The Velocity Measurement Of Two-Phase Flow Based On Empirical Mode Decomposition And Single-channel Blind Source Separation

With the rapid development of science and technology, two-phase flow system is more and more important in the industrial production and human life. It is widely used in petroleum, chemical engineering, metallurgical, electric power, energy, food and some other fields, the effect and status of two phase flow system have become increasingly outstanding. Because of the advancement of demand of measurement, energy conservation and control in industrial production process, the study on the parameters measurement of the two-phase flow is needed more urgent. Scientists have paid a lot of attention on two-phase flow because of the university of its existence and the importance of its application. But now, there is no perfect solution to the problem of the two-phase flow parameters measurement, the velocity of two-phase flow is one of the most important parameters, so the accurate measurement of velocity is very necessary. So this paper proposed two methods to measure two-phase flow velocity using empirical mode decomposition and single-channel blind signal separation.This paper studied about the measurement of two-phase flow velocity, the main work of this paper is as follows:1. On the base of reading mass home and abroad references, the research progress, basic principles and application of the single-channel blind source separation are summarized in this paper. The main testing parameters of the two-phase flow are introduced. The basic principles of measuring methods for two-phase flow velocity are described. The current measurement methods of two-phase flow velocity are analyzed and compared, and the problems of them are indicated.2. In order to solve the blind signal separation (BSS) of the number of sensor observations less than the number of source signals, a single-channel method using empirical mode decomposition (EMD) and fast independent component analysis (FastICA) is proposed. The simulation experiment of single-channel blind signal separation gives the similar coefficients of separation signals obtained by this method and source signals are higher than 0.98.3. The measurement of two-phase flow velocity can be implemented by EMD and spatial filtering effect of capacitance sensors. In order to overcome the disadvantage of the larger error in the measurement of bandwidth when measure the velocity of two-phase flow using the method of spatial filtering, this paper presents a method to measure two-phase flow velocity using EMD and spatial filtering. The relative errors of the velocity obtained by this method are less than 1.2%.4. In order to overcome the interference of noise, single-channel blind source separation is used with it. With its application in de-noising, it can restrain the noise from disturbing the useful signal, meanwhile the basic feature of the two-phase flow signal can be maintained. More accurate transit time can obtained from correlation function curves of two-phase signal after processed. Then two-phase flow velocity can be calculated. Simulation result shows that the maximum relative error of this method is 4.32%.In this paper, the single-channel blind source separation is given by EMD and FastICA. Two methods of the measurement of two-phase flow velocity are proposed, simulation results verified the effectiveness and feasibility of this method, and the measurement accuracy of two-phase flow velocity can be improved.