A Comparative Analysis of Techniques for Electric Vehicle Battery Prognostics and Health Management (PHM)

Paper #:
  • 2011-01-2247

  • 2011-09-13
Rezvani, M., AbuAli PhD, M., Lee, S., Lee, J. et al., "A Comparative Analysis of Techniques for Electric Vehicle Battery Prognostics and Health Management (PHM)," SAE Technical Paper 2011-01-2247, 2011, https://doi.org/10.4271/2011-01-2247.
Batteries are widely used as storage devices and they have recently gained popularity due to their increasing smaller sizes, lighter weights and greater energy densities. These characteristics also render them suitable for powering electric vehicles. However, a key gap exists in that batteries are solely used as storage devices with a lack of information flow. Next-generation battery technologies will constitute the enabling tools that would lead to information-rich batteries, thus allowing the transparent assessment of a battery's health as well as the prediction of a battery's remaining-useful-life (RUL) and its subsequent impact on vehicle mobility. Various methods and techniques have been employed to predict battery RUL in order to improve the accuracy of the State of Charge (SoC) estimation. This paper presents a comparative study of emerging prognostics and health management (PHM) techniques that can give an accurate quantification of the State of Health (SoH) of Li-ion battery cells and predict their remaining useful life. Two models, Adaptive Neural Network (AdNN) and Linear Prediction Error Method (L-PEM), will be used for battery capacity estimation and remaining useful life prediction. Their prediction performance (i.e. accuracy, robustness, sensitivity, etc.) is benchmarked using three Li-ion data sets. It can be concluded that both algorithms can successfully estimate battery capacity one-step-ahead and provide a remaining useful life of the rated capacity which is highly correlated to the battery health. It was observed that Adaptive Neural Networks provide a more accurate capacity estimation of one-step ahead while PEMs showed higher performance in remaining useful life prediction. Such prognostic capabilities applied to everyday battery technologies can digest large amounts of raw battery data (ex. voltage, current, impedance, etc.) and convert it to useful battery health and risk information. This can provide further intelligence to existing on-board Battery Management Systems (BMS) to relate battery health condition to vehicle mobility.
SAE MOBILUS Subscriber? You may already have access.
Members save up to 42% off list price.
HTML for Linking to Page
Page URL

Related Items

Technical Paper / Journal Article