Title: Predictive maintenance strategy of running fault based on ELM algorithm for power transformer

Authors: Qian Wu; Xiaoyi Yang; Renming Deng

Addresses: School of Computer Science and Engineering, Chongqing University of Technology, Chongqing 400054, China ' School of Education, Chongqing Normal University, Chongqing 401331, China ' School of Automation, Chongqing University, Chongqing 400044, China

Abstract: Transformer is the most important core equipment of power system, and once the fault happened, the economic losses and adverse social impacts resulted in faults by the twinkling of eye are difficult to estimate. In order to prevent transformer fault happened, the paper explored a sort of predictive maintenance strategy of transformer fault power transformer supported by the extreme learning machine (ELM) algorithm. In this paper, it made the anatomy of the drawbacks in breakdown maintenance (BM) and preventive maintenance (PM) maintenance system, pointed out the advantage of predictive reliability maintenance (PRM) maintenance system, studied on prediction of the fault diagnosis and predictive algorithm based on ELM, discussed the dynamic fault diagnosis model of power transformer in smart grid environment. A comparative study of different fault pattern prediction algorithms confirmed the rationality and feasibility of the fault prediction algorithm based on ELM. Transformer operation experience shows that the ELM algorithm can provide powerful technical support for the maintenance strategy of transformer fault prediction.

Keywords: equipment of power system; fault happened; transformer fault; predictive maintenance strategy; ELM algorithm; powerful technical support; transformer operation; fault prediction algorithm; powerful technical support; maintenance strategy; transformer fault prediction.

DOI: 10.1504/IJIMS.2018.091999

International Journal of Internet Manufacturing and Services, 2018 Vol.5 No.2/3, pp.297 - 309

Received: 25 Jul 2017
Accepted: 24 Oct 2017
Published online: 24 May 2018 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article