Research Of Wet Gas Measuring Arithmetic Based On Time-varying Series Analysis

Based on a self developed prototype of wet gas meter and experimental data set, a series of works have been carried out. In this thesis, the time-varying series analysis technique was introduced to get the individual flow rate, realizing the online measuring of two-phase flow rate.Analysis the relationship of differential pressure signal which was acquired by the wet gas flow meter and the individual flow rate of two-phase flow. Firstly, the differential pressure signal was pretreated through wavelet transform modulus maximum method, so as to prepare for extracting different flow vectors in different operating conditions. Secondly, introduce the time-varying series models to deal with the differential pressure signal, confirm the data length, basis function, introduce the maximum likelihood estimate to catch the model orders, and impose the recursive least-squares method to catch the parameter. Thirdly, correlation between model parameters and the individual flow rate was investigated by contour map, get the relative eigenvalues. Finally, using the eigenvalues as the input and gas liquid flow rate as output of neuron network, comparing the performance of the BP and RBF network, use the RBF network whose stability is relatively strong to predict the individual flow rate, the relative error are within 10% and 15%.The results show the consanguineous relationship between time-varying AR model parameters and gas liquid flow rate, all the works are potentially useful for the individual flow rate metering of gas-liquid two-phase flow.