An official website of the United States government
Freezing of wind turbines causes loss of wind-generated power. Forecasting or prediction of icing on wind turbine blades based on SCADA sensor data allows taking appropriate actions before icing occurs. This paper presents a newly developed deep learning network model named PCTG (Parallel CNN-TCN GRU) for the purpose of high-accuracy and long-term prediction of icing on wind turbine blades. This model combines three networks, the CNN, TCN, and GRU, in order to incorporate both the temporal aspect of SCADA time-series data as well as the dependencies of SCADA variables. The experimentations conducted by using this model and SCADA data from three wind turbines in a wind farm have generated average prediction accuracies of about 97% for prediction horizons of up to 2 days ahead. The developed model is shown to maintain at least 95% prediction accuracy for long prediction horizons of up to 22 days ahead. Furthermore, for another wind farm SCADA dataset, it is shown that the developed PCTG model achieves over 99% icing prediction accuracy 10 days ahead.
more »
« less
Prediction of Icing on Wind Turbines Based on SCADA Data via Temporal Convolutional Network
Icing on the blades of wind turbines during winter seasons causes a reduction in power and revenue losses. The prediction of icing before it occurs has the potential to enable mitigating actions to reduce ice accumulation. This paper presents a framework for the prediction of icing on wind turbines based on Supervisory Control and Data Acquisition (SCADA) data without requiring the installation of any additional icing sensors on the turbines. A Temporal Convolutional Network is considered as the model to predict icing from the SCADA data time series. All aspects of the icing prediction framework are described, including the necessary data preprocessing, the labeling of SCADA data for icing conditions, the selection of informative icing features or variables in SCADA data, and the design of a Temporal Convolutional Network as the prediction model. Two performance metrics to evaluate the prediction outcome are presented. Using SCADA data from an actual wind turbine, the model achieves an average prediction accuracy of 77.6% for future times of up to 48 h.
more »
« less
- Award ID(s):
- 1916776
- PAR ID:
- 10561780
- Publisher / Repository:
- MDPI
- Date Published:
- Journal Name:
- Energies
- Volume:
- 17
- Issue:
- 9
- ISSN:
- 1996-1073
- Page Range / eLocation ID:
- 2175
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
In cold climates, ice formation on wind turbines causes power reduction produced by a wind farm. This paper introduces a framework to predict icing at the farm level based on our recently developed Temporal Convolutional Network prediction model for a single turbine using SCADA data.First, a cross-validation study is carried out to evaluate the extent predictors trained on a single turbine of a wind farm can be used to predict icing on the other turbines of a wind farm. This fusion approach combines multiple turbines, thereby providing predictions at the wind farm level. This study shows that such a fusion approach improves prediction accuracy and decreases fluctuations across different prediction horizons when compared with single-turbine prediction. Two approaches are considered to conduct farm-level icing prediction: decision fusion and feature fusion. In decision fusion, icing prediction decisions from individual turbines are combined in a majority voting manner. In feature fusion, features of individual turbines are averaged first before conducting prediction. The results obtained indicate that both the decision fusion and feature fusion approaches generate farm-level icing prediction accuracies that are 7% higher with lower standard deviations or fluctuations across different prediction horizons when compared with predictions for a single turbine.more » « less
-
This study proposes an intelligent techno-economic assessment framework for wind energy end users, using a novel dual-input convolutional bidirectional long short-term memory (Dual-ConvBiLSTM) architecture to predict dynamic levelized cost of energy (LCOE). The proposed architecture separates weight matrices for wind supervisory control and data acquisition (SCADA) data and financial data. This allows the model to integrate both data streams at every time step through a custom dual-input cell. This approach is compared with five baseline architectures: Recurrent Neural Network (RNN), LSTM, BiLSTM, ConvLSTM, and ConvBiLSTM, which process data through separate parallel branches and concatenate outputs before final prediction. The Dual-ConvBiLSTM achieves an LCOE estimate of $4.0391 cents/kWh, closest to the actual value of $4.0450 cents/kWh, with a root mean squared error reduction of 51.8% compared to RNN, 47.0% to LSTM, 40.0% to BiLSTM, 36.7% to ConvLSTM, and 34.4% to ConvBiLSTM, demonstrating superior capability in capturing complex interactions between SCADA data and financial parameters. This intelligent framework potentially enhances economic assessment and enables end users to accelerate renewable energy deployment through more reliable financial prediction.more » « less
-
A field campaign was carried out to investigate ice accretion features on large turbine blades (50 m in length) and to assess power output losses of utility-scale wind turbines induced by ice accretion. After a 30-h icing incident, a high-resolution digital camera carried by an unmanned aircraft system was used to capture photographs of iced turbine blades. Based on the obtained pictures of the frozen blades, the ice layer thickness accreted along the blades’ leading edges was determined quantitatively. While ice was found to accumulate over whole blade spans, outboard blades had more ice structures, with ice layers reaching up to 0.3 m thick toward the blade tips. With the turbine operating data provided by the turbines’ supervisory control and data acquisition systems, icing-induced power output losses were investigated systematically. Despite the high wind, frozen turbines were discovered to rotate substantially slower and even shut down from time to time, resulting in up to 80% of icing-induced turbine power losses during the icing event. The research presented here is a comprehensive field campaign to characterize ice accretion features on full-scaled turbine blades and systematically analyze detrimental impacts of ice accumulation on the power generation of utility-scale wind turbines. The research findings are very useful in bridging the gaps between fundamental icing physics research carried out in highly idealized laboratory settings and the realistic icing phenomena observed on utility-scale wind turbines operating in harsh natural icing conditions.more » « less
-
This paper presents regression and classification methods to estimate wind direction in a wind farm from operational data. Two neural network models are trained using supervised learning. The data are generated using high-fidelity large eddy simulations (LES) of a virtual wind farm with 16 turbines, which are representative of the data available in actual SCADA systems. The simulations include the high-fidelity flow physics and turbine dynamics. The LES data used for training and testing the neural network models are the rotor angular speeds of each turbine. Our neural network models use sixteen angular speeds as inputs to produce an estimate of the wind direction at each point in time. Training and testing of the neural network models are done for seven discrete wind directions, which span the most interesting cases due to symmetry of the wind farm layout. The results of this paper are indicative of the potential that existing neural network models have to obtain estimates of wind direction in real time.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.3390/en17092175
