基于AR模型的磨机振动信号特征提取方法

发布时间：2018-08-03 06:44

【摘要】：磨矿过程中的球磨机筒体内部负荷(填充率、料球比)研究是选矿设备节能降耗的重要内容。以试验球磨机为对象,通过采集轴承座振动信号,采用AR模型对振动信号进行特征提取和进行功率谱估计,研究了5种充填率条件下的磨机负荷参数与信号时域特征的相关性,得出随着磨矿过程中筒体中钢球、物料的变化,低频段、高频段的频谱能量值曲线的相应变化规律。研究表明,有量纲时域特征参数可以很好地表征特征信号与负荷状态参数的相关性;无量纲时域特征参数可以很好地解释振动时域信号的波形分布以及振动冲击特征,根据频谱能量分布变化规律,可提取能够表征其对应磨机负荷状态的振动特征,为磨机负荷预测提供依据。
[Abstract]:The study of the inner load (filling ratio, material to ball ratio) of the ball mill in the grinding process is an important part of saving energy and reducing the consumption of the mineral processing equipment. Taking the test ball mill as an object, the vibration signals of the bearing seat are collected, and the feature extraction and power spectrum estimation of the vibration signals are carried out by using AR model. The correlation between the load parameters of the mill and the time-domain characteristics of the signals under five kinds of filling rates is studied. With the change of steel ball and material in the cylinder during the grinding process, the corresponding variation law of the spectrum energy value curve in the low frequency band and the high frequency band is obtained. The results show that the dimensionless time domain characteristic parameters can well characterize the correlation between the characteristic signals and the load state parameters, and the dimensionless time domain characteristic parameters can well explain the waveform distribution and vibration shock characteristics of the vibration time domain signals. According to the law of spectrum energy distribution, the vibration characteristics which can represent the load state of mill can be extracted, which can provide the basis for the load forecasting of mill.
【作者单位】：江西理工大学机电工程学院;
【基金】：国家自然基金项目(51464017) 江西省高等学校科技落地计划项目立项(KJLD13045) 江西省研究生创新专项基金项目(YC2015-S284)
【分类号】：TD453

【相似文献】