English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

State estimation is an important aspect in many robotics applications. Inthis work, we consider the task of obtaining accurate state estimates forrobotic systems by enhancing the dynamics model used in state estimationalgorithms. Existing frameworks such as moving horizon estimation (MHE) and theunscented Kalman filter (UKF) provide the flexibility to incorporate nonlineardynamics and measurement models. However, this implies that the dynamics modelwithin these algorithms has to be sufficiently accurate in order to warrant theaccuracy of the state estimates. To enhance the dynamics models and improve theestimation accuracy, we utilize a deep learning framework known asknowledge-based neural ordinary differential equations (KNODEs). The KNODEframework embeds prior knowledge into the training procedure and synthesizes anaccurate hybrid model by fusing a prior first-principles model with a neuralordinary differential equation (NODE) model. In our proposed LEARNESTframework, we integrate the data-driven model into two novel model-based stateestimation algorithms, which are denoted as KNODE-MHE and KNODE-UKF. These twoalgorithms are compared against their conventional counterparts across a numberof robotic applications; state estimation for a cartpole system using partialmeasurements, localization for a ground robot, as well as state estimation fora quadrotor. Through simulations and tests using real-world experimental data,we demonstrate the versatility and efficacy of the proposed learning-enhancedstate estimation framework.

LEARNEST: LEARNing Enhanced Model-based State ESTimation for Robots using Knowledge-based Neural Ordinary Differential Equations