Machine learning control for drag reduction of a car model in experiment
Abstract
We investigate experimentally a novel model-free in-time control strategy, called
Machine Learning Control (MLC), for aerodynamic drag reduction of a car model. Fluidic
actuation is applied at the trailing edge of a blunt-edged Ahmed body combined with a curved
deflection surface. The impact of actuation on the flow is monitored with base pressure sensors.
Based on the idea of genetic programming, the applied model-free control strategy detects and
exploits nonlinear actuation mechanisms in an unsupervised manner with the aim of minimizing
the drag. Key enabler is linear genetic programming as simple and efficient framework for
multiple inputs (actuators) and multiple outputs (sensors). The optimized control laws comprise
periodic forcing, multi-frequency forcing and sensor-based feedback control. Approximately 33%
base pressure recovery associated with 22% drag reduction is achieved by the optimal control
law for a turbulent flow at Re H ≈ 3 Ã 10 5 based on body height.