The results show that in the simulation of non-linear system model, this framework for Bayesian predictive filter can implement the tracking of simple motion and the orientation prediction.

My main work is applying the particle filter algorithm to random simulate the non-linear Bayesian dynamic models.

Kalman filter is a high efficiency, optimal data process method, which is a realtime recursive filter of linear, non-bias and least square.