A comparison of winter pre-heating requirements for natural displacement and natural mixing ventilation

In winter, natural ventilation can be achieved either through mixing ventilation or upward displacement ventilation (P.F. Linden, The fluid mechanics of natural ventilation, Annual Review of Fluid Mechanics 31 (1999) pp. 201-238). We show there is a significant energy saving possible by using mixing ventilation, in the case that the internal heat gains are significant, and illustrate these savings using an idealized model, which predicts that with internal heat gains of order 0.1 kW per person, mixing ventilation uses of a fraction of order 0.2-0.4 of the heat load of displacement ventilation assuming a well-insulated building. We then describe a strategy for such mixing natural ventilation in an atrium style building in which the rooms surrounding the atrium are able to vent directly to the exterior and also through the atrium to the exterior. The results are motivated by the desire to reduce the energy burden in large public buildings such as hospitals, schools or office buildings centred on atria. We illustrate a strategy for the natural mixing ventilation in order that the rooms surrounding the atrium receive both pre-heated but also sufficiently fresh air, while the central atrium zone remains warm. We test the principles with some laboratory experiments in which a model air chamber is ventilated using both mixing and displacement ventilation, and compare the energy loads in each case. We conclude with a discussion of the potential applications of the approach within the context of open plan atria type office buildings.

Automatic construction and natural-language description of nonparametric regression models

Copyright © 2014, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved. This paper presents the beginnings of an automatic statistician, focusing on regression problems. Our system explores an open-ended space of statistical models to discover a good explanation of a data set, and then produces a detailed report with figures and natural- language text. Our approach treats unknown regression functions non- parametrically using Gaussian processes, which has two important consequences. First, Gaussian processes can model functions in terms of high-level properties (e.g. smoothness, trends, periodicity, changepoints). Taken together with the compositional structure of our language of models this allows us to automatically describe functions in simple terms. Second, the use of flexible nonparametric models and a rich language for composing them in an open-ended manner also results in state- of-the-art extrapolation performance evaluated over 13 real time series data sets from various domains.