Application of the Improved Particle Filter Algorithm to Autonomous Celestial Navigation for Deep-Space Exploration
Jianguo Li · Ceshi jishu xuebao · 2009
Large initial state error and non-Gaussian distribution of state may exist in the deep-space autonomous celestial navigation,so based on the classic six orbit elements,an orbit determination method for deep-space probe is proposed by using starlight angle as measurement information and the unscented particle filter(UPF) algorithm.UPF can gain higher filter accuracy and fewer computation burden by controlling the resampling frequency.Finally,the simulation results indicate that the autonomous celestial method and the improved particle filter algorithm are feasible and efficacious.