M.H Zibaee Nejad - Intelligent Control Systems Lab., Electrical Engineering Dept., Tarbiat Modares University, Tehran, Iran.
Vahid Johari Majd

In this paper, we propose an efficient formulation of output feedback robust model predictive control algorithm for systems with polytopic uncertainty. The formulation is based on quasi-infinite cost function. First, we construct nested invariant ellipsoids and their corresponding state feedback gains; then we design a state estimator and establish closed- loop robust stability by proper choice of design parameters off-line. Next, we calculate the real time control law on-line, regarding to the current estimated state position in the state space. The proposed algorithm reduces conservatism of the available off-line methods due to the use of several Lyapunov functions and quasi infinite cost function. The algorithm is illustrated with an example .

Robust Model Predictive Control (RMPC), output feedback, Linear Matrix Inequalities (LMIs)