摘要
在目前常用的激光粒度仪无模式反演算方法的优点和缺点基础上,为了实现各种方法的稳定性与准确性的统一,本文中提出一种综合的无模式反演算法——二阶段迭法算法,综合采用了改进共轭梯度法和松弛迭代法各自的优点,在稳定性和准确性方面得到了一定程度的改进。经计算机模拟测量误差实验及国家标准颗粒实验的双重验证:该算法对单峰和多峰的分布具有较准确的分辨能力,且对噪声信号的抗干扰能力较强。
Based on the existing inversion methods, which have been used in the particle size analysis by laser scattering based on the Mie theory, in order to satisfy the stability and accuracy in the applications of differen mathods, a hybrid non-model algorithm was proposed to utilize the specificities of both the improved conjugate gradient method and the successive over relaxation technique (i.e., so-called two segments iteration), which enhanced both the stability and the accuracy in a certain extent. The validation through the simulated noise signal counting for the, algorithm was carried out. The simulated results were in a good agreement With the experimental data obtained with nine standard particulate sampies and their mixtures. It was indicated that this hybrid algorithm could satisfy the resolvability for either mono- or multi-dispersed particulate materials and gave a better stability against the irrelevant signals.
出处
《中国粉体技术》
CAS
2008年第3期10-14,共5页
China Powder Science and Technology
关键词
激光粒度分析
Mie散射矩阵
反演算法
laser scattering particle sizing
Mie scattering theory
inversion algorithm