基于似零范数和混合优化的压缩感知信号快速重构算法.pdfVIP

第 40 卷 第 10 期 自 动 化 学 报 Vol. 40, No. 10 2014 年 10 月 ACTA AUTOMATICA SINICA October, 2014 基于似零范数和混合优化的压缩感知信号快速重构算法 伍飞云 1 周跃海 1 童 峰 1 摘 要 欠定系统 (又称超完备系统) 的稀疏信号恢复在压缩感知、源信号分离和信号采集等领域中被广泛研究. 目前这类问 题主要采用 l1 范数约束结合线性规划优化或贪婪算法进行求解, 但这些方法存在收敛速度慢、恢复精度不高等缺陷. 提出一 种快速恢复稀疏信号的算法, 该算法采用一种新的近似 l0 范数代替 l1 范数构造代价函数, 并融合牛顿法和最陡梯度法推导出 寻优迭代式, 以获得似零范数代价函数的最优解. 仿真实验和真实数据实验结果表明, 与经典算法相比, 该算法在能提供相同 精度、甚至更好精度的条件下, 收敛速度更快. 关键词 范数约束, 稀疏信号恢复, 似零范数, 稀疏水声信通, 压缩感知 引用格式 伍飞云, 周跃海, 童峰. 基于似零范数和混合优化的压缩感知信号快速重构算法. 自动化学报, 2014, 40(10): 2145?2150 DOI 10.3724/SP.J.1004.2014.02145 A Fast Sparse Signal Recovery Algorithm Based on Approximate l0 Norm and Hybrid Optimization WU Fei-Yun1 ZHOU Yue-Hai1 TONG Feng1 Abstract Obtaining sparse solutions of under-determined, or over-complete, linear systems of equations has found extensive applications in signal processing of compressive sensing, source separation and signal acquisition. However, the previous approaches to this problem, which generally minimize the l1 norm using linear programming (LP) techniques or greedy methods, are subject to drawbacks such as low accuracy and slow convergence. This paper proposes to replace the l1 norm with a newly de?ned approximate l0 norm (AL0), the optimization of which leads to the derivation of a hybrid approach by incorporating the steepest descent method with the Newton iteration. Numerical s