【摘要】 SF₆断路器中SF₆气体分解产物特性可以为设备的缺陷和故障诊断提供有效依据。通过断路器开断实验获得正常条件和几种缺陷条件下气体分解产物的种类和生成量,并展开深入研究。依据分析结果,提出了利用SO₂、SOF₂、CO、CO₂、CS₂等气体分解产物的峰值比所存在的特定分布关系,分辨断路器缺陷类型。通过选取数值集中、分布稳定的气体分解产物的峰值比作为特征量,采用威布尔分布函数作为拟合函数,矩法估计作为参数估计方法,构造缺陷特征拟合曲线,根据特征量对缺陷状态进行诊断。利用Matlab开发基于熵权法的断路器缺陷诊断程序。以信息熵作为赋权的标准,对不同特征量的有效性和差异性进行评估,根据价值矩阵,给出对应的动态权重。应用缺陷诊断技术对几种缺陷条件和正常开断实验数据进行诊断,结果显示该诊断模型能够准确地分辨出对应的缺陷类型。建立的基于气体产物分析的缺陷诊断综合评价方法,可以为SF6电气设备的缺陷诊断及系统开发提供参考。更多还原

【Abstract】 The characteristic of gas decomposition product of SF₆ circuit breaker can provide effective support for the defect and fault diagnose. In this paper,the type and amount of gas decomposition product under normal condition and several defect conditions were obtained and analyzed by experiments. Based on the results,the distribution relation for the ratios of the peak of gas decomposition product,such as SOF₂,CO,CO₂,CS₂,were used to identify the type of circuit breaker defects. Gas ratios with centralized distribution was selected as eigenvalues and Weibull distribution was adopted as fitting curve to diagnosis and check the defect state,and thus a defect diagnosis method using eigenvalues was realized.Based on entropy method,defect diagnosis program was developed using Matlab. With information entropy as weight determining method,results of different eigenvalue were evaluated and dynamic weight was given to different eigenvalue accordingly. This paper can provide reference to the development of SF₆ power equipment defect diagnosis system.更多还原