Abstract:In this paper we discuss the efficient implementation of RKN-type Fourier collocation methods, which are used when solving second-order differential equations. The proposed implementation relies on an alternative formulation of the methods and the blended formulation. The features and effectiveness of the implementation are confirmed by the performance of the methods on two numerical tests.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to improve the numerical implementation efficiency of the RKN - type Fourier collocation methods for solving second - order differential equations. Specifically, the author proposes an efficient iterative procedure based on these methods and verifies its effectiveness and superiority through numerical tests.
### Background and Problem Description of the Paper
In the past few decades, much research has been devoted to improving the numerical efficiency of implicit methods for solving differential equations. This paper focuses particularly on the numerical solutions of second - order differential equations, that is, equations of the form:
\[ q''(t) = f(q(t)), \quad q(t_0) = q_0, \quad q'(t_0) = q'_0, \quad t \in [t_0, t_{\text{end}}], \]
where \( q(t): \mathbb{R} \to \mathbb{R}^d \) and \( f(q): \mathbb{R}^d \to \mathbb{R}^d \) are analytic functions. The solutions of this type of equations and their derivatives satisfy the formula of variation of constants:
\[
\begin{aligned}
q(t + \mu h) &= q(t) + \mu h q'(t) + h^2 \int_0^\mu (\mu - x) f(q(t + hx)) \, dx, \\
q'(t + \mu h) &= q'(t) + h \int_0^\mu f(q(t + hx)) \, dx.
\end{aligned}
\]
The RKN - type Fourier collocation method is an implicit method for solving the above - mentioned second - order systems. Since implicit methods require iterative solutions, efficient implementation is particularly important. This paper aims to propose an iterative procedure based on a specific structure to efficiently implement these methods.
### Main Contributions
1. **Proposed an efficient iterative procedure**: By reformulating the RKN - type Fourier collocation method and using a mixed formula, an efficient iterative process is proposed.
2. **Numerical tests verified the effectiveness**: The effectiveness and robustness of the proposed method are demonstrated through two numerical tests.
3. **Analyzed the convergence and stability of the method**: The properties such as the convergence, stability, and accuracy of maintaining solutions of the method are studied.
### Numerical Test Results
The article verifies the effectiveness of the proposed method through two specific numerical examples:
- **Perturbed Kepler's Problem**: The CPU time, number of iterations, solution error, Hamiltonian error, and angular momentum error of three methods (RKN - TFC - B, RKN - TFC - F, DIRKN - F) are compared under different step sizes.
- **Hénon - Heiles Model**: Comparisons are also made under different step sizes, showing the advantages of the RKN - TFC - B method in terms of CPU time and error.
### Conclusion
Through the above research, the author concludes that the hybrid implementation of the proposed RKN - type Fourier collocation method is more robust and efficient than the same method and the RKN method using fixed - point iteration.
In addition, the author also points out future research directions, including exploring new splitting techniques, studying the efficient implementation of the trigonometric Fourier collocation method, and considering the influence of different orthogonal bases on the method.