A RECEDING-HORIZON REGULATOR FOR NONLINEAR-SYSTEMS AND A NEURAL APPROXIMATION

A receding-horizon (RH) optimal control scheme for a discrete-time non linear dynamic system is presented. A nonquadratic cost function is co nsidered, and constraints are imposed on both the state and control ve ctors. Two main contributions are reported. The first consists in deri ving a stabilizing regulator by adding a proper terminal penalty funct ion to the process cost. The control vector is generated by means of a feedback control law computed off line instead of computing it on lin e, as is done for existing RH regulators. The off-line computation is performed by approximating the RH regulator by means of a multilayer f eedforward neural network (this is the second contribution of the pape r). Bounds to this approximation are established. Simulation results s how the effectiveness of the proposed approach.