Nonlinear model predictive control enhanced by generalized pointwise min-norm scheme

Nonlinear Model predictive control (NMPC) suffers from the problems of closed loop instability and computation complexity, which greatly limits its application in mechatronic systems involving fast time-varying dynamics. In this paper, a new NMPC enhanced by generalized pointwise min-norm (GPMN) scheme is presented. First, a generalized min-norm control algorithm is developed by introducing a guide function into Freeman’s pointwise min-norm control (PMN) algorithm in order to obtain a stable controller based on a known CLF. Then, the guide function is parameterized according to the Bellman’s optimization principle and the GPMN scheme is further integrated into normal NMPC strategy. As a result, the closed loop stability of NMPC is guaranteed and the real-time applicability is substantially improved. Simulation results have shown the efficiency of the proposed method.