Adaptive Neural Network Control of Fractional-Order Nonlinear Systems via Integer-Order Learning: A Memory-Efficient Approach
Abstract
This letter develops an integer-order (IO) neural-learning approach for a class of fractional-order (FO) nonlinear systems. An initialization-compensated (IC) filtered tracking error yields an exact IO control-relevant error dynamics, while retaining the FO plant dynamics. This enables IO neural-weight adaptation without fractional convolution in the NN learning channels. A neural network (NN) approximates the unknown nonlinearity on a prescribed compact state domain, while a robust discontinuous term compensates for the bounded ideal approximation residual. A nonsmooth Lyapunov analysis establishes semi-global asymptotic tracking. Besides, a reconstruction lemma provides the link between the IC filtered error convergence and the FO tracking errors. The proposed learning mechanism reduces the NN adaptive-law update time by about 99.67%, while preserving tracking performance comparable to a conventional FO adaptive NN controller, with RMSE values of approximately 0.033.