Abstract
The article introduces the design of an online disturbance compensator based on machine learning for quadrotor aircraft. The article presents the state-space models for the quadrotor, which encompass wind disturbances. The machine learning algorithm estimates unmeasurable states, which are linear and angular velocities, and constructs the unknown disturbances. These disturbances are then fed to the controller to compensate for disturbance and deviation in trajectory by varying the rotor speeds of the quadrotor aircraft. To present the simplicity of the proposed system, a simple PD controller is employed to manage the nonlinear modelled quadrotor. For the online training and validation purposes, the Parrot Mambo drone is utilized. The results are provided to demonstrate the effectiveness and advantages of the proposed controller.