Title
Principal component analysis for unsupervised calibration of bio-inspired air flow array sensors Principal component analysis for unsupervised calibration of bio-inspired air flow array sensors
Author
Faculty/Department
Faculty of Applied Economics
Publication type
conferenceObject
Publication
S.l. , [*]
Subject
Computer. Automation
Source (book)
ESANN 2011 proceedings : European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, Bruges (Belgium), 27-29 April 2011
ISBN - Hoofdstuk
978-2-87419-044-5
Carrier
E
Target language
English (eng)
Affiliation
University of Antwerp
Abstract
This paper describes the automatic calibration of a set of air ow sensitive sensors on a robot exposed to unknown random air ow stimuli. This might support the idea that the cricket cercus neural system in the terminal abdominal ganglion is evolved by learning. The algorithm makes use of the singular value decomposition (SVD) and the known reduced model dimension of the system for learning the sensor array setup. The absolute orientation of the array can only be found in function of a reference ow or reference sensor which must be calibrated manually. When only a change in airow measure is needed, the reference sensor can be left uncalibrated.
Handle