摘要
激光自混合干涉技术中颗粒粒度分布反演属于病态问题.为了得到较好的颗粒粒度分布,采用Chahine算法作为非负算法,分别对激光自混合干涉线性和对数形式的功率谱密度进行颗粒粒度分布的反演,比较了两种功率谱密度对应的颗粒粒度分布,并且采用Morrison迭代算法平滑Chahine算法的初始向量,减缓了颗粒粒度分布的振动.仿真结果表明对数形式的功率谱密度反演所得的颗粒粒度分布具有更好的抗噪声能力和稳定性.
Inversion of particle size distribution belongs to ill-condition problem, in order to get relative good distribution,Chahine method was used as nonnegative iterative algorithm for the inversions of linear and logarithmic coordinate frequency power spectral density for estimating particle size distribution （PSD） from self-mixing interference （SMI） signals. The particle size distributions at the two different power spectral densities were also compared. Morrison＇s iterative method was adopted to smooth the initial vector of Chahine method served as a potent approach of alleviating the oscillation in PSD. Simulation results in- dicated that logarithmic form power spectral density had better ability of resisting noise and stability.
出处
《山东理工大学学报（自然科学版）》
CAS
2012年第3期22-27,共6页
Journal of Shandong University of Technology:Natural Science Edition
基金
山东省自然科学基金资助项目(ZR2010FM005)
关键词
激光自混合干涉
功率谱密度
颗粒粒度分布
截断奇异值分解
正则化参数
self-mixing interference
power spectral density
particle size distribution~ truncated singularvalue decomposition
regularization parameter
