Study On A New Method Of Parameters Measurement Of Gas-liquid Two-phase Flow

Two-phase flow widely exists in many industries such as chemical industry, metallurgical industry, energy industry and so on. The research on two-phase flow, which has been one of the frontier subjects concerned deeply in the world, is developing rapidly. However, the inherent complexity of the two-phase mixture causes many difficulties in measuring the parameters of the two-phase flow. In many cases, the measurement technology for two-phase flow has been one of the bottlenecks in the research field of the two-phase flow.Voidage and flowrate are two important parameters of two-phase flow. The dissertation researches the measurement method of the voidage and the flowrate of the oil-gas two-phase flow. The main work of the dissertation includes:1. The basic Genetic Algorithm is improved. Based on the nonlinear scale transform and the Elitist stratage, an improved Genetic Algorithm is proposed. The nonlinear scale transform of the fitness function overcome the local optimum disadvantage of the basic Genetic Algorithm, and the Elitist stratage accelerates the convergence of the Genetic Algorithm. The improved Genetic Algorithm has the advantage of the fast convergence speed and far away from the local optimum. The function optimum and the voidage measurement results of gas-oil two-phase flow both show that the improvement of the Genetic Algorithm is successful. Compared with the basic Genetic Algorithm, the accuracy of the improved Genetic Algorithm improves 15% approximately and the convergence time of the improved Genetic Algorithm improves 25% approximately.2. The Genetic Algorithm is applied to measure the voidage of oil-gas two-phase flow. Based on the Genetic Algorithm. Partial Least Square method and Electrical Capacitance Tomography technique, a new voidage measurement method is proposed. At first, the flow pattern of the oil-gas two-phase flow is identified, and then a relative voidage measurement model is selected to measure the voidage. The flow pattern identificationis achieved by two methods. Both combining the differential pressure fluctuation signal with the signal processing methods and combining the cross-sectional image with the fuzzy pattern identification methods can give a good flow pattern identification result. The voidage measurement model is the linear combination of the capacitance values and is pre-developed before the measurement. The voidage measurement model is developed as following: at first, Genetic Algorithm is used to explore the optimal capacitance combination of which each capacitance give a significant contribution to the voidage measurement, and then Partial Least Square method is further applied to determine the weight coefficient of the contribution of each capacitance to the voidage measurement. Compared with the conventional 2-electrode capacitance method, the proposed voidage measurement method can decrease the influence of the flow pattern on the voidage measurement. Compared with the ECT method, the real-time performance of the proposed voidage measurement method is satisfactory, and the measurement time is less than 0.1 second.3. Combining the voidage measurement method with the Venturi meter, a new flowrate measurement method of oil-gas two-phase flow is proposed. Because the flow pattern identification procedure is considered during the flowrate measurement process, the fluence of the flow pattern on the voidage measurement is reduced. Experimental results show that the proposed flowrate measurement method is successful, and for the different flow patterns, the accuracy of the flowrate measurement is satisfactory. Additionally. Murdock relation and Chisholm relation are suitable for the bubble flow, Separation density model is suitable for the stratified flow, and Murdock relation can give a good measurement result for the slug flow and the churn flow.