Robust flow control and optimal sensor placement using deep reinforcement learning
From MaRDI portal
Abstract: This paper focuses on a drag-reducing control strategy on a 2D-simulated laminar flow past a cylinder. Deep reinforcement learning algorithms have been implemented to discover efficient control schemes, using two synthetic jets located on the cylinder's poles as actuators and pressure sensors in the wake of the cylinder as feedback observation. The present work focuses on the efficiency and robustness of the identified control strategy and introduces a novel algorithm (S-PPO-CMA) to optimise the sensor layout. An energy-efficient control strategy reducing drag by 18.4% at Reynolds number 120 is obtained. This control policy is shown to be robust both to the Reynolds number in the range [100,216] and to measurement noise, enduring signal to noise ratios as low as 0.2 with negligible impact on performance. Along with a systematic study on sensor number and location, the proposed sparsity-seeking algorithm has achieved a successful optimisation to a reduced 5-sensor layout while keeping state-of-the-art performance. These results highlight the interesting possibilities of reinforcement learning for active flow control and pave the way to efficient, robust and practical implementations of these control techniques in experimental or industrial systems.
Recommendations
- Artificial neural networks trained through deep reinforcement learning discover control strategies for active flow control
- Reinforcement-learning-based actuator selection method for active flow control
- Reinforcement-learning-based control of confined cylinder wakes with stability analyses
- Comparative analysis of machine learning methods for active flow control
- Optimal sensor and actuator placement for feedback control of vortex shedding
Cites work
- A data-assimilation method for Reynolds-averaged Navier-Stokes-driven mean flow reconstruction
- A heuristic approach to effective sensor placement for modeling of a cylinder wake
- Active control and drag optimization for flow past a circular cylinder. I: Oscillatory cylinder rotation
- Active control of cylinder wake
- An active flow control strategy for the suppression of vortex structures behind a circular cylinder
- Artificial neural networks trained through deep reinforcement learning discover control strategies for active flow control
- Compressive sensing based machine learning strategy for characterizing the flow around a cylinder with limited pressure measurements
- Control of chaotic systems by deep reinforcement learning
- Control of fluid flow and heat transfer around a square cylinder by uniform suction and blowing at low Reynolds numbers
- Drag force in the open-loop control of the cylinder wake in the laminar regime
- Evaluation of the unsteady RANS capabilities for separated flows control
- Feedback control of vortex shedding using a resolvent-based modelling approach
- Integration of Stiff Equations
- Large-eddy simulation of laminar transonic buffet
- Linear iterative method for closed-loop control of quasiperiodic flows
- Machine learning for fluid mechanics
- Network structure of two-dimensional decaying isotropic turbulence
- Neural networks for modelling and control of dynamic systems. A practitioner's handbook
- Nonlinear dynamics and pattern formation in turbulent wake transition
- Numerical study and physical analysis of the pressure and velocity fields in the near wake of a circular cylinder
- On the generation of a reverse von Kármán street for the controlled cylinder wake in the laminar regime
- On the use of incomplete sensitivities for feedback control of laminar vortex shedding
- Open-loop control of cavity oscillations with harmonic forcings
- Optimal control of the cylinder wake in the laminar regime by trust-region methods and POD reduced-order models
- Optimal rotary control of the cylinder wake in the laminar regime
- Optimal sensor placement for variational data assimilation of unsteady flows past a rotationally oscillating cylinder
- Reconstruction of unsteady viscous flows using data assimilation schemes
- Sensitivity analysis and passive control of cylinder flow
- Sensor and actuator placement trade-offs for a linear model of spatially developing flows
- Simple statistical gradient-following algorithms for connectionist reinforcement learning
- Suboptimal feedback control of vortex shedding at low Reynolds numbers
Cited in
(17)- Deep reinforcement learning for the control of conjugate heat transfer
- Neural-network-based control of discrete-phase concentration in a gas-particle corner flow with optimal energy consumption
- Reinforcement-learning-based control of confined cylinder wakes with stability analyses
- Motor learning at intermediate Reynolds number: experiments with policy gradient on the flapping flight of a rigid wing
- Artificial neural networks trained through deep reinforcement learning discover control strategies for active flow control
- Stabilization of the fluidic pinball with gradient-enriched machine learning control
- Optimal sensor and actuator placement for feedback control of vortex shedding
- Reinforcement-learning-based control of convectively unstable flows
- Reinforcement-learning-based actuator selection method for active flow control
- Comparative analysis of machine learning methods for active flow control
- Self-tuning model predictive control for wake flows
- Deep reinforcement learning-based active control for drag reduction of three equilateral-triangular circular cylinders
- Active control of the flow past a circular cylinder using online dynamic mode decomposition
- Data-driven stabilization of an oscillating flow with linear time-invariant controllers
- Model predictive control of fluid-structure interaction via Koopman-based reduced-order model
- Physics of the vortex gust-airfoil interaction under an optimal mitigation strategy learned through deep reinforcement learning
- Kalman filter-based dynamic mode decomposition for closed-loop flow control with noise-corrupted sensing
This page was built for publication: Robust flow control and optimal sensor placement using deep reinforcement learning
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5853740)
