Novel neural networks for eliminating errors existing in FOG

Rong Zhu; Yanhua Zhang; Qilian Bao

doi:10.1117/12.379962

24 March 2000 Novel neural networks for eliminating errors existing in FOG

Rong Zhu, Yanhua Zhang, Qilian Bao

Proceedings Volume 3936, Integrated Optics Devices IV; (2000) https://doi.org/10.1117/12.379962
Event: Symposium on Integrated Optoelectronics, 2000, San Jose, CA, United States

Abstract

A novel neural network based technique instead of now available measures to raise FOG's measurement precision was proposed in the paper. The technique includes a few of aspects: a series-single-layer neural network used for dealing with the random noises, an advanced learning scheme corresponding with the above network, and an RBF network based method to evaluate the temperature drift. The series- single-layer network, which is composed of two single-layer networks in series, has advantages of simple architecture, fast learning speed, and better performance over conventional BP networks. To conduct the learning of the series-single-layer network, the advanced learning scheme derived from drawing the increments of error into energy function was then developed after referring to the unstable power law noise. Furthermore, in consideration of their good performances, we introduced an RBF network to evaluate and compensate the temperature drift and selected OLS algorithm to construct the network with high speed. The simulation results verified the validity that the proposed technique greatly decreased the errors induced by the different random noises and the biasing drift existing in FOG system.

Citation Download Citation

Rong Zhu, Yanhua Zhang, and Qilian Bao "Novel neural networks for eliminating errors existing in FOG", Proc. SPIE 3936, Integrated Optics Devices IV, (24 March 2000); https://doi.org/10.1117/12.379962

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available