Abstract
Safe and reliable operation of lithium-ion batteries as major energy storage devices influences the service lifetime of several electronic equipment, whose unexpected failures could result in enormous economic and societal losses. Accurate estimation of the state-of-charge (SoC) and state-of-health (SoH) for an operating battery is a critical issue for battery health management, which greatly depends on physical battery models. Due to the variability and uncertainties involved in battery design, manufacturing, and operation, developing a generally applicable battery physical model is a big challenge. This paper presents a generic data-driven approach for lithium-ion battery health management that eliminates the dependency of battery physical models for SoC and SoH estimation. This approach integrates an artificial neural network (ANN) with a dual extended Kalman filter (DEKF) algorithm, where the ANN is trained offline with cyclic battery discharging data to approximate the battery terminal voltage to be used by the DEKF. With the trained ANN, the DEKF algorithm can be employed online for SoC and SoH estimation, where voltage outputs from the trained ANN are used in the DEKF state-space equations to replace the required battery physical model outputs. Experimental results are used to demonstrate the effectiveness of the developed model-free approach for battery health management.
Original language | English (US) |
---|---|
Pages | 3291-3300 |
Number of pages | 10 |
State | Published - 2013 |
Externally published | Yes |
Event | IIE Annual Conference and Expo 2013 - San Juan, Puerto Rico Duration: May 18 2013 → May 22 2013 |
Other
Other | IIE Annual Conference and Expo 2013 |
---|---|
Country/Territory | Puerto Rico |
City | San Juan |
Period | 5/18/13 → 5/22/13 |
Keywords
- Dual extended Kalman filter (DEKF)
- Lithium-ion battery
- Neural Network (NN)
- State-of-Charge (SoC)
- State-of-Health (SoH)
ASJC Scopus subject areas
- Industrial and Manufacturing Engineering