Single drone path planning in complex urban airspace

• Main Author: Hoang, Huong Giang
• Other Authors: Sameer Alam
• Format: Final Year Project
• Language: English
• Published: 2019
• Subjects: Engineering::Aeronautical engineering
• Online Access: http://hdl.handle.net/10356/78864
• Institution: Nanyang Technological University

Path planning is important for developing an Unmanned Aerial System in the context of an urban airspace, to ensure safety and efficiency for operations at lower altitudes. Taking into account safety constraints in such an airspace, path planning needs to generate an optimal travel route, while maintaining sufficient minimum separation from urban infrastructure. The Rapidly-exploring Random Tree (RRT) algorithm is used for path planning in this project due to its wide application in robotic motion planning, and strength in multi-robot collision avoidance. This algorithm computes the trajectory from an initial to a desired end location by creating a collision-free path constructed from nodes and links. This project involves modelling the available urban airspace in San Francisco at several different altitudes, and deterministic routing for a single drone by implementing RRT path planning on each altitude layer. The objective is to generate a collision-free path around buildings, and determine the optimal cruise altitude to minimise energy cost. Subsequently, the performance of RRT is compared with a different path planning approach, namely the Fast Marching Method (FMM). A comparison of RRT and FMM can provide some insight into the differences between using a sampling-based and a grid-based approach to path planning respectively.