A New Sliding Mode Control Based on Neural Networks for a Single-Rotor Helicopter

The quadrotor helicopter is considered one of the most popular drone platforms, due to its simplicity of construction and its large capacity compared to the conventional helicopter. In terms of control system design, the quadrotor helicopter is a typical example of a multivariable, nonlinear and strongly coupled higher-order system. Since some of the dynamics of this system are unmodeled and subjected to unknown external disturbances, it is necessary to design a sophisticated control system capable of handling these difficulties. This chapter investigates a novel sliding mode control based on neural networks for modeling and controlling a single-rotor helicopter. First, we discuss the dynamics of a rotor and helicopter. Secondly, since some dynamics of the system are not modeled, we use the artificial neural networks to model them and inject them into the control law. The stability of our system is ensured by taking into account the Lyapunov function. Finally, we validate our study with simulation results conducted in MATLAB.