基于改进CEEMDAN-DCNN的声发射源识别分类方法
|
谢学斌,刘涛,张欢
|
Identification and Classification Method of Underground AE Source Based on Improved CEEMDAN-DCNN
|
Xuebin XIE,Tao LIU,Huan ZHANG
|
|
表4 DCNN模型主要结构参数
|
Table 4 Main structural parameters of the DCNN model
|
|
网络层 | 输出 | 卷积核尺寸/步长 | padding | 激活函数 |
---|
输入层 | 1×1 024 | - | - | - | 卷积层C1 | 1×64 | 1×3/1 | same | ReLU | DropoutD1 | 1×64 | 比率:0.2 | 卷积层C2 | 1×256 | 1×3/1 | same | ReLU | DropoutD2 | 1×256 | 比率:0.2 | 卷积层C3 | 1×32 | 1×3/1 | same | ReLU | DropoutD3 | 1×32 | 比率:0.2 | 卷积层C4 | 1×32 | 1×3/1 | same | ReLU | DropoutD4 | 1×32 | 比率:0.2 | 卷积层C5 | 1×32 | 1×3/1 | same | ReLU | DropoutD5 | 1×32 | 比率:0.2 | 卷积层C6 | 1×32 | 1×3/1 | same | ReLU | DropoutD6 | 1×32 | 比率:0.2 | 全连接层F1 | 1×512 | - | - | ReLU | dropout | 1×512 | 比率:0.5 | 全连接层F2 | 1×4 | - | - | - |
|
|
|