Design and Optimization of an EV Battery Enclosure Using Machine Learning

In this study, structural optimization of an enclosure under bending and torsional constraints was carried out. Machine learning (ML) approach was used to calculate the objective and constraint functions in the optimization problem. The ML model was trained and validated with data obtained from finite element analyses. The optimization model was then solved by the differential evolution algorithm. Five thicknesses, which are the design parameters in the enclosure, were optimized for minimum mass, and according to the results, the enclosure’s mass decreased by 18.29%.