Work

Unsupervised anomaly detection of industrial building energy consumption

Year: 2024

Type: article

Abstract: Detecting anomalies in building energy consumption can reduce unnecessary energy waste and improve energy efficiency. The role of anomaly detection has become pivotal in industrial buildings because of their high energy consumption and risk of losses caused by abnormal events. Although a considerable amount of data collected through smart meters has indicated the potential for anomaly detection using data mining techniques, such labels are often unavailable in practical situations. Therefore, this study develops an ensemble framework that combines three unsupervised learning algorithms, including Local Outlier Factor, Deep Isolation Forest, and Anomaly Transformer to identify anomalous power consumption, with a particular focus on subsequence anomaly. The transformer-based network is established to precisely impute missing values and enhance the reliability of anomaly detection. The experimental results based on hourly cooling energy consumption in the two observed industrial buildings confirm the effectiveness of the proposed detection method. To better interpret the anomaly detection results, the Extreme Gradient Boosting algorithm is used to construct the relationship between influencing factors and anomalous consumption. The area under the Receiver Operating Characteristic curve is used as a metric for the classification task, and an average of over 0.95 indicates robust classification performance. Weekday and dew point temperature are found to have significant impacts on the electricity usage pattern. The research results can assist in the development of promising solutions for identifying unexpected trends in building energy consumption and supporting efficient energy management.

Source: Energy and Built Environment

DOI: 10.1016/j.enbenv.2024.12.005

Authors: Yi Song, Sennan Kuang, Junling Huang, Da Zhang

Institutions: China Three Gorges Corporation (China), Tsinghua University, Renmin University of China

Language: English

Topic: Building Energy and Comfort Optimization

Subfield: Building and Construction

Field: Engineering

Domain: Physical Sciences

SDGs: Affordable and clean energy, Industry, innovation and infrastructure, Climate action

Open Access status: diamond

Funders: China Three Gorges Corporation, National Natural Science Foundation of China