Comparison of six on-line identification algorithms
Automatica (Journal of IFAC)
Hi-index | 22.15 |
A method is presented for on-line parameter identification of systems in which a parameter vector is premultiplied by a matrix of non-linear state functions. Use is made of delay filters and a multiple observation technique to assure unbiased identification in the presence of measurement noise. A stochastic approximation algorithm is utilized to perform the identification and experimental results are presented for hybrid computer simulation of chemical process models.