Development of control algorithm (PID-LQR) for point to point movement of a nonlinear quadrotor unmanned aerial vehicle

Abstract

In recent years the area of quadrotor UAV has drawn prominent attention of the researchers enabling to develop immense research area in the field of UAVs. A quadrotor has a simple architectural design due to which designing gets simple. Having advantage of uniqueness in shape and simple design it carries numerous kinds of issues regarding its controlling. A quadrotor system is nonlinear in nature, so it requires a suitable controller to ensure its stability during flight. However, researchers have been able to provide some solutions to the issues regarding angular stability, altitude and position control of quadrotor under the challenging conditions such as wind burst and noisy measurements but still they are not successfully resolved. In this thesis report modern control design techniques are discussed and their application in quadrotor control issues are presented. The control techniques are implemented on a longitudinal motion control of quadrotor i.e. issues related while maneuvering on horizontal plane. After imposing few renowned control techniques on the system it was observed that system requires to build a robust control technique for the quadrotor system. An innovative and more robust control technique is proposed for the position controlling quadrotor system. The controller is designed by fusing two distinct control techniques PID and LQR, which is named as PID-LQR. While flying quadrotor can experience two major issues; noises and external disturbance applied on the system. The nature of noises is the noise generated from GPS sensor and the system itself. The nature of disturbance is the disturbance applied externally on the system. The control technique works in such a way that PID is used for tackling a disturbance such as an external push or wind gust and LQR for rejecting noises of the sensor measurements. Furthermore the effectiveness of proposed control technique is also verified by comparing it with autotuned PID and optimized LQR techniques under disturbed and noisy conditions. The simulated results indicate that the proposed method yields a better response as compared to the conventional methods.