USING PATH INTEGRAL SHORT TIME PROPAGATORS FOR NUMERICAL ANALYSIS OF STOCHASTIC HYBRID SYSTEMS

摘要

Algorithms to approximate the evolution of probability density functions for stochastic hybrid systems rely on the knowledge of appropriate short time propagators. It is shown that a path integral propagator known for continuous stochastic systems can be adapted to the hybrid case. With this propagator, the HybPathTree algorithm performs well concerning precision and computational effort, e.g. in reachability analysis.