Algorithm of Fuzzy Track Correlation in Multisensor System Based on Neural Network

The problem of track correlation is transformed to a problem of multi-dimensional assignment in the situation of multiple local nodes in a distributed multisensor data fusion system. Because the problem of multi-dimensional assignment is a typical NP hard problem, a new model of three-dimensional neural network is proposed to deal with it. The experimental results illustrate that this model not only makes the track correlation correct percent high, but also the processing time can not increase exponentially with dimensions