Inversion algorithm of photon correlation spectroscopy by combining linear and nonlinear estimation

Double exponential algorithm of photon correlation spectroscopy (PCS) is realized by the nonlinear estimation, its inversion accuracy is affected by four initial parameters choice. According to linear relationship of the decay linewidth distribution coefficients and exponential terms of autocorrelation function (ACF), a new inversion algorithm combining linear and nonlinear estimation is proposed. By changing four parameters optimization problem into two parameters optimization problem in turns, new algorithm improves the inversion accuracy and reduces two initial parameters. By new and tradition algorithm, ACF of 200nm - 600nm bi-dispersed particles is respectively inversed. For noise-free ACF, inversion error of new algorithm is 0, which is at least better than tradition algorithm 2.01%. For inversion of noise ACF, compared with tradition algorithm, new algorithm can reduce inversion error by 0.93% ~ 6.43%. Therefore, new algorithm is superior to tradition algorithm in inversion accuracy and tolerance of noises.