A Belief-Rule-Based Inference Method for Carrier Battle Group Recognition

This paper proposes a method for carrier battle group recognition with over-the-horizon radar outputs. A newly developed belief-rule-based inference method is employed to deal with uncertainties in both radar outputs and domain knowledge. The recognition belief-rule base is constructed considering the constraints among sub-targets in the carrier battle group and subsequently the inference according to such hierarchical belief-rule base is implemented. Moreover, the parameters training method is designed to update the belief-rule base through accumulated historical data. The simulation results show that the belief-rule-based inference method is effective to address the uncertainties in carrier battle group recognition, and the method performs much better after parameters training.