基于静电传感器阵列和位置指纹的绳状流定位方法研究

Positioning Method Study of Roping Flow Based on Electrostatic Sensor Arrays and Location Fingerprint

由于气固两相流动的复杂性,燃煤电厂一次风管道内煤粉颗粒在管道内分布非常不均匀,尤其在绳状流情况下,颗粒聚集成束并在管道内剧烈摆动,为了实现绳状流质量流量的测量,可先对绳状流进行定位。提出基于静电传感器阵列和位置指纹的绳状流定位方法,通过有限元仿真获得方形管道四壁上静电电极的空间灵敏度,以电极对的相对空间灵敏度为特征、位置坐标为标签构建位置指纹数据库,并使用支持向量回归算法建立定位模型。在重力输送实验台上模拟绳状流进行实验验证,结果表明:在管道处的定位误差最大,当信号时间长度为0.2 s时最大定位误差为4.11 mm。所提出的定位方法有利于消除绳状流位置变化引起的质量流量测量误差。

Due to the complexity of the gas-solid two-phase flow,the distribution of the coal particles within the primary pipes in coal-fired power plants is highly inhomogeneous.Under circumstances of roping flow regime,the particles are highly concentrated into a small region,the position of which changes rapidly.In order to measure the mass flowrate accurately,it is desirable to determine the position of the roping flow.This paper proposed a positioning method of the roping flow using an electrostatic sensor array and the location fingerprint method.The spatial sensitivity of the four electrodes on the walls of the square-shaped pipe was obtained through finite element simulation.A location fingerprint database was constructed using the relative spatial sensitivity of electrode pairs as the feature and the position coordinates as the label.The support vector regression algorithm was used to build a positioning model.The positioning method was experimentally validated using a gravity conveying rig that creates a simulated roping flow.Experimental results show that the positioning error is the largest around the corners of the pipe,which is 4.11 mm when the signal duration is 0.2 s.The positioning error is the largest around the corners of the pipe.The positioning method proposed in this paper can help to reduce the measurement error of the mass flowrate caused by the position variation of the roping flow.