Year: 2023
Type: dissertation
Abstract: State of charge (SOC) estimation is important for marine battery uses. Running on empty battery on a sea is not an ideal situation, neither is overcharging a battery. SOC is the ratio between the current battery charge and the maximum battery capacity. The challenge of SOC estimation is that it is not directly measurable. \n \nThis thesis maps SOC estimation techniques and lithium-ion battery (LiB) models used with them as a literature review. Then a programmable logic controller (PLC) compatible estimation method and model are chosen for implementation. Tests are ran on a LiB to parameterize the selected model and to validate the selected SOC estimation technique so that the absolute error of SOC stays within 20 percentage points. \n \nSteady-state Kalman filter (SSKF) is chosen for implementation because of computational limits. Extended Kalman filter (EKF) is also validated. LiB dynamics are modeled as autoregressive model with exogenous input (ARX). The model is augmented with Coulomb counter and SOC as a function of open circuit voltage (OCV). Both the SSKF and EKF were almost equally accurate, with the EKF being slightly more accurate. The SSKF did not work properly when SOC was under 20 %.
Source: LUTPub (LUT University)
Author: Jantunen, Jyrki
Institution: Lappeenranta-Lahti University of Technology
Language: English
Subfield: Automotive Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7071244991
