基于正交匹配追踪的并行采样信号重构方法  

Parallel Sampling Signal Reconstruction Method Based on Orthogonal Matching Pursuit

作　　者：王琳 洪婉君 林文涛 张紫文 胡忞 易朋兴[1] WANG Lin;HONG Wan-jun;LIN Wen-tao;ZHANG Zi-wen;HU Min;YI Peng-xing(School of Mechanical Science and Engineering,Huazhong University of Science and Technology,Wuhan 430074,China)

机构地区：[1]华中科技大学机械科学与工程学院,湖北武汉430074

出　　处：《仪表技术与传感器》2023年第1期121-126,共6页Instrument Technique and Sensor

基　　金：国家重点研发计划项目(2018YFB2003303)。

摘　　要：为解决从超高速并行采样系统的海量数据中提取有效信息进行存储并准确重构信号这一关键问题,提出了一种基于压缩感知的重构方法——基于粒子群优化的预选分段正交匹配追踪(PPStOMP)算法,从少量低维有效信号中精准恢复原始高维信号进行上位机显示。对改进算法本身的主要输入参数进行仿真测试,得到阈值、步长、最大迭代次数在不同采样率下的推荐取值范围。最后利用图像和采样信号作为原始数据,将改进算法与其他正交匹配跟踪算法进行对比实验,实验结果表明PPStOMP算法的重构性能稳定且良好。It is a key problem to extract effective information from massive data of ultra-high speed parallel sampling system for storage and accurate signal reconstruction.To solve this problem,a reconstruction method based on compressed sensing,preselected piecewise orthogonal matching pursuit algorithm based on particle swarm optimization(PPStOMP),was proposed to accurately recover the original high-dimensional signals from a small number of low-dimensional effective signals for upper computer display.The main input parameters of the improved algorithm were simulated and tested,and the recommended ranges of threshold,step size and maximum iteration times at different sampling rates were obtained.Finally,the improved reconstruction algorithm was compared with other orthogonal matching tracking algorithms by using images and sampling signals as original data.The experimental results verify the reconstruction performance of PPStOMP algorithm is stable and great.

关 键 词：并行采样系统 压缩感知 预正交匹配追踪 粒子群优化算法 重构概率 

分 类 号：TN911[电子电信—通信与信息系统]