Nonlinear output path following control using a two-loop robust model predictive control approach

Abstract

In this paper, output tracking of a geometric path for a nonlinear uncertain system with input and state constraints is considered. We propose an enhanced two-loop model predictive control approach for output tracking of a nonlinear uncertain system. Additionally, we propose an optimal version of output path following control problem to improve the controller synthesis. Satisfaction of the dynamical constraints of a system such as velocity, acceleration and jerk limitations is added to the problem introducing a new augmented system. The recursive feasibility of the proposed method is demonstrated, and its robust stability is guaranteed such that relaxation on the terminal constraint and penalty are achieved. To validate the theoretical benefits of the proposed controller, it is simulated on a SCARA robot manipulator and the results are compared with a two-loop model predictive controller successfully.

Approximate two‐loop robust nonlinear model predictive control with real‐time execution and closed‐loop guarantee

“…Two‐loop MPC is a cascade structured control scheme, which is recently introduced for linear systems, 26 nonlinear systems, 22 and output path following problems 27 . In this scheme, it is supposed that a closed‐loop system, in an industrial application such as aerospace, automotive or mechatronics, is available, and it is prestabilized by using a controller $$$ \mu $$$.…”

Rudder/fin joint anti-rolling control system based on interference model predictive control and sliding mode observer

“…Moradi et al (2019) studied an offline robust model predictive control method (RMPC), and linear matrix inequalities (LMI) were introduced to solve the state and input constraints in the design of vehicle active suspension. Farajzadeh-Devin and Sani (2021) proposed an enhanced dual-loop RMPC without designing terminal constraints and penalty terms, which was successfully applied to the problem of geometric path tracking of uncertain systems. RMPC treats parameter, external, and structural uncertainty as noise, such that all uncertainties can satisfy the constraints.…”