A Lyapunov-based model predictive control strategy with a disturbances compensation mechanism for dual-arm manipulators.

• In many works, applied methods neglect control input and system constraints, which may impact global stabilization if these are not adequately handled. In this paper, the NMPC optimizes the behaviours of the system while ensuring the system constraints and improving the performance of the dual-arm robot. • Eliminating the impacts of external noise and imprecise models on the system performance through the RBFNN and facilitating the process of predicting the optimal behaviours for the NMPC. • Ensuring global stability and the convergence of RBFNN through the updating law based on the Lyapunov stability-taking contraction conditions designed based on the sliding mode controller. • Providing the pseudo-physical model and the comparison of the proposed control strategy with NMPC-RBF and SMC-RBF to highlight the effectiveness and feasibility of this control method. [Display omitted] This paper proposes a Lyapunov-based nonlinear model predictive control (LMPC) - based on adaptive Lyapunov to solve existing problems in nonlinear dual-arm systems such as system constraints and unknown external disturbances. In practice, the constraints tend to adversely affect the system's performance and stability. The nonlinear model predictive control NMPC is considered a promising candidate for handling system constraints while enhancing the robustness of the system. However, the rigour of the modeling procedure has a significant influence on the execution of the NMPC, system convergence cannot be assured in the face of modeling uncertainty. To solve this problem, the proposed controller takes into account external disturbances and unidentified parameters by using an adaptive mechanism constructed via the Radial Basis Function Neural Network (RBFNN). Furthermore, the dominant problem of the NMPC algorithm is the system stability which is considered by the Lyapunov theory backbones by a nonlinear Sliding Mode Control (SMC). The numerical simulations are carried out based on a pseudo-physical model to show the efficiency of the proposed control method. [ABSTRACT FROM AUTHOR]