Open access publications in the SLU publication database

Abstract

In many studies of dynamic systems, the stochastic aspects are as important as the dynamic. It is then important to consider uncertainty in the results. Furthermore, dynamics and stochastics interact because the stochastics excite the dynamics and the dynamics change the conditions for the stochastics. Poisson Simulation is an extension of Continuous System Simulation to model and simulate dynamic and stochastic processes. The power of this idea has been described and verified. However, introducing randomness into a CSS model makes the results stochastic. Therefore, a number of simulation runs followed by a statistical analysis have to be performed. Two cases of the statistical analysis must be distinguished: I. Estimation of output quantities, which is straight-forward, and II. Estimation of parameters, which requires a more complex procedure based on bootstrap methods. This documentation presents methods to solve these problems with the help of specially designed tools for repeated simulation and subsequent statistical analysis. Part I presents the tool StocRes for estimation of output quantities and Part II the tool ParmEst for parameter estimations.