Predictive Control for Euler-Lagrange Systems With Input Saturation and Dead-Zone: A FAS Approach
Xiubo Wang, Guang‐Ren Duan
Abstract
This article presents a predictive control strategy for Euler–Lagrange systems (ELSs) with piece-wise input constraints, includes saturation and dead-zone, by employing fully-actuated system (FAS) approaches. Our comprehensive system design methodology integrates model transformation, controller design, and benchmark applications. Unlike existing methods, the proposed energy-based model transformation (EMT) method first reformulates under-actuated ELSs into (sub-)FASs under general conditions, revealing system controllability and laying a model foundation for controller design. Ulteriorly, the proposed inverse dead-zone predictive cascade control (IDPCC) algorithm employs a cascaded optimization process to redefine new input constraint boundaries, thereby effectively solving the receding horizon optimization problem and circumventing complex nonlinear piece-wise input constraints. By combining IDPCC algorithm with EMT methods not only enhances the solvability of optimization problems in ELSs but also ensures the stability of the corresponding closed-loop system. Finally, the effectiveness of the model transformation methods is validated through four benchmark examples, including simulations of the RTAC system.