|
Abstract: The main goal of this thesis is to obtain reliable outputs,
which requires robustness relative to the noise included in input data
and to the sensor deterioration or even the missing of the sensor.
Without of any pre-defined knowledge concerning sensors，the multi-sensor
data level fusion model based on artifical neuron can be used for the
estimation of fused data with minimum mean square errors through
observed data so as to calibrate the fusion neuron. Simulation results
show that the fused data are much more sensitive，accurate，reliable than
that of single sensor data .
Key words: Multi-sensor data Data level fusion Artifical neuron Mean
square error Fusion calibration
CLC No: TB526
国家自然科学基金(59990472)和国家“九五”攀登B(PD9521908z1)资助项目. Received 20020228,
received in revised form 20030123
|