Adaptive robust H ∞ neural network fixed-time control for non-strict feedback systems with deadzone and input saturation utilising a reduced observer and command filtered backstepping technique

This article discusses fixed-time neural network adaptive robust H∞ control with observer for unknown nonstrict-feedback nonlinear systems, taking into account input saturation, dead zone, and external disturbance. A novel function is suggested for estimating nonlinear input functions which offers a more precise approximation when compared to past techniques. Assuming the system's output is accessible as a state variable, the proposed state observer is simplified and reduced in order accordingly. The command filtered is implemented to prevent the issue of virtual control input derivation in the backstepping technique. A new methodology is presented that employs the H∞ index to tackle external disturbances in a non-strict feedback system, while also considering input characteristics and integrating an observer. Two illustrative examples are provided to showcase the effectiveness of this method. The findings suggest that the proposed strategy can successfully attain the desired tracking performance, ensure stability over a specified duration, and effectively counteract disturbances.