Multiphase Flow Detection Based On CFD Information And Data Dimension Reduction Method

Multiphase flow and its transfer process exist widely in all aspects of production and life,involving many basic scientific issues such as flow,heat transfer,mass transfer,chemical and biological reactions.Multiphase flow detection technology is of great significance to the development of national economy or national defense science and technology,human health,ecological or environmental changes,protection,sustainable development and utilization.However,the fundamental natural law of multiphase flow and its mathematical description method have not been grasped by people.With the development of science and technology,a series of multiphase flow detection technologies have been developed,but due to the complexity of the multiphase flow process,various methods have certain defects.Since the 1960s,Computational Fluid Dynamics(CFD)has calculated numerical solutions of fluid flow through discrete methods.With the development of computer technology,it has gradually become one of the mature and powerful methods for studying fluid flow,especially complex fluid problems such as multiphase turbulent flow.In this paper,a new multiphase flow detection method is proposed by using the existing mature CFD software ANSYS Fluent and combining with the mathematical algorithm.Therefore,the main content of this paper includes the introduction of multiphase flow and CFD basic principle,and puts forward a new process tomography(PT)method based on computational fluid dynamics(CFD)and matrix decomposition for flow field reconstruction.The CFD sample database is first established by CFD simulation in this method,and the sample base matrix is extracted from the sample database using matrix factorization method including Singular value decomposition(SVD)and Non-negative matrix factorization(NMF).Then,the basis matrix can be used to inversely solve the sample matrix with a small amount of sensor measurement data to achieve quick reconstruction of the flow field,thereby reducing the sampling points required for reconstruction and achieving the dimension reduction reconstruction.From the reconstruction imaging results,it can be seen that for different flow field process parameters,the mathematical model proposed in this paper has a good reconstruction effect,avoids the calculation of complex multiphase flow sensitive fields,effectively improve the speed,and can be competent for the reconstruction of complex multiphase flow field in different situations.The main innovation of this paper is that it combines CFD numerical simulation and sparse sensor measurement,realizes the reconstruction of flow field distribution image by using SVD,NMF and dimension reduction algorithm,overcomes the inherent shortcomings of some multiphase flow detection methods,such as the calculation of complex sensitive field of electrical capacitance tomography(ECT),the detection difficulty of large-scale and complex space multiphase flow,and provides a new way for multiphase flow detection.