Based on analyzing divided difference filter(DDF) and Gaussian sum filter(GSF), a GSF-based DDF algorithm
is developed for nonlinear dynamic state space(DSS) models with non-Gaussian noise, which is suitable for the filtering
problem of nonlinear/non-Gaussian systems. When the likelihood function appeares at the tail of the transfer probability
density, the proposed algorithm can improve the precision of nonlinear/non-Gaussian filtering compared with the traditional
particle filter(PF). Experiments show that the proposed method works well in the filtering for DSS models with non-Gaussian
noise.