Stokes特征值问题的一种基于残差型后验误差估计的新的自适应混合有限元法-计算数学专业论文.docxVIP

II II 万方数据 Abstract The a posterior error estimates and adaptive algorithms of finite element methods are the main streams in the academic circle. From the perspective of methodology, one has concluded that adaptive methods are the optimal discretization methods when using finite element methods to solve differential equations. The a posterior error estimates are the theoretical foundations of adaptive finite element methods. The idea of a posterior error estimates and adaptive finite element methods was first proposed by Babuska and Rheinboldt in 1978, and was great developed so far and thus produced many important approaches of a posterior error estimates such as residual types and recovery types. Par- tial differential equation eigenvalue problems are also hot topics focused on by numerous scholars. In recent years, they attached much importance to adaptive finite element meth- ods for partial differential equation eigenvalue problems. Under the above background, in this paper we aim to discuss a new adaptive mixed finite element algorithm for the Stokes eigenvalue problem. In this paper, we combine mixed finite element method, multi-scale discretization and Rayleigh quotient iteration to propose a new adaptive algorithm based on residual type a posterior error estimates for the Stokes eigenvalue problem. Both reliability and efficiency of the error indicator are proved. In addition, we combine higher order finite elements and the shifted inverse power method to propose a modified algorithm. The efficiency of both algorithms are also investigated by using Chen’s iFEM package [28]. Numerical results are satisfying. Keywords: Stokes eigenvalue problem, mixed finite element, Rayleigh quotient iteration, a posterior error estimates, adaptive algorithm. PAGE 10 PAGE 10 万方数据 Introduction Adaptive methods and residual type a posterior estimates, which were first proposed by Babuska and Rheinboldt [1, 2], have gained an enormous attention in scientific engineer- ing co