TY - JOUR
T1 - State of charge estimator based on tractable extended state observers for supercapacitor packs
AU - Sandoval-Chileño, M. A.
AU - Lozada-Castillo, N.
AU - Cortez, R.
AU - Luviano-Juárez, A.
AU - Vazquez-Arenas, J.
N1 - Publisher Copyright:
© 2024 Elsevier Ltd
PY - 2024/4/30
Y1 - 2024/4/30
N2 - A state of charge estimator (SOCE) is proposed for a single and a supercapacitor pack aimed at easy implementation and low computational cost extended state observers. This SOCE reduces modeling errors by implementing extended states using the classic supercapacitor equation. The proposed method involves a simple characterization of supercapacitors, which are tested using different discharge currents up to 40 [A] and subjected to added measurement noise. The proposed estimator is compared against different performance indexes, including the Unscented Kalman filter (UKF), unscented Kalman filter with parameter adaptation (PA), Extended Kalman Filter (EKF), Kalman filter with open-circuit voltage (OCV), Particle Filter (PF), and H-infinity estimator. This comparison was conducted using discharge profiles under different experimental conditions, where the SOCE herein proposed leads to the lowest error among all documented indexes and a performance improvement between 5.53 [%] and 99.90 [%]. Similarly, SOCE facilitates the characterization of the model inputs, reducing the preprocessing time while improving the execution time by 37 [%], and its estimation error by up to 86.77 [%]. These features enable the robust online implementation of embedded systems, improving SOC estimation.
AB - A state of charge estimator (SOCE) is proposed for a single and a supercapacitor pack aimed at easy implementation and low computational cost extended state observers. This SOCE reduces modeling errors by implementing extended states using the classic supercapacitor equation. The proposed method involves a simple characterization of supercapacitors, which are tested using different discharge currents up to 40 [A] and subjected to added measurement noise. The proposed estimator is compared against different performance indexes, including the Unscented Kalman filter (UKF), unscented Kalman filter with parameter adaptation (PA), Extended Kalman Filter (EKF), Kalman filter with open-circuit voltage (OCV), Particle Filter (PF), and H-infinity estimator. This comparison was conducted using discharge profiles under different experimental conditions, where the SOCE herein proposed leads to the lowest error among all documented indexes and a performance improvement between 5.53 [%] and 99.90 [%]. Similarly, SOCE facilitates the characterization of the model inputs, reducing the preprocessing time while improving the execution time by 37 [%], and its estimation error by up to 86.77 [%]. These features enable the robust online implementation of embedded systems, improving SOC estimation.
KW - Extended state observers
KW - Online estimator
KW - State of charge
KW - Supercapacitor pack
UR - http://www.scopus.com/inward/record.url?scp=85186266878&partnerID=8YFLogxK
U2 - 10.1016/j.est.2024.111086
DO - 10.1016/j.est.2024.111086
M3 - Artículo
AN - SCOPUS:85186266878
SN - 2352-152X
VL - 85
JO - Journal of Energy Storage
JF - Journal of Energy Storage
M1 - 111086
ER -