摘要
空中目标识别是防空作战辅助决策的关键环节之一。根据空中目标的各种属性,建立基于BP神经网络原理的空中目标识别模型。利用MATLAB神经网络train函数训练采集的样本数据,得到稳定的权值和阈值,为后继的目标识别提供依据。该模型利用神经网络的记忆、联想、容错功能,进一步提高空中目标识别的稳定性和可信度。降低了个别传感器误判而造成的目标识别错误概率。
The efficiency of aerial target identification is one of key process in combat decision-aid for anti-air operation. According to each attribute of air-targets, this paper establishes the identification model on the theory of BP neural network, Using the train function of MATLAB BP neural network to train the data adopted, it gains the stated weight and threshold to provide basis for later coming targets. This model enhances the stability and reliability of the result because of BP Neural Network＇s ability of remembrance association and accommodating error, and reduces the error rate coming from some sensors misjudgment.
出处
《指挥控制与仿真》
2007年第3期30-32,共3页
Command Control & Simulation
作者简介
陈慎刚（1980-），男，湖南岳阳人，硕士研究生，研究方向为海军合同战术效能分析。
许腾（1966-），男，教授，博士生导师。
