Gaussian Mixture Filtering with Nonlinear Measurements Minimizing Forward Kullback-Leibler Divergence

Research on monitoring and analysis technology of power data flow interface in external interaction scenarios

“…In formula (1), ( ) proportion of the distance between the two items in the filtering model [6][7]. By using the KL distance of the data, the similarity calculation formula for two types of items can be obtained, as shown in formula (2):…”

Dynamic rEvolution : Adaptive state estimation via Gaussian processes and iterative filtering

“…This approach aims to minimize some divergence measure, typically the kl divergence, between the true and an assumed posterior [118,119]. This approach has been used to develop variational Gaussian filters [119], but has also recently been extended to Gaussian sum filters [120]. These approaches are generally applicable to any ssm but typically involve complex high-dimensional optimization problems.…”

“…Extending the methods to Gaussian mixture filtering could be one avenue of exploration. While iterated filtering has been developed for Gaussian mixture filtering [120], dynamically iterated analogs are still missing and should be straightforwardly available through the lens of Paper D. Exploring connections to the recently proposed Bayesian recursive update filter (bruf) [138] may also be of interest. In [138], the bruf is pointed out to (essentially) be a single particle implementation of a particle flow filter.…”