论文部分内容阅读
为了提高动态光散射(DLS)粒径反演精度,考虑到粒径分布(PSD)的非负性及截断奇异值反演(TSVD)法的抗干扰性,比较了信赖域法(Trust)及内点牛顿法(IPN)非负约束的TSVD反演的特性,并结合两者特性,提出了一种Trust与IPN相结合的截断奇异值(Trust-IPN-TSVD)粒径反演方法。该方法继承了Trust-TSVD及IPN-TSVD方法的优点。通过模拟200～500nm单峰、100～700nm双峰分布颗粒Trust、IPN及Trust-IPN三种TSVD方法的反演结果可以看出:相对于Trust-TSVD,Trust-IPN-TSVD最多分别可改善单峰、双峰分布颗粒反演PSD峰值误差、相对误差为1.68%、0.2461,1.41%、0.0587,且它的PSD更平滑;相对于IPN-TSVD,Trust-IPN-TSVD最多可改善单、双峰分布颗粒反演PSD峰值误差、相对误差为4.52%、3.710,9.47%、0.4229,且它的PSD明显变窄。因此,Trust-IPN-TSVD法的反演PSD具有较高的精度及较好的平滑性,更符合理论分布。最后实测颗粒的反演结果证实了该结论。
In order to improve the accuracy of DLS particle size inversion, taking into account the nonnegativeness of particle size distribution (PSD) and the anti-interference of truncated singular value inversion (TSVD) method, In this paper, we propose a Trust-IPN-TSVD particle size inversion method based on the combination of Trust and IPN. This method inherits the advantages of Trust-TSVD and IPN-TSVD methods. By inverting the results of three TSVD methods, namely, Trust, IPN and Trust-IPN, which simulate 200-500 nm single peak and 100-700 nm bimodal distribution particles, it can be seen that the Trust-IPN-TSVD can improve single The peak-bimodal distributions of PSD inversion peak errors, relative errors were 1.68%, 0.2461, 1.41%, 0.0587, and its PSD was smoother. Compared with IPN-TSVD, Trust-IPN- The distribution of particle inversion PSD peak error, the relative error of 4.52%, 3.710,9.47%, 0.4229, and its PSD was significantly narrower. Therefore, the inversion PSD of Trust-IPN-TSVD method has higher accuracy and better smoothness, which is more in line with the theoretical distribution. The inversion of the measured particles confirms this conclusion.