Konferenční příspěvek

Suzdaleva E., Nagy I., Dungl M.

Proceedings of the 14th Applied Stochastic Models and Data Analysis Conference (ASMDA2011)

Faculty of Economics of the University of Rome “La Sapienza”

Rome

2011

Ing. Evgenia Suzdaleva, Ph.D. (ÚTIA - Oddělení adaptivních systémů)

state estimation, online filtering, hybrid systems

The paper deals with online state estimation for dynamic hybrid systems with mixed continuous and discrete states. The proposed solution is based on a decomposed version of the state-space model and Bayesian filtering. Specialization to Gaussian linear dynamic and multinomial state-space models is described. Experiments with real data illustrating the presented approach are provided.