Review of sampling-data analysis methods in nonlinear optimum filtering algorithm

Based on the Bayesian parameters estimation theory, prospecting from the reducing amount of calculation and improving computational efficiency, the deterministic sampling methods such as the GaussHermite filtering, extended Kalman filtering, and Sigma-points Kalman filtering algorithms and the random sampling methods of particles filtering algorithms were reviewed, the research field and development direction of Bayesians optimal theory were presented, which could design the new efficient and high-precision particle filtering algorithm from the sampling function design, resampling technique and Gauss approximate method. At the same time, designing the box particle filtering algorithms became the new idea of the particle filtering algorithm.