摘要:This paper deals with fractional non-linear system identification, where the Hammerstein Controlled Auto-Regression (HCAR) model is considered. The identification process is derived for the regression form of the HCAR system based on the Over-parametrization principle and the Key-term separation principle. Levenberg-Marquardt algorithm combined with each of these principles is developed to identify the fractional HCAR system. Various simulations test the efficiency of the optimization method based on these principles.