基于SMOTE和NGO-GBDTT的变压器故障诊断方法
Li-Zhong Wang1, Jian-Fei Chi1, Ye-Qiang Ding1
1State Grid Zhejiang Power Co., Ltd, Hangzhou Linping Power Supply Company, Hangzhou, 311199, China.
这项研究介绍了一种使用合成少数人过量采样技术 (SMOTE) 和北方戈斯霍克优化 (NGO) 的新型变压器故障诊断方法,以提高梯度增强决策树 (GBDT) 的性能. 该方法有效地解决了不平衡的数据,提高了故障识别的准确性,减少了错误判断.
科学领域:
- 电气工程 电气工程
- 人工智能的人工智能
- 机器学习 机器学习
背景情况:
- 变压器故障诊断对于电力系统可靠性至关重要.
- 在故障诊断中不平衡的数据集导致模型准确性较低.
- 现有的方法在培训数据不足和少数样本错误分类方面扎.
研究的目的:
- 提出一种先进的变压器故障诊断方法.
- 为了减轻不平衡样本对诊断准确性的影响.
- 为了提高渐变增强决策树 (GBDT) 模型的性能.
主要方法:
- 使用合成少数人过量采样技术 (SMOTE) 来进行数据平衡.
- 采用非编码比率方法用于多维特征构造.
- 应用光梯度增强机 (LightGBM) 用于特征选择.
- 使用北方戈肖克优化 (NGO) 算法优化GBDT参数.
主要成果:
- 提出的方法有效地扩大了少数群体的样本,减少了错误判断.
- 与其他集成模型相比,实现了高故障识别精度.
- 证明了低误判率和稳定的诊断性能.
- 通过NGO算法成功优化GBDT参数.
结论:
- SMOTE和NGO-GBDT方法为变压器故障诊断提供了一个强大的解决方案.
- 这种方法显著提高了识别变压器故障的准确性和可靠性.
- 该技术在处理不平衡的数据集方面是有效的,这是该领域的一个常见挑战.
更多相关视频
03:31Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
Published on: December 15, 2023
10:52Design, Instrumentation and Usage Protocols for Distributed In Situ Thermal Hot Spots Monitoring in Electric Coils using FBG Sensor Multiplexing
Published on: March 8, 2020
相关概念视频
Power System Three-Phase Short Circuits
Bus Impedance Matrix
In the first circuit, all machine voltage sources are short-circuited, leaving only the prefault voltage source at the fault location. The positive-sequence bus impedance matrix can be determined by solving the nodal equations,...
Transformers in Distribution System
Distribution substation transformers come in various ratings and typically use mineral oil for insulation and cooling. To prevent moisture and air from entering the oil, some transformers use an inert gas like nitrogen to fill the...
Transformers with Off-Nominal Turns Ratios
Fault Types
For line-to-line faults occurring between phases B and C, the...
Three-Phase Short Circuit—Unloaded Synchronous Machine
This behavior occurs due to the magnetic flux produced by the short-circuit armature currents. Initially, these currents follow high-reluctance paths but eventually shift to...
