Closed-Loop Stabilizing MPC for Discrete-Time Bilinear Systems

Model-based predictive control (MPC) strategies for discrete-time bilinear systems are developed, which incorporate the guarantee of nominal closed-loop stability. It is demonstrated how the structure of the bilinear model can be exploited, both for a MPC strategy which is based on the combined use of an end-point weighting and end-point inequality constraint, and for a MPC strategy which uses the closed-loop paradigm, deploying perturbations on an off-line designed control law. These two MPC algorithms are compared by means of a numerical example.