Abstract
In this paper, we present a framework to construct general stochastic Runge–Kutta Lawson schemes. We prove that the schemes inherit the consistency and convergence properties of the underlying Runge–Kutta scheme, and confirm this in some numerical experiments. We also investigate the stability properties of the methods and show for some examples, that the new schemes have improved stability properties compared to the underlying schemes.
Similar content being viewed by others
Notes
The vectorization of a \(d\times d\) matrix \(A=\{a_{i,j}\}_{i,j=1}^d\) is the \(d^2\)-dimensional vector given by \(\text {vec}(A)=(a_{1,1},a_{2,1},\dots ,a_{d,d-1},a_{d,d})^\top \).
References
Arara, A.A., Debrabant, K., Kværnø, A.: Stochastic B-series and order conditions for exponential integrators. In: Numerical Mathematics and Advanced Applications, Lecture Notes in Computational Science and Engineering, pp. 419–427. Springer (2019). https://doi.org/10.1007/978-3-319-96415-7_37
Arnold, L.: Stochastic differential equations: theory and applications. Wiley-Interscience [John Wiley & Sons], New York-London-Sydney (1974). Translated from the German
Biscay, R., Jimenez, J.C., Riera, J.J., Valdes, P.A.: Local linearization method for the numerical solution of stochastic differential equations. Ann. Inst. Statist. Math. 48(4), 631–644 (1996). https://doi.org/10.1007/BF00052324
Buckwar, E., Kelly, C.: Towards a systematic linear stability analysis of numerical methods for systems of stochastic differential equations. SIAM J. Numer. Anal. 48(1), 298–321 (2010). https://doi.org/10.1137/090771843
Buckwar, E., Kelly, C.: Non-normal drift structures and linear stability analysis of numerical methods for systems of stochastic differential equations. Comput. Math. Appl. 64(7), 2282–2293 (2012). https://doi.org/10.1016/j.camwa.2012.02.059
Buckwar, E., Sickenberger, T.: A comparative linear mean-square stability analysis of Maruyama- and Milstein-type methods. Math. Comput. Simul. 81(6), 1110–1127 (2011). https://doi.org/10.1016/j.matcom.2010.09.015
Buckwar, E., Sickenberger, T.: A structural analysis of asymptotic mean-square stability for multi-dimensional linear stochastic differential systems. Appl. Numer. Math. 62(7), 842–859 (2012). https://doi.org/10.1016/j.apnum.2012.03.002
Burrage, K., Burrage, P.M.: High strong order explicit Runge-Kutta methods for stochastic ordinary differential equations. Appl. Numer. Math. 22(1–3), 81–101 (1996). https://doi.org/10.1016/S0168-9274(96)00027-X. (Special issue celebrating the centenary of Runge-Kutta methods)
Burrage, K., Burrage, P.M.: Order conditions of stochastic Runge-Kutta methods by \(B\)-series. SIAM J. Numer. Anal. 38(5), 1626–1646 (2000). https://doi.org/10.1137/S0036142999363206
Carbonell, F., Jimenez, J.C.: Weak local linear discretizations for stochastic differential equations with jumps. J. Appl. Probab. 45(1), 201–210 (2008). https://doi.org/10.1239/jap/1208358962
Cohen, D.: On the numerical discretisation of stochastic oscillators. Math. Comput. Simul. 82(8), 1478–1495 (2012). https://doi.org/10.1016/j.matcom.2012.02.004
Debrabant, K., Kværnø, A.: B-series analysis of stochastic Runge-Kutta methods that use an iterative scheme to compute their internal stage values. SIAM J. Numer. Anal. 47(1), 181–203 (2008/09). https://doi.org/10.1137/070704307
Debrabant, K., Kværnø, A., Mattsson, N.C.: Lawson schemes for highly oscillatory stochastic differential equations and conservation of invariants. Preprint (2019). arXiv:1909.12287
Debrabant, K., Kværnø, A., Mattsson, N.C.: Matlab code: Runge-Kutta Lawson schemes for stochastic differential equations (2020). https://doi.org/10.5281/zenodo.4062482
Erdoğan, U., Lord, G.J.: A new class of exponential integrators for SDEs with multiplicative noise. IMA J. Numer. Anal. 39(2), 820–846 (2019). https://doi.org/10.1093/imanum/dry008
Higham, D.J.: Mean-square and asymptotic stability of the stochastic theta method. SIAM J. Numer. Anal. 38(3), 753–769 (2000). https://doi.org/10.1137/S003614299834736X
Jimenez, J.C., Shoji, I., Ozaki, T.: Simulation of stochastic differential equations through the local linearization method. A comparative study. J. Stat. Phys. 94(3–4), 587–602 (1999). https://doi.org/10.1023/A:1004504506041
Kloeden, P.E., Platen, E.: Numerical solution of stochastic differential equations. In: Applications of Mathematics, vol. 23, 2nd edn. Springer-Verlag, Berlin (1999). https://doi.org/10.1007/978-3-662-12616-5
Komori, Y., Burrage, K.: A stochastic exponential Euler scheme for simulation of stiff biochemical reaction systems. BIT 54(4), 1067–1085 (2014). https://doi.org/10.1007/s10543-014-0485-1
Komori, Y., Cohen, D., Burrage, K.: Weak second order explicit exponential Runge–Kutta methods for stochastic differential equations. SIAM J. Sci. Comput. 39(6), A2857–A2878 (2017). https://doi.org/10.1137/15M1041341
Komori, Y., Mitsui, T., Sugiura, H.: Rooted tree analysis of the order conditions of ROW-type scheme for stochastic differential equations. BIT 37(1), 43–66 (1997). https://doi.org/10.1007/BF02510172
Lawson, J.D.: Generalized Runge-Kutta processes for stable systems with large Lipschitz constants. SIAM J. Numer. Anal. 4, 372–380 (1967). https://doi.org/10.1137/0704033
Milstein, G.N.: Numerical integration of stochastic differential equations, Mathematics and its Applications, vol. 313. Kluwer Academic Publishers Group, Dordrecht (1995). Translated and revised from the 1988 Russian original https://doi.org/10.1007/978-94-015-8455-5
Milstein, G.N., Repin, Y.M., Tretyakov, M.V.: Numerical methods for stochastic systems preserving symplectic structure. SIAM J. Numer. Anal. 40(4), 1583–1604 (2002). https://doi.org/10.1137/S0036142901395588
Rößler, A.: Rooted tree analysis for order conditions of stochastic Runge–Kutta methods for the weak approximation of stochastic differential equations. Stoch. Anal. Appl. 24(1), 97–134 (2006). https://doi.org/10.1080/07362990500397699
Rößler, A.: Second order Runge–Kutta methods for Stratonovich stochastic differential equations. BIT 47(3), 657–680 (2007). https://doi.org/10.1007/s10543-007-0130-3
Rößler, A.: Second order Runge–Kutta methods for Itô stochastic differential equations. SIAM J. Numer. Anal. 47(3), 1713–1738 (2009). https://doi.org/10.1137/060673308
Rößler, A.: Strong and weak approximation methods for stochastic differential equations—some recent developments. In: Recent developments in applied probability and statistics, pp. 127–153. Physica, Heidelberg (2010). https://doi.org/10.1007/978-3-7908-2598-5_6
Saito, Y., Mitsui, T.: Stability analysis of numerical schemes for stochastic differential equations. SIAM J. Numer. Anal. 33(6), 2254–2267 (1996). https://doi.org/10.1137/S0036142992228409
Shampine, L.F., Reichelt, M.W.: The MATLAB ODE suite. SIAM J. Sci. Comput. 18(1), 1–22 (1997). https://doi.org/10.1137/S1064827594276424. Dedicated to C. William Gear on the occasion of his 60th birthday
Acknowledgements
Nicky Cordua Mattsson would like to thank the SDU e-Science centre for partially funding his PhD and the Department of Mathematics at the Norwegian University of Science and Technology for kindly hosting him during his visit. The authors would like to thank two anonymous reviewers for very detailed reading and the resulting many helpful comments.
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by David Cohen.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix
Appendix
1.1 Platen strong order 1.5 SL scheme
Writing the explicit order 1.5 strong scheme by Platen for \(M=1\) [18, Eq. 11.2.1] in the form (2.9), we see that the coefficients \(z_i^{m,n}\) of the scheme are given by
\(m=0\) | \(m=1\) | |
---|---|---|
\(i=1\) | \(z^{0,n}_1 = \frac{1}{2}h\) | \(z^{1,n}_1 = \Delta W^n - \frac{1}{h}[\Delta W^n h-\Delta Z^n]\) |
\(i=2\) | \(z^{0,n}_2 = \frac{1}{2\sqrt{h}}\Delta Z^n + \frac{1}{4}h\) | \(z^{1,n}_2 = \frac{1}{h}\left[ \frac{1}{2}\Delta W^n h-\Delta Z^n - \frac{1}{4} \{\frac{1}{3}(\Delta W^n)^2 - h\}\Delta W^n\right] \) |
\(i=3\) | \(z^{0,n}_3 =-\frac{1}{2\sqrt{h}}\Delta Z^n + \frac{1}{4}h\) | \(z^{1,n}_3 = \frac{1}{h}\left[ \frac{1}{2}\Delta W^n h-\Delta Z^n + \frac{1}{4} \{\frac{1}{3}(\Delta W^n)^2 - h\}\Delta W^n\right] \) |
\(i=4\) | \(z^{0,n}_4 = 0\) | \(z^{1,n}_4 = \frac{1}{4h} \left[ \frac{1}{3}(\Delta W^n)^2 - h\right] \Delta W^n\) |
\(i=5\) | \(z^{0,n}_5 = 0\) | \(z^{1,n}_4 = -\frac{1}{4h} \left[ \frac{1}{3}(\Delta W^n)^2 - h\right] \Delta W^n\) |
where
with \(U ^n\sim \mathcal {N}(0,1)\). Similarly, the coefficients of \(Z_{ij}^{m,n}\) are given by
\(j=1\) | \(j=2\) | \(j=3\) | \(j=4\) | \(j=5\) | |
---|---|---|---|---|---|
\(i=1\) | |||||
\(m=0\) | \(Z^{0,n}_{1,1} = 0\), | \(Z^{0,n}_{1,2} = 0\), | \(Z^{0,n}_{1,3} = 0\), | \(Z^{0,n}_{1,4} = 0\), | \(Z^{0,n}_{1,5} = 0\) |
\(m=1\) | \(Z^{1,n}_{1,1} = 0\), | \(Z^{1,n}_{1,2} = 0\), | \(Z^{1,n}_{1,3} = 0\), | \(Z^{1,n}_{1,4} = 0\), | \(Z^{1,n}_{1,5} = 0\) |
\(i=2\) | |||||
\(m=0\) | \(Z^{0,n}_{2,1} = h\), | \(Z^{0,n}_{2,2} = 0\), | \(Z^{0,n}_{2,3} = 0\), | \(Z^{0,n}_{2,4} = 0\), | \(Z^{0,n}_{2,5} = 0\) |
\(m=1\) | \(Z^{1,n}_{2,1} = \sqrt{h}\), | \(Z^{1,n}_{2,2} = 0\), | \(Z^{1,n}_{2,3} = 0\), | \(Z^{1,n}_{2,4} = 0\), | \(Z^{1,n}_{2,5} = 0\) |
\(i=3\) | |||||
\(m=0\) | \(Z^{0,n}_{1,1} = h\), | \(Z^{0,n}_{1,2} = 0\), | \(Z^{0,n}_{1,3} = 0\), | \(Z^{0,n}_{1,4} = 0\), | \(Z^{0,n}_{1,5} = 0\) |
\(m=1\) | \(Z^{1,n}_{1,1} = -\sqrt{h}\), | \(Z^{1,n}_{1,2} = 0\), | \(Z^{1,n}_{1,3} = 0\), | \(Z^{1,n}_{1,4} = 0\), | \(Z^{1,n}_{1,5} = 0\) |
\(i=4\) | |||||
\(m=0\) | \(Z^{0,n}_{2,1} = h\), | \(Z^{0,n}_{2,2} = 0\), | \(Z^{0,n}_{2,3} = 0\), | \(Z^{0,n}_{2,4} = 0\), | \(Z^{0,n}_{2,5} = 0\) |
\(m=1\) | \(Z^{1,n}_{2,1} = \sqrt{h}\), | \(Z^{1,n}_{2,2} = \sqrt{h}\), | \(Z^{1,n}_{2,3} = 0\), | \(Z^{1,n}_{2,4} = 0\), | \(Z^{1,n}_{2,5} = 0\) |
\(i=5\) | |||||
\(m=0\) | \(Z^{0,n}_{2,1} = h\), | \(Z^{0,n}_{2,2} = 0\), | \(Z^{0,n}_{2,3} = 0\), | \(Z^{0,n}_{2,4} = 0\), | \(Z^{0,n}_{2,5} = 0\) |
\(m=1\) | \(Z^{1,n}_{2,1} = \sqrt{h}\), | \(Z^{1,n}_{2,2} = -\sqrt{h}\), | \(Z^{1,n}_{2,3} = 0\), | \(Z^{1,n}_{2,4} = 0\), | \(Z^{1,n}_{2,5} = 0\) |
Using the definitions of the coefficients \(c_{m}^{n,i}\) and \(\Delta L_{i}^n\) we calculate
\(m=0\) | \(m=1\) | \(\Delta L_i^n\) | ||
---|---|---|---|---|
\(i=1\) | \(c^{n,1}_0=0\) | \(c^{n,1}_1=0\) | \(\implies \) | \(\Delta L_1^n = 0\) |
\(i=2\) | \(c^{n,2}_0=h\) | \(c^{n,2}_1=\sqrt{h}\) | \(\implies \) | \(\Delta L_2^n = (A_0 -\gamma ^\star A_1^2)h + A_1 \sqrt{h}\) |
\(i=3\) | \(c^{n,3}_0=h\) | \(c^{n,3}_1=-\sqrt{h}\) | \(\implies \) | \(\Delta L_3^n = (A_0 -\gamma ^\star A_1^2)h - A_1 \sqrt{h}\) |
\(i=4\) | \(c^{n,4}_0=h\) | \(c^{n,4}_1=2\sqrt{h}\) | \(\implies \) | \(\Delta L_4^n = (A_0 -\gamma ^\star A_1^2)h + 2A_1 \sqrt{h}\) |
\(i=5\) | \(c^{n,5}_0=h\) | \(c^{n,5}_1=0\) | \(\implies \) | \(\Delta L_5^n = (A_0 -\gamma ^\star A_1^2)h\) |
Finally, using the definitions of \(c_{m}^{n}\) and \(\Delta L^n\) we calculate \(c^{n}_0=h\), \(c^{n}_1=\Delta W^n\) and thus \(\Delta L^n = (A_0 -\gamma ^\star A_1^2)h + A_1\Delta W^n\).
With all the parameters of the scheme (2.12) in place, the resulting Platen strong order 1.5 SL scheme is given by
and
1.2 Platen weak order 2.0 SL scheme
Following the same steps as above, for the explicit order 2 weak scheme by Platen for \(M=1\) [18, Eq. 15.1.1] as underlying scheme we obtain
and thus the corresponding Platen weak order 2.0 SL scheme is given by
Rights and permissions
About this article
Cite this article
Debrabant, K., Kværnø, A. & Mattsson, N.C. Runge–Kutta Lawson schemes for stochastic differential equations. Bit Numer Math 61, 381–409 (2021). https://doi.org/10.1007/s10543-020-00839-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10543-020-00839-8
Keywords
- Systems of stochastic differential equations
- Stochastic Runge–Kutta
- Stochastic Lawson
- Mean-square stability