Advanced data analytics for building energy modeling and management
- PDF / 401,439 Bytes
- 2 Pages / 612 x 808 pts Page_size
- 70 Downloads / 189 Views
Advanced data analytics for building energy modeling and management Editorial
Cheng Fan1, Fu Xiao2 (), Da Yan3 (Guest Editors) 1. Sino-Australia Joint Research Center in BIM and Smart Construction, College of Civil and Transportation Engineering, Shenzhen University, Shenzhen, China 2. Department of Building Services Engineering, The Hong Kong Polytechnic University, Hong Kong, China 3. Building Energy Research Center, School of Architecture, Tsinghua University, Beijing, China
© Tsinghua University Press and Springer-Verlag GmbH Germany, part of Springer Nature 2020
The data-driven approach has been widely adopted in fault detection and diagnosis and optimal control of building energy systems for improving building performance since the 1980s, owing to the increasing demands for convenient and computationally efficient yet accurate models for online and automated applications. Conventional data-driven methods in the building field usually utilize a small amount of data (such as several hours or several days) which results in the applicability and generalizability of the data-driven models/knowledge being largely restricted by the scope of training data. Meanwhile, conventional data-driven methods still rely much on domain knowledge, such as selecting the model inputs and classifying data, which limits the capability of the data-driven approach in discovering new knowledge from data. The advance of the data-driven approach is restricted by data sources and the power of data analytics techniques. Nowadays, buildings are not only energy intensive, but also information/data intensive. The wide adoption of information technologies, including both wired and wireless Internet and Internet of Things (IoT) in modern buildings has facilitated the collection and storage of a huge amount of operational data of building systems and equipment, as well as more environment and human behavior-related data. The effective use of the big building data becomes a challenge to and an opportunity for modern buildings facing the increasing complexity and integration of building and building energy systems, the increasing dynamic interactions among buildings, systems, occupants and power grids, and the increasing requirements of sustainability and intelligence. E-mail: [email protected]
This topical issue features the most recent work done by the active researchers in the building field on using advanced data analytics for building energy modeling and management. The authors adopted a diversity of advanced data mining and machine learning techniques, including neural networks, text mining, unsupervised clustering, fuzzy string matching algorithm, various regression techniques, support vector machine, boosting tree, random forest, kernel entropy component analysis, XGBboost algorithm, ensemble learning, reinforcement learning, transfer learning, etc., to discover novel data-driven knowledge from building data. The data sources include building automation systems, smart meters, social networks, public data sets and experiments. The
Data Loading...
