Stable strong order 1.0 schemes for solving stochastic ordinary differential equations (Q1759582): Difference between revisions

A class of stable numerical methods (SSO1), an extension of the balanced method and the Milstein method, for solving systems of stiff Itô stochastic differential equations of the form \[ dX(t)= f(X(t))\,dt+ \sum^d_{i=1} g_i(X(t))\,dW^i(t),\quad X(t_0)= X_0, \] is presented. A theorem is proved that gives conditions under which strong convergence of order 1 to the exact solution is achieved. The theorem is then extended to establish conditions under which a generalized balanced scheme has strong order of convergence \(p\). A procedure for choosing the parameters in the SSO1 method to minimize global error is given. An examination of asymptotic stability is performed that shows that the SSO1 method surpasses the Milstein method, semi-implicit Milstein methods, and the balanced method.