Konferenční příspěvek

Proceedings of the 6th International PhD Workshop on Systems and Control

2005

Ing. Lenka Pavelková, Ph.D. (ÚTIA - Oddělení adaptivních systémů)

Autoregressive model with exogenous inputs (ARX) is a widely used black-box type model underlying adaptive predictors and controllers. The output noise is usually supposed to have the Gaussian distribution with zero mean value. These models are good algorithmically processed, they have a wide range of exploitation and their usage give reasonable results. Light tails of the normal distribution imply that its unbounded support can often be accepted as a reasonable approximation of reality, which is mostly bounded. In some cases, however, this assumption is unrealistic or do not fit a subsequent processing. Then, techniques dealing with unknown-but-bounded equation errors are used. In the paper the output noise is assumed to have the uniform distribution.