基于Kohonen网络的典型绝缘缺陷局部放电模式识别

针对变电站环境下局部放电识别所面临的不可控干扰多、现有识别方法初始参数难确定的问题，在设计符合变电站放电特点的典型缺陷，并采集多个样本数据的基础上，结合统计特征参数提取方法，基于具有自组织竞争识别、抗干扰性强特点的Kohonen网络，得出了一种适用于干扰环境下局部放电识别的新方法。通过探究Kohonen网络竞争层节点数对识别效果的影响，得出了针对样本数据的最佳识别参数，并将此网络与常用的模式识别算法在同等条件下进行对比，证明了其面对多种放电识别时的高稳定性与高识别率，验证了其用于变电站环境下局部放电识别时的优良性能。

Partial discharge recognition of typical insulation defect based on Kohonen network

Aiming at the problem that the uncontrollable interference faced by the partial discharge identification in substation, and the initial parameters of the existing identification method are difficult to determine, in the design of defects that meet the discharge characteristics of the substation, multiple sample data were collected, and combined with the statis-tical characteristic parameters extraction method, based on the Kohonen network with self-organizing competition recognition and strong anti-interference characteristics, a new method suitable for partial discharge identification in substation was presented. By exploring the influence of the Kohonen network''s parameters on its recognition effect, the recognition effect is optimized. Then by comparing this network with the commonly used pattern recognition algorithm under the same conditions, the high stability and high recognition rate of Kohonen network was proved, and its excel-lent performance in partial discharge identification of substation was verified.