A Two-Stage Bayesian optimisation for Automatic Tuning of an Unscented Kalman Filter for Vehicle Sideslip Angle Estimation

A. Bertipaglia; B. Shyrokau; Mohsen Alirezaei; R. Happee

doi:10.1109/IV51971.2022.9826998

A Two-Stage Bayesian optimisation for Automatic Tuning of an Unscented Kalman Filter for Vehicle Sideslip Angle Estimation

A. Bertipaglia, B. Shyrokau, Mohsen Alirezaei, R. Happee

Intelligent Vehicles

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

1 Citation (Scopus)

53 Downloads (Pure)

Abstract

This paper presents a novel methodology to auto-tune an Unscented Kalman Filter (UKF). It involves using a Two-Stage Bayesian Optimisation (TSBO), based on a t-Student Process to optimise the process noise parameters of a UKF for vehicle sideslip angle estimation. Our method minimises performance metrics, given by the average sum of the states’ and measurement’ estimation error for various vehicle manoeuvres covering a wide range of vehicle behaviour. The predefined cost function is minimised through a TSBO which aims to find a location in the feasible region that maximises the probability of improving the current best solution. Results on an experimental dataset show the capability to tune the UKF in 79.9% less time than using a genetic algorithm (GA) and the overall capacity to improve the estimation performance in an experimental test dataset of 9.9% to the current state-of-the-art GA.

Original language	English
Title of host publication	Proceedings of the 2022 IEEE Intelligent Vehicles Symposium (IV)
Publisher	IEEE
Pages	670-677
ISBN (Electronic)	978-1-6654-8821-1
DOIs	https://doi.org/10.1109/IV51971.2022.9826998
Publication status	Published - 2022
Event	2022 IEEE Intelligent Vehicles Symposium (IV) - Aachen, Germany Duration: 5 Jun 2022 → 9 Jun 2022

Conference

Conference	2022 IEEE Intelligent Vehicles Symposium (IV)
Country/Territory	Germany
City	Aachen
Period	5/06/22 → 9/06/22

Bibliographical note

Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care
Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

Keywords

Training
Intelligent vehicles
Measurement uncertainty
Gaussian processes
Cost function
Bayes methods
Kalman filters

Access to Document

10.1109/IV51971.2022.9826998

A_Two-Stage_Bayesian_optimisation_for_Automatic_Tuning_of_an_Unscented_Kalman_Filter_for_Vehicle_Sideslip_Angle_EstimationFinal published version, 1.04 MB

Cite this

@inproceedings{5ad8312ae7df410abb0ecd1e97856d3c,

title = "A Two-Stage Bayesian optimisation for Automatic Tuning of an Unscented Kalman Filter for Vehicle Sideslip Angle Estimation",

abstract = "This paper presents a novel methodology to auto-tune an Unscented Kalman Filter (UKF). It involves using a Two-Stage Bayesian Optimisation (TSBO), based on a t-Student Process to optimise the process noise parameters of a UKF for vehicle sideslip angle estimation. Our method minimises performance metrics, given by the average sum of the states{\textquoteright} and measurement{\textquoteright} estimation error for various vehicle manoeuvres covering a wide range of vehicle behaviour. The predefined cost function is minimised through a TSBO which aims to find a location in the feasible region that maximises the probability of improving the current best solution. Results on an experimental dataset show the capability to tune the UKF in 79.9% less time than using a genetic algorithm (GA) and the overall capacity to improve the estimation performance in an experimental test dataset of 9.9% to the current state-of-the-art GA.",

keywords = "Training, Intelligent vehicles, Measurement uncertainty, Gaussian processes, Cost function, Bayes methods, Kalman filters",

author = "A. Bertipaglia and B. Shyrokau and Mohsen Alirezaei and R. Happee",

note = "Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.; 2022 IEEE Intelligent Vehicles Symposium (IV) ; Conference date: 05-06-2022 Through 09-06-2022",

year = "2022",

doi = "10.1109/IV51971.2022.9826998",

language = "English",

pages = "670--677",

booktitle = "Proceedings of the 2022 IEEE Intelligent Vehicles Symposium (IV)",

publisher = "IEEE",

address = "United States",

}

Bertipaglia, A , Shyrokau, B, Alirezaei, M & Happee, R 2022, A Two-Stage Bayesian optimisation for Automatic Tuning of an Unscented Kalman Filter for Vehicle Sideslip Angle Estimation. in Proceedings of the 2022 IEEE Intelligent Vehicles Symposium (IV). IEEE, pp. 670-677, 2022 IEEE Intelligent Vehicles Symposium (IV), Aachen, Germany, 5/06/22. https://doi.org/10.1109/IV51971.2022.9826998

A Two-Stage Bayesian optimisation for Automatic Tuning of an Unscented Kalman Filter for Vehicle Sideslip Angle Estimation. / Bertipaglia, A.; Shyrokau, B.; Alirezaei, Mohsen et al.
Proceedings of the 2022 IEEE Intelligent Vehicles Symposium (IV). IEEE, 2022. p. 670-677.

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - A Two-Stage Bayesian optimisation for Automatic Tuning of an Unscented Kalman Filter for Vehicle Sideslip Angle Estimation

AU - Bertipaglia, A.

AU - Shyrokau, B.

AU - Alirezaei, Mohsen

AU - Happee, R.

N1 - Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

PY - 2022

Y1 - 2022

N2 - This paper presents a novel methodology to auto-tune an Unscented Kalman Filter (UKF). It involves using a Two-Stage Bayesian Optimisation (TSBO), based on a t-Student Process to optimise the process noise parameters of a UKF for vehicle sideslip angle estimation. Our method minimises performance metrics, given by the average sum of the states’ and measurement’ estimation error for various vehicle manoeuvres covering a wide range of vehicle behaviour. The predefined cost function is minimised through a TSBO which aims to find a location in the feasible region that maximises the probability of improving the current best solution. Results on an experimental dataset show the capability to tune the UKF in 79.9% less time than using a genetic algorithm (GA) and the overall capacity to improve the estimation performance in an experimental test dataset of 9.9% to the current state-of-the-art GA.

AB - This paper presents a novel methodology to auto-tune an Unscented Kalman Filter (UKF). It involves using a Two-Stage Bayesian Optimisation (TSBO), based on a t-Student Process to optimise the process noise parameters of a UKF for vehicle sideslip angle estimation. Our method minimises performance metrics, given by the average sum of the states’ and measurement’ estimation error for various vehicle manoeuvres covering a wide range of vehicle behaviour. The predefined cost function is minimised through a TSBO which aims to find a location in the feasible region that maximises the probability of improving the current best solution. Results on an experimental dataset show the capability to tune the UKF in 79.9% less time than using a genetic algorithm (GA) and the overall capacity to improve the estimation performance in an experimental test dataset of 9.9% to the current state-of-the-art GA.

KW - Training

KW - Intelligent vehicles

KW - Measurement uncertainty

KW - Gaussian processes

KW - Cost function

KW - Bayes methods

KW - Kalman filters

UR - http://www.scopus.com/inward/record.url?scp=85135370465&partnerID=8YFLogxK

U2 - 10.1109/IV51971.2022.9826998

DO - 10.1109/IV51971.2022.9826998

M3 - Conference contribution

SP - 670

EP - 677

BT - Proceedings of the 2022 IEEE Intelligent Vehicles Symposium (IV)

PB - IEEE

T2 - 2022 IEEE Intelligent Vehicles Symposium (IV)

Y2 - 5 June 2022 through 9 June 2022

ER -

A Two-Stage Bayesian optimisation for Automatic Tuning of an Unscented Kalman Filter for Vehicle Sideslip Angle Estimation

Abstract

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this