| 摘要: |
| 为实现对海杂波的抑制，根据海杂波混沌动态特性，利用广义回归神经网络(GRNN)进行海杂波预测再对消，最后引入时间窗方差滤波。分析对McMaster大学IPIX雷达含目标实测数据的处理结果，原始数据信杂比小于等于0 dB，只采用GRNN预测对消后信杂比提高但仍有短时海杂波尖峰的影响，经过方差滤波后短时尖峰基本消失，最终信杂比提高到约11.67 dB。故所提方法对海杂波有很好的抑制效果，能够检测出湮没在海杂波中的小目标。 |
| 关键词: 海杂波抑制 广义回归神经网络 时间窗 方差滤波 预测 对消 |
| DOI： |
|
| 基金项目: |
|
| Sea clutter suppression based on GRNN and time-window variance filtering |
| BI Jing-zhang,LIU Rong,ZHOU Xi-chen,REN Yuan |
| () |
| Abstract: |
| According to the chaotic dynamics of sea clutter,generalized regression neural network(GRNN) is used for sea clutter prediction and cancellation,and time-window variance filtering is applied to suppress sea clutter.Based on the analysis of processing results of radar data with target measured by Intelligent Pixel-Processing(IPIX) radar of McMaster University,the signal to clutter ratio(SCR) is not more than 0 dB.There are short-time sea clutter peaks after GRNN′s prediction and cancellation while the SCR is improved,which can almost all be removed through variance filtering.Finally,the SCR is improved to about 11.67 dB.It is concluded that the proposed method has good cancellation effect to sea clutter,which can detect small target in sea clutter. |
| Key words: sea clutter suppression GRNN time-window variance filtering prediction cancellation |
