선형계통의 파라미터 추정을 위한 최적 입력의 설계

Design of the optimal inputs for parameter estimation in linear dynamic systems

Optimal input design problem for linear regression model with constrained output variance has been considered. It is shown that the optimal input signal for the linear regression model can also be realized as an ARMA process. Monte-Carlo simulation results show that the optimal stochastic input leads to comparatively better estimation accuracy than white input signal.