Robust Optimization of an Electric Bus Body Frame Based on the Mesh Morphing Technology

The traditional design optimization of the bus body frame are mainly limited to the optimization of the thickness of the parts. In this work, we perform the optimization design of the bus body frame by optimizing the sectional shape of the tube beams based on the mesh morphing technology. Several groups of finite element analysis are performed for the body frame and the sectional sizes of the rectangular tube beams of the chassis and the side structure of the body that have a greater impact on the body performance are selected for optimization. The mesh morphing technology is used to establish shape design variables for the selected tube beams, and the design variables are comprised of the length, width, and thickness of the sections of the selected tube beams. Based on the entropy weight method and the order preference by similarity to the ideal solution (TOPSIS) comprehensive weight method, the design variable with a higher comprehensive contribution is obtained. Next, the multi-objective optimization of the structure of the bus body frame is carried out by combining the Latin hypercube experimental design, radial basis function neural network (RBFNN) and the multi-objective non-dominated sorting genetic algorithm II (NSGA-II). Finally, the six-sigma robust design optimization is performed to improve the reliability and the sigma level of response. After robust optimization, finite element analysis is carried out to verify and evaluate the optimization results. It is shown that the prediction errors are all within 0.3%. The stiffness of the bus body frame has been improved remarkably, and the lightweight rate of bus body substructure reaches 5.9%.