Abstract:Flapping flight has undoubtedly played a significant role in many of the aeronautical engineering achievements in the past several centuries. Flapping motions of wings/fins are used by animals (such as birds, insects and bats) to generate lift or thrust to keep them aloft in the surrounding fluid. Researches in this area are not only motivated by the fundamental interest of understanding the mechanism of flapping flight, but also the development of Micro Air Vehicles (MAVs) based on biomimetics. In this lecture, we will address our efforts towards both experimental studies and virtual testing of flapping flight with our High End Digital Prototyping (HEDP) system. The aim of our research is to use physical and biological models to drive the engineering design of these vehicles. Firstly, two types of flapping-wing Unmanned Air Vehicles (UAVs) with spatial four-bar drive mechanism will be presented, including the wind tunnel experiment and prototype flight test results. Secondly, various computation approaches have been adopted to simulate the flapping motions, including traditional moving mesh techniques based on Arbitrary Lagrangian Eulerian (ALE) formulation, and Lattice Boltzmann method with immersed boundary methods. For a single wing-system, we have investigated the influences of various parameters on both the symmetric flapping and asymmetric flapping with bird-like or turtle-like flapping mode. For a tandem wing system, we have investigated the interaction between the hind wing and the vortex shedding off the fore wing and learned how to utilize the flow mechanism to achieve better aerodynamic performances. Moreover, Dynamic Mode Decomposition (DMD) algorithm has been adopted to reveal the underlying physics due to its capability and accuracy in analyzing the complex transient systems. Finally, the flight control design of a dragonfly-like Flapping Wing Micro Aerial Vehicle (FWMAV) is presented, which has incorporated a Linear-Quadratic Regulator (LQR) method and an Iterative Learning Tuning (ILT). In general, we expect that these virtual testing by computational simulations and flight control can drive the design of our next generation flapping-wing UAVs in the future.

Flightmare: A Flexible Quadrotor Simulator

FastSim: A Modular and Plug-and-Play Simulator for Aerial Robots

Autonomous Flights in Dynamic Environments with Onboard Vision

TIE: an Autonomous and Adaptive Terrestrial-Aerial Quadrotor

QuadSwarm: A Modular Multi-Quadrotor Simulator for Deep Reinforcement Learning with Direct Thrust Control

Towards Virtual Testing of Flapping Flight

Learning to Fly in Seconds

Pegasus Simulator: An Isaac Sim Framework for Multiple Aerial Vehicles Simulation

RotorTM: A Flexible Simulator for Aerial Transportation and Manipulation

Quad-rotor Flight Simulation in Realistic Atmospheric Conditions

MAT-Fly: An Educational Platform for Simulating Unmanned Aerial Vehicles Aimed to Detect and Track Moving Objects

Learning to Fly -- a Gym Environment with PyBullet Physics for Reinforcement Learning of Multi-agent Quadcopter Control

VisFly: An Efficient and Versatile Simulator for Training Vision-based Flight

Agilicious: Open-source and open-hardware agile quadrotor for vision-based flight

An Augmented Reality Visualization System for Simulated Multirotor Aerial Vehicles

Back to Newton's Laws: Learning Vision-based Agile Flight via Differentiable Physics

Demonstrating Agile Flight from Pixels without State Estimation

FlightGoggles: A Modular Framework for Photorealistic Camera, Exteroceptive Sensor, and Dynamics Simulation

An Open-source Hardware/Software Architecture and Supporting Simulation Environment to Perform Human FPV Flight Demonstrations for Unmanned Aerial Vehicle Autonomy

Deep Reinforcement Learning-based Quadcopter Controller: A Practical Approach and Experiments

XTDrone: A Customizable Multi-Rotor UAVs Simulation Platform