Noisy Data Processing In Dynamic Light Scattering

Dynamic light scattering (DLS) technology can get the particle size by studying the scattering intensity signal fluctuations over time. In DLS data processing, due to the existence of ill conditioned equation, scattering light signal even with very small noise will seriously affect the accuracy of the measurement result, we should try to avoid the interference of noise. In the process of measurement, the existence of the noise is always inevitable, and in recent years, with the application of DLS technology, the demand of measurements under the strong noise environment get more and more urgent, so studying the effect of noise on measurement results and choosing appropriate technicals deal with the noisy DLS data are very important to obtain accurate measurement results. This paper mainly studied the selection of regularization initial model, regularized inversion of DLS data considering noise as an independent variate, the influence of noise to the bimodal particle size inversion and the effect of angle calibration noise on unimodal and bimodal particle size distribution (PSD) results. The main researches of this paper consists of,1. Analyzed the influence of initial model on the PSD inversion results. The narrow and wide simulation distributions of 90nm and 250nm were inversed by the smallest, the flattest and the smoothest initial model respectively. The inversion results show that the initial model has almost no influence on the inversion results under the noise level of 0. With the increase of noise level, although the errors of peak and PSD value inversed by the three initial models are all increased, the increases by using flattest model and smoothest model are less than using the smallest model obviously. When the noise level is greater than 0.01, the better particle size distribution results can be obtained by using the flattest model than the smallest and the smoothest models, and more accurate particle peak values can be got by using the smoothest model than the flattest and the smallest models.2. The PSDs inversion using a modified regularization algorithm was studied. Noises in the DLS data were considered as an independent variate in the original equation, and the coefficient matrix changed accordingly to fit the idea, then according to the dimension of coefficient matrix the bad PSD data was piked out from the recovered PSDs. PSDs inversion results under different noise level show that the modified regularization algorithm considers noise as an independent variate alleviate the influence of noise in the data and improve the inversion accuracy of low SNR DLS data obviously.3. The influence of noise on bimodal PSD and the angle calibration noise in multiangle dynamic light scattering (MDLS) were studied. The inversion results show that in the process of inversion bimodal small PSD the antinoise ability of regularization inversion algorithm superior to the antinoise ability of Chahine inversion algorithm. Using MDLS technology to inverse bimodal big PSD, the inversion results show that with the increase of number of scattering angle, the inversion results more and more good, and influence by noise decreased with the increase of the number of scattering angle, when the number of scattering angle is six, inversion results is best and influenced by noise smallest, when the number of scattering angle is more than six, inversion result from bad to worse instead, antinoise ability is becoming more and more weak too. The angle calibration noise in MDLS has certain influence of bimodal PSDs and the bimodal PSDs containing small particles are more easily effected by angle calibration noise.Although DLS technique is widely used, but the presence of noise in the technology and its influence on measurement results has always been a problem of concern, in this paper, the research of noisy data processing, makes every effort to achieve the accurate measurement of DLS technology.