基于机器学习的煤矿输送带RFID健康监测方法  

Machine Learning Based RFID Enabled Health Monitoring of Coal Mine Conveyor Belt

作　　者：郭文升 李建平 GUO Wensheng;LI Jianping(Ulan Mulun Coal Mine,CHN Energy Shendong Coal Group Co.,Ltd.,Ordos 017000,China)

机构地区：[1]国家能源集团神东煤炭集团乌兰木伦煤矿,内蒙古鄂尔多斯017000

出　　处：《机械与电子》2025年第4期64-68,共5页Machinery & Electronics

摘　　要：提出了一种基于机器学习的煤矿输送带健康监测方法,利用RFID技术通过分析接收信号强度指示器数据,建立了贝叶斯正则化人工神经网络模型。通过物理实验验证了该模型在不同裂纹类型和宽度下的有效性。实验结果表明,该模型在裂纹检测任务中达到了97.2%的高准确率,显著优于传统的机器学习算法如支持向量机、决策树和线性判别分析。研究表明,所提健康监测方法不仅在裂纹检测中具有很高的准确性和泛化能力,还能够实现实时监测。This paper proposes a machine learning based health monitoring method for coal mine conveyor belts using RFID technology by analyzing the received signal strength indicator data and establishing a Bayesian regularized artificial neural network model.The validity of the model under different crack types and widths was verified by physical experiments.Experimental results show that this BRANN model achieves a high accuracy of 97.2%in the crack detection task,which significantly outperforms traditional machine learning algorithms such as support vector machines,decision trees and linear discriminant analysis.It is shown that the proposed health monitoring method not only has high accuracy and generalization ability in crack detection,but also can realize real-time monitoring.

关 键 词：机器学习 贝叶斯正则化人工神经网络 RFID 健康监测 裂纹检测 

分 类 号：TD528[矿业工程—矿山机电] TP391.4[自动化与计算机技术—计算机应用技术]