priority to fresh air volume. Compared to the traditional control strategy, the MPC strategy increases ventilation hours by 56.3%, from 366 h to 572 h. With the MPC model, self-adaptive adjustment of window openness can be achieved based on hour-ahead weather forecasting parameters, facilitating the climatic self-adaptive intelligent operation of passive buildings in the future.