基于BP神经网络的炮兵战场目标识别模型

战场目标的识别是一个相当复杂的过程,为了实现识别的自动化和计算机化,采用BP神经网络方法构造数学模型,选取合适的输入、输出及隐性结点,通过反复的学习和测试就可以得到符合实际的结果,从而为指挥员判断敌情提供决策依据.选取常用的音响、闪光、烟尘、机动和配置五种目标特征信息作为输入结点,通过多次仿真测试,说明运用BP神经网络进行战场目标识别是可行的,这也为情报处理自动化系统的实现提供了一个可靠的方法.

Battleground Target Recognition Model for Artillery Based on BP Neural Net

Battleground target recognition is a considerably complex process.In order to achieve the automation and computerization of recognition,we adopt BP neural net means to structure mathematical model.By means of selecting appropriate the import and export and recessiveness connection,and studying and testing again and again,we may obtain results that accord with the practice.Thereby it can supply the decision-making gist for the commander in deciding the enemy's situation.In this paper,we select five normal kinds of target feature message,which are the acoustics and flashing light and smoke dust and mobility and disposition,as the import point,and find that it is feasible to utilize BP neural net to carry on the target discrimination in battleground by means of imitation testing time and again.So the paper supplies a dependable means for achieving the automatic information handling system.