Radial Basis Function with PID for Quadcopter: Disturbed Trajectory Tracking

Uncertainties and perturbations are the ‘enemies’ of a flying robot without a proper and reliable controller. The easiest controller to implement is the PID controller which its gains only need to be tuned properly either using Ziegler-Nichol's Method or manual tuning. Both are time consuming and single-acting PID controllers cannot adapt in various situations with only single-tuning. Being in different situations requires re-tuning, which is unfavorable during physical flight mode. In this paper, single-acting PID controller will be transformed into a hybridized mode which includes the action of Radial Basis Function (RBF) Neural Network. The objective is to help a quadcopter to survive various uncertainties and perturbations with a self-tuned algorithm. An RBF network is one kind of Artificial NN but with simpler network design and more accurate local approximation. The performance of the proposed work is proved through simulations using MATLAB/Simulink. Different situations are presented to test the final system, which are the wind disturbance and trajectory tracking. Results are presented in this paper using visual simulation and ISE performance index. After comparison between the proposed work and ZN-tuned PID controller is made, RBFPID controller wins the deal.