Ventilated cavity flows behind a backward facing step with a combination computational fluid dynamics and Error Back Propagation algorithm

作     者：He, Zanyang Sun, Tiezhi Zou, Li Jiang, Yichen Duan, Lei 

作者机构：Shanghai Jiao Tong Univ Sch Naval Architecture Ocean & Civil Engn Shanghai 200240 Peoples R China Dalian Univ Technol Sch Naval Architecture State Key Lab Struct Anal Ind Equipment Dalian 116024 Peoples R China Shanghai Jiao Tong Univ State Key Lab Ocean Engn Shanghai 200240 Peoples R China 

出 版 物：《OCEAN ENGINEERING》 (海洋工程)

年 卷 期：2022年第260卷

核心收录：

学科分类：07[理学] 0707[理学-海洋科学] 0824[工学-船舶与海洋工程] 0814[工学-土木工程] 

基　　金：National Key Research and Development Program of China [2017YFE0132000] National Natural Science Foundation of China Outstanding Youth Scientist Foundation of Dalian [2019RJ05] Natural Science Foundation of Liaoning Province of China [2020MS106] Liao Ning Revitalization Talents Program [XLYC1908027] Dalian High-level Talent Innovation Support Program (Youth Technology Star) [2020RQ003] Fundamental Research Funds for the Central Universities [DUT21LK25, DUT20TD108, DUT20LAB308] 

主　　题：Ventilated cavity flows Backward facing step Error Back Propagation algorithm DES model Optimize drag reduction 

摘      要：Ventilated cavity flow is of a great practical interest for the features that may reduce the drag around the area near the cavity. The present paper demonstrates the prediction of ventilated cavity flows on the backward facing step, which establishes the prediction model of drag reduction by the numerical simulation and artificial intelligence algorithm. The results of the computational fluid dynamics are compared with available experimental data to validate the precision, which shows good agreement. This paper numerically analyzes the influence of the number, location and part ventilation volume of injectors on the drag reduction of ventilated cavity flows. With the same air-entrant coefficient, the number and location of injectors both have a significant impact on drag reduction performance. And the part ventilation volume of the injectors may also have a limited impact. Accordingly, the prediction model of drag reduction is established by the Error Back Propagation algorithm (BP algorithm) that has accurate algorithms and fast calculation to save time and computing resources. Meanwhile, the results indicate that BP algorithm can reasonably predict the performance of ventilated cavity flows, which means the model can optimize the drag reduction on ventilated cavity flows and it can be useful in engineering.