Robust backstepping control of side-by-side unmanned helicopter based on fuzzy neural networks extended state...

The combined influence of mass-inertia variations and aerodynamic disturbances generates coupled uncertainties that substantially degrade the positioning and attitude stability of side-by-side unmanned helicopters. To mitigate these issues, this study develops a control strategy integrating a fuzzy neural network-enhanced extended state observer with robust backstepping control (FNNESO-RBSC). The core innovation lies in utilizing the FNN to accurately estimate the time-varying total disturbance within the ESO framework, overcoming the limitation of conventional fixed-gain ESO. This advanced observation mechanism enables dynamic disturbance compensation within the backstepping control law. The flight control system is subsequently developed and validated through hardware-in-loop (HIL) simulations. Experimental results demonstrate that the proposed controller achieves superior performance compared with conventional active disturbance rejection control (ADRC) and adaptive neural network ESO-based finite-time convergent sliding mode control (ANNESO-FTCSMC) methods, exhibiting enhanced disturbance rejection, improved robustness, and higher trajectory tracking accuracy across varying operational scenarios.