A1231
Title: Penalized linked component analysis for spatial-temporal burst detection in water distribution systems
Authors: Shenghao Xia - Bowling Green State University (United States) [presenting]
Abstract: Detecting bursts from spatial-temporal (ST) hydraulic data is critical for water distribution system management. Conventional anomaly detection methods, such as statistical process control and basis expansion, are inefficient and inaccurate in detecting and localizing bursts from data, which are continuously collected from multiple potential locations. These limitations can delay response and lead to substantial economic losses from water loss and subsequent infrastructure repair. A new method based on penalized linked component analysis is proposed, which extracts ST anomaly features by differentiating the commonly shared normal features and individual anomaly features. Penalization is incorporated in the algorithm to increase the sensitivity of location detection and reduce the rate of false alarms. The effectiveness of the proposed method is demonstrated with a simulated case study.
