Robust iterative learning control design for batch processes with uncertain perturbations and initialization

Abstract A robust iterative learning control (ILC) scheme for batch processes with uncertain perturbations and initial conditions is developed. The proposed ILC design is transformed into a robust control design of a 2‐D Fornasini–Marchsini model with uncertain parameter perturbations. The concepts of robust stabilities and convergences along batch and time axes are introduced. The proposed design leads to nature integration of an output feedback control and a feedforward ILC to guarantee the robust convergence along both the time and the cycle directions. This design framework also allows easy enhancement of the feedback and/or feedforward controls of the system by extending the learning information along the time and/or the cycle directions. The proposed analysis and design are formulated as matrix inequality conditions that can be solved by an algorithm based on linear matrix inequality. Application to control injection packing pressure shows the proposed ILC scheme and its design are effective. © 2006 American Institute of Chemical Engineers AIChE J, 2006