ABSTRACT
This chapter discusses least squares (LS) approaches to parameter estimation of nonlinear systems. Most real-life and practical systems have nonlinear characteristics, and for accurate modelling/prediction and control, these characteristics should not be ignored. If the type of nonlinearity is known, then only certain or all unknown parameters of the dynamic system need to be estimated. The Gaussian least squares differential correction method for nonlinear parameter estimation is based on differential correction technique, and the algorithm can be used to estimate the initial conditions of states as well as parameters of a nonlinear dynamic model. The estimation before modeling methodology for parameter estimation is essentially a two-step approach. In the first step, the extended Kalman filter is used for state estimation. In the second step of LS regression analysis, one can evolve the most suitable and detailed mathematical model, the parameters of which are estimated using LS.
