Surrogate-Based Optimization of Tandem Twin-Rotor Wind Turbines Accounting for Wake Interactions

Wake interaction in a tandem twin-rotor turbine couples power capture with thrust on the shared support. This study develops a time-resolved surrogate for upstream IEA Wind 22 MW and downstream NREL 5 MW rotors. It covers spacing ratios $0.01\leq L/D\leq0.20$ and tip-speed ratios $4\leqλ\leq11$. Vortex-method simulations of 160 configurations under turbulent inflow train per-time-node XGBoost models to predict 600 s power, thrust, and torque histories. A CNN-Attention-LSTM network extends these histories, and a genetic algorithm balances the system power and thrust coefficients. Ten-fold cross-validation gives median power-history $R^2$ values of 0.995 upstream and 0.958 downstream for unseen configurations, but the first stage cannot extrapolate in tip-speed ratio. With fixed pitch and prescribed rotor speed, the coefficients apply where both rotors reach the simulated tip-speed ratio below rated wind speed. Increasing spacing raises upstream power by up to 6% and lowers downstream power by up to 62%. System power changes by less than 4% at fixed $λ$. The high-power end of the Pareto front lies near $λ=10$. At $L/D=0.20$, the system power coefficient is highest. This configuration has the lowest peak thrust among the high-power candidates.

Authors

Publication Details

Journal
arXiv (Cornell University)
Published
2026-11-16
DOI
https://doi.org/10.48550/arxiv.2611.21185
Primary Topic
Wind Energy Research and Development
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Surrogate-Based Optimization of Tandem Twin-Rotor Wind Turbines Accounting for Wake Interactions

Yinong Tian, Shan Wang
arXiv (Cornell University)
Wind Energy Research and Development
preprint

Surrogate-Based Optimization of Tandem Twin-Rotor Wind Turbines Accounting for Wake Interactions

Yinong Tian, Shan Wang
preprint en

Abstract

Wake interaction in a tandem twin-rotor turbine couples power capture with thrust on the shared support. This study develops a time-resolved surrogate for upstream IEA Wind 22 MW and downstream NREL 5 MW rotors. It covers spacing ratios $0.01\leq L/D\leq0.20$ and tip-speed ratios $4\leqλ\leq11$. Vortex-method simulations of 160 configurations under turbulent inflow train per-time-node XGBoost models to predict 600 s power, thrust, and torque histories. A CNN-Attention-LSTM network extends these histories, and a genetic algorithm balances the system power and thrust coefficients. Ten-fold cross-validation gives median power-history $R^2$ values of 0.995 upstream and 0.958 downstream for unseen configurations, but the first stage cannot extrapolate in tip-speed ratio. With fixed pitch and prescribed rotor speed, the coefficients apply where both rotors reach the simulated tip-speed ratio below rated wind speed. Increasing spacing raises upstream power by up to 6% and lowers downstream power by up to 62%. System power changes by less than 4% at fixed $λ$. The high-power end of the Pareto front lies near $λ=10$. At $L/D=0.20$, the system power coefficient is highest. This configuration has the lowest peak thrust among the high-power candidates.

arXiv (Cornell University)
Wind Energy Research and Development
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon