Power Output Of Wind Turbine Research Articles

Gradient boosting-based approach for short- and medium-term wind turbine output power prediction

Being able to predict the output power of wind turbines and wind farms is crucial in the process of integrating such stochastic energy sources with power systems. To support stakeholders in short- and mid-term wind power prediction, a novel data-driven Machine Learning based approach is proposed. This approach relies on three Gradient Boosting (GB) regressor implementations. The novelty of our approach is also manifested in the fact, that it is respectively based on the use of MERRA-2 reanalysis data and GEOS FP meteorological forecasts in models training and wind power prediction. It makes the short- and mid-term prediction unique in enriching the results even for time horizon of 240 h with resolution of 1 h. The data preprocessing and cleaning, feature engineering, and training, testing and validation of the models are presented in details. The performances of the models and prediction accuracy are evaluated relying on a few absolute and relative error measures. The proposed methodology is implemented in the output power prediction of a wind turbine located in Poland. The results of predictions are compared with other Machine Learning algorithms. The results show that proposed GB implementations can capture accepted accuracy of prediction and outperform other investigated algorithms.

Research on complex wake interference of aligned rotors considering the precone variation of the upstream wind turbine

To understand the interaction wake effects of aligned horizontal-axis wind turbines, a refined and low-cost wake simulation framework is developed by combining the actuator line (AL) with a local dynamic Smagorinsky (LDS) model. Firstly, the published performance of the NREL-5MW wind turbine is used to corroborate the accuracy of the AL-LDS coupling model. Secondly, the research is extended to predict the interference between two tandem rotor wakes. The influence of the precone characteristics on the overall wake and power output of the tandem wind turbines is explored by modifying the upstream wind turbine and introducing the latest third-generation Ω vortex identification method in the post-processing procedure. According to results, the tandem wind turbines have complex wake interference, with the precone angle variation always reducing the global output power, and leading to an unstable downstream turbine wake. This is detrimental to system lifespan due to high performance fluctuation and stress and as well may have ecological implications due to sediment alteration. From results, an upwind turbine with precone angle (negative) can lead to higher downstream but lower upstream turbine performance, compared with a non-tilted turbine. Despite the optimal precone angle of the upstream wind turbine is 0°, a -2.5° proves more advantageous owing to the large flexible deformation of the turbine wingtips. The findings can serve as a continuous incentive for improving wind farm numerical simulation.

Optimization of wind farm portfolios for minimizing overall power fluctuations at selective frequencies – a case study of the Faroe Islands

Abstract. Hourly modeled wind turbine power output time series – modeled based on outputs from the mesoscale numerical weather prediction system Weather Research and Forecasting Model (WRF) – are used to examine the spatial smoothing of various wind farm portfolios located on a complex isolated island group with a surface area of 1400 km2. Power spectral densities (PSDs), hourly step-change functions, and duration curves are generated, and the 5th and 95th percentiles and the standard deviations of the hourly step-change functions are calculated. The spatial smoothing is identified from smaller high-frequency PSD amplitudes, lower hourly fluctuations, and more flat duration curves per installed wind power capacity, compared with single wind turbine outputs. A discussion on the limitation of the spatial smoothing for the region is included, where a smoothing effect is observed for periods of up to 1–2 d, although it is most evident at higher frequencies. By maximizing the smoothing effect, optimal wind farm portfolios are presented with the intention of minimizing overall wind power fluctuations. The focus is mainly on the smoothing effect on the 1–3 h timescale, during which the coherency between wind farm power outputs is expected to be dependent on how the regional weather travels between local sites, thereby making optimizations of wind farm portfolios relevant – in contrast to a focus on either lower or higher frequencies on the scale of days or minutes, respectively, during which wind farm power output time series are expected to be either close to fully coherent due to the same weather conditions covering a small region or not coherent as the turbulences in separate wind farm locations are expected to be uncorrelated. Results show that an optimization of the wind farm capacities at 14 pre-defined wind farm site locations has a minimal improvement on the hourly fluctuations compared with a portfolio with equally weighted wind farm capacities. However, choosing optimized combinations of individual wind farm site locations decreases the 1–3 h fluctuations considerably. For example, selecting a portfolio with four wind farms (out of the 14 pre-defined wind farm site locations) results in 15 % lower 5th and 95th percentiles of the hourly step-change function when choosing optimal wind farm combinations compared with choosing the worst wind farm combinations. For an optimized wind farm portfolio of seven wind farms, this number is 13 %. Optimized wind farm portfolios consist of distant wind farms, while the worst portfolios consist of clustered wind farms.

Wind turbine output power prediction and optimization based on a novel adaptive neuro-fuzzy inference system with the moving window

This study focuses on predicting the output power of wind turbines (WTs) using the wind speed and WT operational characteristics. The main contribution of this work is a model identification method based on an adaptive neuro-fuzzy inference system (ANFIS) through multi-source data fusion on a moving window (MoW). The proposed ANFIS-MoW-based approach was applied to data in different time series windows, namely the very short-term, short-term, medium-term and long-term time horizons. Data collected from a 30-MW wind farm on the west coast of Nouakchott (Mauritania) were used in the computational analysis. In comparison to nonparametric models from the literature and models employing artificial intelligence machine learning techniques, the proposed ANFIS-MoW model demonstrated superior predictions for the output power of the WT with the fusion of very few data collected from different WTs. Moreover, for various time series windows (TSW) and meteorological conditions, additional benchmarking demonstrated that the ANFIS-MoW-based method outperformed five existing ANFIS-based models, including grid partition (ANFIS-GP), subtractive clustering (ANFIS-SC), fuzzy c-means clustering (ANFIS-FCM), genetic algorithm (ANFIS-GA), and particle swarm optimization (ANFIS-PSO). The results indicated that the suggested methodology is a promising soft-computing tool for accurately estimating the WT output power for WTs' sustainability through better control of their operation.

An Integration of Genetic Feature Selector, Histogram-Based Outlier Score, and Deep Learning for Wind Turbine Power Prediction

ABSTRACT During the last decades, the importance of clean energy resources is being increased. Wind is one of the most significant clean energy resources. Forecasting the output power of wind turbines is important for turbine control to improve power grids’ performance and maintenance. In this study, a novel method for predicting the power of wind turbines is proposed based on integrating data preprocessing, re-sampling, feature selection (genetic algorithm), and outlier detection (Histogram-Based Outlier Score) techniques to prepare the data for the deep learning (DL) algorithms. The results show that, after removing features chosen by the genetic algorithm (GA) method, the mean absolute error (MAE) reduced considerably to 333.7. Integrating Histogram-Based Outlier Score (HBOS) with genetic algorithm (GA) significantly decreased the MAE to 488. Comparing the results with benchmark machine learning algorithms, namely Light Gradient Boosting Machine (LGBM), Extreme Gradient Boosting Regression (×GBR), K-Nearest Neighbor (KNN), Support Vector Regression (SVR), and Recurrent Neural Networks (RNN) models, shows a remarkable improvement in the accuracy of turbine power prediction for about 78.7, 944.9, 104.7, 1456.6, and 17.1 in mean absolute error (MAE), respectively.

Potential of wind generators as additional source of energy in hybrid renewable energy systems located in Baltic region

Study of the joint daily variability of wind speed and solar activity in the region of Gulf of Finland and its impact on renewable energy sources. The influence of external factors on the variability of total output power of ground-based wind turbines and solar photovoltaic panels is investigated. To a small extent, wind turbines are able to complement the generation of renewable energy sources: 15% –30% wind turbines and 70% –85% solar photovoltaics is preferable ratio of renewable energy sources. Comparing two similar climate regions in different scenarios lead economic feasibility strongly corelate with fuel prices. Standalone diesel generators can be quite effective at first, but in the long-term it is not viable solution due to fuel costs. Combined wind-PV renewable energy system is more profitable, but it has higher initial cost.

Demand Side Management Strategy for Multi-Objective Day-Ahead Scheduling Considering Wind Energy in Smart Grid

Distributed energy resources (DERs) and demand side management (DSM) strategy implementation in smart grids (SGs) lead to environmental and economic benefits. In this paper, a new DSM strategy is proposed for the day-ahead scheduling problem in SGs with a high penetration of wind energy to optimize the tri-objective problem in SGs: operating cost and pollution emission minimization, the minimization of the cost associated with load curtailment, and the minimization of the deviation between wind turbine (WT) output power and demand. Due to climatic conditions, the nature of the wind energy source is uncertain, and its prediction for day-ahead scheduling is challenging. Monte Carlo simulation (MCS) was used to predict wind energy before integrating with the SG. The DSM strategy used in this study consists of real-time pricing and incentives, which is a hybrid demand response program (H-DRP). To solve the proposed tri-objective SG scheduling problem, an optimization technique, the multi-objective genetic algorithm (MOGA), is proposed, which results in non-dominated solutions in the feasible search area. Besides, the decision-making mechanism (DMM) was applied to find the optimal solution amongst the non-dominated solutions in the feasible search area. The proposed scheduling model successfully optimizes the objective functions. For the simulation, MATLAB 2021a was used. For the validation of this model, it was tested on the SG using multiple balancing constraints for power balance at the consumer end.

Active blade pitch control and stabilization of a wind turbine driven PMSG for power output regulation

A common strategy in controlling a permanent magnet synchronous generator (PMSG) driven by a wind turbine is the maximization of output power of the wind turbine itself. A control strategy must be adopted, is to deliver a desired reduced amount of power whenever it is required. In order to realize the direct control of wind turbine output power across a wide range of wind speeds, a linearized parameter varying dynamic model of the nonlinear wind turbine system including wind disturbances is developed and used in this paper. The stability of the wind turbine system is analyzed and a blade pitch controller is designed, based on the linearized, parameter-varying, model-predictive control and is validated. Thus, the wind turbine is regulated in a way that the generator delivers the demanded power output to the load. Moreover, the blade pitch control system also performs the key function of augmenting the stability of the wind turbine, for the right choice of the gains.

Study on Aerodynamic Characteristics of a Savonius Wind Turbine with a Modified Blade

In order to improve the static start-up problem of Savonius wind turbines, a Savonius wind turbine with a modified blade is proposed. It was obtained by twisting the half-cylindrical blades of the basic Savonius wind turbine by 70°. The aerodynamic performance of the wind turbine before and after the modification was compared. Firstly, the static torque coefficient of two wind turbines at different azimuth angles were obtained by means of three-dimensional numerical simulation. The static flow field around the wind turbine was analyzed. Then, the output power and speed characteristics of a spiral Savonius wind turbine under different incoming wind speeds were evaluated in the wind tunnel. The results show that, compared with the Savonius wind turbine with half-cylindrical blades, the spiral wind turbine could start at any azimuths in one rotation cycle. The reverse torque was eliminated. The static torque coefficient fluctuation range was reduced by 10%. The start-up performance was effectively improved. This investigation could provide guidance for the improvement of start-up characteristics of Savonius wind turbines.

A stochastic agent-based cooperative scheduling model of a multi-vector microgrid including electricity, hydrogen, and gas sectors

With increasing hydrogen usage, hydrogen subsystem should be considered in the multi-energy chain and a model is required that assumes current energy infrastructures, while preserving independent energy subsystems. In this study, a stochastic agent-based model is introduced for the coordinated scheduling of multi-vector microgrids considering interactions between electricity, hydrogen, and gas agents. Power to hydrogen (P2H) through electrolysis, hydrogen to power (H2P) through fuel cells, hydrogen to gas (H2G) through methanation, and gas to power (G2P) through distributed generation (DG) units are modeled to present the interactions among energy agents. The interactions in terms of shared variables and coupling constraints are described using augmented Lagrangian relaxation (ALR) and alternating direction method of multipliers (ADMM) to obtain three correlated optimization problems, preserving the privacy of energy sectors with minimum data exchange. An iterative process is accomplished among energy sectors to reach a consensus. Uncertainties in the wind turbine (WT) and photovoltaic (PV) power output, hydrogen vehicles (HVs), demands, and prices are captured using a stochastic method. To evaluate the proposed method, case studies are conducted using a multi-energy microgrid. The results verify that the microgrid is well scheduled and the interactions are accurately modeled, representing the effectiveness of the proposed method.

Active Disturbance Rejection Control for Wind Turbine Fatigue Load

With the participation of wind power in grid frequency modulation, the fatigue load of the wind turbine increases accordingly. A new control method that considers both fatigue load and output power of wind turbine (WT) is proposed in this paper. A linear active disturbance rejection control (LADRC) is designed and applied for the pitch angle in the wind turbine load reduction control. The particle swarm optimization (PSO) algorithm is used to optimize the parameters of the wind turbine controller, and the total variation of the wind turbine shaft torque and tower bending moment is added to construct a new objective function to further reduce the fatigue load of the wind turbine. The design-optimized controller is validated on a 5 MW wind turbine in SimWindFarm. The simulation results show that the LADRC controller can accurately track the reference power of the wind turbine, reduce the pitch angle fluctuation of the wind turbine, reduce the fatigue load of the wind turbine, and improve the service life of the wind turbine.

Power prediction of wind turbine in the wake using hybrid physical process and machine learning models

Precise power prediction for wind turbines under wake effects is requisite for wind farm wake control to increase the energy production and economic benefits. Wind farm wake effects that have spatial and temporal characteristics significantly complicate the physical modelling. To overcome the modelling difficulties, this paper proposes two new coupled structures of physical process and machine learning to predict wind turbine output power under wake effects. Their structures are fully presented and verified by comparisons to the pure physical model, neural network model alone and existing physical-guided neural network models. One of the new models that couples physical model and transfer learning approach shows the best prediction performance. It uses data generated from physical model and small size real data to establish the model, which makes it adaptive to data with different distribution. The results confirm that a robust hybrid physical and machine learning model can simultaneously inherit the advantages from developments of physical models and machine learning approaches. Important insights into HPML models for output power prediction of WTs under wake effects are provided.

Power Prediction of Wind Turbine Based on The Presumed Shape of Power Curve

An accurate model of power plays a crucial role in turbine energy assessment, wind turbine condition monitoring, estimation of wind energy potential, warranty formulations, power forecasting, wind turbine selection, optimization of the operational cost and expansion of windfarm. To achieve all these, algorithms of linear and cubic law models are used to predict the output power of BWC Excel 10 wind turbine. The comparative results show that the considered models can approximate and satisfactorily predicts the output power of wind turbines when compared with fundamental equation of wind turbine that depends on stringent factors like air density, turbine blade parameters, mechanical and control issues etc to yield similar results.

Wind resource assessment considering the influence of humidity

The current study presents a wind resource assessment (WRA) approach by combining existing approaches, including wind probability density estimation based on hourly wind speed frequency, wind power density (WPD) and wind energy density (WED), wind turbine (WT) power output and power curve modeling, and annual energy production (AEP). Wind probability density investigation employed various probability density functions (PDF), including parametric probability density functions such as Weibull, Normal, and Gamma, and non-parametric distribution, including Kernel Density Estimator (KDE). The present study also models the influence of humidity on air density for estimating WPD, WT power output, and AEP. The current study validated the proposed approach by conducting case studies for selected sites of remote Indonesian archipelago islands. AEP estimation proposed by this study can assist the site-turbine fitting design, especially for relatively moist locations.

An Efficient Estimation of Wind Turbine Output Power Using Neural Networks

Wind energy is a valuable source of electric power as its motion can be converted into mechanical energy, and ultimately electricity. The significant variability of wind speed calls for highly robust estimation methods. In this study, the mechanical power of wind turbines (WTs) is successfully estimated using input variables such as wind speed, angular speed of WT rotor, blade pitch, and power coefficient (Cp). The feed-forward backpropagation neural networks (FFBPNNs) and recurrent neural networks (RNNs) are incorporated to perform the estimations of wind turbine output power. The estimations are performed based on diverse parameters including the number of hidden layers, learning rates, and activation functions. The networks are trained using a scaled conjugate gradient (SCG) algorithm and evaluated in terms of the root mean square error (RMSE) and mean absolute percentage error (MAPE) indices. FFBPNN shows better results in terms of RMSE (0.49%) and MAPE (1.33%) using two and three hidden layers, respectively. The study indicates the significance of optimal selection of input parameters and effects of changing several hidden layers, activation functions, and learning rates to achieve the best performance of FFBPNN and RNN.

A parallel Archimedes optimization algorithm based on Taguchi method for application in the control of variable pitch wind turbine

Archimedes optimization algorithm (AOA) is a recent metaheuristic algorithm that offers several advantages, including a few parameters, an easy-to-understand interface, and easy implementation. Still, some drawbacks exist, e.g., lack of diversity for search-exploring capacity, drop-trap local optimum. This study suggests a new variant of AOA based on the parallel and Taguchi method (TPAOA) for the global optimization problems and the wind turbine parameter adjust-tuning variable pitch controller problem. The parallel mechanism with communication strategy and the Taguchi orthogonal combination deal with the AOA’s drawbacks. The experimental results show that the proposed algorithm is more competitive than the other algorithms under the CEC2017 test suite. The wind turbine problem of parameter tuning difficulty of variable pitch controller is solved by applying the TPAOA. Compared solution results show that the TPAOA proves the feasibility smooth the output power of wind turbines and reducing the impact of wind speed fluctuations on the power grid, which has high feasibility.

A New Wind Turbine Power Performance Assessment Approach: SCADA to Power Model Based with Regression-Kriging

Assessment of the wind turbine output power (WTG OP) during the operation and maintenance is one of the key indicators of operation quality evaluation. It is often carried out in the form of the wind speed-power curve. This form only considers the wind speed, and it is usually measured according to relevant IEC standards, e.g., IEC 61400-12, which has problems such as long measurement duration and harsh conditions. This study proposes a WTG OP assessment method based on SCADA data by using the regression-kriging algorithm. The influences of wind shear, turbulence intensity, and air density on the WTG OP were analyzed. Two regression-kriging output power models were built based on SCADA data (i.e., SCADA2power model) and wind resource parameters from met mast (i.e., wind2power model). According to the evaluation of the simulation result, it was found that the results of the two models are basically consistent. Based on the evaluation of historical data under normal operating conditions, the goodness of fitting output power of the two models is 99.9%. This shows that the regression-kriging-based wind turbine power performance assessment method based on SCADA data has an accurate prediction and the potential of general application in WTG OP evaluation.

Wind process pattern forecasting based ultra-short-term wind speed hybrid prediction

Wind power has received extensive attention due to its superiorities of clean and pollution-free. However, because of the randomness and volatility of wind power, accurate power prediction is needed to help its consumption. Wind speed is the key to wind power prediction, but traditional prediction method cannot accurately grasp the wind speed variation trend and the traditional wind process partition method has some defects, an ultra-short-term wind speed hybrid prediction method based on wind process pattern forecasting is proposed in this paper. Firstly, a wind process (WP) division method considering the influence of wind speed on the operation state, mode switch and output power of wind turbine is proposed. Secondly, according to the operating characteristics of the wind turbine, all the WPs is classified into different wind process patterns (WPP), and the effectiveness of the classification is verified. Then, the Adaboost algorithm is used to forecast the WPP of the next 4 h. Finally, the wind speed hybrid prediction model of each pattern is established, the corresponding model is automatically selected based on WPP to predict the wind speed. Simulation results show that the proposed model can reliably forecast future WPP and the prediction accuracy is better than conventional models.

An economic investigation of the wind-hydrogen projects: A case study

Hydrogen as an energy carrier can play a significant role in reducing environmental emissions if it is produced from renewable energy resources. This research aims to assess hydrogen production from wind energy considering environmental, economic, and technical aspect for the East Azerbaijan province of Iran. The economic assessment is performed by calculation of payback period, levelized cost of hydrogen, and levelized cost of electricity. Since uncertainty in the power output of wind turbines may affect the payback period, all calculations are performed for four different turbine degradation rates. While it is common in the literature to choose the wind turbine based on a single criterion, this study implements Multi-Criteria Decision-Making (MCDM) techniques for this purpose. The results of Step-wise Weight Assessment Ratio Analysis illustrates that economic issue is the most important criterion for this research. The results of Weighted Aggregated Sum Product Assessment shows that Vestas V52 is the most suitable wind turbine for Ahar and Sarab cities, while Eovent EVA120 H-Darrieus is a better choice for other stations. The most suitable location for wind power generation is found to be Ahar, where it is estimated to annually generate 2914.8 kWh of electricity at the price of 0.045 $/kWh, and 47.2 tons of hydrogen at the price of 1.38 $/kg, which result in 583 tons of CO2 emission reduction.

Active Power Consensus Control for Wind Turbines with Time Delays

The distributed active power control problem is explored by equating wind turbines to multi-agent systems in this paper. Both time delays and unknown topological relations are considered in the proposed model. Using the graph discovery algorithm, the algebraic connectivity of the graph is found in advance. Further, a proportional control protocol is proposed based on the adjustable margin of different wind turbines. Moreover, the proposed distributed controller not only handles the problem of supply-demand balance between the wind farm and power grid but also regulates the output power of individual wind turbines based on their capacities. Finally, simulations are performed on the wind turbines to illustrate the validity of the proposed method.