Smart Skin separation control using distributed-input distributed-output - multi-modal actuators - and machine learning

Aug 4, 2026

Efficient flow separation control brings major economic and environmental value for transportation systems. This study applies a machine learning approach to reduce flow separation in Smart Skin, a flow control device with distributed-input and distributed-output (DIDO) capability. Smart Skin uses 30 hybrid actuator units combining height-adjustable vortex generators and mini-jet actuators on a backward-facing ramp, with distributed pressure taps for flow-state monitoring. Parameter–performance mapping is complex, so PSO-TPME is used to optimize high-dimensional control parameters, yielding results that outperform parametric-study baselines and suggest a strong future for distributed, sensor-actuated machine learning flow control.

LanguageEnglish

CategoryTechnology

FormatPDF

Pages28

Reading Time1 h 11 min

AccessFree to download

Introduction

Passive flow separation control

Active flow control and actuator types

Synergy of active and passive control

Sensor-driven closed-loop flow control

arXiv :2311 .08116v1 [ ee ss . SY] 14 Nov 2023 Under consideration for publication in J. Fluid Mech. 1 Banner appropriate to article type will appear here in typeset article Smart Skin separation control using distributed-input distributed-output, multi-modal actuators, and machine learning Songqi Li1† 1 School of Mechanical Engineering and Automation, Harbin Institute of Technology, 518055 Shenzhen, PR China (Received xx; revised xx; accepted xx) Efficient flow separation control represents significant economic benefit. This study applies a machine learning algorithm to minimize flow separation in Smart Skin, a flow control device that features distributed-input and distributed-output (DIDO) . Smart Skin comprises 30 hybrid actuator units, each integrating a height-adjustable vortex generator and a mini-jet actuator. These units are deployed on a backward-facing ramp to reduce flow separation in a distributed manner. To monitor the flow state, distributed pressure taps are deployed around the multi-modal actuators. Parametric studies indicate that the mapping between control parameters and separation control performance is complex. To optimize separation control, a cutting-edge variant of the particle swarm optimization (PSO-TPME, Shaqarin & Noack 2023) are used for the control parameters in the Smart Skin. This algorithms is capable of achieving a fast optimization in high-dimensional parameter spaces. The results demonstrate the efficiency of PSO-TPME, and the optimized solution significantly outperform the best result from the parametric study. These findings represent a promising future of machine learning-based flow control using distributed actuators and sensors. Key words: Flow control; distributed-input distributed-output; multi-modal actuator; machine learning 1. Introduction Turbulence control offers substantial economic and environmental benefits in the realm of transportation vehicles. In the case of high-speed trains, The aerodynamic drag takes up 75-80 % of the total drag when the train is running at a speed of 300 km h −1 (Baker 2014; Raina et al. 2017; Zhang et al. 2018) . This proportion increases to over 90 % when the train speed reaches 400 km h −1 (Yang et al. 2012; Yu et al. 2021) . In passenger cars, aerodynamic pressure drag is the predominant factor (Sudin et al. 2014; Altaf et al. 2014; Geropp & Odenthal 2000) . The contribution of aerodynamic pressure drag corresponds to 90 % of the total aerodynamic drag, with 80 % of this contribution originating from the rear part of the † Email address for correspondence: [lisongqi@hit.edu.cn](lisongqi@hit.edu.cn) 2 car (Kourta & Gilliron 2012) . One of the main sources of aerodynamic drag in passenger cars is the flow separation near the vehicle’s rear end, where the detached flow often results in significant energy losses (Hucho & Sovran 1993) . A wide range offlow control techniques have been successfully employed in both academic and industrial flow configurations to mitigate flow separation (Brunton & Noack 2015; Gad-el Hak 1996). Passive control devices, including vortex generators (Selby et al. 1992; Lin et al. 1994), dimples (Lake et al. 2000; Ballerstein & Horst 2023), flaps (Camacho-Snchez et al. 2023), and transverse grooves (Mariotti etal. 2017), have demonstrated their effectiveness for separation control in various flow configurations. In a recent study by Viswanathan (2021), a parametric analysis was conducted to explore the effect of different vortex generators on car drag reduction under varying yawing angles. In recent decades, the rapid advancement of actuators has made active flow control a prominent research area. A comprehensive overview of active actuators can be found in Cattafesta & Sheplak (2011) . Fluidic oscillators (Raghu 2013; Metka & Gregory 2015), synthetic jets (Gilarranz et al. 2005; Kim & Kim 2009), plasma actuators (Post & Corke 2006; Roupassov et al. 2009), pulsed jets (Li et al. 2017; Fan et al. 20

FAQs

Why is PSO-TPME applied to optimize separation control?

The mapping between control parameters and separation-control performance is complex and difficult to tune. PSO-TPME is used to efficiently optimize high-dimensional control parameters and improve performance over parametric studies.

Aug 4, 2026

Upload by Ava Thompson