Handling Individual Pitch Control within an Actuator Disk framework: verification against the Actuator Line method and application to wake interaction problems

Moens, Maud; Coquelet, Marion; Trigaux, François; Chatelain, Philippe

doi:10.1088/1742-6596/2265/2/022053

Cited by 3 publications

(7 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Even though it would be interesting to train the NN in realistic turbulent flow fields, it is quite clear that the cost of the LES used in this work is prohibitive. However, multiple compromises can be envisioned for the training in a higher fidelity environment such as (1) generating synthetic turbulence data bases, typically using the Mann algorithm 50 or atmospheric boundary layer precursor simulations, 61 to train the NN in turbulent wind at affordable cost; (2) using dynamic wake models like the one proposed in Lejeune 62 to produce much more realistic meandering waked flow conditions, (3) using LES at coarser spatio‐temporal resolutions, which implies using disk approaches to model the rotor and thus handling IPC within an actuator disk framework, as proposed in Moens, 63 or (4) use the low‐fidelity environment to first train the NN and then refine the training in a LES framework. The best option should be chosen based on some cost‐benefit analysis, which leaves a variety of perspectives open for the present methodology.…”

Section: Resultsmentioning

confidence: 99%

“…With an improved load sensing and an enlarged action space, the missing element is a realistic training environment. Multiple possibilities towards that purpose were proposed before, such as using wake models, 62 synthetic, 50 or precursor‐generated 61 turbulence or disk‐based LES 63 …”

Section: Discussionmentioning

confidence: 99%

“…Multiple possibilities towards that purpose were proposed before, such as using wake models, 62 synthetic, 50 or precursor-generated 61 turbulence or disk-based LES. 63 To conclude on a more general level, this work has demonstrated that a wind turbine pursuing a certain objective can learn how to behave in the simple flow it is subjected to and exploit that knowledge when evolving in more complex flows. Including an extended action and a more realistic training environment to the framework could lead RL-IPC to outperform conventional IPC, in terms of both loads and pitch activity, especially in the case of flows like the meandering wakes encountered in wind farms.…”

Section: Discussionmentioning

confidence: 99%

“…rotor and thus handling IPC within an actuator disk framework, as proposed in Moens,63 or (4) use the low-fidelity environment to first train the NN and then refine the training in a LES framework. The best option should be chosen based on some cost-benefit analysis, which leaves a variety of perspectives open for the present methodology.…”

mentioning

confidence: 99%

See 3 more Smart Citations

A reinforcement‐learning approach for individual pitch control

et al. 2022

View full text Add to dashboard Cite

Individual pitch control has shown great capability of alleviating the oscillating loads experienced by wind turbine blades due to wind shear, atmospheric turbulence, yaw misalignment, or wake impingement. This work presents a novel controller structure that relies on the separation of low-level control tasks and high-level ones. It is based on a neural network that modulates basic periodic pitch angle signals. This neural network is trained with reinforcement learning, a trial and error way of acquiring skills, in a low-fidelity environment exempt from turbulence. The trained controller is further deployed in large eddy simulations to assess its performances in turbulent and waked flows. Results show that the method enables the neural network to learn how to reduce fatigue loads and to exploit that knowledge to complex turbulent flows.When compared to a state-of-the-art individual pitch controller, the one introduced here presents similar load alleviation capacities at reasonable turbulence intensity levels, while displaying very smooth pitching commands by nature.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

A reinforcement‐learning approach for individual pitch control

et al. 2022

View full text Add to dashboard Cite

show abstract

“…However, this procedure is also implemented to be performed online, during the simulation. It is relevant for large wind farm simulations, in order to avoid saving a large amount of AD data and post‐processing calculations, but it is particularly important for load alleviation strategies, for example, applying an individual pitch control scheme within an AD framework 51 or future global wind farm control investigations.…”

Section: Methodsmentioning

confidence: 99%

Assessment of an actuator disk‐based approach for the prediction of fatigue loads in a perspective of wind farm‐scale application

2022

View full text Add to dashboard Cite

This work aims at assessing the loads and the fatigue estimates computed using an advanced actuator disk (AD) method coupled to large eddy simulation, at a resolution appropriate for large wind farm calculations. In order to compute pertinent fatigue loads, blade trajectories are replicated through the disk, and the AD aerodynamic forces are interpolated onto these "virtual blades" at each time step. This approach, denoted AD-B, is verified against a Vortex-Particle Mesh (VPM) method coupled to immersed lifting lines at a fine resolution, through simulations of an isolated rotor in a turbulent wind. Two different methods are used to evaluate the fatigue damages: the widely used Rainflow counting (RFC) procedure and the spectral Dirlik's approach (DK), both combined with a Palgrem-Miner rule. In the present work, the DK counting method is considered to further investigate the potential of extrapolating some loading data in the high frequencies, uncaptured by the AD-B model at a coarse resolution. The results show that the fatigue estimates computed using the RFC procedure are very similar for the VPM and the AD-B methods, thus without any modeling of the unresolved scales for the disk. Indeed, the AD-B model captures the loads of middle and large amplitudes that contribute the most to the rotor damage. The use of extrapolated loading data makes the AD-B fatigue estimates closer to the VPM ones, but the DK counting method globally leads to results that are quite different from those obtained using the RFC procedure. Further investigations are thus required for the combination of extrapolated loading information and spectral counting methods.

show abstract

Simulating the helix wake within an actuator disk framework: verification against discrete-blade type simulations

Coquelet

Moens

Duponcheel

et al. 2023

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

Dynamic flow control strategies are raising interest for wake mitigation purposes. Among the different strategies, the so-called helix one relies on individual pitch control (IPC). The numerical simulation of the helix is thus readily performed by means of discrete-blade capturing methods. Yet, if this control strategy is considered at the scale of wind farms, the resolution required by such methods becomes prohibitive and actuator disk (AD) models should be envisioned. It is however not trivial to translate IPC strategies to an AD framework which by definition considers rotor-averaged effects. This work assesses the ability of an AD method to simulate the helix strategy by comparing it to a higher fidelity approach relying on a discrete-blade capturing model. Results show that the disk-type approach supplemented with a disk-adapted IPC scheme is able to capture both the forced motion of the wake at low turbulence and the faster wake recovery at moderate turbulence. From a quantitative perspective, the disk-type approach predicts bigger power gains, compared to those foreseen by the discrete-blade type approach, for a downstream turbine in the wake of a helix-operated one.

show abstract

Handling Individual Pitch Control within an Actuator Disk framework: verification against the Actuator Line method and application to wake interaction problems

Cited by 3 publications

References 24 publications

A reinforcement‐learning approach for individual pitch control

A reinforcement‐learning approach for individual pitch control

Assessment of an actuator disk‐based approach for the prediction of fatigue loads in a perspective of wind farm‐scale application

Simulating the helix wake within an actuator disk framework: verification against discrete-blade type simulations

Contact Info

Product

Resources

About