Cubature Particle Filter And Its Application In GNSS Positioning

The state of estimation of nonlinear systems has caught the attention of many researchers,and becomes a hot research field with great theoretical value and application field.Because the particle filter is not limited to model characteristics and noise distribution,the particle filter algorithm has its own advantages in such problems,and has been widely applied in many fields such as navigation,statistical signal processing,economics and so on.This paper mainly focuses on the particle filter algorithm and its applications in the navigation system point positioning.The research is mainly concentrated in the following contents:Particle filter theory and theorem have been introduced in detail.Bayesian theory and Monte Carlo algorithm have been introduced.Apply the particle filter to satellite navigation and positioning,the results show that the particle filter has high precision than the classical least squares algorithm.Considering how to design better proposal distribution,take cubature kalman filter as a new proposal distribution function.By virtue of new observations,the accuracy and stability of algorithm are evidently enhanced.The particle degeneracy can be avoided.Numerical example with these filters shows its feasibility and effectiveness.In order to improve the precision when the dynamic model of filter contains errors,adaptive cubature particle filter is put up.Design a new proposal distribution function which contains an adaptive factor.The application in GNSS dynamic point positioning shows that the new method can improve the accuracy.