A comparative study for the boundary control of a reaction-diffusion process: MPC vs backstepping.

摘要

In this paper, we consider a reaction-diffusion process described by a linear parabolic partial derivative equation (PDE). Two radically different control approaches are compared: the model predictive control (MPC) and the backstepping approach. The stabilization of the unstable reaction-diffusion process is first studied. Then, to deal with parameter uncertainties, an adaptive backstepping controller is developed and compared to a model predictive controller based on an internal model control (IMC) structure. Simulation results illustrate the efficiency of the two approaches in terms of precision and computational time.