极低信噪比条件下主动声纳回波信号的提取

Extraction of echo signals from active sonar under extremely low SNRby using spare decompositon

为优化在极低信噪比条件下对主动声纳回波信号的提取效果,引入了基于非线性理论的稀疏分解方法,并在LDW粒子群算法的基础上,通过改进算法中一个惯性因子的方法对粒子群算法进行了优化。用Matlab软件模拟仿真实验论证表明:改进算法不仅能够加快算法的收敛速度,同时又不降低算法提取信号的性能;在信噪比不低于-32dB的情况下,该方法具有较好的回波提取效果。

In order to optimize the effect of extracting echo signals from the active sonar under the condition of the extremely low SNR, this paper introduces a signal-decomposed method based on the non-linear spare decomposition and optimized the particle group algorithm based on LDW by impro- ving an inertial factor in the algorithm. The Matlab software is used for simulation and experiment. The results show that the improved algorithm can not only increase its convergent speed but also maintain the performance of signal extraction. When SNR is more than --32db, this method functions very well in the extraction of echoes.