Setting the PID Controller for Controlling Quadrotor Flight: a Gradient Approach

Abstract

The paper deals with the design of an autopilot for a quadrotor based on a PID controller. A mathematical description of the control object in form of the system of second-order differential equations is also presented. Setting the parameters of the PID controller is an iterative procedure. The paper investigates three optimal tuning algorithms based on minimizing the quality integral: coordinate descent, gradient descent, and steepest descent. During the research, a number of advantages of the gradient search algorithm are established. Gradient search is used as the main design method; on its basis, the autopilot model is synthesizing. Some simulation results are presented, which are confirmed by the Simulink autopilot model.