基于小波包分析及支持向量机的超音速目标识别方法研究  

Identifying supersonic projectiles based on wavelet packet analysis and SVM method

作　　者：字正华[1] 石庚辰[1] 

机构地区：[1]北京理工大学机电工程学院,北京100081

出　　处：《物理实验》2004年第8期12-14,18,共4页Physics Experimentation

摘　　要：提出了基于小波包分析及支持向量机的超音速目标识别方法 .通过 5 .5 6mm ,7.6 2mm和 12 .7mm三种枪弹试验获取信号 ,用小波包分解激波信号 ,统计每个频带的能量特征 ,用支持向量机方法训练测试样本 ,获得了很好的分类效果 .仿真结果表明基于超音速飞行体产生的激波信号来识别目标是可行的 .A identifying method of supersonic projectiles based on wavelet packet analysis and support vector machine(SVM) methods is proposed. By experimental analysis for 5.56 mm, 7.62 mm and 12.7 mm projectiles, decomposed signals by wavelet packet analysis and learning and training samples with method of SVM are gotten, effective classifications are arrived. It is confirmed that identifying supersonic projectiles based on shock wave signal is feasible.

关 键 词：小波包分析 支持向量机 激波信号 超音速目标识别 

分 类 号：O429[理学—声学]