Figure 1.
The longest orbits we are able to validate, with a total number of coefficient of approximately 14000. In blue for $p = 1$ , in green for $p = 2$ and in red for $p = 3$ . The initial value is given by (22)
-
Previous Article
Underlying one-step methods and nonautonomous stability of general linear methods
- DCDS-B Home
- This Issue
-
Next Article
Partitioned second order method for magnetohydrodynamics in Elsässer variables
September
2018, 23(7): 2825-2858.
doi: 10.3934/dcdsb.2018164
Polynomial interpolation and a priori bootstrap for computer-assisted proofs in nonlinear ODEs
| 1. | Technical University of Munich, Faculty of Mathematics, 85748 Garching b. München, Germany |
| 2. | Département de Mathématiques et de Statistique, Université Laval, 1045 avenue de la Médecine, Québec, QC, G1V 0A6, Canada |
| 3. | McGill University, Department of Mathematics and Statistics, 805 Sherbrooke St W, Montreal, QC, H3A 0B9, Canada |
In this work, we introduce a method based on piecewise polynomial interpolation to enclose rigorously solutions of nonlinear ODEs. Using a technique which we call a priori bootstrap, we transform the problem of solving the ODE into one of looking for a fixed point of a high order smoothing Picard-like operator. We then develop a rigorous computational method based on a Newton-Kantorovich type argument (the radii polynomial approach) to prove existence of a fixed point of the Picard-like operator. We present all necessary estimates in full generality and for any nonlinearities. With our approach, we study two systems of nonlinear equations: the Lorenz system and the ABC flow. For the Lorenz system, we solve Cauchy problems and prove existence of periodic and connecting orbits at the classical parameters, and for ABC flows, we prove existence of ballistic spiral orbits.
Keywords:
Computer-assisted proof,
nonlinear ODEs,
Picard-like operator,
piecewise polynomial interpolation,
ABC flows.
Mathematics Subject Classification:
Primary: 65G20, 65P99, 65D30, 37M99, 37C27.
Citation:
Maxime Breden, Jean-Philippe Lessard. Polynomial interpolation and a priori bootstrap for computer-assisted proofs in nonlinear ODEs. Discrete & Continuous Dynamical Systems - B,
2018, 23
(7)
: 2825-2858.
doi: 10.3934/dcdsb.2018164
![]()
References:
show all references
References:
Figure 2.
A validated periodic orbit of the Lorenz system, whose period $\tau$ is approximately $11.9973$ . We used two iterations of a priori bootstrap, that is $p = 3$ , for the validation. If we want to minimize the total number of coefficients to do the validation, we can take $k = 3$ and $m = 602$ (which makes $7225$ coefficients in total), and we then get a validation radius of $1.5627\times 10^{-4}$ . It is possible to get a significantly lower validation radius, at the expense of a slight increase in the total number of coefficients: for instance with $k = 5$ and $m = 495$ (which makes $8911$ coefficients in total), we get a validation radius of $4.7936\times 10^{-9}$
Figure 3.
Validated connecting orbit for the Lorenz system, with parameters $(\sigma, \beta, \rho) = (10, \frac{8}{3}, 28)$ . The local stable manifold of the origin is in blue, the local unstable manifold of $\left(\sqrt{\beta(\rho-1)}, \sqrt{\beta(\rho-1)}, \rho-1\right)$ in yellow, and the green connection between them (of length $\tau\simeq 17.3$ ) is validated using polynomial interpolation, with a priori bootstrap ($p = 3$ ). The proof gives a validation radius of $r = 3.1340\times 10^{-5}$
Figure 4.
These are the orbits that are described in Theorem 7.1. The color varies from blue for $A = 1$ to red for $A = 0.1$ . Each proof was done with $p = 2$ , $k = 2$ and $m = 50$ , and gave a validation radius varying from $r = 4.8313\times 10^{-8}$ to $r = 7.4012\times 10^{-6}$
Figure 5.
This is the orbit that is described in Theorem 7.2. The proof was done with $p = 2$ , $k = 2$ and $m = 300$ , and gave a validation radius $r = 4.0458\times 10^{-6}$
Figure 6.
An example of the situation described just above, with $k = 11$ and $i_0 = 9$ . The black squares represent the other Chebyshev points $x_{i}^k$
Table 1.
Comparisons for $p = 1$ . $\tau_{max}$ is the longest integration time for which the proof succeeds, and $r$ is the associated validation radius, that is a bound of the distance (in $\mathcal{C}^0$ norm) between the numerical data used and the true solution
| | |
|||
| 1 | 2333 | 13998 | ||
| 2 | 1556 | 14004 | 0.64 | |
| 3 | 1167 | 14004 | 0.58 |
| | |
|||
| 1 | 2333 | 13998 | ||
| 2 | 1556 | 14004 | 0.64 | |
| 3 | 1167 | 14004 | 0.58 |
Table 2.
Comparisons for $p = 2$ . $\tau_{max}$ is the longest integration time for which the proof succeeds, and $r$ is the associated validation radius, that is a bound of the distance (in $\mathcal{C}^0$ norm) between the numerical data used and the true solution
| | |
|||
| 1 | 2333 | 13998 | 0.97 | |
| 2 | 1556 | 14004 | ||
| 3 | 1167 | 14004 | 5.5 | |
| 4 | 933 | 13995 | 4.9 |
| | |
|||
| 1 | 2333 | 13998 | 0.97 | |
| 2 | 1556 | 14004 | ||
| 3 | 1167 | 14004 | 5.5 | |
| 4 | 933 | 13995 | 4.9 |
Table 3.
Comparisons for $p = 3$ . $\tau_{max}$ is the longest integration time for which the proof succeeds, and $r$ is the associated validation radius, that is a bound of the distance (in $\mathcal{C}^0$ norm) between the numerical data used and the true solution
| | |
|||
| 2 | 1556 | 14004 | 5.6 | |
| 3 | 1167 | 14004 | ||
| 4 | 933 | 13995 | ||
| 5 | 778 | 14004 | 8.0 | |
| 9 | 467 | 14010 | 7.9 | |
| 19 | 233 | 13980 | 6.9 |
| | |
|||
| 2 | 1556 | 14004 | 5.6 | |
| 3 | 1167 | 14004 | ||
| 4 | 933 | 13995 | ||
| 5 | 778 | 14004 | 8.0 | |
| 9 | 467 | 14010 | 7.9 | |
| 19 | 233 | 13980 | 6.9 |
Table 4.
Minimal number of coefficients needed to validate the orbit of length $\tau = 2$ , starting from $u_0$ given in (22), and the associated validation radius provided by the proof. We repeat that in this example, we put the emphasis on getting an existence result with a minimal amount of computational power, which of course results in less sharp validation radius, especially in situations where the $Y$ bound is the limiting factor
| |
||||
| no proof | ||||
| no proof | ||||
| no proof | ||||
| |
||||
| no proof | ||||
| no proof | ||||
| no proof | ||||
Table 5.
The intervals $[\tau^-_A, \tau^+_A]$ , for $A = 0.1, 0.2, \ldots, 1$ in which the period $\tau_A$ of the solution described in Theorem 7.1 is proved to be
| |
|||||||||||
| |
|||||||||||
| [1] |
A. Aschwanden, A. Schulze-Halberg, D. Stoffer. Stable periodic solutions for delay equations with positive feedback - a computer-assisted proof. Discrete & Continuous Dynamical Systems, 2006, 14 (4) : 721-736. doi: 10.3934/dcds.2006.14.721 |
| [2] |
Chiara Caracciolo, Ugo Locatelli. Computer-assisted estimates for Birkhoff normal forms. Journal of Computational Dynamics, 2020, 7 (2) : 425-460. doi: 10.3934/jcd.2020017 |
| [3] |
István Balázs, Jan Bouwe van den Berg, Julien Courtois, János Dudás, Jean-Philippe Lessard, Anett Vörös-Kiss, JF Williams, Xi Yuan Yin. Computer-assisted proofs for radially symmetric solutions of PDEs. Journal of Computational Dynamics, 2018, 5 (1&2) : 61-80. doi: 10.3934/jcd.2018003 |
| [4] |
Thomas Wanner. Computer-assisted equilibrium validation for the diblock copolymer model. Discrete & Continuous Dynamical Systems, 2017, 37 (2) : 1075-1107. doi: 10.3934/dcds.2017045 |
| [5] |
Maciej J. Capiński, Emmanuel Fleurantin, J. D. Mireles James. Computer assisted proofs of two-dimensional attracting invariant tori for ODEs. Discrete & Continuous Dynamical Systems, 2020, 40 (12) : 6681-6707. doi: 10.3934/dcds.2020162 |
| [6] |
Piotr Zgliczyński. Steady state bifurcations for the Kuramoto-Sivashinsky equation: A computer assisted proof. Journal of Computational Dynamics, 2015, 2 (1) : 95-142. doi: 10.3934/jcd.2015.2.95 |
| [7] |
Flaviano Battelli, Michal Fečkan. Blue sky-like catastrophe for reversible nonlinear implicit ODEs. Discrete & Continuous Dynamical Systems - S, 2016, 9 (4) : 895-922. doi: 10.3934/dcdss.2016034 |
| [8] |
Primitivo B. Acosta-Humánez, J. Tomás Lázaro, Juan J. Morales-Ruiz, Chara Pantazi. On the integrability of polynomial vector fields in the plane by means of Picard-Vessiot theory. Discrete & Continuous Dynamical Systems, 2015, 35 (5) : 1767-1800. doi: 10.3934/dcds.2015.35.1767 |
| [9] |
Jan Bouwe van den Berg, Elena Queirolo. A general framework for validated continuation of periodic orbits in systems of polynomial ODEs. Journal of Computational Dynamics, 2021, 8 (1) : 59-97. doi: 10.3934/jcd.2021004 |
| [10] |
Craig Cowan. Supercritical elliptic problems involving a Cordes like operator. Discrete & Continuous Dynamical Systems, 2021, 41 (9) : 4297-4318. doi: 10.3934/dcds.2021037 |
| [11] |
Daniel Visscher. A new proof of Franks' lemma for geodesic flows. Discrete & Continuous Dynamical Systems, 2014, 34 (11) : 4875-4895. doi: 10.3934/dcds.2014.34.4875 |
| [12] |
Fang Qin, Ying Jiang, Ping Gu. Three-dimensional computer simulation of twill woven fabric by using polynomial mathematical model. Discrete & Continuous Dynamical Systems - S, 2019, 12 (4&5) : 1167-1178. doi: 10.3934/dcdss.2019080 |
| [13] |
V. Rehbock, K.L. Teo, L.S. Jennings. Suboptimal feedback control for a class of nonlinear systems using spline interpolation. Discrete & Continuous Dynamical Systems, 1995, 1 (2) : 223-236. doi: 10.3934/dcds.1995.1.223 |
| [14] |
Lucio Boccardo, Daniela Giachetti. A nonlinear interpolation result with application to the summability of minima of some integral functionals. Discrete & Continuous Dynamical Systems - B, 2009, 11 (1) : 31-42. doi: 10.3934/dcdsb.2009.11.31 |
| [15] |
Jaume Llibre, Yilei Tang. Limit cycles of discontinuous piecewise quadratic and cubic polynomial perturbations of a linear center. Discrete & Continuous Dynamical Systems - B, 2019, 24 (4) : 1769-1784. doi: 10.3934/dcdsb.2018236 |
| [16] |
Hebai Chen, Jaume Llibre, Yilei Tang. Centers of discontinuous piecewise smooth quasi–homogeneous polynomial differential systems. Discrete & Continuous Dynamical Systems - B, 2019, 24 (12) : 6495-6509. doi: 10.3934/dcdsb.2019150 |
| [17] |
Benoît Pausader, Walter A. Strauss. Analyticity of the nonlinear scattering operator. Discrete & Continuous Dynamical Systems, 2009, 25 (2) : 617-626. doi: 10.3934/dcds.2009.25.617 |
| [18] |
Hsin-Yuan Huang. Schubart-like orbits in the Newtonian collinear four-body problem: A variational proof. Discrete & Continuous Dynamical Systems, 2012, 32 (5) : 1763-1774. doi: 10.3934/dcds.2012.32.1763 |
| [19] |
Wenmin Sun, Jiguang Bao. New maximum principles for fully nonlinear ODEs of second order. Discrete & Continuous Dynamical Systems, 2007, 19 (4) : 813-823. doi: 10.3934/dcds.2007.19.813 |
| [20] |
Christoph Bandt, Helena PeÑa. Polynomial approximation of self-similar measures and the spectrum of the transfer operator. Discrete & Continuous Dynamical Systems, 2017, 37 (9) : 4611-4623. doi: 10.3934/dcds.2017198 |
2019 Impact Factor: 1.27
Tools
Metrics
Other articles
by authors
[Back to Top]