摘要
有限等距常数是压缩感知测量矩阵的重要参数之一,例如采用正交匹配追踪精确重构稀疏信号须保证有限等距常数满足一定的条件。但有限等距常数的求解是NP难问题,限制了理论结果的实际应用。将有限等距常数求解视为组合优化问题,采用模拟退火算法得到局部最优解,该解是有限等距常数的下限估计值。实验结果表明估计结果稳定,并发现常见几类测量矩阵有限等距常数大于1,这意味着这些矩阵不满足有限等距性质,此现象需进一步研究解释。
Restricted Isometry Constants(RIC)is one of the most important parameters for Compressed Sensing(CS)measurement matrices.For example,RIC should meet some conditions to ensure exact recovery of sparse signals with orthogonal matching pursuit.However,it is a NP-hard problem to solve out RIC which limits application of theoretical results.It is considered as combinatorial optimization problem to solve out RIC.A local optimum is found by simulated annealing algorithm which is a lower limit of RIC.Experiments show that the estimation is stable,and discover a phenomenon that RICs of several common measurement matrices are greater than one.This phenomenon means that these matrices can not satisfy Restricted Isometry Property and requires further study to make an explanation.
作者
贾彬彬
刘俊莹
JIA Bin-bin;LIU Jun-ying(College of Electrical and Information Engineering,Lanzhou University of Technology,Lanzhou 730050 China;Key Laboratory of Gansu Advanced Control for Industrial Processes,Lanzhou University of Technology,Lanzhou 730050 China;National Demonstration Center for Experimental Electrical and Control Engineering Education,Lanzhou University of Technology,Lanzhou 730050 China)
出处
《自动化技术与应用》
2019年第2期5-7,27,共4页
Techniques of Automation and Applications
基金
甘肃省自然科学基金(编号1610RJYA007
编号1610RJYA026)
甘肃省工业过程先进控制重点实验室开放课题(编号XJK201517)
关键词
压缩感知
测量矩阵
有限等距常数
模拟退火算法
compressed sensing
measurement matrices
Restricted Isometry Constants
simulated annealing algorithm
