Stabilization of the fluidic pinball with gradient-enriched machine learning control
From MaRDI portal
Abstract: We stabilize the flow past a cluster of three rotating cylinders, the fluidic pinball, with automated gradient-enriched machine learning algorithms. The control laws command the rotation speed of each cylinder in an open- and closed-loop manner. These laws are optimized with respect to the average distance from the target steady solution in three successively richer search spaces. First, stabilization is pursued with steady symmetric forcing. Second, we allow for asymmetric steady forcing. And third, we determine an optimal feedback controller employing nine velocity probes downstream. As expected, the control performance increases with every generalization of the search space. Surprisingly, both open- and closed-loop optimal controllers include an asymmetric forcing, which surpasses symmetric forcing. Intriguingly, the best performance is achieved by a combination of phasor control and asymmetric steady forcing. We hypothesize that asymmetric forcing is typical for pitchfork bifurcated dynamics of nominally symmetric configurations. Key enablers are automated machine learning algorithms augmented with gradient search: explorative gradient method for the open-loop parameter optimization and a gradient-enriched machine learning control (gMLC) for the feedback optimization. gMLC learns the control law significantly faster than previously employed genetic programming control. The gMLC source code is freely available online.
Recommendations
- 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
- Machine learning control -- taming nonlinear dynamics and turbulence
- Artificial neural networks trained through deep reinforcement learning discover control strategies for active flow control
- Robust flow control and optimal sensor placement using deep reinforcement learning
Cites work
- A Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output from a Computer Code
- A generalized mean-field model of the natural and high-frequency actuated flow around a high-lift configuration
- A hierarchy of low-dimensional models for the transient and post-transient cylinder wake
- A physics-based approach to flow control using system identification
- A Simplex Method for Function Minimization
- Active turbulence control for drag reduction in wall-bounded flows
- Actuators for active flow control
- Artificial intelligence control of a turbulent jet
- Artificial neural networks trained through deep reinforcement learning discover control strategies for active flow control
- Cluster-based feedback control of turbulent post-stall separated flows
- Cluster-based network model
- Control of Flow Over a Bluff Body
- DYNAMICS AND CONTROL OF HIGH-REYNOLDS-NUMBER FLOW OVER OPEN CAVITIES
- FEEDBACK CONTROL OF COMBUSTION OSCILLATIONS
- Feedback shear layer control for bluff body drag reduction
- Frequency selection by feedback control in a turbulent shear flow
- scientific article; zbMATH DE number 1034218 (Why is no real title available?)
- Linear feedback stabilization of laminar vortex shedding based on a point vortex model
- Linear Genetic Programming
- Linear genetic programming control for strongly nonlinear dynamics with frequency crosstalk
- Low-order model for successive bifurcations of the fluidic pinball
- Machine learning control -- taming nonlinear dynamics and turbulence
- Metric for attractor overlap
- Numerical simulations of flow past three circular cylinders in equilateral-triangular arrangements
- Recursive dynamic mode decomposition of transient and post-transient wake flows
- Reduced-order modelling of the flow around a high-lift configuration with unsteady Coanda blowing
- The wake of a cylinder performing rotary oscillations
- Wave packets and turbulent jet noise
Cited in
(13)- 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
- Explorative gradient method for active drag reduction of the fluidic pinball and slanted Ahmed body
- From snapshots to manifolds – a tale of shear flows
- Comparative analysis of machine learning methods for active flow control
- Machine-learning a perfect bending soccer goal shot
- Self-tuning model predictive control for wake flows
- Jet mixing enhancement with Bayesian optimization, deep learning and persistent data topology
- Turbulence control: from model-based to machine learned
- Actuation manifold from snapshot data
- Data-driven stabilization of an oscillating flow with linear time-invariant controllers
- Orbital cluster-based network modelling
- Bayesian optimization for active flow control
This page was built for publication: Stabilization of the fluidic pinball with gradient-enriched machine learning control
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4987910)
