Gaussian Mixture Filter in Hybrid Navigation
Simo Ali‐Löytty, Niilo Sirola · 2007
This paper studies the Gaussian Mixture Filter (GMF) in hybrid navigation and is divided into three parts focusing on different problems. The first part deals with approximating the likelihood function as a Gaussian mixture. The second part gives a new way to robustify GMF by adding more components, and the third part concentrates on different ways to reduce the number of mixture components so that the algorithm is fast without losing significant information. Finally, we compare GMF with other filters using simulations.