Session: SYMP S-5: Mechano-Intelligent Robots

Paper Number: 140469

140469 - Mechanical Intelligence for Coordination and and Control of Wingbeats in Bioinspired Flapping Wing Robots 

The flight muscles of insects come in two forms: 1) synchronous muscle which requires a signal from the brain to actuate, and 2) asynchronous muscle which actuates from a stretch response. The stretch-response of asynchronous muscle is called delay-stretched actuation (DSA), and it enables high-frequency wing oscillations that adapt to the system’s mechanical resonance. The majority of bioinspired flapping wing robots use synchronous actuation with a sinusoidal signal generating wingbeats. More recently our group has been developing actuation methods that emulate delayed stretch activation using DC motors, and thus the flapping wing robot possesses adaptive and responsive wingbeat flapping. However, prior studies of DSA for robotics have only explored single wings in a bench-top lab test. In this work we will describe the development of a two-wing flapping wing robot that uses two independent actuators driven through DSA feedback. Inspired by the insect exoskeleton we achieve wingbeat coordination through a compliant linkage that couples left and right actuator-wing pairs. Thus, this robot possess emergent wingbeat dynamics that rely on autonomous feedback driven oscillations (DSA), and mechanically intelligent compliant structures. 

 

In this study, we first describe the implementation of DSA on a micro DC motor. Without using traditional sensors such as encoders, we developed a velocity sensing method (which is essential for DSA) by utilizing the linear relationship between back-emf voltage and angular velocity of DC motors. In bench-top tests we verified the properties of DSA such as self-oscillation, self-adaptivity. Also we tested the frequency and amplitude that the DSA can generate in a physical environment with different intrinsic DSA parameters settings. To characterize the system we designed an experiment to identify the stiffness and inertia of the robot’s transmission and wing and verified the accuracy of our experiments by comparing the results with corresponding resonance frequency.

To achieve flight, pairs of wings must flap in tight coordination. In asynchronous insects the flapping wing coordination is not derived from neural signals, but instead through compliant coupling of the left-right muscle pairs through the elastic thorax. Thus, we built and analyzed a two-wing system with two individual DSA drive motors and physically connected them with an elastic linkage system. With the help of this thorax-like elastic linkage, we observed nice synchronized wingbeats which showed excellent robustness to external perturbation. We studied the dependence of wingbeat coupling on the properties of the elastic coupling mechanics. Further, we analyzed the influence of elastic connection for wingbeat generation from unbalanced wings to identify robust mechanically intelligent solutions for wingbeat coordination.

Presenting Author: Rundong Yang Mechanical and aerospace engineering, University of California, San Diego

Presenting Author Biography: I am a PhD student in UC San Diego, majoring in Mechanical engineering. My research focused on properties of active matters and bio-inspired flapping wing systems.

Authors:

Rundong Yang
Ellen Liu
Ethan Wold
Simon Sponberg
Nicholas Gravish