摘要
本文提出了一种以线性神经网络为基础测量气、固两相流速度的新方法。首先通过对两个传感器之间的流动噪声输送过程进行分析 ,建立两个传感器信号之间的数学模型。并以此为基础通过线性神经网络测量流动噪声的渡越时间 ,进而测出流速。实验结果表明 :此方法可以克服常用的相关估计方法中存在的随机误差较大、分辨率较低的缺点 ,为两相流流速的测量提供了一种有效的手段。
In this paper, a new measurement method of flow velocity in gas solid two phase flow is presented by using linear neural networks. Through analyzing flow noise's delivering process between two sensors, the mathematics model is built between two sensor signals. The flow noise transit time estimation is given on the basis of mathematics model and linear neural networks, then the flow velocity of gas solid two phase flow is measured. The experiment result demonstrated that the method could overcome the drawbacks of the bigger random error and the lower resolution in the cross correlation way. A effective means are presented for measuring the flow velocity.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2000年第4期343-345,共3页
Chinese Journal of Scientific Instrument
基金
辽宁省教育委员会科研项目
关键词
线性神经网络
固体速度测量
气固两相流
Linear neural networks Transit time Flow signal of gas sclid two phase
