Open AccessPartial discharge localisation methodology for power transformers based on improved acoustic propagation route search algorithm
Author(s) -
Wang YanBo,
Fan YuHang,
Qin ShaoRui,
Chang DingGe,
Shao XianJun,
Mu HaiBao,
Zhang GuanJun
Publication year - 2018
Publication title -
iet science, measurement and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.418
H-Index - 49
eISSN - 1751-8830
pISSN - 1751-8822
DOI - 10.1049/iet-smt.2018.5092
Subject(s) - transformer , partial discharge , algorithm , computer science , search algorithm , shortest path problem , path (computing) , voltage , engineering , electrical engineering , graph , theoretical computer science , programming language
The localisation of partial discharge (PD) sources using the acoustic emission (AE) technique hasattracted increasing research attention. The complicated propagation routes andwave‐type conversion in power transformer can induce considerable localisationerror. In this paper, the catadioptric phenomenon of AE wave propagation isexplained in detail. With the incident angle varying, the wave type of thedirect wave in tank wall could convert and the velocity will change accordingly.A transformer model is established in which each node refers to a suspected PDposition and a shortest route search algorithm is proposed to calculate theshortest path between two nodes in this model. However, the fastest route ismost significant to PD localisation rather than the shortest route when TDOAmethod is used. As a result, an improved propagation route search (IPRS)algorithm, which can recreate the propagation process and calculate the fastestAE routes, is proposed to localise the PD origin. To verify the feasibility ofthe IPRS algorithm, localisation experiments were performed in 35 and 110 kVtransformers, respectively. Compared with other present localisation methods,such as the Chan algorithm, the genetic algorithm and the imperial competitivealgorithm, the proposed algorithm can effectively reduce localisation error.