Title :
A neural approach to data fusion
Author :
Chowdhury, Fahmida N.
Author_Institution :
Dept. of Electr. Eng., Michigan Technol. Univ., Houghton, MI, USA
Abstract :
A neural approach to data fusion is proposed. We assume that remote sites process local sensor data, and the fusion center does not have covariance information. A neural network consisting of one neuron for each component of the measurement vector is proposed as the fusion center, provided it has been trained with past data. This is an alternative to the standard approach of estimating the covariances explicitly. To demonstrate the idea, some simulation results are shown
Keywords :
learning (artificial intelligence); neural nets; sensor fusion; covariances; data fusion; local sensor data; measurement vector; neural network; Covariance matrix; Estimation error; Filters; Genetic expression; History; Neural networks; Neurons; Sensor fusion; State estimation; Target tracking;
Conference_Titel :
American Control Conference, Proceedings of the 1995
Conference_Location :
Seattle, WA
Print_ISBN :
0-7803-2445-5
DOI :
10.1109/ACC.1995.529797
