Towards Progressively Detecting Faults in a Large Power System: A Visual Analytics Approach

Electricity serves as the foundation of everyday life. Grid operators are obliged to provide stable power supply to end users. However, given the scale and complexity of modern nationwide grids, it has become increasingly challenging to locate faults quickly and accurately during power failure. Previous locating techniques mainly focus on data analyzing methods, while little work is done using visual analytics approach. This paper presents a progressive visual analytics procedure (i.e., overview, detecting faulted bus candidates, locating buses, and looking into bus details) to detect fault location and figure out possible faulted buses. In particular, an efficient clustering algorithm associated with a novel visualization for time-varying series as well as multiple and coordinated views is developed to locate faulted buses intuitively and quickly. Case studies and expert feedback have indicated that our approach is contributing to real-world applications.