| 基于BP神经网络的海上目标检测 |
| |
| 引用本文： | 宗成阁,国磊.基于BP神经网络的海上目标检测[J].东南大学学报(自然科学版),2006(Z1). |
| |
| 作者姓名： | 宗成阁 国磊 |
| |
| 作者单位： | [1]哈尔滨工业大学电子与信息技术研究院 [2]哈尔滨工业大学电子与信息技术研究院 哈尔滨 |
| |
| 摘 要： | 为了减少展宽的一阶海杂波谱对空间超分辨谱估计技术的影响,提高MUSIC算法的检测性能,用BP神经网络实现海杂波预测和对消.利用神经网络的可对任意非线性函数模拟的特性,对展宽的海杂波进行模拟.用模拟后的结果实现一阶海杂波的对消,来满足MUSIC算法的应用条件.最后,用MUSIC算法分辨海上目标的方位信息.实验结果表明,对消前后目标背景噪声子空间特征值发生改变,对消后更接近于MUSIC算法的假设条件,提高了MUSIC算法的检测性能,扩大了MU-SIC算法的应用范围,实现在非高斯噪声背景条件下应用MUSIC算法检测目标.
|
| 关 键 词： | 神经网络 海杂波 高频地波雷达 |
Detection of sea-target based on BP neural network |
| |
| Zong Chengge Guo Lei.Detection of sea-target based on BP neural network[J].Journal of Southeast University(Natural Science Edition),2006(Z1). |
| |
| Authors: | Zong Chengge Guo Lei |
| |
| Abstract: | To decrease the infection of the widened first-order sea clutter echo which changes the noise space of the sea targets,and improve the resolution and direction of arrival(DOA) estimated accuracy of the MUSIC algorithm,the characteristic of the artificial neural network is utilized,which can simulate any non-linear functions to simulate the widened first-order sea clutter.The influence of the sea clutter is weakened by the obtained result,and the azimuth information of the signals is obtained by the MUSIC algorithm.Test results illustrate that the eigenvalue of noise space changes before and after sea clutter eliminating,and the performance of the MUSIC algorithm is improved.It becomes possible to use the MUSIC algorithm in non-gauss noise space. |
| |
| Keywords: | artificial neural network sea clutter high frequency surface wave radar(HFSWR) |
| 本文献已被 CNKI 维普 等数据库收录！ |
|
