摘要
瞄准新一代激光对地近炸引信对地面目标精确识别的急切需求,基于激光引信弹目交会回波进行KNN算法的地面目标识别探索研究。首先通过雷达方程建立激光近炸引信的弹目交会回波方程;其次对回波数据进行归一化处理并将数据集划分为训练数据集和测试数据集;然后提取回波数据的强散射点个数和等效目标尺寸特征;最后将得到的特征子集进行KNN算法的目标识别训练和测试。实验结果表明,基于KNN算法的识别方法准确率高,平均识别精度为86.3%。该算法能够较好地实现坦克、雷达车及装甲车三种不同地面目标的精确识别。
Aiming at the urgent demand for accurate recognition of ground target by the new-generation of laser proximity fuze,an exploration study on ground target recognition using KNN algorithm based on the echo of the mission-target encounter of laser fuze was conducted.Firstly,the echo equation of mission-target encounter was established by the radar equation.Secondly,the echo data was processed by normalization,and the data was divided into the training dataset and the test dataset.Then,the data was subjected to feature extraction,such as the number of strong scattering points,equivalent target size and effective length.Finally,the obtained feature subset was used for target recognition training and testing based on KNN algorithm.The experimental results indicated that the recognition method based on the KNN algorithm had high accuracy,with an average recognition accuracy of 86.3%.The KNN algorithm could better realize the accurate recognition of three different ground targets of tanks,radar vehicles and armored vehicles.
作者
何海军
田博
胡鹏飞
李林豪
王伟
李铁
李岗
HE Haijun;TIAN Bo;HU Pengfei;LI Linhao;WANG Wei;LI Tie;LI Gang(Xi'an Institute of Electromechanical Information Technology,Xi'an 710065,China;Science and Technology on Electromechanical Dynamic Control Laboratory,Xi'an 710065,China)
出处
《探测与控制学报》
北大核心
2025年第4期134-140,共7页
Journal of Detection & Control
基金
机电动态控制重点实验室开放课题基金资助(6142601012301)。
作者简介
何海军(1991-),男,甘肃天水人,博士,工程师。
