Parameter estimation with improved model prediction for over-parametrized nonlinear systems

Weifeng Chen, Baojia Wang, Lorenz T. Biegler

Abstract: Highlights•New method proposed for parameter estimation of ill-conditioned nonlinear systems•Approach is based on transformed parameters from SVD•reduced Hessian and moment criterion are applied to determine transformed parameter subsets•values of inestimable parameters are determined through Monte Carlo Sampling•demonstrated on two nonlinear dynamic systems with improved accuracy and reduced computational cost