摘要
针对现有煤岩识别模型和方法准确率低、稳定性差、难以在工程实践中获得应用的问题,提出了基于经验模式分解(EMD)与深度卷积神经网络(DCNN)的混合智能识别方法。首先,应用EMD对采煤过程中的振动信号进行分解,得到一系列的本征模式分量(IMF)。然后利用DCNN进行IMF信息的融合,并自动提取特征信息。最后使用Softmax实现煤岩分界的智能识别。工程应用试验数据表明,该方法能够有效、准确地实现煤岩分界的识别,并具有良好的稳定性。
In view of the problems that the existing coal and rock identification model and methods have low accuracy and poor stability,which are difficult to be applied in engineering practice,a hybrid intelligent identification method based on empirical mode decomposition(EMD)and deep convolutional neural network(DCNN)was proposed.Firstly,EMD was applied to decompose the vibration signals during the coal mining process to obtain a series of intrinsic mode components(IMF).Then,DCNN was used to fuse IMF information and automatically extract feature information.Finally,Softmax was used to achieve intelligent recognition of coal and rock boundaries.The engineering application test data shows that this method can effectively and accurately identify the coal and rock boundary,and has good stability.
作者
李雄
沈良
田亚锋
尹家宽
王立阳
杨东晨
慕礼洋
朱益军
Li Xiong;Shen Liang;Tian Yafeng;Yin Jiakuan;Wang Liyang;Yang Dongchen;Mu Liyang;Zhu Yijun(Qinglongsi Coal Mine Branch,CHN Energy Yulin Energy Co.,Ltd.,Yulin 719408,China;Beijing Branch,Changzhou Research Institute Co.,Ltd.,China Coal Technology and Engineering Group,Beijing 100013,China)
出处
《煤矿机械》
2024年第1期58-60,共3页
Coal Mine Machinery