Capacity Estimation of Lithium-Ion Batteries Using Electrochemical Impedance Spectroscopy and Optimized Extreme Learning Machine

The fast and accurate estimation of lithium-ion battery capacity is crucial for ensuring their safe and reliable operation and for studying the optimal utilization of retired batteries. In this paper, we propose a new method for estimating the available capacity of lithium-ion batteries based on their electrochemical impedance spectroscopy (EIS). Firstly, features are extracted from the EIS and its distribution of relaxation time during battery aging, and a health feature set is constructed using an improved mutual information technique for feature selection. Then, the capacity of the lithium-ion battery is estimated using the improved whale optimization algorithm-extreme learning machine (IWOA-ELM). It is demonstrated through experimental results that the capacity can be accurately estimated using the IWOA-ELM method, with the root mean square error controlled within ±1%.