基于近似模型的电动汽车前风窗玻璃除霜性能优化

Optimization of Electric Vehicle Defrosting Performance of Windscreen Based on Approximate Model

建立某款电动汽车的除霜仿真三维模型。定义该车型除霜出风口的格栅偏转角度为设计变量,通过拉丁超立方设计选定30组试验样本点,目标函数为前风窗玻璃内表面的部分区域平均风速。拟合得到Kriging近似模型,利用多岛遗传算法对该近似模型进行优化设计。研究结果表明:优化后前挡风玻璃的视野A区和A'区的近壁面平均速度分布均匀,低速区减少,通过除霜瞬态模拟分析发现,该款车型的除霜性能得到提高,满足国标对汽车的除霜性能要求。

A three-dimensional model of defrosting simulation for aelectric vehicle was established.The deflection angle of the defrosting grid was defined as the design variable. Through the Latin hypercube design,30 groups of test sample points were selected,and the objective function was the velocity magnitude of the inner surface of the front windscreen glass. Fitting to obtain Kriging approximation model,the multi-island genetic algorithm was used to optimize the design of the approximate model. The results show that the distributionof air’s velocity near the wall is uniform and the region at low speed is less. The defrosting efficiency of the car is improved by the transient simulation defrosting,which meets the defrosting performance requirements of the national standard.