Attitude Stabilization for a Micromechanical Flying Insect via Ocelli Output Feedback

Luca Schenato
(Professor S. Shankar Sastry)
DARPA, (NSF) KDI ECS 9873474, and (ONR) MURI N00014-98-1-0671

Unmanned air vehicles (UAV) have been a very active area of research. Despite recent remarkable achievements obtained with fixed and rotary aircrafts [1], their use in many tasks is still limited by their maneuverability and size. However, the extraordinary flight capabilities of insects have inspired the design of small UAVs with flapping wings mimicking real flying insects. Their unmatched maneuverability, low fabrication cost, and small size make them very attractive for cost-critical missions in environments that are unpenetrable for larger size UAVs. In order to fabricate a micromechanical flying insect (MFI) eventually capable of stable flight, it is necessary to design and develop sensors capable of estimating the orientation of the insect. We are currently investigating the use of ocelli, a set of biologically inspired sensors, capable of estimating the orientation. The ocelli are four wide-angle photoreceptors placed on the head of the insect. They are oriented in such a way they collect light from different region of the sky (see Figure 1). The intensity of the light from the sky is a function of the latitude only, therefore, in principle, it is possible to estimate the orientation of the insect by opportunely combining the output of the four photoreceptors. Our goal is to use this information to design stabilizing controllers for hovering and maneuvering flight modes for an MFI.


Figure 1: Ocelli modeling: every photoreceptor collects light intensity from a specific area of the sky.

[1]
D. H. Shim, H. J. Kim, H. Chung, and S. Sastry, "Multi-Functional Autopilot Design and Experiments for Rotorcraft-based Unmanned Aerial Vehicles," Digital Avionics Systems Conf., Daytona Beach, FL, October 2001.

More information (http://robotics.eecs.berkeley.edu/~lusche) or

Send mail to the author : (lusche@eecs.berkeley.edu)


Edit this abstract