A comparison of the target tracking in marine navigational radars by means of GRNN filter and numerical filter

The article presents research results on the application of artificial GRNN neural network in the process of radar tracking in marine navigational radars. A comparison has been presented of the tracking process by means of a numerical filter implemented in a real radar device and a neural filter prepared by the authors. The test situation was registered on a real vessel and next introduced into a simulator estimating the vector of target movement by means of a GRNN filter. Results have been presented for various GRNN filters.