An Improved Robust Predictive Control Approach Based on Generalized 3rd Order S-PARAFAC Volterra Model Applied to a 2-DoF Helicopter System

This paper proposes a generalized robust predictive control approach based on 3(rd) order S-PARAFAC Volterra Model. The main idea is to use the predictive control law based on the latter model with taking into account the parameter and operation uncertainties resulting from the measurement noise and the robust identification technique named Unknown But Bounded Error (UBBE). One of the advantages is the convexity of the objective function with respect to the parameter uncertainty set. The min-max optimization problem becomes thus simpler, by minimizing the objective function in the worst-case only over the set of uncertain models. This work proposes a new generalized nonlinear robust control algorithm dedicated for uncertain processes. In this algorithm we optimize a quadratic criterion and take into account the physical and geometrical constraints due to parameter uncertainties. The performances of this work are illustrated and validated on a benchmark as a continuous stirred-tank reactor system (CSTR) and on an experimental 2-DoF helicopter system.