Parameter identification of chaotic systems is a classical inverse problem in nonlinear science.Due to the high efficiency and robustness of intelligent optimization algorithm,parameter identification of chaotic system based on intelligent optimization algorithm has become a hot topic in recent years.This paper focuses on the parameter identification of fraction-al-order chaotic system.Based on the classical sparrow search algorithm,an improved sparrow search algorithm which intro-duced a random disturbance and the Lévy local flight search is proposed.The identified systems were selected as fractional-order Chen chaotic system and fractional-order Lorenz chaotic system,and simulation were carried out in six different intelli-gent optimization algorithms.The results demonstrate that the improved sparrow search algorithm has faster convergence rate and higher identification accuracy,and it can ensure that the error of identified parameters are less than 10-4.The re-search results can provide a new solution for intelligent synchronization control of nonlinear systems.