Performance Comparison of the Distributed Extended Kalman Filter and Markov Chain Distributed Particle Filter (MCDPF)

The phenomenon of simple harmonic motion is more vigilantly explained using a simple pendulum. The angular motion of a pendulum is linear in nature. But the analysis of the motion along the horizontal direction is non-linear. To estimate this, several algorithms like the Kalman filter, Extended Kalman Filter etc. are adopted. Here in this paper, Particle filter is chosen which is a method to form Monte Carlo approximations to the solutions of Bayesian filtering equations. Sequential importance resampling based Particle filters are used where the filtering distributions are multi-nodal or consist of discrete state components since under these circumstances the Bayesian approximations do not always work well.

Download Full-text

Performance comparison of particle swarm optimization and extended Kalman filter methods for tracking in non-linear dynamic systems

2017 International Conference on Control, Automation and Diagnosis (ICCAD) ◽

10.1109/cadiag.2017.8075641 ◽

2017 ◽

Cited By ~ 2

Author(s):

Nesrine Amor ◽

Souad Chebbi

Keyword(s):

Particle Swarm Optimization ◽

Kalman Filter ◽

Extended Kalman Filter ◽

Dynamic Systems ◽

Particle Swarm ◽

Performance Comparison ◽

Swarm Optimization ◽

Linear Dynamic Systems ◽

Linear Dynamic ◽

Filter Methods

Download Full-text

A New Linear Regression Kalman Filter with Symmetric Samples

Symmetry ◽

10.3390/sym13112139 ◽

2021 ◽

Vol 13 (11) ◽

pp. 2139

Author(s):

Xiuqiong Chen ◽

Jiayi Kang ◽

Mina Teicher ◽

Stephen S.-T. Yau

Keyword(s):

Kalman Filter ◽

Linear Regression ◽

Particle Filter ◽

Normal Distribution ◽

Extended Kalman Filter ◽

Nonlinear Filtering ◽

Unscented Kalman Filter ◽

Standard Normal Distribution ◽

Leibler Divergence ◽

Standard Normal

Nonlinear filtering is of great significance in industries. In this work, we develop a new linear regression Kalman filter for discrete nonlinear filtering problems. Under the framework of linear regression Kalman filter, the key step is minimizing the Kullback–Leibler divergence between standard normal distribution and its Dirac mixture approximation formed by symmetric samples so that we can obtain a set of samples which can capture the information of reference density. The samples representing the conditional densities evolve in a deterministic way, and therefore we need less samples compared with particle filter, as there is less variance in our method. The numerical results show that the new algorithm is more efficient compared with the widely used extended Kalman filter, unscented Kalman filter and particle filter.

Download Full-text

A Comparison of Extended Kalman Filter, Sigma-Point Kalman Filter, and Particle Filter in GPS/INS Sensor Fusion

AIAA Guidance, Navigation, and Control Conference ◽

10.2514/6.2010-8332 ◽

2010 ◽

Cited By ~ 14

Author(s):

Jason Gross ◽

Yu Gu ◽

Srikanth Gururajan ◽

Brad Seanor ◽

Marcello Napolitano

Keyword(s):

Kalman Filter ◽

Particle Filter ◽

Sensor Fusion ◽

Extended Kalman Filter ◽

Sigma Point

Download Full-text

Performance Comparison of Inertial Measurement Units Fused With Odometry in Extended Kalman Filter for Dead-Reckoning Navigation

Volume 5A: 43rd Mechanisms and Robotics Conference ◽

10.1115/detc2019-98184 ◽

2019 ◽

Author(s):

Kevin Carey ◽

Benjamin Abruzzo ◽

David P. Harvie ◽

Christopher Korpela

Keyword(s):

Kalman Filter ◽

Extended Kalman Filter ◽

Autonomous Vehicle ◽

Dead Reckoning ◽

Ground Truth ◽

Performance Comparison ◽

Inertial Measurement Units ◽

Ground Vehicle ◽

Inertial Measurement ◽

Measurement Units

Abstract This paper aims to aid robot and autonomous vehicle designers by providing a comparison between four different inertial measurement units (IMUs) which could be used to aid in vehicle navigation in a GPS-denied or inertial-only scenario. A differential-drive ground vehicle was designed to carry the multiple different IMUs, mounted coaxially, to enable direct comparison of performance in a planar environment. The experiments focused on the growth of pose error of the ground vehicle originating from the odometry senors and the IMUs. An extended Kalman Filter was developed to fuse the odometry and inertial measurements for this comparison. The four specific IMUs evaluated were: CNS 5000, Xsens 300, Microstrain GX5-35, and Phidgets 1044 and the ground truth for experiments was provided by an Optitrack motion capture system (MCS). Finally, metrics for choosing IMUs, merging cost and performance considerations, are proposed and discussed. While the CNS 5000 has the best objective error specifications, based on these metrics the Xsens 300 exhibits the best absolute performance while the Phidgets 1044 provides the best performance-per-dollar.

Download Full-text