System for bearing-only contact state estimation using recurrent neural networks

A neural network for generating a first set of signals representative of a bearing out, a bearing rate out and a norm range rate out for an observation leg in response to target bearing and ownship speed and course information for an associated observation leg, said neural network comprising:

• an input layer comprising a plurality of input nodes, a hidden layer comprising a plurality of hidden nodes, and an output layer comprising a plurality of output nodes;
• said input nodes including a number of nodes which comprise means for receiving said target bearing and ownship speed and course information and a number of nodes for receiving a signal representative of intermediate network state information in a delayed state;
• said hidden nodes receiving from said input nodes a second set of signals representative of said bearing information, said ownship speed and course information, and said intermediate network state information in a delayed state and processing said second set of signals and said intermediate network state information in a delayed state in response to a weight information associated with each input node and each respective hidden node in relation to a predetermined non-linear function and generating a third set of signals which are supplied to said output nodes and a signal representative of a new intermediate network state information;
• and
• said output nodes receiving said third set of signals and and using said third set of signals to generate said first set of signals.