Wind farm layout optimization with load constraints using surrogate modelling

Riva, Riccardo; Liew, Jaime; Friis-Møller, Mikkel; Dimitrov, Nikolay Krasimirov; Barlas, Emre; Réthore, Pierre-Élouan; Beržonskis, Arvydas

doi:10.1088/1742-6596/1618/4/042035

Cited by 24 publications

(23 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…12 A more recent approach integrated generic fatigue loading on structural elements as a design constraint in TOPFARM. 13 Another approach considered fatigue loads on blades and AEP in a multiobjective optimization function. 10 Additionally, previous approaches to layout optimization commonly treat each turbine's coordinates as individual optimization variables.…”

Section: Wind Plant Layout Optimization Load Simulation and Surrogate Modelingmentioning

confidence: 99%

“…Layout optimization based on structural component damage equivalent loading constraints were integrated by Riva et al into the MDAO-based TOPFARM. 13 The integration of blade loads into layout optimization was explored by Stanley et al using CCBlade; 48 they found that AEP is similar when optimizing a 10-turbine array with and without total damage constraints.…”

Section: Wind Plant Layout Optimization Load Simulation and Surrogate Modelingmentioning

confidence: 99%

See 1 more Smart Citation

Reliability‐based layout optimization in offshore wind energy systems

et al. 2021

View full text Add to dashboard Cite

Existing methods for optimizing wind array layouts typically use power or cost objectives and rarely consider reliability-based objectives. Component and system failure rates, however, are dependent on location-specific wind conditions, are influenced by array layout and wake interactions, and have a direct and significant impact on capital costs, operational costs, and power production. Although wind power plant models exist that calculate wind loads with sufficient resolution to capture component loading dynamics from wind conditions, they are computationally expensive and thus not suitable for research applications requiring many evaluations, particularly optimization. This study describes the development of computationally efficient, reliability-based layout optimization methods, enabling us to explore the relationship between component reliability and layout optimization. These methods include the surrogate modeling of the planet bearing life based on varying wind conditions simulated in FAST.Farm and the formulation of reliability-based objectives based on failure cost and power production models. Through demonstration of this method, we explore how wind conditions, objective functions, and capacity density influence reliability-based layout optimization. Results indicate that considering reliability alongside power production can reduce failure costs associated with replacement costs and downtime whilemaintaining or improving power production. Our conclusions highlight the opportunity for wind power plant developers to integrate reliability and operational expenditures alongside performance and capital expenditure objectives in plant design and development to improve plant performance and costs.

show abstract

Section: Wind Plant Layout Optimization Load Simulation and Surrogate Modelingmentioning

confidence: 99%

Section: Wind Plant Layout Optimization Load Simulation and Surrogate Modelingmentioning

confidence: 99%

Reliability‐based layout optimization in offshore wind energy systems

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Neural network models have been used to build efficient surrogate models based on measured operational data or high-fidelity simulation data. These surrogate models are capable of estimating wake-induced effects on wind turbine power [1,2,3], loads [4,5], damage and failures [6]. These surrogate models, among other applications, enable efficient optimization of wind farm layout and control strategy [5,7].…”

Section: Introductionmentioning

confidence: 99%

Modelling of turbine power and local wind conditions in wind farm using an autoencoder neural network

Dou

Dimitrov

2022

J. Phys.: Conf. Ser.

Self Cite

View full text Add to dashboard Cite

Neural network models have been widely applied in wind energy to enable data-driven modelling of turbine’s power, loads, damage, and other quantities in wind farm. They have shown advantages in the flexibility, robustness, and efficiency, as well as the ease of application. An important task in data-driven modelling of wind farm is the characterization of wake-induced effects within wind farm. Because of the large number of wind turbines in wind farm and the broad range of conditions required for power output assessment, it has become relevant to use computationally efficient surrogate models that are based on parameterizations of the wind farm layout. Previous studies on surrogate models of wake-induced effects have focused on parameterizations based on manually extracted features from the wind farm geometry. The present study applies an autoencoder neural network to automatically learn the features representing the wind farm layout. Specifically, two autoencoders are studied and compared. The autoencoder provides dimension reduction and feature learning that allow representing the entire wind farm layout in just a few latent variables, thus allowing efficient and accurate surrogate models of wake-induced effects. The outputs of the autoencoder, together with any ambient wind condition variables, serve as inputs to a standard feedforward neural network that predicts the wake-induced effects for any turbine in the wind farm. Our results show that the autoencoder provides an automatic and effective way to parameterize the wind farm geometry, and facilitates the building of efficient surrogate models that are capable of accurate modeling of wake-induced effects in a wind farm with arbitrary layout.

show abstract

“…Of those papers, only three consider the entire plant [8][9][10], while the other papers consider only a single upstream turbine [11][12][13]. There is a similar dearth of research literature for loads-aware plant-layout optimization: only two papers have considered loads during plant-layout optimization [14,15]. Of those two papers, one presented only a single, simple optimization layout [14] and the other considered the fatigue loads for only one load channel: the blade-root edgewise moment [15].…”

Section: Introductionmentioning

confidence: 99%

“…There is a similar dearth of research literature for loads-aware plant-layout optimization: only two papers have considered loads during plant-layout optimization [14,15]. Of those two papers, one presented only a single, simple optimization layout [14] and the other considered the fatigue loads for only one load channel: the blade-root edgewise moment [15]. The lack of robust publications in this area makes it difficult to draw concrete conclusions about whether the potential increase in plant production is negated by a correlated increase of wind turbine loads.…”

Section: Introductionmentioning

confidence: 99%

The Importance of Wake Meandering on Wind Turbine Fatigue Loads in Wake

2021

View full text Add to dashboard Cite

Considering loads when optimizing wind-farm layouts or designing farm-control strategies is important, but the computational cost of using high-fidelity wake models in the loop can be prohibitively high. Using simpler models that consider only the spatial variation of turbulence statistics is a tempting alternative, but the accuracy of these models with respect to the aeroelastic response is not well understood. This paper therefore highlights the effect of replacing wake meandering with spatially varying statistics (“profile functions”) in the inflow to a downstream turbine. Profile functions at different downstream and lateral locations are extracted from a large-eddy simulation with an upstream turbine and compared with two lower-fidelity models: one that prescribes both the mean and standard deviation of the turbulence and one that prescribes only the mean. The aeroelastic response of an NREL 5 MW wind turbine is simulated with the three different wake-model fidelities, and various quantities of interest are compared. The mean values for the power and rotor speed for the medium-and low-fidelity model match well, but the accuracy of the fatigue loads varies greatly depending on the load channel. Prescribing the profile function for the standard deviation is only beneficial for the tower-base fore-aft moment; all other DELs had similar accuracies for both the medium- and low-fidelity models. The paper concludes that blade DELs can be estimated using these simple models with some accuracy, but care should be taken with the load channels related to the shaft torsion and tower-base fore-aft bending moment.

show abstract

Wind farm layout optimization with load constraints using surrogate modelling

Cited by 24 publications

References 9 publications

Reliability‐based layout optimization in offshore wind energy systems

Reliability‐based layout optimization in offshore wind energy systems

Modelling of turbine power and local wind conditions in wind farm using an autoencoder neural network

The Importance of Wake Meandering on Wind Turbine Fatigue Loads in Wake

Contact Info

Product

Resources

About