基于信息融合技术的磨煤机负荷特征信息提取方法

Information Fusion Technology Based Feature Extraction Method for Ball Mill Load Measurement

磨煤机噪声信号含有能够表征磨煤机负荷的重要信息,是判断磨煤机运行状态的重要信号,但噪声信号各频段对负荷表征的灵敏度不同,仅靠噪声信号能量难以获取准确可靠的负荷变化信息。为此,研究出一种新的磨煤机负荷特征信息提取方法:利用信息融合技术中的多传感器一致性检测原理,将对负荷反应灵敏的多个磨煤机噪声特征频段提取出来,再按照最小二乘方法融合特征频段功率标定值,得到表征磨煤机负荷的特征信息。通过现场采集的磨煤机噪声信号,对所研究的方法进行了仿真验证,结果表明该方法有效可靠。

The noise signals of ball mills contain important information on mill loads and are critical for mill condition determination However. because of the load sensitivity variety of their frequency spectrum,it is difficult to acquire the accurate and reliable information of load change. In this paper, a novel load feature extraction method is proposed. This method adopts the multi sensor consistency check algorithm to extract the mill noise characteristic frequency bandsthat are sensitive to the load response and obtain the mill load characteristic information by following the least square principle to fuse the scale value of characteristic frequency bands. The noise signals are gathered on site and used to take the simulation test on the new approach. The test results prove that the new method is effective and reliable.