Metamodel-Based Optimization of a Lithium-Ion Battery Cell for Maximization of Energy Density with Evolutionary Algorithm

Abstract

The demand for lithium-ion batteries is gradually increasing, not only in the field of portable electronic devices but also in high-capacity energy storage systems and electric vehicles, owing to such advantages as a high capacity and high power. Accordingly, power and capacity improvements of lithium-ion batteries are required. To achieve such improvements, it is very effective to optimize the electrode thickness, porosity, separator thickness, and active particle size. In this study, maximization of the specific energy density is achieved using metamodel-based global optimization. The capacity and power of the design variables are calculated using an electrochemical-thermal model. The sampling points for the generation of a metamodel are obtained using an optimal Latin hypercube design, and a radial basis function-based metamodel is generated. Furthermore, the global optimal solution is obtained using an evolutionary algorithm. As the results of the optimization show, unnecessary mass reduction, increased negative/positive capacity ratio, and decreased internal resistance, thereby improving the specific energy density without changing the specific power density. (C) 2019 The Electrochemical Society.