针对海杂波信号因混有噪声而难以提取的特点,提出基于改进的集成经验模态分解(MEEMD)的海杂波去噪方法。文中提出的MEEMD 在补充的集成经验模态分解(CEEMD)的基础上,利用排列熵和Savitzky-Golay 滤波对CEEMD 分解后的固有模态函数进行处理,最后在经验模态分解分解重构后得到削噪后的信号。以IPIX 雷达实测得到的海杂波数据进行仿真实验,结合最小二乘支持向量机建立混沌序列的单步预测模型,从预测误差中检测淹没在海杂波背景中的微弱信号,并用均方根误差判断去噪效果。仿真结果表明,文中所提出的MEEMD 算法对模式混淆有很好的抑制效果,去噪后得到的均方根误差为0. 000 847,比去噪前的均方根误差0. 012 2 降低了两个数量级。

In view of the characteristics of sea clutter signals which are difficult to extract due to mixed noises, an improved global mean empirical mode decomposition (MEEMD) based denoising method for sea clutter is proposed. The MEEMD proposed in this paper is based on the CEEMD decomposition and uses permutation entropy and Savitzky-Golay filtering to deal with it. Finally, the denoised signal is obtained after EMD decomposition and reconstruction. In order to get the IPIX radar sea clutter data simulation experiment, combined with least squares support vector machine to establish single chaotic sequence prediction model, the prediction errors from detection of weak signal submerged in the sea clutter wave in the background, and the root mean square error of judgment denoising effect. The simulation results show that the proposed MEEMD has good inhibition effect on mode confusion, the root mean square error after denoising was 0. 000 847, 0. 012 2 more than the root mean square error of denoising before two orders of magnitude lower.

TN911.7

国家自然科学基金资助项目(61671248);江苏省高校自然科学研究重大项目(15KJA460008);江苏省“信息与通信工程冶优势学科资助,江苏省研究生科研与实践创新计划项目(KYLX160955)