Multi-dimensional Taylor network-based adaptive command-filter control for nonlinear systems with actuator faults, dead-zone, and saturation

This research aims to develop an adaptive control method for strict-feedback nonlinear systems using multidimensional Taylor networks (MTNs) to address input dead zone, saturation, and actuator faults. MTNs are utilised to estimate uncertain system functions and to enhance reliable control performance. Additionally, a command filter technique is introduced to circumvent the ‘explosion of complexity’ issue during derivative computation for virtual control laws and further improve reliable control performance. The paper addresses critical non-smooth nonlinearities, such as input saturation and dead zone, by approximating them with non-affine smooth functions and transforming them into an affine form using the mean value theorem. Using Lyapunov stability theory and the backstepping method, an adaptive controller is designed to ensure bounded signals in the closed-loop system. Additionally, it achieves excellent performance with small tracking errors through careful selection of control parameters. A simulation example is provided to validate the effectiveness of the proposed method.