Data-Driven State of Health Estimation for Lithium-Ion Batteries Based on Universal Feature Selection

Abstract

A simple yet effective health indicator (HI)-based data-driven model forecasting the state of health (SOH) of lithium-ion batteries (LIBs) and thus enabling their efficient management is developed. Five HIs with high physical significance and predictive power extracted from voltage, current, and temperature profiles are used as model inputs. The generalizability and robustness of the proposed ridge regression–based linear regularization model are assessed using three NASA datasets containing information on the behavior of batteries over a wide range of temperatures and discharge rates. The maximum mean absolute error, maximum root-mean-square error, and maximum mean absolute percentage error of the SOH for the three groups of batteries are determined as 0.7%, 0.86%, and 2.1%, respectively. Thus, the developed model exhibits high accuracy in estimating the SOH of LIBs under multiworking conditions and is sufficiently robust to be applicable to low-quality datasets obtained under other conditions.

A self‐adaptive, data‐driven method to predict the cycling life of lithium‐ion batteries

“…However, the charging rate is much larger than the discharging rate in most cases, which can induce a quicker degradation of a battery's cycling life in the charging process than in the discharging process. [45][46][47][48] An overall consideration of both charging and discharging information is supposed to achieve a high-accuracy prediction of battery SOH and unveil the underlying degradation mechanisms. [49][50][51][52] Finally, the temperature control also impacts the electrochemical polarization and then leads to the degradation trajectory fluctuation in batteries.…”

“…Second, the information behind the discharging process was considered in the current model. However, the charging rate is much larger than the discharging rate in most cases, which can induce a quicker degradation of a battery's cycling life in the charging process than in the discharging process 45–48 . An overall consideration of both charging and discharging information is supposed to achieve a high‐accuracy prediction of battery SOH and unveil the underlying degradation mechanisms 49–52 .…”

Modeling Rate Dependent Volume Change in Porous Electrodes in Lithium-Ion Batteries

“…Considering the design of modules and packs for electric vehicles, the intercalation induced expansion and aging induced expansion of the active material and electrodes must be taken into account. 31 In the module designs for the Chevrolet Volt 32 and Chevrolet Bolt, 33 the cell is a pouch 34 type format and the volume change is mitigated using foam inserts within a rigid structural module design. When considering a module with cylindrical cells, the volume change is generally assumed to be contained by the can and deformation of the can is ignored when considering the design of the module and pack, however, when considering prismatic cells and pouch cells, the volume change in these cells cannot be ignored and must be considered due to the expansion seen in the direction perpendicular to the orientation of the electrodes.…”

State-of-Health Estimation of Lithium-ion Batteries Based on Singular Value Decomposition and an Improved Group Method of Data Handling

“…As an important energy storage device, lithium-ion batteries are widely used in many fields such as electric vehicles and portable electronic products due to their long life, high power, wide operating temperature and other advantages. [1][2][3] Lithium-ion batteries will gradually age during use, which is a non-linear process. There are many factors that cause the degradation of lithium-ion batteries, such as the reduced lithium inventory, the excessive charging and discharging, and the temperature changes.…”

“…[37][38][39] This paper has the following contributions: (1) SVD is used for feature extraction, which reduces the complexity of feature extraction while avoiding the influence of human experience. (2) The PSO algorithm is used to optimize the key parameters of GMDH and a PSO-GMDH estimation model is established. Compared with GMDH without PSO optimization, the proposed model has a better network structure and effectively improves the accuracy of SOH estimation.…”