Multifeature-based importance weighting for the PHD SLAM filter
Keith Yu Kit Leung, Felipe Inostroza, Martin David Adams · IEEE Transactions on Aerospace and Electronic Systems · 2016
The probability-hypothesis-density simultaneous localization and mapping filter is a random-finite-set estimation method that incorporates the probability-hypothesis-density filter within a Rao-Blackwellized particle filter, and was developed for navigation and mapping problems. However, the filter tends to diverge due to the existing importance-weighting methods used in the Rao-Blackwellized particle filter. This article introduces a new importance-weighting method that drastically improves the robustness of the probability-hypothesis-density simultaneous localization and mapping filter. Performance evaluations are conducted using both simulations and real experimental data sets.