Control of Quadrotor using a Neural Network based PID

Abstract

The control of Unmanned Aerial Vehicles (UAV) is an important task in mobile robotics because of its extensive use and the advantages over other kind of mobile robots like ground vehicles. In this paper, a PID controller for a quadrotor based on an Artificial Neural Network (ANN) is proposed. The neural network can handle the limitations of the PID when controlling complex non linear systems. The neural network consists in a single neuron with three inputs and their weights represent the gains of the PID (K <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">p</sub> , K <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">i</sub> and K <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">d</sub> ). Simulations and experiments with a Quanser Qball-X4 are presented.