| Home | KSAE | E-Submission | Sitemap | Contact Us |  
top_img
International Journal of Automotive Technology > Volume 23(5); 2022 > Article
International Journal of Automotive Technology 2022;23(5): 1419-1426.
doi: https://doi.org/10.1007/s12239-022-0124-9
DRAG REDUCTION PREDICTION OF AHMED MODEL WITH TRAVELING WAVE BASED ON BP NEURAL NETWORK
Xingjun Hu , Jinglong Zhang , Yufei Luo , Jingyu Wang , Pengzhan Ma , Wei Lan , Chunbo Dong
State Key Laboratory of Automotive Simulation and Control, Jilin University
PDF Links Corresponding Author.  Jingyu Wang  , Email. wangjy@jlu.edu.cn
ABSTRACT
n this paper, a traveling wave model is proposed to explore its influence on the aerodynamic drag of a Ahmed model, the experimental and numerical results of aerodynamic drag coefficient CD for the Ahmed model are in good agreement. Then by defining the aerodynamic benefit coefficient ΔCD as the evaluation index for the orthogonal experiment, range analysis is conducted to determine the influences of the amplitude A, wavelength λ and frequency ω of the wave and the vehicle speed u on ΔCD. After the analysis it can been found that λ has the least importance among these parameters, hence A, ω and u are used to construct the 105 samples for training the BP neural network to predict ΔCD, results show that ΔCD obtained from the neural network is significantly affected by the parameters of traveling wave. The prediction accuracy of the network is furtherly verified by another 15 samples which are also built on A, ω and u, and the corresponding data overlap rate of ΔCD is 96 %, so it can be concluded that the BP neural network constructed in this paper is accurate enough to predict ΔCD.
Key Words: Ahmed model, Traveling wave model, BP neural networks, Drag reduction prediction
Editorial Office
21 Teheran-ro 52-gil, Gangnam-gu, Seoul 06212, Korea
TEL: +82-2-564-3971   FAX: +82-2-564-3973   E-mail: manage@ksae.org
About |  Browse Articles |  Current Issue |  For Authors and Reviewers
Copyright © The Korean Society of Automotive Engineers.                 Developed in M2PI
Close layer
prev next