神经网络的坦克目标威胁评估

Threat Assessment for Tank based on Neural Network

在未来高技术局部战争条件下,坦克不得不面临一个动态复杂的战场环境,这就要求坦克能根据战场环境实时地进行目标威胁评估及自主射击。以坦克的自主作战为背景,提出了一种基于神经网络的目标威胁评估方法。首先提出了包含来袭目标性质、指战员指令、保卫目标的重要度、我方武器性能值等一套威胁评估指标体系;然后采用神经网络方法进行威胁评估。神经网络方法的优势主要体现在:①考虑了诸因素的相互关联;②充分考虑了战术信息;③权系数对具体环境的适应性;④方法的通用性。实例结果表明,神经网络方法科学合理,为坦克目标威胁的评估提供了新思路。神经网络方法具有普遍意义,能推广应用于其它武器平台的目标威胁评估。

In the coming high tech partial combat, Tanks will have to act in a dynamic complex environment,which will constantly require changes in the threat assessment and uninfluenced fire. In this paper,we present an approach to the threat assessment based on the neural network in the context of the uninfluenced combat of the tank. At first, we put forward a set of index system, which including the property of aggressive threat, directorial instruction, the significance of protective target and the performance of our weapon,to the threat assessment. At second ,we implement the threat assessment with neural network. There are four advantages to the neural network,which is： ①considers the conjunction of these factors , ②considers the tactical information, ③the adaptability to dynamic complex environment, ④general usage of this method, An example suggests that the method is scientific and reasonable. The research provides a new idea for threat assessment of tanks. Apparently, the nemml network can be popularized to other threat assessment.