ADAPTIVE LQ CONTROL USING REDUCED HAMILTONIAN FOR CONTINUOUS-TIME SYSTEMS WITH UNMATCHED INPUT DISTURBANCES.

This paper develops an adaptive LQ based disturbance rejection solution for continuous-time possibly nonminimum-phase systems with uncertainties from both dynamics and unmatched input disturbances, using a reduced Hamiltonian function for performance improvement. Such a reduced Hamiltonian LQ control framework is constructed by using a control-separation design and a reduced cost function, with which a two-component controller structure is employed for the desired control objectives: one component dealing with disturbance rejection and one component dealing with system stabilization and output regulation. Such a new control design is first developed for nominal (uncertainty-free) control-separation LQ solutions using reduced Hamiltonian under different relative degree conditions and for finite-time and infinite-time cases and is demonstrated to have the capability to improve system performance as compared with traditional LQ solutions. Based on the developed new controller structure, the adaptive disturbance rejection problem is solved by developing a stable adaptive parameter estimation algorithm for the two-component control law, to ensure desired system stabilization and output regulation in the presence of system and disturbance uncertainties. Simulation results are shown to verify the effectiveness of our proposed LQ based disturbance rejection control schemes. [ABSTRACT FROM AUTHOR]