摘要
由于反舰导弹在速度、机动性等方面的显著潜力,已日益成为舰船的主要威胁。尽快识别反舰导弹类型对缩短系统反应时间,正确预测目标运动具有重要意义。针对从不同目标传感器提取的末制导雷达辐射源参数和弹道特性参数,对目标数据进行相关处理,应用径向基神经网络(RBFNN)分别对反舰导弹模式识别,仿真中充分考虑了各种误差干扰并进行容错性处理,仿真结果表明该算法的有效性。最后将两部分识别结果通过D-S证据理论进行综合决策,进一步提高了系统识别决策的可信度。
For the potential of speed and maneuverability of antiship missile, it is coming to be the main threatener of ship in nowadays. Recognizing the style of antiship missile as soon as possible will be more important to shorten the response time and forecast the movement of the targets exactly. In this article , the pattern recognition of antiship missile is studied from two aspects, the style of the radar and trajectory characteristic using Radial Basis Neural Network. Considering that there must be some error in the pattern eigenvector , we fully improve the tolerance in training. The simulation results show that using neural network to identify the pattern of antiship missile is effective and reliable. Then make an integrated decision using,D-S evidence theory to improve the reliability of the recognition.
出处
《火力与指挥控制》
CSCD
北大核心
2007年第11期42-45,共4页
Fire Control & Command Control
基金
中国船舶工业集团公司船舶系统工程部课题
