Hybrid filtering
The signal \(x(t)\in \mathbb{R}^{n_1}\) and observation \(y(t)\in\mathbb{R}^{n_2}\) are governed by the following linear hybrid system: \[ \begin{aligned} \dot x(t) &= A(t,\theta(t)) x(t)+ A_1(t,\theta(t))+ u(t,\theta(t)),\\ y(t) &= H(t,\theta(t)) x(t)+ H_1(t,\theta(t))+ v(t,\theta(t)).\end{aligned} \] Here \(u\) and \(v\) are some disturbances, \(\theta(t)\) is a hidden switching process the unknown probability distribution of which is among a finite number of candidate distributions. The most probable trajectory approach is used to derive a finite-dimensional recursive filter. For the entire collection see [Zbl 1005.00048].
- Optimal filtering of discrete-time hybrid systems
- Trajectory encoding for systems with irregular observations
- The salted Kalman filter: Kalman filtering on hybrid dynamical systems
- Bayesian filtering of stochastic hybrid systems in discrete-time and interacting multiple model
- scientific article; zbMATH DE number 4190664 (Why is no real title available?)
- Two-time-scale Hybrid Filters: Near Optimality
- Filtering for hybrid systems with Markovian and deterministic switching parameters(∗)
- A finite-dimensional filter for hybrid observations
- Hybrid filtering for linear systems with non-Gaussian disturbances
- scientific article; zbMATH DE number 7639971 (Why is no real title available?)
This page was built for publication: Hybrid filtering
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4438229)