Discrete-time stochastic system identification and its application to
sediment transport modelling
Dr H. Chen (CUED)
The parameter identification problem of discrete-time stochastic
system is discussed in the talk, dealing with the case when the
dynamic system noise is
coloured and its variance function may be unbounded in time.
It will be proved, under weak conditions, that the extended least
squares estimate of the system parameters has strong consistency and
fast convergence rate.
Modelling of sediment transport dynamics by using system
identification is also presented as an application of the
methodology.
Back to
Control Seminars Page.