基于大数据分析的变压器油溶乙炔气体浓度自动化检测技术研究  

Research on Automatic Detection Technology of Acetylene Gas Concentration in Transformer Oil Based on Big Data Analysis

作　　者：盛吉 高明 钱雨峰 李兴 SHENG Ji;GAO Ming;QIAN Yufeng;LI Xing(Yangzhou Power Supply Branch,State Grid Jiangsu Electric Power Co.,Ltd.,Yangzhou 225000,China)

机构地区：[1]国网江苏省电力有限公司扬州供电分公司,扬州225000

出　　处：《自动化与仪表》2024年第7期75-78,83,共5页Automation & Instrumentation

基　　金：国网江苏省电力科技有限公司科技项目(J2023157)。

摘　　要：为了有效掌握当前变压器故障状态,该文研究基于大数据分析的变压器油溶乙炔气体浓度自动化检测技术。利用奇异谱分析方法对油溶乙炔时间序列分解和重构,将上述时间序列划分成训练集和测试集,并将这2个数据集作为输入,使用大数据分析算法中的卷积神经网络建立变压器油溶乙炔气体浓度自动化检测模型,将训练集和测试集映射、特征提取,使用模型的输出层输出气体浓度自动化检测结果。实验表明,该方法具备较强的变压器油溶乙炔时间序列奇异谱分析能力,同时可在不同测试环境下实现变压器油溶乙炔气体浓度自动化检测结果,应用性较好。In order to effectively grasp the current fault state of transformer,the automatic detection technology of transformer oil-soluble acetylene gas concentration based on big data analysis was studied.The time series of oil-soluble acetylene is decomposed and reconstructed by using the singular spectrum analysis method,the above time series is divided into a training set and a test set,and the two datasets are taken as inputs,and the convolutional neural network in the big data analysis algorithm is used to establish an automatic detection model of transformer oil-soluble acetylene gas concentration,the training set and the test set are mapped and feature extracted,and the output layer of the model is used to output the automatic detection results of gas concentration.Experiments show that this method has a strong ability to analyze the singular spectrum of transformer oil-soluble acetylene time series,and can realize the automatic detection results of transformer oil-soluble acetylene gas concentration in different test environments,and has good applicability.

关 键 词：大数据分析 变压器 油溶乙炔 气体浓度 自动化检测技术 奇异谱分析 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]