效果视频利用CNN网络输电线路故障诊断(Python代码压缩包带有数据集和代码解压缩可直接运行)_哔哩哔哩_bilibili环境库要求TensorFlow2.4.0版本以上即可1.数据集介绍将故障区分为具体的不同类型单相短路故障、两相接地短路故障、两相相间故障、三相相间短路故障。这里随意举出每种类别的两个样本进行展示。GCBAIaIbIcVaVbVc1001-151.2918124-9.67745156385.800162260.400749853-0.132934945-0.2678149071001-336.1861826-76.2832619518.328896580.312731934-0.123633156-0.1890987791011-343.4870147104.56275133.7942853090.2720425010.011317575-0.2833600761011-339.1254001105.4293167-0.2672412250.277820540.021756839-0.299577378011019.38615173-785.553797768.7279081-0.210406869-0.00201120.212418069011018.47841651-783.8619173767.9410527-0.217651204-0.002604510.2202557140111506.5917463374.8825788-879.3449970.042029705-0.025636401-0.0163933050111495.1384715387.4159615-880.42530960.042107683-0.025103056-0.0170046272.模型CNN网络每类故障有1000个样本3.效果平均识别准确率为 98.053%混淆矩阵F1-score4.对项目感兴趣的可以关注最后一行