Artificial neural network enabled P2D model deployment for end-of-line battery cell characterization

The production chain of lithium-ion battery cells is an intricate process with manifold process-product interdependencies leading to a complex product, the battery cell. In order to assess the quality and the performance of the battery cell in the end-of-line characterization, battery cell models need to be fit to experimental data in order to estimate unmeasured physical parameters of the cell. This procedure is laborious and error prone to the model complexity and the large number of physical parameters and is therefore presently not suitable as an end-of line test in a large-scale battery cell production. In this paper, a battery cell model is used to generate data by varying measured and unmeasured parameters in order to train an artificial neural network model for a fast and reliable diagnostics. The presented artificial neural network model is capable of fitting the battery cell model in less than a second to make it more suitable for end-of-line characterization.