纳米颗粒粒径测量的非负最小二乘法算法分析及改进  

Analysis and Improvement on Non-negative Least Square Algorithm in Size Measurement for Nano-particles

作　　者：周述苍[1] 杨燕婷[1] 周誉昌[1] 

机构地区：[1]广东工业大学实验教学部,广东广州510006

出　　处：《中国粉体技术》2009年第5期72-75,共4页China Powder Science and Technology

摘　　要：介绍用于光子相关光谱法颗粒粒径分布计算中的非负最小二乘法,并对其进行改进,增加自动搜索反演范围功能。通过数值模拟分析,比较了在无噪声和多种噪声水平影响下非负最小二乘法和自动搜索反演范围的最小二乘法对单分散和双分散颗粒系的反演结果。研究表明:自动搜索反演范围的最小二乘法用于颗粒粒径的计算能获得更加准确的结果。A non-neqative least square （NNLS） algorithm was presented for inversion computation of particles size distribution in photon correla tion spectroscopy. An auto-adjusting range NNLS （AAR-NNLS） algorithm was developed based on the NNLS algorithm. The numerical tests were performed,which simulated the mono-disperse particulate systems and dual-disperse paniculate systems with zero and nonzero noise poilutions. The comparisons were made between inverse results of NNLS algorithm and those of AAR-NNLS algorithm. The study showed that the results of nano-particles measured by AAR-NNLS algorithm were more accurate.

关 键 词：光子相关光谱法 粒径测量 非负最小二乘法 

分 类 号：O436.2[机械工程—光学工程]