基于CNN-LSTM 的煤矿设备故障特征识别模型优化

打开文本图片集

中图分类号：TQ515；TH133.3文献标识码：A

文章编号：1001-5922（2025）06-0135-04

Optimization of coal mine equipment fault feature recognition model based on CNN-LSTM

MI Yanjun，ZHANG Hou，XU Hongliang，YANG Zhenhua，HU Wei（CHN ENERGY Shaanxi Deyuan Fugu Energy Co.，Ltd.，Sandaogou Colliery，Yulin ，Shaanxi China）

Abstract：Bearing is the core component of coal mine equipment. The special service environment leads to high failure rate，which affects the production efficiency of enterprises. The fault feature recognition accuracy of tradition⁃ al coal mine bearings is low，the anti-noise performance is poor，and the fault is not found in time，which causes huge economic losses to the enterprise. Combining the advantages of long short-term memory network in processing and predicting time series and the advantages of convolutional neural network in capturing local features of data，a fault feature recognition model of coal mine bearing equipment based on CNN-LSTM is proposed. The experimental data results show that the proposed model has higher bearing fault feature recognition accuracy than the LSTM mod⁃ el and CNN model，and the fault feature recognition model has good anti-noise performance. This has certain practi⁃ cal application value for identifying the fault characteristics of coal mine bearing equipment and reducing the fail⁃ ure rate of coal mine equipment.

Key words：long short-term memory network；convolutional neural network；coal mine bearing；fault feature recog⁃ nition

恶劣的服役环境使得煤矿轴承面临极端温度、重载荷、污染等挑战，这导致煤矿轴承故障频发，有效增加了设备的维修难度和维修成本[1]。（剩余5585字）