Development of recycling strategy for large stacked systems: Experimental and machine learning approach to form reuse battery packs for secondary applications

Secondary battery utilization is one of the most promising strategies to solve the problem of battery recycling in the future. The objective of this research is to provide practical solutions for the screening and regrouping of retired lithium batteries. Firstly, a systematic clustering method is proposed. The method is mainly divided into three stages: (1) Fast screening technology of voltage and internal resistance (2) Retired battery status of health (SOH) detection (3) Retired battery clustering method based on self-organizing maps (SOM) neural network. Secondly, a validation experiment was performed. This experiment covers the collection, disassembly of retired battery packs, retired batteries SOH detection, classification, and reassembly of new reuse battery packs. Results show that our proposed screening scheme can quickly identify the initial state of retired batteries and provide a solid basis for further decision-making. After battery test and intelligent SOM screening, the inconsistency of capacity and internal resistance of retired battery pack has been reduced. In addition, the experimental results show that the capacity and potential cycle numbers of reuse pack manufactured by SOM clustering are 25% and 50% more than those of reuse pack manufactured by randomly selected retired batteries. Thus, it proves that the proposed screening method was efficient for retired battery second use. Moreover, the original consistency of retired battery pack has significant impacts on batteries reuse. Thus, the reuse strategies should consider applying for spent battery packs which already has maintained some level of reasonable consistency.