基于属性约简与BP神经网络的舰艇目标威胁评估方法

Threat Assessment Method of Naval Vessel Target Based on Attribute Reduction and BP Neural Network

通过建立属性约简与反向传播算法(Back Propagation,BP)神经网络的舰艇威胁评估模型,对舰艇目标威胁程度进行评估判断.主要通过属性约简给出了舰艇威胁评估的典型作战因素,验证基于BP神经网络解决非线性舰艇威胁评估的技术可行性,并建立了多目标舰艇威胁评估指标体系,为导入大量作战数据验证舰艇威胁评估算法奠定理论基础.通过验证数据,给出了舰艇威胁评估结果,实验结果表明BP神经网络能有效地解决作战态势评估中的非线性问题,威胁评估准确、稳定,对实现智能作战威胁评估具有重要意义,为辅助决策技术的实现提供了技术途径和理论基础.

To assist commanders to make a correct judgment on the threat degree of naval vessel targets,this paper establishes a naval vessel threat assessment model based on attribute reduction and(Back Propagation,BP)neural network.Through attribute reduction,typical military operational factors of naval vessel threat assessment are given to verify the technical feasibility of solving the nonlinear naval vessel threat assessment based on BP neural network,and meanwhile,a multi-objective naval vessel threat assessment indicator system is established,which lays a theoretical foundation for importing a large number of combat data to verify the algorithm of naval vessel threat assessment.The results of naval vessel threat assessment are given by validating data.The experiments show that BP neural network can effectively solve the non-linear problem in battle situation assessment accurately and stably,which is greatly significant to realize intelligent battle threat assessment.Moreover,it also provides a technical approach and theoretical basis for the realization of assistant decision-making technology.