Adaptive Sliding Mode Model-Free Control for Aircraft Attitude Control with System Uncertainty

This paper addresses the issue of attitude control for dual-control aircraft with model uncertainty, presenting a novel adaptive model-free sliding mode control (ASMMFC) algorithm based on long-short-term memory (LSTM) neural network. In comparison to existing adaptive sliding mode control algorithms utilizing neural networks, ASMMFC relaxes the underlying assumptions, thus broadening its scope of applicability. In the process of control law design, the LSTM neural network is used to fit the system term, so that it no longer depends on the prior information of the model. The online update law of neural network weights is derived by the Lyapunov method, and the stability of the system is proved. In the simulation process, the ASMMFC is applied to the aircraft attitude control scenario with model uncertainty and time-varying parameters. Simulation results show that the performance of the ASMMFC is better than other existing adaptive sliding mode control methods.