基于深度学习和多模型融合的局部放电模式识别方法

局部放电与电力设备的绝缘状态息息相关，准确识别局部放电类型对于保障电网运行具有重要意义。文中提出一种基于深度学习和多模型融合的局部放电模式识别方法。首先，设计并搭建开关柜内4类典型局部放电缺陷模型，采集局部放电相位分布(phase resolved partial discharge，PRPD)图谱并建立样本集；其次，搭建基于迁移学习的深度残差网络，对局部放电缺陷进行识别；最后，利用Sugeno模糊积分将深度残差网络(deep residual net￣work，DRN)模型与传统识别模型进行融合。实验结果表明：迁移学习模型相比于无迁移学习模型有着更好的更新能力和泛化性能；融合模型与单一模型相比具有更高的识别准确率。所提方法能够准确识别局部放电缺陷类型，对于电力设备的运维检修具有一定的参考价值。

A partial discharge pattern recognition method based on deep learning and multi-model fusion

Partial discharge is closely related to the insulation state of power equipment. Accurate identification of partial discharge types is of great significance to ensure the operation of the power grid. A partial discharge pattern recognition method based on deep learning and multi-model fusion is proposed in this paper. Firstly,four kinds of typical partial discharge defect models in the switchgear are designed and built,and the phase resolved partial discharge (PRPD) spectrum is collected to estalish sample set. Then,a deep residual network based on transfer learning is built to identify partial discharge defects. Finally,the deep residual network (DRN) model is fused with the traditional recognition model by Sugeno fuzzy integral. The experimental results show that the transfer learning model has better update ability and generalization performance than the non-transfer learning model does,and the fusion model has higher recognition accuracy than the single model does. The proposed method can accurately identify the types of partial discharge defects,and has a certain reference value for the operation and maintenance of power equipment.