基于神经网络的威胁评估研究

威胁评估是战场损伤模拟研究的重要内容之一,对于开展更为深入的战场损伤模拟研究是非常重要的.威胁评估涉及的因素较多,且难于建立一种合理的评估准则和评估数学模型.该文根据战场损伤模拟中威胁评估的非线性特点,以及神经网络在解决非线性复杂问题所具有的独特优势,提出了基于神经网络的威胁评估方法,讨论了威胁评估神经网络模型的构建方法,并从某战损试验数据中选取训练样本,进行了实例分析,证明了该方法的可行性和有效性.

Threat Assessment Based on Neural Network

Threat assessment is one of the most important researching contents of battlefield damage simulation.The factors influencing threat assessment are excessive in the number.In addition,it's very hard to establish a reasonable assessment rule or an assessment model of mathematics.Considering the nonlinear characteristics of threat assessment in battlefield damage simulation as well as the advantages of neural network in solving nonlinear complex problems,this paper brings forward an assessment method based on neural network and discusses the method of threat assessment modeling using neural network,which has been proved feasible and effective through analysis of living example.