基于倒谱距离的采煤机煤岩截割振动信号识别

发布时间：2018-05-15 11:29

  本文选题：煤炭开采 + 采煤机　； 参考：《工矿自动化》2017年01期

【摘要】：在基于煤岩截割振动信号分析的煤岩界面识别过程中,针对常规时频域分析方法对噪声敏感、振动信号能量变化适应性差等问题,提出一种基于倒谱距离的采煤机煤岩截割振动信号识别方法。通过分析振动传感器采集的采煤机不同负载状况下的截割振动信号,得出结论:与采煤机割岩状态相比,割煤状态下得到的振动信号与空载状态下的标准信号的倒谱距离更大;割岩状态下振动信号的倒谱距离呈明显的周期性,且周期为滚筒旋转1周的时间,而割煤状态下的振动信号无此特征。工业试验结果表明,该方法在煤岩硬度差大于10 MPa时,识别准确率达75%。
[Abstract]:In the process of coal and rock interface recognition based on the analysis of coal and rock cutting vibration signal, the conventional time-frequency domain analysis method is sensitive to noise and the adaptability of vibration signal energy change is poor. A method based on cepstrum distance for vibration signal recognition of coal cutting in coal mining machine is proposed. By analyzing the cutting vibration signals of shearer under different load conditions collected by vibration sensor, it is concluded that the cepstrum distance between the vibration signals obtained in coal cutting state and the standard signals in no-load state is larger than that in cutting rock state of shearer. The cepstrum distance of vibration signal in the state of rock cutting is obviously periodic, and the period is that the drum rotates for one week, but the vibration signal under the condition of coal cutting has no such characteristic. The industrial test results show that the recognition accuracy of this method is 75 when the hardness difference of coal and rock is more than 10 MPa.
【作者单位】： 中煤科工集团上海研究院;天地科技股份有限公司上海分公司;
【基金】：天地科技技术创新基金资助项目(KJ-2015-TDSH-01)
【分类号】：TD421.6
，

本文编号：1892274