中国电气工程学报（英文） ›› 2024, Vol. 10 ›› Issue (1): 35-47.doi: 10.23919/CJEE.2024.000052

• • 上一篇    下一篇

  

  • 收稿日期:2023-12-15 修回日期:2024-01-12 接受日期:2024-02-01 出版日期:2024-03-25 发布日期:2024-04-10

Simulation of Ultrasonic Propagation in Transformers within Thermal Fields and Intelligent Methodology for Hot-spot Temperature Recognition*

Dongxin He1, Dechao Yang1, Xinhua Guo2, Jiefeng Liu3,*, Haoxin Guo1, Qingquan Li1, Gilbert Teyssedre4   

  1. 1. School of Electrical Engineering, Shandong University, Jinan 250061, China;
    2. School of Mechanical and Electronic Engineering, Wuhan University of Technology, Wuhan 430070, China;
    3. School of Electrical Engineering, Guangxi University, Guangxi 530004, China;
    4. UPS, INPT, CNRS, LAPLACE (Laboratoire Plasma et Conversion d'Energie), University of Toulouse, Toulouse 31100, France
  • Received:2023-12-15 Revised:2024-01-12 Accepted:2024-02-01 Online:2024-03-25 Published:2024-04-10
  • Contact: * E-mail: jiefengliu2018@gxu.edu.cn
  • About author:Dongxin He received the B.S. degree in Electrical Engineering from Shandong University and a Ph.D. degree from North China Electric Power University in Beijing. In 2015, he travelled to the LAPLACE Laboratory at the University of Toulouse, France, as a Visiting Scholar. He is currently an Associate Professor at Shandong University, Jinan, China. His research interests include space-charge characteristics in insulation materials, such as cable insulation and power device encapsulation insulation. He has been selected for the Young Talent Lifting Project of the China Association for Science and Technology (CAST).
    Dechao Yang was born in the Hebei Province, China. He received the B.S. degree from Shandong University, Shandong, China, in 2023. He is currently pursuing a Master's degree in Electrical Engineering at Shandong University, Shandong, China. His current research interests are primarily focused on the use of ultrasonic sensing technology combined with multiphysics simulations to investigate and analyze hotspot temperatures inside converter transformers.
    Xinhua Guo received his Ph.D. in Information Processing from the Tokyo Institute of Technology, Tokyo, Japan, in 2014. From May 2014 to September 2016, he served as an Assistant Professor at the School of Mechanical and Electronic Engineering, Wuhan University of Technology, China. He has been an Associate Professor since October 2016. His research interests include acoustic sensing, imaging, and control methods.
    Jiefeng Liu was born in Hebei, China in 1985. He received his M.S. and Ph.D. degrees in Electrical Engineering from Chongqing University, Chongqing, China in 2011 and 2015, respectively. In 2018, he joined Guangxi University, where he serves as an Associate Professor at the School of Electrical Engineering. He has authored and co-authored over 90 papers published in journals and conferences. Dr. Liu's research interests include insulation condition assessment and fault diagnosis of transformers.
    Haoxin Guo was born in Yantai, China. He received his B.S. degree from Qingdao University, Qingdao, China in 2022. Currently, he is pursuing his Master's degree in Electrical Engineering at Shandong University, Shandong, China. His research interests focus on machine learning algorithms for monitoring internal hot spot temperatures in converter transformers based on ultrasonic thermometry.
    Qingquan Li was born in Laiwu city, Shandong Province, China, in 1969. He received the Ph.D. degree in Electrical Engineering from Xi'an Jiaotong University, Xi'an, China, in 2003. Currently, he is a Professor at Shandong University, Jinan, China. His research interests include the lightning protection and grounding technology, the high voltage insulation and measurement technology, and the detection and diagnosis techniques for electrical equipment.
    Gilbert Teyssedre was born in May 1966 in Rodez, France. He received his Engineer degree in Materials Physics and graduated in Solid State Physics in 1989 at the National Institute for Applied Science (INSA). Then he joined the Solid State Physics Lab in Toulouse and obtained his Ph.D. degree in 1993 for work on ferroelectric polymers. He entered the CNRS in 1995 and has been working since then at the Electrical Engineering Lab (now LAPLACE) in Toulouse. His research activities concern the development of luminescence techniques in insulating polymers with focus on chemical and physical structure, degradation phenomena, space charge and transport properties. He is currently Research Director at CNRS and is leading a team working on the reliability of dielectrics in electrical equipment.
  • Supported by:
    * National Natural Science Foundation of China (U1966209, 52277155 and 2021CXGC010210).

PDF

68

关键词: Ultrasonic temperature measurement, non-destructive testing, oil-immersed transformers, machine learning

Abstract: Hot-spot temperature of transformer windings is a crucial indicator of internal defects. However, current methods for measuring the hot-spot temperature of transformers do not apply to those already in operation and suffer from data lag. This study introduces a novel inversion method that combines ultrasonic sensing technology, multiphysics simulation, and the K-nearest neighbors algorithm. Leveraging the penetrative ability and temperature sensitivity of ultrasonic sensing, a detailed physical field simulation model was established. This study extensively investigates the characteristics of ultrasonic wave signals inside transformers. The investigation includes different temperature fields, ranging from 40 ℃ to 110 ℃ at 10 ℃ intervals, and various ultrasonic wave emitter conditions. By extracting the key features of the acoustic signals, such as the peak time, propagation time, and peak amplitude, an accurate inversion of the winding hot-spot temperature is successfully achieved. The results demonstrate that this method achieves a high accuracy rate (98.57%) in inverting the internal winding hot-spot temperatures of transformers, offering an efficient and reliable new approach for measuring winding hot-spot temperatures.

Key words: Ultrasonic temperature measurement, non-destructive testing, oil-immersed transformers, machine learning