Bi-level optimization for cross-sectional size of bus body

A bi-level optimization method, integrating both local line search and overall multi-objective optimization, is proposed aiming to provide a solution for lightweight design of integral bus bodies. On the first level, the lightest structure under the strength condition is obtained via implementing line-search process in local bar models of the body frame based on the section libraries. On the second level, the design variables are screened by means of sensitivity analysis, and then the bus body structure is optimized by using the multi-objective genetic algorithm. This approach is implemented into an integral bus body frame during the concept design stage. It is verified that the obtained structure scheme is 10.57% lighter than the target bus model and the major mechanical performances are also better than the target one.