Design and Implementation of a Fuzzy Controller and Minimum-Acceleration and Minimum-Snap Trajectory Generation for Smooth and Fast Path Tracking by a Quadrotor UAV

سعید جهانبانی فرد¹

1) دانشجوی کارشناسی ارشد

IOI Code : ABGZ-ABGKEB
Publication : The third international conference on new approaches in engineering and basic sciences(icnabs.ir)
Abstract :
    In recent years, precise control of multirotor UAVs has become one of the key topics in robotics and autonomous systems. In many practical applications, a UAV must be able to follow a predefined trajectory with suitable speed, high motion smoothness, and low tracking error. This study focuses on the design and evaluation of a control and trajectory-generation framework for a quadrotor UAV in real-world conditions, with the goal of achieving accurate, smooth, and fast tracking of predefined trajectories. In the control section, the main strategy is to maintain a quasi-hover condition with a yaw rate close to zero throughout the motion. To this end, a fuzzy proportional–derivative (PD) controller was used in the position loop to generate commanded accelerations based on position and velocity errors, and then the required thrust and desired roll and pitch angles were computed from the reference accelerations. The attitude control is also implemented in a fuzzy framework to track the desired angles and reduce angular errors and angular-rate errors; however, in the real experiment, the attitude controller was already available and the main focus was on tuning the position-control gains. In the trajectory-generation section, two piecewise polynomial methods were examined: a minimum-acceleration trajectory based on third-order polynomials and a minimum-snap trajectory based on seventh-order polynomials. Trajectory timing was performed using a speed profile based on the lengths of the path segments so that movement would be faster on longer segments and slower in regions with tighter turns. For the minimum-snap trajectory, two types of constraints were compared: “continuity of derivatives at the intermediate waypoints” and “zero derivatives at the intermediate waypoints.” The former enables smoother and faster trajectories, while the latter facilitates more accurate waypoint passage and greater safety near obstacles. The tracking results, in both 3D and 2D views, show a small steady-state error and confirm the accuracy of the proposed method.

Keywords : quadrotor uav trajectory tracking fuzzy control attitude control position control