Numerical simulation models of fluid power systems typically have a large scale of different time constants and so can be considered as numerically stiff systems.
The Rosenbrock method is a differential equation integration algorithm with suitable accuracy and stability properties. In this paper the implementation of the Rosenbrock method as part of a time-domain fluid power system simulation package is presented.
Adaptive time step size control makes it possible to achieve an optimal computational effort with controllable integration error based on user-defined error tolerance.
Computation of examples show considerable improvement in accuracy and computational speed over the previously used integration method.