Search

Thermal comfort is one of the most important factors for occupant satisfaction and, as a result, for the building energy performance. Decentralized heating and cooling systems, also known as “Personal Environmental Comfort Systems” (PECS), have attracted significant interest in research and industry in recent years. While building simulation software is used in practice to improve the energy performance of buildings, most building simulation applications use the PMV approach for comfort calculations. This article presents a newly developed building controller that uses a holistic approach in the consideration of PECS within the framework of the building simulation software Esp-r. With PhySCo, a dynamic physiology, sensation, and comfort model, the presented building controller can adjust the setpoint temperatures of the central HVAC system as well as control the use of PECS based on the thermal sensation and comfort values of a virtual human. An adaptive building controller with a wide dead-band and adaptive setpoints between 18 to 26 °C (30 °C) was compared to a basic controller with a fixed and narrow setpoint range between 21 to 24 °C. The simulations were conducted for temperate western European climate (Mannheim, Germany), classified as Cfb climate according to Köppen-Geiger. With the adaptive controller, a 12.5% reduction in end-use energy was achieved in winter. For summer conditions, a variation between the adaptive controller, an office chair with a cooling function, and a fan increased the upper setpoint temperature to 30 °C while still maintaining comfortable conditions and reducing the end-use energy by 15.3%. In spring, the same variation led to a 9.3% reduction in the final energy. The combinations of other systems were studied with the newly presented controller.

Solar radiation data is essential for the development of many solar energy applications ranging from thermal collectors to building simulation tools, but its availability is limited, especially the diffuse radiation component. There are several studies aimed at predicting this value, but very few studies cover the generalizability of such models on varying climates. Our study investigates how well these models generalize and also show how to enhance their generalizability on different climates. Since machine learning approaches are known to generalize well, we apply them to truly understand how well they perform on different climates than they are originally trained. Therefore, we trained them on datasets from the U.S. and tested on several European climates. The machine learning model that is developed for U.S. climates not only showed low mean absolute error (MAE) of 23 W/m2, but also generalized very well on European climates with MAE in the range of 20 to 27 W/m2. Further investigation into the factors influencing the generalizability revealed that careful selection of the training data can improve the results significantly.