基于多源信号与混合注意力的伺服阀故障诊断
|
|
杨晨,严建文,李磊,李贵闪
|
Fault diagnosis of servo valve based on multi-source signal and hybrid attention
|
|
Chen YANG,Jianwen YAN,Lei LI,Guishan LI
|
|
| 表 5 DC-MTF-Net模型与先进模型的4种评价指标对比 |
| Tab.5 Comparison of four evaluation metrics between DC-MTF-Net model and advanced models |
|
| % | | 模型 | 准确率
(标准差) | 精确率
(标准差) | 召回率
(标准差) | F1分数
(标准差) | | A | 99.54
(±1.00) | 99.55
(±0.96) | 99.54
(±1.00) | 99.53
(±0.99) | | A5 | 83.37
(±3.51) | 82.88
(±4.12) | 83.37
(±3.51) | 82.35
(±4.50) | | A6 | 88.80
(±3.87) | 89.98
(3.51) | 88.80
(±3.87) | 88.89
(±4.05) | | A7 | 83.94
(±2.15) | 84.78
(±1.87) | 83.94
(±2.15) | 83.61
(±2.48) | | A8 | 84.87
(±17.32) | 84.92
(±21.11) | 84.87
(±17.32) | 83.06
(±20.18) | | A9 | 60.20
(±22.66) | 59.81
(26.63) | 60.20
(±22.66) | 56.71
(±26.02) | | A10 | 72.75
(±15.82) | 75.17
(±17.83) | 72.75
(±15.82) | 71.19
(±17.43) | | A11 | 77.32
(±16.32) | 80.02
(±15.34) | 77.32
(±16.32) | 75.74
(±18.19) |
|
|
|