Drag Reduction Of Low-drag Ahmed Vehicle Model Using Combined Actuations

With the largest number of cars in the world,China’s energy and environmental problems are becoming more and more serious.Reducing the drag of vehicle is an important way to solve these problems.More than 60%of the driving drag experienced by high-speed cars is aerodynamic drag.Flow control technology,which includes active control and passive control,can control the wake structure of a car,and it has been proven to be an effective method to reduce the aerodynamic drag.This paper takes the 35°Ahmed vehicle body as the research object,aiming to achieve the goal of automobile drag reduction through the application of active flow control.The control is performed by 5 groups of micro-jet array actuators arranged at the rear of the car,and the control means include steady jet and pulse jet.The experiment was carried out in a low-speed wind tunnel,and the Reynolds number was1.67×10~5.The experiment of single control of each group of actuators tested the influence of jet parameters on the drag and the pressure on the back of each group of actuators at different jet angles,and the single control experiments obtained the highest drag reduction of 7%.Therefore,the jet angle of each group of actuators in the coming combined control case is determined.We propose an Adaptive sampling method,and refers to exploratory gradient method(EGM)’s"sampling-optimization"framework,and obtains the Weighted Sampling Simplex Algorism(WSSA).Based on simulation results,this new method has a faster convergence rate in optimum searching.Based on this algorithm,a set of machine learning experimental control system was developed.Using this control system,the drag reduction control experiments based on combined steady jet and combined pulsed jet have been carried out successively,and the maximum drag reduction of about 18%and 20%are obtained respectively.From the comparison of the streamline diagram and pressure distribution diagram before and after the control,it is found that the two optimal control case make the flow separation of the upper shear layer delay from the upper edge of the slant to the lower edge of the slant,reduce the size of the recirculation zone,and effectively improve the pressure at the rear of the car model.Our experiments have proved that by combining multiple actuators arranged at the rear of the car model,the mutual coupling of multiple controls can be used to achieve better results than a single control.The machine learning method has achieved good results in the car drag reduction control in this paper,which shows that the machine learning method has broad application prospects for the optimal control problem of the multi-input control system.