Optimal Designing Back Stepping Controller for Quadrotor By using Reinforcement Learning Automata Method

Unmanned aerial vehicles (UAV) are increasingly used in civilian and military application because of their ability to hover, fly in very low altitude and confined spaces. In this paper, analyzing, designing and simulation of optimal controller of UAVs with four rotors in six degree of freedom is investigated that is known as Quadrotor. Because of simple design and maintance they become popular device. On the other hand, designing low level controller for quadrotor due to model nonlinearities and inherent instabilities is a challenge. Controller whitch was used in this paper is back stepping controller. This controller is optimized by Reinforcement Learning Automata (RLA) and then is applied to attitude and altitude controllers. Results showed that RLA method has better operation than Matlab toolbox.