Target tracking is an important issue in underwater surveillance systems. The tracking systems in sea warfare utilize passive sonar to have bearing only information contaminated with noise, which is assumed here as additive zero mean Gaussian noise. In underwater warfare two dimensional target motion analysis is familiar. The Kalman Filter is used to obtain the target parameters with the help of bearing data coming from sensor. The error in target parameters of velocity, range, heading and bearing are estimated. For some of the scenarios the errors are unacceptable to real time combat systems. Hence alternative methods are surveyed and ANN (Artificial Neural Network) is coupled with Kalman filter to reduce the creeping errors in the solution in spite of Kalman adaptive filters exist. The network selected for this purpose is backpropagation neural network. The network is pre-trained using different inputs to predict the said target parameters. The simulation results are presented and comparative studies are conducted. The ANN provides the adaptive capability the filter model needs.


    Access

    Access via TIB

    Check availability in my library


    Export, share and cite



    Title :

    Artificial neural network embedded Kalman Filter bearing only passive target tracking


    Additional title:

    Zielverfolgung im Wasser unter Verwendung eines Kalman-Filters und eines neuronalen Netzwerks


    Contributors:


    Publication date :

    1999


    Size :

    4 Seiten, 6 Bilder, 8 Quellen


    Type of media :

    Conference paper


    Type of material :

    Print


    Language :

    English