空间滤波、小波变换和神经网络在固体速度测量中的应用

The Application of Spatial Filter, Wavelet Transform and Neural Networks to Solid Velocity Measurement

介绍一种利用空间滤波、小波变换和神经网络处理两相流信号的基本方法,首先研究电容传感器的空间滤波效应,并找出固体速度和带宽之间的关系,然后利用小波变换的滤波特性,确定传感器的带宽;通过神经网络确定固体速度和带宽之间的系数。进而求出固体速度。最后给出对实际测量信号的处理结果。

The essential method for processing two phase flow signals based on spatial filter, wavelet transform and neural networks is presented. Firstly, spatial filtering effects of capacitance sensors are studied, and the relationship in sloid velocity and band width is found. Secondly, the capacitance sensor width of pass-band is gained by using the characteristics of wavelet transform filtering. Thirdly, the coefficient in solid velocity and bandwidth is obtained by neural networks, furthermore solid velocity is solved. Finally, The processing results of practical measurement signals are given.