Aerodynamic Drag Reduction Of High Drag Vehicle Model Based On Artificial Intelligence And Combined Actuators

At the era of the severe environment pollution,as well as the shortage of resources,the researching focused on the drag reduction of auto models emerges great importance,which can efficiently diminish the fuel consumption by reducing the aerodynamic drag.Based on such consideration,aerodynamic drag reduction of 25 ° Ahmed-car-model is studied in this paper.First,the influence of unsteady frequency on a single actuator is measured via the traditional open-loop method.Then,the control method based on artificial intelligence and steady blowing were also explored,20 pressure taps on the back were applied to calculate the evaluation function in the experiment process.Finally,the unsteady experiment of the combination of artificial intelligence actuation is researched.The value of the evaluation function is the only evaluation standard in the process of artificial intelligence.Due to the relationship between drag and back pressure,the lower the value of evaluation function is,the better the effect will be.According to the comparison of the value of each evaluation function in each generation,the optimal individual in each generation could be automatically selected,and it can be eliminated generation by generation.The experiment of unsteady actuation of a single actuator proves that with a Reynolds number of 1.67 × 105 and a duty cycle of 0.4,the frequency has a different degree of influence on each actuator,and their drag reduction ranges are different.Through the optimization process of eleven generation of artificial intelligence’s steady combination of air blowing,it was found that the flow of each actuator was constantly changing,and the flow of each actuator did not change significantly until the ninth generation of optimization was started.The best blowing flow rate combination of the actuators S1,S2,S3,and S4 are 1.81 × 10-3,1.20 × 10-3,4.20 × 10-2,and 4.60 × 10-3,respectively.Under the optimal flow combination,the rear of the car model achieved a 27% pressure recovery.The drag reduction rate at this pressure recovery measured by a force balance is 23%.Compared with the traditional manual adjustment,the artificial intelligence genetic algorithm is used to find a better blowing combination,which achieves greater pressure recovery and better drag reduction effects.This shows that the three classical processes in artificial intelligence genetic algorithms,namely cross mutation and replication,can be well applied to the research of aerodynamic drag reduction of a car model,and can well solve the problem of multiple inputs and multiple outputs in aerodynamic research of a car model.The experiments used artificial intelligence genetic algorithms to study the effects of unsteady blowing combination.After nine generations of optimization,the best blowing method was found in the ninth generation.When the optimal frequency of actuator S1 is 250 Hz and the duty cycle is 0.5;the optimal frequency of actuator S2 is 500 Hz and the duty cycle is 0.5;the optimal frequency of actuator S3 is 500 Hz and the duty cycle is 0.3;Actuator S4 is normally closed;Actuator S5 is normally open,18% pressure recovery and 12% drag reduction are obtained.Compared with the steady blowing combination,the unsteady blowing combination has a worse pressure recovery and drag reduction effect.This is mainly because the vortex structure on the back of the car model is very sensitive to volume flow rate,but not sensitive to frequency and duty cycle.