A bias-free least-squares parameter estimator for continuous-time state-space models

A bias-compensation method for continuous-time MIMO state-space model identification is presented. The Poisson moment functional approach is used to handle the time-derivative problem. The conventional least-squares algorithm, an instrumental variable algorithm and the proposed bias-free algorithm are applied to the parameter estimation of a simulated system under different noise levels via Monte Carlo simulations. The numerical study illustrates the performances of the proposed method