Error Processing of Sparse Identification of Nonlinear Dynamical Systems via $L_{\mathrm{\infty }}$ Approximation

This paper deals with the error processing problem of sparse identification of nonlinear dynamical systems(SINDy) through introducing the $L_{\mathrm{\infty }}$ approximation to take place of the former $L_2$ approximation. The motivation is that the $L_{\mathrm{\infty }}$ approximation could better describe the error phenomenon in the SINDy, which consists of the derivative approximation error and the measurement noise. Then, an iterative thresholding algorithm is proposed to solve the reformulated problem. 3 scenarios of possible errors are considered in the experiment. The results show that the $L_{\mathrm{\infty }}$ approximation performs better or at least equal than the $L_2$ approximation in face of different error cases. Hence, it is reasonable to consider the $L_{\mathrm{\infty }}$ approximation in the applications of the SINDy.