Multiple Turbines Research Articles

Multi-step ahead forecasting of wind vector for multiple wind turbines based on new deep learning model

Obtaining wind speed and direction predictions with high accuracy is of vital significance for the utilization of wind energy. This research focuses on three problems: how to realize multiple station prediction, multi-step ahead prediction and wind vector prediction. Firstly, a 4-D time series variable scene is proposed for multi-step ahead forecasting of multiple station wind vector. Then, a wind vector prediction model (CM-G) based on Convolutional Minimum Gate Memory Neural Network and Graph Convolution Neural Network is constructed to improve the prediction accuracy. Further, three multi-step ahead forecasting modes are compared and their advantages, disadvantages and application scenarios are analyzed. Finally, the model and method proposed in this study are applied to two datasets in Tibet, China. The experimental results and conclusions are as follows: (1) The prediction accuracy of CM-G proposed in this study is higher than that of the existing state-of-the-art model in 90 % of the metrics. (2) From single-step to multi-step ahead forecasting, the prediction accuracy of wind vector gradually decreases. (3) In wind vector prediction, the prediction accuracy of wind speed is higher than that of wind direction. (4) The direction of accuracy improvement for multiple stations is relatively consistent with the average wind direction.

Novel control of PV-wind-battery powered standalone power supply system based LSTM based ANN

Integrated wind-photovoltaic (PV) based standalone electric power supply systems are widely used for various applications. A battery storage system is needed to provide continuous power supply to loads despite changes in loads, wind speed, and solar irradiance. Power quality is crucial in these hybrid systems, as the battery needs to charge from surplus power when generation exceeds the load and discharge to meet load demand. A bidirectional DC to DC converter is used to connect the battery to the network, and maximum power point tracking devices with proper algorithms are incorporated for optimal utilization of PV and wind turbines. Multiple PV systems and wind turbines are considered for proper power supply system ratings. Long short-term memory (LSTM) based artificial neural network (ANN) controllers are implemented for various control units in the hybrid standalone power system. The proposed control techniques improve power quality under various situations. Results are presented using MATLAB/Simulink to evaluate the performance of the proposed method.

Lattice Boltzmann simulation for wake interactions of aligned wind turbines using actuator line model with turbine control

A wind turbine wake causes a decrease in wind speed and an increase in turbulence intensity. The wind turbine wake interaction is essential for predicting the power output of a wind farm consisting of many wind turbines. This research proposes a CFD method able to reproduce wake interactions and power outputs of multiple wind turbines with high speed and accuracy. Large eddy simulations with the lattice Boltzmann method are used for fluid calculations, specifically for large-scale CFD simulations. The wind turbines are represented using an actuator line model. Optimal power generation efficiency is achieved by controlling the rotor speed and blade pitch angle. Large-scale simulations of eight aligned wind turbines are conducted using 1.75 billion grid points and 40 GPUs. We compare two cases with and without control to investigate the effect of turbine control on wake and power output. Both the instantaneous and mean streamwise velocities confirm that the turbine control reduces the wake velocity deficit of the downwind wind turbine. High-speed inflow of wind to the downstream turbines augments their power output. With implementation of turbine control, the power outputs of the downstream turbines agree well with the observation data obtained in an earlier study. The results demonstrate the importance of controlling the rotational speed and pitch angle for actuator line simulations.

Efficient prediction of tidal turbine fatigue loading using turbulent onset flow from Large Eddy Simulations

To maximise the availability of power extraction from a tidal stream site, tidal turbines need to be able to operate reliably when located within arrays. This requires a thorough understanding of the operating conditions, which include turbulence, velocity shear due to bed proximity and roughness, ocean waves and due to upstream turbine wakes, over the range of flow speeds that contribute to the loading experienced by the devices. High-fidelity models such as Large Eddy Simulation (LES) can be used to represent these complex flow conditions and turbine device models can be embedded to predict loading. However, to inform micro-siting of multiple turbines with an array, the computational cost of performing multiple simulations of this type is impractical. Unsteady onset conditions can be generated from the LES to be used in an offline coupling fashion as input to lower-fidelity load prediction models to enable computationally efficient array design. In this study, an in-house Blade Element Momentum (BEM) method is assessed for prediction of the unsteady loads on the turbines of a floating tidal device with unsteady inflow developed with the in-house LES solver DOFAS. Load predictions are compared to those obtained using the same unsteady inflow to the commercial tool Tidal Bladed and from an Actuator Line Model (ALM) embedded in the LES solver. Estimates of fatigue loads differ by up to 3% for mean thrust and 11% for blade root bending moment for a turbine subject to a turbulent channel flow. When subjected to more complex flows typical of a turbine wake, the predictions of rotor thrust fatigue differ by up to 10%, with loads reduced by the inclusion of a pitch controller.

Effects of wind shear and thrust coefficient on the induction zone of a porous disk: A wind tunnel study

AbstractNeglecting the velocity reduction in the induction zone of wind turbines can lead to overestimates of power production predictions, and, thus, of the annual energy production for a wind farm. An experimental study on the induction zone associated with wind turbine operations is performed in the boundary‐layer test section of the BLAST wind tunnel at UT Dallas using stereo particle image velocimetry. This experiment provides a detailed quantification of the wind speed decrease associated with the induction zone for two different incoming flows, namely, uniform flow and boundary layer flow. Operations of wind turbines in different regions of the power curve are modeled in the wind tunnel environment with two porous disks with a solidity of 50.4% and 32.3%, which correspond to thrust coefficients of 0.71 and 0.63, respectively. The porous disks are designed to approximate the wake velocity profiles previously measured for utility‐scale wind turbines through scanning wind LiDARs. The results show that the streamwise velocity at one rotor diameter upwind of both disks decreases 1% more for the boundary layer flow than for the uniform flow. Further, the effect of shear in front of the disk with a higher thrust coefficient can be observed until 1.75 rotor diameter upwind of the disk, whereas for the disk with a lower thrust coefficient, the effect of shear becomes negligible at 1.25 rotor diameter upwind. It is found that at one rotor diameter upwind, for both incoming flows, the disk having a higher thrust coefficient has 2% more velocity reduction than the lower‐thrust‐coefficient disk. The results suggest that the variability in wind shear and rotor thrust coefficient, which is encountered during typical operations of wind turbines, should be considered for the development of improved models for predictions of the rotor induction zone, the respective cumulative effects in the presence of multiple turbines, namely, wind farm blockage, and more accurate predictions of wind farm power capture.

Transfer learning applications for autoencoder-based anomaly detection in wind turbines

Anomaly detection in wind turbines typically involves using normal behaviour models to detect faults early. Normal behaviour models are often implemented through the use of neural networks, of which autoencoders are particularly popular in this field. However, training autoencoder models for each turbine is time-consuming and resource intensive. Thus, transfer learning becomes essential for wind turbines with limited data or applications with limited computational resources. This study examines how cross-turbine transfer learning can be applied to autoencoder-based anomaly detection. Here, autoencoders are combined with constant thresholds for the reconstruction error to determine if input data contains an anomaly. The models are initially trained on one year’s worth of data from one or more source wind turbines. They are then fine-tuned using small amounts of data from the target wind turbine. Three methods for fine-tuning are investigated: adjusting the entire autoencoder, only the decoder, or only the threshold of the model. The performance of the transfer learning models is compared to baseline models that were trained on one year’s worth of data from the target wind turbine. The results of the tests conducted in this study indicate that models trained on data of multiple wind turbines do not improve the anomaly detection capability compared to models trained on data of one source wind turbine. In addition, modifying the model’s threshold can lead to comparable or even superior performance compared to the baseline, whereas fine-tuning the decoder or autoencoder further enhances the models’ performance.

Design and Fabrication of Wind Turbine Utilization for Water Pumping and Power Generation

bstract: Windmills have been a symbol of sustainable energy for centuries, and their relevance continues to grow in the modern era. This abstract explores the dual functionality of windmills for both pumping and power generation. Windmills have evolved from their traditional agricultural use to become versatile renewable energy sources, capable of harnessing wind energy for various applications. For pumping, windmills are employed to lift water from wells or reservoirs, providing a reliable source of water for irrigation and domestic use, especially in remote or off-grid areas. Their mechanical design, often featuring rotating blades, efficiently converts wind energy into rotational motion, driving a pump mechanism. In the realm of power generation, windmills are pivotal in the transition to cleaner energy sources. Modern wind turbines are designed with advanced aerodynamics and materials, enabling them to capture substantial kinetic energy from the wind. This energy is then transformed into electrical power through generators. Wind farms, comprised of multiple turbines, contribute significantly to electricity grids, reducing carbon emissions and dependence on fossil fuels. This abstract underscores the dual role of windmills as sustainable solutions for pumping and power generation, emphasizing their versatility and contribution to a greener future.

Multi-objective layout optimization for wind farms based on non-uniformly distributed turbulence and a new three-dimensional multiple wake model

Wind farm layout optimization (WFLO) can reduce the turbulence intensity on the wind turbines while maximizing power generation. To accurately compute turbulence intensity distribution in the wind farm, we utilize the innovative three-dimensional cosine-shaped turbulence intensity (3D-COTI) model. By integrating the three-dimensional cosine-shaped linear entrainment (3DCLE) wake model previously developed by the authors, we introduce a new multiple wind turbine wake model (3DCLE-M). This model considers the wake superposition effect, enabling precise calculation of power generation. Building upon the 3D-COTI and 3DCLE-M models, we propose a multi-objective WFLO approach to maximize power generation and minimize streamwise turbulence intensity in the wind-turbine-swept area. The results indicate that under maximized power generation, this approach significantly reduces turbulence intensity and its non-uniform distribution in the wind-turbine-swept area. In two ideal cases, the maximum turbulence intensity in all wind-turbine-swept areas of the optimized layout is 25 % and 3.8 % lower than that of the selected benchmark layout. By optimizing the layout of the Horns Rev offshore wind farm, a new layout can be obtained with increased total power output by 1.58 % and the reduced total duration of high-fatigue-load wind turbines under high turbulence intensity by 25 % compared to the benchmark.

Predictive capability of an improved AD/RANS method for multiple wind turbines and wind farm wakes

The actuator disc (AD) model combined with the Reynolds-averaged Navier-Stokes (RANS) model has been demonstrated as the most effective approach for simulating wind farm wakes due to its computational efficiency and acceptable accuracy. However, challenges remain in defining the reference inflow wind speed for the AD model and identifying a robust turbulence model for the RANS method. To solve these issues, two strategies for establishing the AD model (AD-up1D model and AD-local model), along with four turbulence models (standard k-ε, SST k-ω, linear pressure-strain RSM and modified RSM), are comprehensively validated here to predict multiple wakes within two representative wind farms. The temporal variability of multiple turbines’ power output and the significant impact of wind direction on the power is particularly emphasized. Overall, this study indicates that the proposed AD-local/Mod RSM method facilitates a realistic wake recovery in the cumulated wake flow and agrees well with the field measurements. It consistently performs well, never ranking last in any test cases, with percentage deviations ranging from 1.8% to 9.0%. Most importantly, this study evaluates the strengths and weaknesses of each AD/RANS method and provides recommendations for wind farm design purposes.

A Data-Driven Approach for Generalizing the Laminar Kinetic Energy Model for Separation and Bypass Transition in Low- and High-Pressure Turbines

Abstract No common laminar kinetic energy (LKE) transition model has to date been able to predict both separation-induced and bypass transition, both phenomena commonly found in low-pressure turbines and high-pressure turbines. Here, a data-driven approach is adopted to develop a more general LKE transition model suitable for both transition modes. To achieve this, two strategies are adopted. The first is to extend the computational fluid dynamics (CFD)-driven model training framework for simultaneously training models on multiple turbine cases, subject to multiple objectives. By increasing the training data set, different transition modes can be considered. The second strategy employed is the use of a newly derived set of local non-dimensionalized variables as training inputs to reduce the search space. Because one of the training turbine cases is characterized by strong unsteady effects, for the first time an unsteady solver is utilized during the CFD-driven training, and the time-averaged results are used to calculate the cost function as part of the model development process. The results show that the data-driven models do perform better, in terms of their predictions of pressure coefficient, wall shear stress, and wake losses, than the baseline model. The models were then tested on two previously unseen testing cases, one at a higher Reynolds number and one with a different geometry. For both testing cases, stable solutions were obtained with results improved over the predictions using the baseline models.

Multi-Wind Turbine Wind Speed Prediction Based on Weighted Diffusion Graph Convolution and Gated Attention Network

The complex environmental impact makes it difficult to predict wind speed with high precision for multiple wind turbines. Most existing research methods model the temporal dependence of wind speeds, ignoring the spatial correlation between wind turbines. In this paper, we propose a multi-wind turbine wind speed prediction model based on Weighted Diffusion Graph Convolution and Gated Attention Network (WDGCGAN). To address the strong nonlinear correlation problem among multiple wind turbines, we use the maximal information coefficient (MIC) method to calculate the correlation weights between wind turbines and construct a weighted graph for multiple wind turbines. Next, by applying Diffusion Graph Convolution (DGC) transformation to the weight matrix of the weighted graph, we obtain the spatial graph diffusion matrix of the wind farm to aggregate the high-order neighborhood information of the graph nodes. Finally, by combining the DGC with the gated attention recurrent unit (GAU), we establish a spatio-temporal model for multi-turbine wind speed prediction. Experiments on the wind farm data in Massachusetts show that the proposed method can effectively aggregate the spatio-temporal information of wind turbine nodes and improve the prediction accuracy of multiple wind speeds. In the 1h prediction task, the average RMSE of the proposed model is 28% and 33.1% lower than that of the Long Short-Term Memory Network (LSTM) and Convolutional Neural Network (CNN), respectively.

Wind power forecasting method of large-scale wind turbine clusters based on DBSCAN clustering and an enhanced hunter-prey optimization algorithm

As the large-scale grid connection of wind turbines poses challenges to the safe and stable operation of the power grid, it is necessary to forecast the power of wind turbine clusters. However, with the rapid increase in the installed capacity of wind turbines, simultaneously improving the accuracy and efficiency of large-scale wind-turbine cluster power prediction models has become a challenge. In this research, a novel hybrid power-forecasting model for wind-turbine clusters has been designed using density-based spatial clustering of applications with noise (DBSCAN) and an enhanced hunter-prey optimization algorithm (ENHPO). First, the Pearson correlation coefficient was used to select multiple variables that significantly affected the wind power. Second, multiple wind turbines are clustered into different groups using the DBSCAN clustering algorithm, and ENHPO is employed to optimize the DBSCAN parameters to strengthen the clustering performance. Finally, one wind turbine with a high correlation was selected as the representative wind turbine in each group, and power prediction was achieved using the multivariate long short-term memory. Simulation results of four datasets in different seasons show that, compared with other clustering methods, such as fuzzy C-means, balanced iterative reduction, and clustering using hierarchies, K-means, and density peak clustering, the prediction accuracy and prediction efficiency of the proposed hybrid prediction model are improved by 21.85% and 18.07%, respectively, on average. The experimental results effectively demonstrate that the designed model can enhance the accuracy and efficiency of power forecasting simultaneously.

Condition Monitoring of Wind Turbine Drivetrain Bearings

Abstract As an extremely valuable and reliable energy source, the renewable energy is developing around the world at an unprecedented pace. By the end of 2022, the global installed capacity of offshore wind turbines is expected to reach 46.4 GW, among which 33.9 GW in Europe. Efficiencies in Operations and Maintenance (O&M) offer potential to achieve significant cost savings as it accounts for around 20%–30% of overall offshore wind farm costs. A recent study of approximately 350 offshore wind turbines indicates that gearboxes might have to be replaced as early as 6.5 years. Therefore sensing and condition monitoring (CM) systems are needed in order to obtain reliable information on the state and condition of different critical parts. The development and use of such technologies will allow companies to schedule actions at the right time. The reduced costs of O&M enable the wind energy at a competitive price thus strengthening productivity of the wind energy sector. At the academic level, a plethora of methodologies have been proposed during the last decades focusing toward early and accurate fault detection. Among others, Envelope Analysis is one of the most important methodologies, where the envelope of the vibration signal is estimated, usually after filtering around a selected frequency band excited by impacts due to the faults. Different tools, such as Kurtogram, have been proposed in order to accurately select the optimum filter parameters (center frequency and bandwidth). Cyclostationary analysis and corresponding methodologies, i.e., the cyclic spectral correlation and the cyclic spectral coherence, have been proved as powerful tools for CM. On the other hand the application, test, and evaluation of such tools in general industrial cases is still rather limited. Therefore the main aim of this paper is the application and evaluation of advanced diagnostic techniques and diagnostic indicators, including the Enhanced Envelope Spectrum and the Spectral Flatness on real-world vibration data collected from vibration sensors on gearboxes in multiple wind turbines over an extended period of time of nearly four years. The diagnostic indicators are compared with classical statistic time and frequency indicators, i.e., Kurtosis, Crest Factor, etc. and their effectiveness is evaluated based on the successful detection of two failure events.

Analysis of Wind Turbine Wake Dynamics by a Gaussian-Core Vortex Lattice Technique

The development and deployment of the next generation of wind energy systems calls for simulation tools that model the entire wind farm while balancing accuracy and computational cost. A full-system wind farm simulation must consider the atmospheric inflow, the wakes and consequent response of the multiple turbines, and the implementation of the appropriate farm-collective control strategies that optimize the entire wind farm’s output. In this article, we present a novel vortex lattice model that enables the effective representation of the complex vortex wake dynamics of the turbines in a farm subject to transient inflow conditions. This work extends the capabilities of our multi-physics suite, CODEF, to include the capability to simulate the wakes and the high-fidelity aeroelastic response of multiple turbines in a wind farm. Herein, we compare the results of our GVLM technique with the LiDAR measurements obtained at Sandia National Laboratories’ SWiFT facility. The comparison shows remarkable similarities between the simulation and field measurements of the wake velocity. These similarities demonstrate our model’s capabilities in capturing the entire wake of a wind turbine at a significantly reduced computational cost as compared to other techniques.

Computational Fluid Dynamics Analysis of Flow Characteristics and Heat Transfer Variabilities in Multiple Turbine Blade Cooling Channels

Abstract In the realm of gas turbine engine technology, turbine blades are persistently subjected to extreme thermal conditions due to prolonged exposure to high-temperature gases. Given this operating environment, devising effective cooling strategies is paramount for ensuring the turbine’s safety and longevity. Despite the critical nature of this issue, existing literature scarcely addresses cooling channels that incorporate factors like variable cross-sectional attributes and rotational effects. In this study, Computational Fluid Dynamics (CFD) and numerical methodologies were employed to analyze the flow mechanism and performance of three distinct cooling channels characterized by variable cross-sectional features. The results indicate a marginal thermal performance improvement of up to 3% for U-shaped cooling channels, specifically at a Reynolds number of 20,000, in comparison to Net cooling channels. Conversely, jet impingement channels outperform U-shaped channels by an impressive margin of over 11%. Additionally, at a Reynolds number of 60,000, jet impingement cooling channels manifest a significant upsurge of nearly 25% in the ratio between average Nusselt number, and the reference Nusselt number when compared with U-shaped and Net cooling channels.

Optimization Analysis of Multiple Steam Turbines and Condensers in Paper Mill Power Plant

The cooling tower water’s influence on turbines and condensers thermal efficiency were researched in this paper and the condenser working condition impact to turbines thermal efficiency was also discussed. Based on the basic model of turbines, condensers and cooling water, the optimization model for the system of turbines, condensers, cooling water and cooling water towers fan was proposed. One paper mill power plant which has multiple turbines, condensers and cooling water towers was taken as a case to verify the optimization model, the result showed that the optimization model was practical and useful for paper mill power plant; it can improve power plant efficiency and bring good economical performance for paper

Wind turbine wake superposition under pressure gradient

We investigate the effect of pressure gradient on the cumulative wake of multiple turbines in wind tunnel experiments spanning across a range of adverse pressure gradient (APG), zero pressure gradient (ZPG), and favorable pressure gradient (FPG). Compared to the upstream-most turbine, the in-wake turbines exhibit lower (higher) wake velocity in APG (FPG) than in the ZPG. The maximum velocity deficit shows a lesser difference for the in-wake turbine between different cases compared to the upstream-most one. This is linked to the effect of the wake of the upstream turbine. Conversely, the wake width varies more for the in-wake turbines. A new analytical approach to model the cumulative wake velocity deficit is proposed. This approach extends the application of the analytical pressure gradient model to multiple turbine wakes. Specifically, the new approach explicitly accounts for the effect of the pressure gradient induced by the wake of the upstream turbine on the wake of the downstream one. The new method is compared to the linear summation approach and experimental data. It agrees well with the experiments and outperforms the linear summation approach.

Data-driven Anomaly Detection Method Based on Similarities of Multiple Wind Turbines

Data-driven Anomaly Detection Method Based on Similarities of Multiple Wind Turbines

Supervised learning-based multi-site lean blowout prediction for dry low emission gas turbine

Current dry low emission (DLE) gas turbines have extremely low nitrogen oxides (NOx) emissions, allowing them to comply with stringent environmental regulations. The ultra-low temperatures required to achieve such low emissions also increase the risk of a lean blowout (LBO). DLE gas turbines are particularly vulnerable to LBO, which can lead to system instability, higher carbon oxides (CO) emissions, damaged components, substantial financial losses and environmental damages. In this regard, this current study proposes a novel supervised learning-based prediction technique with efficient performance on out-of-distribution data from multiple DLE gas turbine plants (multi-site). The proposed prediction approach exploits the competitive advantages of both adaptive boosting (AdaBoost) and Linear support vector machine (LSVM) to improve the generalization capability as well as prediction accuracy. The proposed algorithm was trained and tested using a real-world DLE gas turbine dataset from six different sites. The result indicates that the proposed model consistently achieving LBO prediction accuracy rates above 99.9% and Mathew’s correlation coefficient (MCC) score above 0.9 across all datasets. All lean-premixed gas turbines could benefit from the developed Ada-LSVM, as it is an accurate model with a high generalization performance that can be used across multiple sites.

Collective large-scale wind farm multivariate power output control based on hierarchical communication multi-agent proximal policy optimization

Wind power is becoming an increasingly vital source of renewable energy worldwide. However, controlling power generation in wind farms faces significant challenges due to the inherent complexity of these systems. To address this issue and maximize power output, we propose a novel communication-based multi-agent deep reinforcement learning approach for large-scale wind farm control. We introduce a multivariate power model for wind farms to analyze the impact of wake effects on power output. This model considers controllable variables such as axial induction factor, yaw angle, and tilt angle. To coordinate the continuous controls in large-scale wind farms, we propose the hierarchical communication multi-agent proximal policy optimization (HCMAPPO) algorithm. The wind farm is divided into multiple wind turbine aggregators (WTAs), and neighboring WTAs can exchange information through hierarchical communication to optimize power output. Our simulation results demonstrate that the multivariate HCMAPPO approach significantly increases wind farm power output compared to traditional PID control, coordinated model-based predictive control, and the multi-agent deep deterministic policy gradient algorithm. The HCMAPPO algorithm can be trained using a thirteen-turbine wind farm environment and effectively applied to large-scale wind farms. Importantly, as the wind farm scale increases, there is no significant increase in wind turbine blade fatigue damage from wake control. This multivariate HCMAPPO control strategy enables the collective maximization of power output in large-scale wind farms.