Reference:
A. Moradvandi,
R.E.F. Lindeboom,
E. Abraham, and
B. De Schutter,
"Models and methods for hybrid system identification: A systematic
survey," Proceedings of the 22nd IFAC World Congress,
Yokohama, Japan, pp. 95-107, July 2023.
Abstract:
Dynamical systems and processes that either exhibit non-smooth
behaviors (e.g. through logic control or natural phenomena) or work in
different modes of operation are usually represented using hybrid
systems models, i.e. mathematical models that combine continuous
dynamics with discrete-event dynamics. Identification of a hybrid
system includes finding switching patterns and identification of model
parameters to obtain a data-driven model. This survey paper provides a
systematic review of models (how to parameterize the system) and
methods (how to identify unknown parameters) proposed for hybrid
system identification with an exposition of recent advances and
developments, and further research directions.