Machine learning and energy efficient building design

Machine learning and energy efficient building design

With increasing interest in sustainable design, the issue of energy-efficiency in the building design process is receiving ever more attention from designers and researchers. The construction industry has to find its way of reducing national greenhouse gas emissions. Key players are reluctant to spend big on energy efficiency and here’s where machine learning can help.

In order to achieve energy efficiency, most buildings’ architects will consider the shape and orientation of the building and its solar protections, high performing building envelope: thorough insulation, high performing glazing windows, and air-sealed construction and high performance controlled ventilation through mechanical insulation, and heat recovery. In 1996 Lysen introduced the design concept Trias Energetica, which suggests an organized approach to reduce the dependence on fossil fuels. This concept mainly focuses on ways that deal with energy to achieve savings, reduction of dependence and environmental benefits, while maintaining the building’s comfort and construction progress.

A high percentage of newly-constructed commercial office buildings experience energy consumption that exceeds specifications and system failures after being put into use. This problem is even worse for older buildings.Several software packages have been developed to accurately simulate building energy consumption (e.g. Energy-Plus, DOE-2, and Green Building Studio) and to support decision makers at the design phase to produce more energy efficient structures. However, in reality, simulation tools are mainly used to validate the performances of the final de- sign of a building rather than exploring multiple design possibilities.

Energy modeling programs provide users with key building performance indicators such as energy use and demand, temperature, humidity, and costs. Even a simple energy modeling run generates hundreds of pages of data. Examples of building features simulated include the estimated energy costs in terms of building orientation, HVAC system, lighting efficiency and control, roof and wall insulation and construction, glazing type, water usage, day-lighting and so on. Such volumes of data are simply beyond human abilities to identify the best combination of building components (insulation, windows, doors, etc.) and systems (heating and cooling systems, ventilation, etc.) during the building design process. Evaluating building energy modeling outputs clearly overwhelms the traditional methods of data analysis such as spreadsheets and ad-hoc queries.

Leon Wu, Gail Kaiser, David Solomon and other researchers used a new approach, ‘predictive building energy optimization’, which uses machine learning and automated online evaluation of historical and real-time building data to improve efficiency and reliability of building operations without requiring large amounts of additional capital investment. Their machine learning approach uses a predictive model to generate accurate energy demand forecasts and automated analyses that can guide optimization of building operations. In parallel, an automated online evaluation system monitors efficiency at multiple stages in the system workflow and provides building operators with continuous feedback.

A prototype was implemented in a large commercial building in Manhattan. Predictive machine learning model applies support vector regression to the building’s historical energy use and temperature and wet-bulb humidity data from the building’s interior and exterior in order to model performance for each day. This predictive model closely approximates actual energy usage values, with some seasonal and occupant-specific variability, and the dependence of the data on day-of-the-week makes the model easily applicable to different types of buildings with minimal adjustment.

The development of energy-efficient buildings is a sustainable vision that entails huge challenges for environmental and technical innovation. It has consequences for all professions, not the least for architectural design and building engineering, since it is here that the full complexity of building performance analysis has to be addressed and managed throughout the design process that can be achieved with help of machine learning techniques.

Author: AI Business