基于重采样降噪与主成分分析的宽卷积深度神经网络风机故障诊断方法
|
|
刘展, 包琰洋, 李大字
|
Fault Diagnosis Method of Wind Turbines Based on Wide Deep Convolutional Neural Network With Resampling and Principal Component Analysis
|
|
Zhan LIU, Yanyang BAO, Dazi LI
|
|
表3 消融实验结果
|
Tab. 3 Results of ablation experiment
|
|
| 风机编号 | 训练集 | 测试集 |
|---|
| 第1组 | 第2组 | 第3组 | 第4组 | 第1组 | 第2组 | 第3组 | 第4组 |
|---|
| 01 | 0.956 | 0.942 | 0.998 | 0.985 | 0.956 | 0.948 | 0.984 | 0.976 | | 02 | 0.957 | 0.949 | 0.997 | 0.987 | 0.949 | 0.943 | 0.988 | 0.975 | | 03 | 0.957 | 0.937 | 0.998 | 0.992 | 0.956 | 0.941 | 0.989 | 0.978 | | 04 | 0.952 | 0.938 | 0.998 | 0.989 | 0.956 | 0.945 | 0.983 | 0.973 | | 05 | 0.954 | 0.951 | 0.996 | 0.986 | 0.965 | 0.941 | 0.984 | 0.974 | | 06 | 0.968 | 0.949 | 0.998 | 0.983 | 0.957 | 0.938 | 0.987 | 0.972 | | 07 | 0.953 | 0.943 | 0.996 | 0.987 | 0.954 | 0.946 | 0.985 | 0.974 | | 08 | 0.959 | 0.942 | 0.998 | 0.985 | 0.955 | 0.942 | 0.982 | 0.975 | | 09 | 0.951 | 0.945 | 0.997 | 0.988 | 0.953 | 0.943 | 0.986 | 0.968 | | 10 | 0.957 | 0.943 | 0.997 | 0.990 | 0.955 | 0.943 | 0.984 | 0.973 |
|
|
|