| GA-BP神经网络在GIS典型缺陷局部放电识别中的应用 |
| |
| 引用本文： | 吕启深,吴磊,梁基重,杨益公.GA-BP神经网络在GIS典型缺陷局部放电识别中的应用[J].电瓷避雷器,2011(4):34-38,42. |
| |
| 作者姓名： | 吕启深 吴磊 梁基重 杨益公 |
| |
| 作者单位： | 1. 深圳供电局,广东深圳,518020
2. 西安交通大学电气工程学院,陕西西安,710049 |
| |
| 摘 要： | 基于BP算法的神经网络和基于实数编码遗传算法的神经网络用于模式识别有其各自的优缺点.为此提出利用GA-BP混合算法神经网络,对GIS典型缺陷的局部放电进行模式识别.以超声局部放电探测仪对不同缺陷包括金属突出物、悬浮电位缺陷和绝缘子内部气泡引起的局部放电信号进行了测量,并从测量结果中提取PRPD谱图特征值作为训练样本和测...
|
| 关 键 词： | 神经网络 实数编码 误差反向传播算法 遗传算法 |
Application of GA-BP-NN in PD Pattern Recognition of GIS Typical Defects |
| |
| Lü Qi-shen,WU Lei,LIANG Ji-chong,YANG Yi-gong.Application of GA-BP-NN in PD Pattern Recognition of GIS Typical Defects[J].Insulators and Surge Arresters,2011(4):34-38,42. |
| |
| Authors: | Lü Qi-shen WU Lei LIANG Ji-chong YANG Yi-gong |
| |
| Affiliation: | Lü Qi-shen1,WU Lei2,LIANG Ji-chong2,YANG Yi-gong1(1.Shenzhen Power Supply Bureau,Shenzhen,518020,2.School of Electrical Engineering,Xi'an Jiaotong University,Xi'an,710049,China) |
| |
| Abstract: | The pattern recognitions using neural network based on real-code BP algorithms and on genetic algorithms have its own advantages and disadvantages.It is proposed to use the neural network based on GA-BP hybrid Algorithm to identify the PD of GIS typical defects.Three defect of the PD signals,including metal protrusion,insulator void and floating discharge,are measured by using AIA(acoustic insulation analyzer),and the characteristic parameter of the PRPD plot in the measurement results is collected as the t... |
| |
| Keywords: | neural network real-code error back-propagation algorithm GA algorithms |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
