On the problem of weight factors and functions in fusion applications

K. Bendjilali, F. Belkhouche

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Sensor fusion plays an important role in many applications. The abundance of sensory data from different sources does not guarantee better quality information. It is reasonable to assume that not all sensors have the same levels of accuracy and precision. This needs to be taken into consideration when sensor fusion or integration is performed. In many applications it is difficult to measure the sensors' accuracy and precision, which could be dynamic variables. This paper explores different approaches to address the problems related to competitive sensor fusion. The paper focuses on optimal weight assignment and nonlinear weight functions using artificial neural networks as well as simultaneous fusion and calibration of multi-sensor systems. Optimal weight assignment is formulated as a constrained optimization problem and solved using the Karush-Kuhn-Tucker conditions to minimize the variance. Single and multi-layer artificial neural networks are used to determine linear and nonlinear weight functions for sensor fusion. Although they require training data, neural networks are particularly effective when little statistical information about the sensors is available. Several examples are presented to demonstrate the proposed methods.

Original languageEnglish (US)
Title of host publication2019 53rd Annual Conference on Information Sciences and Systems, CISS 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781728111513
DOIs
StatePublished - Apr 16 2019
Event53rd Annual Conference on Information Sciences and Systems, CISS 2019 - Baltimore, United States
Duration: Mar 20 2019Mar 22 2019

Publication series

Name2019 53rd Annual Conference on Information Sciences and Systems, CISS 2019

Conference

Conference53rd Annual Conference on Information Sciences and Systems, CISS 2019
CountryUnited States
CityBaltimore
Period3/20/193/22/19

All Science Journal Classification (ASJC) codes

  • Information Systems

Fingerprint Dive into the research topics of 'On the problem of weight factors and functions in fusion applications'. Together they form a unique fingerprint.

Cite this