Study on Degradation Model of Varistors Based on Machine Learning
Metal Oxide Varistor (MOV) is a crucial component in Surge Protection Device (SPD) , which is widely used in the lightning protection field of railways.Therefore, the working status of MOV determines the actual effect of lightning protection.Status-sensing and degradation-assessment of MOV can effectively avoid surge intrusion caused by SPD failure.In this paper, we conducted a degradation experiment of MOV, and during the entire life cycle of MOV several degradation-sensitive parameters are collected.First, the change trend of each parameter is analyzed, and then the degradation model of MOV is constructed by k-Nearest Neighbor (kNN) and Linear Regression method.Finally, the verification is carried out.The model constructed in this paper has a high accuracy rate, and provides a more comprehensive method for condition monitoring of MOV.The model realizes the condition monitoring of component level in lightning protection system and provides a reference method for intelligent operation and maintenance of lightning protection system in railway signal and communication department.
Metal Oxide Varistor (MOV) degradation-sensitive parameters degradation model
Tong, Xiao Shaoyun, Jin
Protection department of Signal & Communication Research Institute CHINA ACADEMY OF RAILWAY SCIENCES CORPORATION LIMITED Beijing, China
国际会议
4th International Lightning Protection Symposium (ILPS2018第四届雷电防护国际研讨会)
深圳
英文
103-107
2018-10-24（万方平台首次上网日期，不代表论文的发表时间）