基于最大似然估计-Hilbert法的截齿侧向载荷特征识别  

Feature recognition of side load on conical pick cutting coal based on maximum likelihood estimation with Hilbert transformation

作　　者：刘春生[1] 韩飞[1] 任春平[1] 王庆华[1] 于信伟[1] 

机构地区：[1]黑龙江科技大学机械工程学院,哈尔滨150022

出　　处：《黑龙江科技大学学报》2015年第3期299-303,共5页Journal of Heilongjiang University of Science And Technology

基　　金：国家自然科学基金面上项目(51274091)

摘　　要：为探求镐型截齿旋转截割煤岩侧向载荷的统计规律及时频谱特性,实现其特征参量的定量提取与有效识别,利用多截齿参数可调式旋转截割实验台开展煤岩截割实验。采用最大似然估计法与希尔伯特变换Hilbert法,给出锐齿、棱齿和钝齿侧向载荷统计特征与时频谱特征的内在关联。结果表明:在截齿旋转角为零度、截槽对称的实验条件下,截齿侧向载荷概率密度函数曲线服从正态分布规律,截齿侧向载荷的方向具有明显的不确定性,绕y轴正、负半轴波动。其均值接近于零,不利于截齿的自回转运动。二维时频谱表明截割能量集中在5 Hz以内的低频区域,且可有效识别侧向载荷的时频特征。This paper is a targeted effort to investigate the statistical law and time-frequency spectrum characteristic of the side load on conical pick cutting coal and thereby achieve quantitative extraction and effective recognition these characteristic parameters. The investigation involves experiment on coal rock cutting using multi-picks parameter adjustable rotary cutting test bench; and adoption of the maximum likelihood estimation with Hilbert transformation to obtain the internal connection between the statistical law and time-frequency spectrum characteristic of sharp,blunt and edges picks. The result shows that under the experimental condition where the rotation angle of pick is zero and the kerf is symmetric,the curve of probability density function of side load conforms to normal distribution; the side load has obvious uncertain direction and fluctuates around y axis,and has the mean value close to zero,which goes against self rotary motion. The 2D time-frequency spectrum shows that the cutting energy is concentrated in the low frequency region within 5 Hz,allowing an effective identification of the time-frequency characteristics of lateral load.

关 键 词：镐型截齿 最大似然估计-Hilbert 侧向载荷 特征识别 

分 类 号：TD421.61[矿业工程—矿山机电]