Development of the finite and infinite interval learning control theory

This thesis centers on the theories of Finite Interval Learning (FIL) and Infinite Interval Learning (IIL) for nonlinear systems with deterministic uncertainties. Considering the existence of non-smooth nonlinearties in real systems, Iterative Learning Control (ILC) is extended to systems with deadzone/backlash, which provides a simple way to deal with such high nonlinearities. Benefiting from Composite Energy Function (CEF), the robust FIL is further proposed to handle norm-bounded uncertainties. In conventional FIL, only uniform desired trajectory is considered. To overcome the limitation, a new FIL approach is introduced to enable the learning from different control tasks. To widen the application of FIL, Fuzzy Logic Learning Control scheme is proposed, in which FIL is added to a Fuzzy Logic Controller directly. By taking advantage of CEF, the FIL is extended to the IIL for both parametric and norm-bounded uncertainties. To facilitate the practical application, a kind of observer-based learning control algorithm is outlined for nonlinear systems with parametric uncertainties.

Author dc:creator

• XU JING

• learning control, finite time interval, infinite time interval,uncertainty,contraction mapping principle,composite energy function