Date of Award

8-2024

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

School

Mathematics and Natural Sciences

Committee Chair

Dr. Huiqing Zhu

Committee Chair School

Mathematics and Natural Sciences

Committee Member 2

Dr. Haiyan Tian

Committee Member 2 School

Mathematics and Natural Sciences

Committee Member 3

Dr. C.S. Chen

Committee Member 3 School

Mathematics and Natural Sciences

Committee Member 4

Dr. James Lambers

Committee Member 4 School

Mathematics and Natural Sciences

Abstract

A localized Hermite method of approximate particular solutions (LHMAPS) is presented in this dissertation. This method is designed to improve the accuracy of the localized method of approximate particular solutions (LMAPS) for solving partial differential equations. LMAPS is a strong-form method that defines local radial basis function approximations to the solution values at a group of collocation points before applying the differential operator to this approximation function. Based on the LMAPS's local scheme, LHMAPS seeks Hermite-type local approximations based on both function values and derivatives. Numerical experiments validate the superior accuracy of the proposed method to the LMAPS by solving the 2D Poisson equation, Helmholtz equation, and convection-reaction-diffusion equation. Also, when the convection term is dominating in the convection-reaction-diffusion problem, we employ the up-winding technique to control the oscillatory nature of the numerical solution. Again, numerical results show that the proposed method has better accuracy than LMAPS and RBF-HFD.

Available for download on Saturday, February 01, 2025

Share

COinS