Select
Optimization of Vehicle Aerodynamic Drag Based on Discrete Adjoint Method and Surrogate Model
He Yansong, Cao Lipeng, Zhang Zhifei, Li Yun, Chen Zhao
2020, 42 (11 ):
1577-1584.
doi: 10.19562/j.chinasae.qcgc.2020.11.018
The discrete adjoint method and the surrogate model are introduced to propose a fast and effective optimization strategy for reducing vehicle aerodynamic drag in this paper. Firstly, through the aerodynamic modelling and simulation analysis on vehicle, the changing tendency of drag coefficient and measured point pressure are obtained, which are then compared with the results of wind tunnel test. They agree well, showing that the accuracy of model meets requirements. Then with drag coefficient as objective, the discrete adjoint method is adopted to conduct sensitivity analysis on vehicle surface, and the components with higher sensitivity such as front bumper, rear-view mirror, tail wing and rear bumper are determined to be the objects of optimization. Next, the sample space is constructed by Hammersley design of experiment and the sample point model are parameterized by using mesh free form deformation technology with the corresponding drag coefficients calculated. Finally, the surrogate model is constructed by Kriging interpolation method, and the multi-island genetic algorithm is selected to perform a global optimization on surrogate model. The results of optimization indicate that the drag coefficient of vehicle reduces by 3.29%.
References |
Related Articles |
Metrics