An unknown fault identification method based on PSO-SVDD in the IoT environment

• Main Authors: Erbao Xu, Yan Li, Lining Peng, Mingshun Yang, Yong Liu
• Format: Article
• Language: English
• Published: Elsevier 2021-08-01
• Series: Alexandria Engineering Journal
• Subjects: Equipment fault diagnosis; Box transformer substation; PSO-SVDD
• Online Access: http://www.sciencedirect.com/science/article/pii/S1110016821001496

When a new fault occurs, how to determine whether the new fault is a known fault or an unknown fault outside the fault pattern base. If a new unknown fault is identified, adding the unknown fault to the fault pattern base for adaptive updating the fault diagnosis model has become a new problem in the field of fault diagnosis. In order to solve this problem, we take Box transformer substation (BTS) widely used in power distribution equipment as an example, propose an unknown fault identification method. First, through the construction of the IoT framework including the perception layer, transmission layer and application layer, real-time data collection and online monitoring for the BTS can be realized. Then, using Support Vector Data Description (SVDD) as the unknown fault identification method, and optimizing the relevant parameters by Particle Swarm Optimization (PSO) algorithm, so that BTS can identify unknown faults with a timely and effective manner. Meanwhile, through the retraining of the model, the adaptive update of the existing fault diagnosis model is achieved. Finally, the validity of the designed method is verified by an example.