Variable Trial Length AILC for Nonlinear Systems with Non-Square Control Gain Matrices

This paper presents a novel adaptive iterative learning control (AILC) algorithm tailored for generic nonlinear systems with non-square control gain matrices. By integrating the Rayleigh quotient into the composite energy function, the algorithm guarantees convergence. Additionally, it employs piecewise parametric updating laws to accommodate trial length variations and employs error-tracking methods to handle iteration-varying initial states, thereby enhancing robustness. The algorithm demonstrates improved computational efficiency and learning efficacy, making it suitable for diverse control systems, as validated through a numerical simulation.