An Analysis of the Milstein Scheme for SPDEs without a Commutative Noise Condition
Claudine von Hallern,Andreas Rößler
DOI: https://doi.org/10.48550/arXiv.1910.03543
2019-10-09
Abstract:In order to approximate solutions of stochastic partial differential equations (SPDEs) that do not possess commutative noise, one has to simulate the involved iterated stochastic integrals. Recently, two approximation methods for iterated stochastic integrals in infinite dimensions were introduced in C. Leonhard and A. Rößler: Iterated stochastic integrals in infinite dimensions: approximation and error estimates, Stoch. Partial Differ. Equ. Anal. Comput., 7(2): 209-239 (2019). As a result of this, it is now possible to apply the Milstein scheme by Jentzen and Röckner: A Milstein scheme for SPDEs, Found. Comput. Math., 15(2): 313-362 (2015) to equations that need not fulfill the commutativity condition. We prove that the order of convergence of the Milstein scheme can be maintained when combined with one of the two approximation methods for iterated stochastic integrals. However, we also have to consider the computational cost and the corresponding effective order of convergence for a meaningful comparison with other schemes. An analysis of the computational cost shows that, in dependence on the equation, a combination of the Milstein scheme with both of the two methods may be the preferred choice. Further, the Milstein scheme is compared to the exponential Euler scheme and we show for different SPDEs depending on the parameters describing, e.g., the regularity of the equation, which one of the schemes achieves the highest effective order of convergence.
Probability,Numerical Analysis