Electrical Life Prediction for Contact of SF<sub>6</sub> Circuit Breaker Based on LSTM-<italic style="font-style: italic">MIV</italic> Neural Network

MA Feiyue; LI Ao; WU Chengwei; YU Jiaying; CHEN Lei; NI Hui

doi:10.13296/j.1001-1609.hva.2024.02.006

您当前的位置：

首页 >

文章列表页 >

Electrical Life Prediction for Contact of SF₆ Circuit Breaker Based on LSTM-MIV Neural Network

Research & Analysis | 更新时间：2025-04-26

- Electrical Life Prediction for Contact of SF₆ Circuit Breaker Based on LSTM-MIV Neural Network
- High Voltage Apparatus Vol. 60, Issue 2, Pages: 69-77(2024)
- 作者机构：
  
  1.国网宁夏电力有限公司电力科学研究院，银川　 750001
  2.西安交通大学电气工程学院，西安　 710049
- 作者简介：
- 基金信息：
  
  Science and Technology Project of State Grid Ningxia Electrical Power Co., Ltd.(5229DK20055)
- DOI：10.13296/j.1001-1609.hva.2024.02.006
  CLC：
- Received：18 September 2023，
  
  Revised：15 November 2023，
  
  Published：16 February 2024
- 稿件说明：
移动端阅览
MA Feiyue, LI Ao, WU Chengwei, et al. Electrical Life Prediction for Contact of SF₆ Circuit Breaker Based on LSTM-MIV Neural Network[J]. High voltage apparatus, 2024, 60(2): 69-77.
DOI：

MA Feiyue, LI Ao, WU Chengwei, et al. Electrical Life Prediction for Contact of SF₆ Circuit Breaker Based on LSTM-MIV Neural Network[J]. High voltage apparatus, 2024, 60(2): 69-77. DOI： 10.13296/j.1001-1609.hva.2024.02.006.

摘要

文中基于LW30-40.5型SF

断路器全寿命烧蚀试验数据，采用归一化异常指数

反映断路器触头剩余电寿命，0表示电寿命完备，1表示电寿命耗尽。根据断路器分闸动态电阻—行程曲线，提出了将动态电阻—行程曲线上行程大于8 mm且电阻值大于500 μΩ的第一个峰值点作为断路器主触头接触阶段与弧触头接触阶段分界点。据此得到反映断路器触头异常指数

的5个特征参数，采用皮尔逊相关系数法对特征参数进行独立性分析。通过LSTM神经网络建立了基于特征参数与异常指数

之间的回归预测模型，

回归预测均方差和绝对平均误差分别为0.014 921和0.013 053，采用

MIV

方法确定了各特征参数对异常指数

影响权重。

Abstract

Based on the lifetime ablation test data of LW30-40.5 SF

circuit breaker

normalized anomaly index

is adopted in this paper to reflect the remaining electrical life of contact of the circuit breaker

where 0 indicates complete electrical life and 1 indicates

exhaustion of electrical life. According to the dynamic resistance-stroke curve of the circuit breaker

the first peak point with a stroke greater than 8 mm and a resistance greater than 500 μΩ on the dynamic resistance-stroke curve is proposed as the dividing point between the contacting stage of main contact of and the contacting stage of arc contact of the circuit breaker. Based on this

five characteristic parameters reflecting the abnormal index

of contact of the circuit breaker are obtained. The Pearson correlation coefficient method is used to analyze the characteristic parameters independently. LSTM neural network is used to set up a regression prediction model on the basis of relationship between the characteristic parameters and the anomaly index

. The mean square error and absolute mean error of the regression prediction of the

index are 0.014 921 and 0.013 053

respectively. The

MIV

method is used to determine the influence weights of each characteristic parameter on the anomaly index

关键词

Keywords

references

Views

139

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Electrical Endurance Assessment System of SF₆ Circuit Breaker for Switching Capacitor Bank Based on Current Monitoring and Dynamic Resistance Analysis

Study on Corrosion Characteristics of Arc Contact of SF₆ Circuit Breaker in Switching Operation

Study on Gas Decomposition Product Characteristics of SF₆ Circuit Breaker During Simulated Ablation Process

Study on Closing Pre-breakdown Characteristics of SF₆ Circuit Breaker During Contacts Deterioration

Research on Ablation Characteristics of Copper-tungsten Arc Contacts of High-voltage Circuit Breakers Based on Optical Measurement System

Related Author

WU Chengwei

DENG Junbo

LI Ao

CHEN Bangfa

HE Zilan

DAI Yuannan

DIAO Muxin

JIANG Qijian

Related Institution

State Key Lab of Electrical Insulation and Power Equipment, Xi'an Jiaotong University

State Key Laboratory of Electrical Insulation and Power Equipment, Xi'an Jiaotong University

Foshan Supply Bureau of Guangdong Power Grid Co., Ltd.

State Grid Ningxia Electrical Power Co., Ltd.

Power Research Institute of State Grid Ningxia Electrical Power Co., Ltd.

⁰

Electrical Life Prediction for Contact of SF6 Circuit Breaker Based on LSTM-MIV Neural Network

DOI：10.13296/j.1001-1609.hva.2024.02.006

摘要

Abstract

关键词

Keywords

references

Electrical Life Prediction for Contact of SF₆ Circuit Breaker Based on LSTM-MIV Neural Network