Page 89 - 摩擦学学报2025年第8期

第 8 期 王大刚, 等: 钢丝绳摩擦损伤定量识别研究 1187

表 8 损伤钢丝绳机器学习数据库
Table 8 Machine learning database of damaged wire rope
Neural network input data
Serial number Loss
F 1 F 2 F 3 F 4 F 5 F 6 F 7
1 0.008 0.404 0.305 0.060 0.488 1.000 0.720 0.1%
2 0.000 0.121 0.185 0.433 0.481 0.688 0.110 0.2%
3 0.075 0.177 0.176 0.211 0.408 0.665 0.441 0.2%
4 0.443 0.478 0.096 0.106 0.402 0.444 0.331 0.2%
5 0.456 0.674 0.147 0.240 0.721 0.355 0.474 0.3%
6 0.415 0.499 0.800 0.067 0.366 0.506 0.202 0.3%
7 0.098 0.318 0.216 0.403 0.437 0.590 0.194 0.3%
8 0.336 0.609 0.055 0.067 0.357 0.385 0.643 0.3%
9 0.325 0.562 0.426 0.247 0.408 0.076 0.036 0.3%
10 0.048 0.193 0.118 0.295 0.734 0.470 0.221 0.4%
··· ··· ···
270 0.778 0.357 0.204 0.565 0.126 0.354 0.863 8.0%

(a) 0.1%~1.0%

(b) 1.1%~2.0%

(c) 2.1%~3.0%

(d) 3.1%~4.0%

(e) 4.1%~5.0%

(f) 5.1%~6.0%

(f) 6.1%~7.0%

(g) 7.1%~8.0%
Fig. 12 Wavelet time-frequency diagram of signal after noise reduction with different cross-sectional loss
图 12 不同截面积损失降噪后信号小波时频图