利用神经网络估算汽车风阻系数

Estimate Drag Coefficient of Automobile with Neural Network

汽车风阻是整车重要的指标之一,其对整车行驶阻力产生直接影响,在全行业节能减排背景下,汽车风阻系数目标设计已经被视为前期开发的重要指标之一。开发前期,了解对标车相关设计指标,针对性地提出车型相应的设计开发目标。工程上可以由造型输入通过CFD仿真获得前期风阻系数,根据设计目标给出优化方案。然而对标车的风阻数据往往需要风洞试验获得,需要一定的周期和较高的成本。本文探讨了一种基于神经网络的汽车风阻系数估算方法,是一种快速且容易实现的方法。

Automobile wind resistance is one of the important indicators of the vehicle,its direct impact on vehicle driving resistance,for the industry energy conservation and emissions reduction,target setting of automobile wind resistance coefficient has been one of the important indexes in the early stage.At the beginning of the development,understanding of the related design index,setting a directed target of the designed car,engineering obtains the wind resistance coefficient by CFD simulation with modelling,and provide optimization scheme according to the designed target,but the standard car wind resistance often obtain by wind tunnel test which needs certain time and even high cost.This paper suggests a method to estimate the drag coefficient of automobile based on neural network which is a fast and easy method to realize.