摘要
提出一种基于支持向量机(support vector machine,SVM)的电流互感器(current transducer,CT)二次侧饱和电流补偿算法。以最近1周期故障电流采样数据的归一化值作为输入向量,以故障后5个周期的电流数据作为训练样本,利用SVM来建立CT二次侧饱和电流与一次侧电流之间的非线性关系,进而对饱和电流进行精确补偿。仿真分析表明,该方法在各种CT饱和条件下均能有效补偿,对于相同的训练样本,其补偿精度要高于神经网络方法。
Current transducer malfunction of protective and （CT） saturation will lead to control devices. Based on support vector machine （SVM）, an algorithm for compensating secondary saturation current of CT is proposed, in which the normalized value of sampling data of fault current within the latest period is taken as input vector; by use of SVM a non-linear model describing the complex relation between primary current and secondary saturated current of CT is established; and then the accurate compensation of saturated current is performed. Simulation results show that the proposed method can effectively compensate the saturation current of CT under various saturation conditions, and as for the same training sample the compensation accuracy by the proposed method is better than that by neural network approaches.
出处
《电网技术》
EI
CSCD
北大核心
2008年第4期86-90,共5页
Power System Technology
关键词
电流互感器
支持向量机
电气测量
current transducer
support vector machine electric measurement
