Collision avoidance algorithm for UAVs with MATLAB

• Main Author: Chan, Boone-Wy
• Other Authors: Low Kin Huat
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Mechanical engineering
• Online Access: https://hdl.handle.net/10356/141756
• Institution: Nanyang Technological University

The utilisation of Unmanned Aerial Vehicles (UAVs) for commercial, recreational, and scientific purposes is commonplace in today’s world. As a result, the urban airspace is becoming more densely populated. It is therefore prudent to invest research into path planning and collision avoidance techniques for UAVs. This report explores one such popular path planning method: The Rapidly-exploring Random Tree (RRT) algorithm. Widely used in robotic motion planning, this algorithm has tremendous potential in multi-robot collision avoidance. By selecting random points within the boundaries of a search space and attempting to grow the tree to these points, the algorithm is able to effectively search the environment and eventually plot a collision-free path from an initial seed location to a desired goal location. RRT* is an optimised version of the conventional RRT algorithm, capable of producing a smoother and shorter path to the destination location. In this report, MATLAB was used to implement an RRT* algorithm capable of path planning and collision avoidance within a dynamic virtual environment to simulate a densely populated airspace containing other UAVs. The results of the simulation were then evaluated based on runtime, path length, and the number of re-planning attempts required to arrive at the goal. The simulation results yielded insights into the optimal parameters with which the RRT* algorithm would perform. Techniques with which the algorithm performance might be improved were also investigated.