(NEPTUNE) Real-time state awareness via nerve-like sensing system for autonomous fly-by-feel aerial vehicle
Abstract
ABSTRACT: The next generation intelligent vehicles will be able to ~react~ during operation through aricher ~perception~ of its environment, leading to a more elaborate ~interpretation~ of itscurrent state. This project builds on the sensor network technology that has long beenmatured at the Structures and Composites Lab (SACL) at Stanford University. One of the mainapplications SACL has been focusing on is development of sensor network equipped UAV wing.The data obtained through sensor network have unique properties of (i) having been collectedfrom multiple locations on the wing, and (ii) providing information on different characteristicsof the wing through the use different modality sensors, such as piezoelectric, strain, andtemperature.Our first objective in this project is bringing global discernment capability to flying vehicle byfusing local information collected by individual sensors on the wing, leading to a high~levelinterpretation of its existing flight state. The level of insight we propose to obtain will be madepossible by processing the multi~sensory & multi~modality data via advanced physics~assistedlearning and inference algorithms. Such a robust knowledge on the operational conditionspaves the way for the second objective: We propose to develop a real~time active controlalgorithm to achieve advanced flight objectives using the domain knowledge of interpretedstates. The flight objectives to be designated include stall avoidance, minimizing vibrationunder turbulence, and lift/drag ratio maximization. The wing control variables include angleof attack, airfoil shape, and freestream airspeed. We will describe the problem curation andcustomer discovery processes, provide several technical objectives formulated from end userfeedback, and describe a series of proposed Minimum Viable Products we plan to test underthe NEPTUNE program for next generation intelligent vehicles as described in the proposedMission Model Canvas.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Mar 11, 2020
- Source ID
- N000142012211
Entities
People
- Fu-Kuo Chang
Organizations
- Office of Naval Research
- Stanford University
- United States Navy