基于多源数据融合的悬臂式掘进机截割载荷预测  

Cutting Load Prediction of Cantilever Roadheader Based on Multi-source Data Fusion

作　　者：赵红美[1] 杨珍明[1] Zhao Hongmei;Yang Zhenming(Tangshan Polytechnic College,Tangshan 063299,China)

机构地区：[1]唐山工业职业技术学院,河北唐山063299

出　　处：《煤矿机械》2020年第10期199-201,共3页Coal Mine Machinery

摘　　要：针对巷道掘进工作环境恶劣、截割载荷识别难度大的问题,提出了一种基于多源数据融合的掘进机截割载荷预测方法。以掘进机截割部在实际工况中采集的振动信号、电流信号以及温度信号为多源数据基础,采用小波包奇异值分解算法提取出多源数据的特征向量,通过WPA-RBF神经网络预测模型预测不同工况下截割不同硬度岩壁的截割载荷。仿真结果表明,基于多源数据融合的特征提取方法,可以对不同工况下截割不同硬度等级岩壁的载荷状态进行准确预测,确保了悬臂式掘进机在工况突变状况下的截割稳定性。Aiming at the problems of the harsh working environment of tunnel driving and the difficulty of cutting load identification,a prediction method for cutting load of roadheader based on multi-source data fusion was proposed.Based on the multi-source data of the vibration signal,current signal and temperature signal collected by the roadheader cutting components under actual working conditions,the wavelet packet singular value decomposition algorithm was used to extract the eigenvectors of the multi-source data and through the WPA-RBF neural network prediction model,predicted the cutting load for cutting different hardness rock wall under different working conditions.The simulation results show that the feature extraction method based on multi-source data fusion can accurately predict the load state of cutting rock walls with different hardness levels under different working conditions and ensure the cutting stability of the cantilever roadheader under sudden change of working conditions.

关 键 词：悬臂掘进机 载荷预测 多源数据融合 RBF神经网络 

分 类 号：TD421.5[矿业工程—矿山机电]