Abstract:
This thesis addresses the development of trajectory planning algorithms for unmanned aerial vehicles (UAVs). It focuses on the motion planning for Quadrotor type and contributes in the both stages: the path planning and the trajectory planning. In path planning, a technique is proposed to reduce the collision checking time in cluttered environment for sampling-based path planning. The method disassociates the dependence of time complexity with number of obstacles, by performing a pre-processing that stores information about the obstacles, then using them in the collision checking queries. Also, we propose a new variant of Rapidly
Exploring Random Trees Star (RRT*) named NP-RRT* to deal mainly with narrow passages in the workspace and to speed up the convergence rate to optimal solution, also, to minimize the memory consumption. In the trajectory planning, we implement the Particle Swarm Optimization (PSO) to minimize the snap and distance trajectory for Quadrotor. The minimum snap polynomial trajectories are the natural choice for Quadrotors, since the motor commands and attitude accelerations of the vehicle are proportional to the snap of the path.
