A data mining approach for continuous battery cell manufacturing processes from development towards production

Abstract

Battery cells are central components of electric vehicles determining their operational characteristics, such as driving range, power output, and safety. Automotive OEMs undertake the necessary efforts to ensure the integration of safe and high-performance battery cells in their electrified fleets. In addition, an increased sustainable awareness of their customers and governmental policies force them to not only focus on operational goals, but rather on environmental aspects as well. Especially, battery cell manufacturing is associated with various negative environmental impacts (e.g. carbon dioxide emission). Therefore, this study develops a concept facilitating the development of novel continuous processes in battery cell manufacturing by enabling virtual experiments and an automatic optimization of economic and ecologic targets. Virtual experiments are enabled by training data-driven models that transfer the gained knowledge from development to large-scale production. The concept includes an inline-capable controller adjusting set points of process parameters with respect to a cost model quantifying product quality and environmental aspects. The validity of the proposed concept is demonstrated with data acquired from real battery cell production chain covering a continuous mixing process.

Keywords

• Battery cell manufacturing
• Continuous processing
• Extrusion
• Rheology
• Deep learning