thinit (System Identification Toolbox)

Set initial values for the parameters to be estimated.

th = thinit(th0)
th = thinit(th0,R,pars,sp)

Description

This function randomizes initial parameter estimates for model structures th0 in the theta format that correspond to state-space models. th is the same model structure as th0, but with a different nominal parameter vector. This vector is used as the initial estimate by pem.

The parameters are randomized around pars with variances given by the row vector R. Parameter number k is randomized as pars(k) + e*sqrt(R(k)), where e is a normal random variable with zero mean and a variance of 1. The default value of R is all ones, and the default value of pars is the nominal parameter vector in th0.

Only models that give stable predictors are accepted. If sp = 'b', only models that are both stable and have stable predictors are accepted.

sp = 's' requires stability only of the model, and sp = 'p' requires stability only of the predictor. Sp = 'p' is the default.

A maximum of 100 trials are made by thinit. It may be difficult to find a stable predictor for high order systems just by trial and error. An alternative is then to compute the Kalman filter predictor for a randomized model.