巡航导弹智能自主在线目标识别选择是巡航导弹智能自主作战的重要特征之一.主要针对在巡航导弹弹载计算机计算能力弱、反应时间较短的情况下,如何快速地对目标进行识别分类.采用模板匹配方法进行自动目标识别的大量工作在地面进行,降低了弹载计算机的计算量.为了提高在线智能自主目标识别精度和速度,提出了基于相似度竞争网络,以继承竞争网络在模式识别方面快速简单的优点,同时针对竞争网络抗干扰能力差,对模板类型以外目标无法剔除的不足进行改进,很好解决了巡航导弹在线智能自主目标识别的算法问题.最后通过仿真对算法进行验证.%Intelligent self - selecting attacking target (ISAT) on route is one of the most important characters of the cruise missile's intelligent self - attacking. The paper was Mainly for the problem how to quickly identify and classify targets under tha case that onboard computer in cruise missile calculation ability is weak and the reaction time is shorter. Lots of automatic target recognition work using template matching method were done on ground to reduce the calculation amount of on - board computer. Similar competition neural network ( SCNN) was proposed for improving the precision and the speed of the online intelligent autonomous target recognition. The SCNN inherits the competitive neural network's advantages of fast and simple in pattern recognition, at the same time, the problem of poor antijamming ability of the competition networks and the problem that goals o outside the template type cannot be eliminated, were well solved. At the end of the paper,the algorithm was verified by simulation.
展开▼
