Authors:

Darin Tsai

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

In aerial robotics, propeller drones rose to popularity for their simplicity in dynamics and ease of control. However, its capabilities fall short to flapping wing vehicles (FWV) in areas of flight safety, maneuverability, and better adaptability to unknown environments. In contrast to propeller drones' wide range of support software, the complicated dynamics of FWVs often hinder the development of fundamental packages that serve as a baseline for critical features such as automated control and trajectory optimization. This inherent engineering challenge incentivized this study's creation of open-sourced Generalized Robo-Aerial Simulation Setup (GRASS), an open-source package platform for underactuated FWV systems, which have fewer control inputs than their degrees-of-freedom. GRASS lowers the difficulties of entrance into this field by assisting researchers in developing and testing FWVs with minimal resources required. It's an all-encompassing software that integrates complex 3D models, tuneable flapping aerodynamics, and control, enabling researchers to leverage the ever-decreasing computational cost to their advantage. GRASS is written in C++ and Python with a workflow that allows user import of 3D models for simulation in Gazebo, a 3D dynamic simulator. Additionally, bidirectional communication between Gazebo and ROS2, a robot software toolkit, is implemented to isolate control, data collection, and analysis, while simultaneously enabling robot hardware testing. Incorporating the blade element method into GRASS, a 10-degree-of-freedom FWV with simple control and real-time data streaming is simulated with the goal of achieving autonomous hovering. GRASS serves as a comprehensive strategy for the fusion of multi-disciplinary subteams, streamlining the continuous integration and development of FWV.

Source:

Purdue University / 2023

Topics:

No topics listed

Co-authors:

Darin Tsai