Simulation and solution of stochastic systems with a component‐based software design

Often uncertainties occur in the numerical simulation of real‐world problems. A useful method is stochastic modelling. In this context Monte Carlo (MC) methods and the stochastic Galerkin method are well known and frequently used. In the field of component‐based software systems a component can be seen as an independent software or as a part of a software system. A corresponding architecture allows clearness, flexibility and reusability. The Component Template Library ( CTL ) is a so called middleware to realise component‐based software systems, and fits to the requirements of scientific computing. In this paper we present the design of a software system for the stochastic simulation using MC and stochastic Galerkin methods based on the CTL. For each stochastic method a number of components are realised to allow parallel features. (© 2008 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)