Abstract
The present study uses an Ising machine to solve a multi-constrained, multi-objective optimization problem of car-body structure design. While Ising machines require explicit formulation in the context of QUBO, vehicle performance, such as impact performance, is generally a black box. Therefore, we apply FMQA, which approximates the black-box function using FM and solves QUBO for optimization. In FMQA, the difficulty lies in the appropriate weighting of evaluation functions and the suppression of local search. In this study, we propose their resolution and compare our method with conventional optimization methods to verify the usefulness of quantum (-inspired) computing in real problems.
