Autonomous control of a quadcopter

Abstract

As research into robotics accelerates into the 21st century, growing interest has been shown into quadcopter UAVs, especially in the field of automatic control. Being an underactuated, multivariable and nonlinear system, the quadrotor system is a perfect test-bed for the design of various linear and nonlinear control techniques. This dissertation studies the design and implementation of a fully autonomous control system for a quadcopter UAV. In particular, this project outlines the effectiveness of PID control techniques for most practical control applications. The practicability of cascade control, used in the controller design to reduce the complexity of a large control system is also investigated. The quadcopter used in this project is a custom build, primarily comprising of a QAV400 quadcopter frame and a Pixhawk Mini flight controller, among other crucial components making up the system. The flight controller was loaded with the PX4 software framework, which is an open source framework capable of being modified according to the application. Having an open hardware and software platform was crucial for this project and others to follow.