by Lotfi, B and Huang, L
Abstract:
This paper proposes an approach for velocity and position estimation through acceleration measurements. In this approach, the noises of acceleration signal are removed through a smoothing process based on diffusion equation, a special partial differential equation (PDE). Then a mechanism is designed to remove abnormal deviations in the signal before and after integration. The proposed approach can substantially reduce the errors (drifts) accumulated from the integration of acceleration signal. Its effectiveness and superiority over other popular methods are validated by simulation and experimental results.
Reference:
An approach for velocity and position estimation through acceleration measurements (Lotfi, B and Huang, L), In Measurement: Journal of the International Measurement Confederation, Elsevier, volume 90, 2016.
Bibtex Entry:
@article{lotfi2016anmeasurements,
author = "Lotfi, B and Huang, L",
journal = "Measurement: Journal of the International Measurement Confederation",
month = "Aug",
pages = "242--249",
publisher = "Elsevier",
title = "An approach for velocity and position estimation through acceleration measurements",
volume = "90",
year = "2016",
abstract = "This paper proposes an approach for velocity and position estimation through acceleration measurements. In this approach, the noises of acceleration signal are removed through a smoothing process based on diffusion equation, a special partial differential equation (PDE). Then a mechanism is designed to remove abnormal deviations in the signal before and after integration. The proposed approach can substantially reduce the errors (drifts) accumulated from the integration of acceleration signal. Its effectiveness and superiority over other popular methods are validated by simulation and experimental results.",
doi = "10.1016/j.measurement.2016.04.011",
issn = "0263-2241",
keyword = "Double integration",
keyword = "Drift",
keyword = "Inertial navigation systems",
language = "eng",
pii = "S0263224116300689",
day = "1",
}