Wind Speed Series Research Articles

A hybrid approach based on the Gaussian process with t-observation model for short-term wind speed forecasts

Accurate wind speed prediction is a significant factor in improving and optimizing wind power production. Particularly, reliable short-term wind speed forecasting contributes to the real-time optimization of wind farm operation. However, this short-term forecasting task remains challenging due to the strong stochastic nature and dynamic uncertainty of wind speed. This paper proposes a hybrid model that consists of the Empirical Wavelet Transform (EWT), Expectation Propagation (EP) algorithm and Gaussian process regression with the Student-t Observation Model (GPR-t) for short-term wind speed forecasting. The proposed approach firstly extracts meaningful information from a short-term wind speed series and subsequently models the inherent uncertainty and the dynamic features of the wind speed time-series. Additionally, the wind speed series presents a time-varying characteristic. Thus, this study adopts a moving window approach in the prediction processes, thereby permitting the proposed model to respond quickly to the dynamic characteristic of wind speed. To examine the forecasting performance of the suggested hybrid model, the validation of the proposed model is performed against several other existing models with half-hour and hourly wind speed data obtained from a windmill farm located in northwestern China. The computational results demonstrate that the proposed hybrid approach generates satisfactory point predictive accuracy and interval forecasting performance.

Homogenization of Tianjin monthly near-surface wind speed using RHtestsV4 for 1951\u20132014

Historical Chinese surface meteorological records provided by the special fund for basic meteorological data from the National Meteorological Information Center (NMIC) were processed to produce accurate wind speed data. Monthly 2-min near-surface wind speeds from 13 observation stations in Tianjin covering 1951–2014 were homogenized using RHtestV4 combined with their metadata. Results indicate that 10 stations had significant breakpoints—77% of the Tianjin stations—suggesting that inhomogeneity was common in the Tianjin wind speed series. Instrument change accounted for most changes, based on the metadata, including changes in type and height, especially for the instrument type. Average positive quantile matching (QM) adjustments were more than negative adjustments at 10 stations; positive biases with a probability density of 0.2 or more were mainly concentrates in the range 0.2 m s−1 to 1.2 m s−1, while the corresponding negative biases were mainly in the range −0.1 to −1.2 m s−1. Here, changes in variances and trends in the monthly mean surface wind speed series at 10 stations before and after adjustment were compared. Climate characteristics of wind speed in Tianjin were more reasonably reflected by the adjusted data; inhomogeneity in wind speed series was largely corrected. Moreover, error analysis reveals that there was a high consistency between the two datasets here and that from the NMIC, with the latter as the reference. The adjusted monthly near-surface wind speed series shows a certain reliability for the period 1951–2014 in Tianjin.

Wind power prediction using hybrid autoregressive fractionally integrated moving average and least square support vector machine

Precise prediction of wind power can not only conduct wind turbine's operation, but also reduce the impact on power systems when wind energy is injected into the grid. A hybrid autoregressive fractionally integrated moving average and least square support vector machine model is proposed to forecast short-term wind power. The proposed hybrid model takes advantage of the respective superiority of autoregressive fractionally integrated moving average and least square support vector machine. First, the autocorrelation function analysis is used to detect the long memory characteristics of wind power series, and the autoregressive fractionally integrated moving average model is applied to forecast linear component of wind power series. Then the least square support vector machine model is established to forecast nonlinear component of wind power series by making use of wind speed, wind direction and residual error series of the autoregressive fractionally integrated moving average model. Finally, the prediction of wind power is obtained by integrating the prediction results of autoregressive fractionally integrated moving average and least square support vector machine. Compared with other models, the results of two examples demonstrate that the proposed hybrid model has higher accuracy of wind power prediction in terms of three performance indicators.

A Hybrid Method for Short-Term Wind Speed Forecasting

The accuracy of short-term wind speed prediction is very important for wind power generation. In this paper, a hybrid method combining ensemble empirical mode decomposition (EEMD), adaptive neural network based fuzzy inference system (ANFIS) and seasonal auto-regression integrated moving average (SARIMA) is presented for short-term wind speed forecasting. The original wind speed series is decomposed into both periodic and nonlinear series. Then, the ANFIS model is used to catch the nonlinear series and the SARIMA model is applied for the periodic series. Numerical testing results based on two wind sites in South Dakota show the efficiency of this hybrid method.

Combination of Satellite Images and Numerical Model for the State Followed the Coast of the Bay of Bejaia-Jijel

Bay of Bejaia-Jijel extends over a length of 100 km with the presence of the ports, the beaches of rivers that
 discharge of the bay. This area characterized by strong economic activity, namely tourism and fisheries.
 However, severe erosion, high hydrodynamic activity and significant silting of ports affect this bay. Hence, the
 interest of this study, which tries to explain these phenomena, based on multisource data with multitasking
 models. First, we developed an algorithm that can convert satellite images in coastline vector. This technique
 applied to three images LANDSAT TM and OLI sensed in 1987, 2011; 2015. The multi-temporal monitoring
 coastlines show that the region suffered severe erosion. This erosion is 4.6 m / year for the period of 1987/2011
 and 1.5 m / year for the period 2011/2015. To explain this phenomenon we interested to do a study of
 hydrodynamics using the SWAN software. We used a long time series of wind speed and direction to discern
 extreme cases in the region. For maximum wind, the significant wave height recorded very high values and a
 very active orbital current with a speed that exceeds 0.7 m / s. Numerical modelling has allowed us to explain
 the erosion but does not explain the speed difference coastline. To find the explanation of erosion speed
 difference between the two periods we consulted the administrative archive of the region. In the archive, the
 number of authorized hourglasses is, 12 in the 90s, but 36 quarries operate haphazardly. In this period the
 balance of suspended matter is completely unbalanced which promotes erosion. After 2008, these quarries were
 closed therefore the balance of the suspended material has improved and consequently coastal erosion has
 decreased.

A novel probabilistic wind speed forecasting based on combination of the adaptive ensemble of on-line sequential ORELM (Outlier Robust Extreme Learning Machine) and TVMCF (time-varying mixture copula function)

The uncertainty and nonstationary of wind speed have compelled the power system operators and researchers to search for more accurate and reliable techniques to implement wind speed forecasting (WSF). In allusion to this phenomenon, this paper presents an adaptive ensemble of model for the probabilistic WSF, which is based on combination of the adaptive ensemble of on-line sequential ORELM (OS-ORELM) and the time-varying mixture copula function (TVMCF) to perform multi-step WSF. An OS-ORELM with forgetting mechanism based on Cook’s distance (λCDFF OS-ORELM) serves as a basic WSF model and an on-line ensemble using ordered aggregation (OEOA) technique is employed to improve the prediction performance. In the data pre-processing period, the Bernaola Galvan algorithm (BGA) is employed to partition the raw wind speed series into segments and the adaptive variational mode decomposition (AVMD) is used to decompose each segment into sub-series with different sub-band. Each transformed sub-series is well-modeled with the application of λCDFF OS-ORELM-OEOA, which is optimized by modified crisscross optimization algorithm (CSO). Eventual forecast results are obtained through aggregate calculation. Then the probabilistic prediction intervals (PIs) of wind speed are established in a TVMCF framework by modeling the conditional forecasting error. Case studies using the real wind speed data from the National Renewable Energy Laboratory (NREL) demonstrate that the proposed model can not only improves point forecasts compared with benchmark methods, but also constructs higher quality of probabilistic PIs.

Dynamics of solar wind speed: Cycle 23

A statistical signal processing approach has been made to study the dynamics of the speed of steady flow of hot plasma from the corona of sun known as solar wind generated in Solar Cycle 23. A long time series of solar wind speed of length 2492days from 1st Jan, 1997 to 28th October, 2003 collected from Coordinated Heliospheric Observations (COHO) data base at NASA’s National Space Science Data Center (NSSDC) is investigated for this purpose. Detection of nonlinearity and chaos in dynamics of solar wind speed is the prime objective of this work. In the present analysis delay vector variance (DVV) method is used to detect the existence of nonlinearity within the dynamics of solar wind speed. To explore the signature of the chaos in it multiple statistical methodologies like ‘0–1’ test, the correlation dimension analysis, computation of Information Entropy of the time series and Largest Lyapunov Exponent method have been applied. It has been observed that though the coronal plasma i.e. solar wind flow rate has a nonlinear dynamics but without any chaos. The absence of chaos indicates a probable regular behaviour of the series. The unit magnitude of the Correlation dimension indicates the presence of the deterministic component of the series. Embedding Dimension obtained argues that the deterministic component has dimension of six. The nearly zero value of the Lyapunov exponent claims that the system is conservative and exhibits Lyapunov stability. These revelations establish that not only the solar wind speed alone but the solar wind-magnetosphere coupling is also contributing towards the complexity of the magnetospheric plasma dynamics.

Modeling and forecasting energy flow between national power grid and a solar–wind–pumped-hydroelectricity (PV–WT–PSH) energy source

The structure of modern energy systems has evolved based on the assumption that it is the demand side which is variable, whilst the supply side must adjust to forecasted (or unforecasted) changes. But the increasing role of variable renewable energy sources (VRES) has led to a situation in which the supply side is also becoming more and more unpredictable. To date, various approaches have been proposed to overcome this impediment. This paper aims to combine mixed integer modeling with an Artificial Neural Networks (ANN) forecasting method in order to predict the volume of energy flow between a local balancing area which is using PV–WT–PSH and the national power system (NPS). Calculations has been performed based on the hourly time series of wind speed, irradiation and energy demand. The results indicate that both probabilistic and ANN models generate comparably accurate forecasts; however, the opportunity for improvement in the former appears to be significantly greater. The mean prediction error (for a one hour ahead forecasts) for the best model was 0.15MWh, which amounts to less than 0.2% of a mean hourly energy demand of the considered energy consumer. The proposed approach has huge potential to reduce the impact of VRES on the NPS operation as well as can be used to facilitate the process of their integration and increase their share in covering energy demand.

Climate trends on the extreme winds in Brazil

The main source of electricity in Brazil is from hydro, which has about 65.2% share of the country’s electric energy matrix. However, over the last decade the wind energy increased from 19 MW to 2.2 GW. Since wind is an intermittent energy source, heavily determined by the weather and climatic conditions, and important effects on wind power generation can be expected in the mid and long term, in particular related to the impacts of extreme winds. The IPCC AR5 (Intergovernmental Panel on Climate Change) indicates changes in wind speed at the surface in some regions of the world, and increased wind strength in mid-latitude regions. This study scrutinizes future scenarios of extreme winds in Brazil by applying trend analysis techniques on a 50-year historical series of observational wind speed and meteorological parameters at 10 m height in Brazil. Embracing techniques of cluster analysis it was possible to characterize six main regions with macro climatic similarities. To assess the goodness fit distribution, we designate two stations per homogenous region, taking as criteria the stations with better performance in the qualification process to determine the wind distribution pattern in each region applying the Kolmogorov-Smirnov test (KS) and the lowest standard error (SE). After evaluating the frequency distribution of wind speed, the best fit result for the frequency distribution of maximum wind speed is the Gumbel model. The analysis of climatic trends performed by Mann-Kendall test revealed that in minimum wind speed series is not conclusive because it shows disparate results between homogeneous regions. On the other hand, the analysis of climatic trends of maximum wind speed presents 100% positive trends in Group#1, an equal number of stations with not significant trends and positive trends for Group#2, 36.8% more stations with positive trends than negative trends for Group#3 and 20% of stations with more negative trends than stations with positive trends for Group#4. This way, based in these results, is possible assert that there are an increase in the maximum extreme wind in Brazil, mainly in mid-latitudes.

Impact of climate change on dynamic behavior of offshore wind turbine

ABSTRACTGlobal warming is expected to change the wind and wave patterns at a significant level, which may lead to conditions outside current design criteria of monopile supported offshore wind turbine (OWT). This study examines the impact of climate change on the dynamic behavior and future safety of an OWT founded in clay incorporating dynamic soil–structure interaction. A statistical downscaling model is used to generate the time series of future wind speed and wave height at local level. The responses and fatigue life of OWT are estimated for present and future periods and extent of change in design is investigated at offshore location along the west coast of India. Wind speed, wave height, and wave period data are collected from the buoy deployed by Indian National Centre for Ocean Information Services and the future climate for the next 30 years is simulated using the general circulation model corresponding to Special Report on Emission Scenarios A1B scenario. The OWT is modeled as Euler–Bernoulli beam and soil–structure interaction is incorporated using nonlinear p-y springs. This study shows that changes in design of OWT are needed due to increased responses owing to the effect of climate change. Fatigue life is found to be decreased because of climate change.

Short-Term Wind Speed Prediction Using EEMD-LSSVM Model

Hybrid Ensemble Empirical Mode Decomposition (EEMD) and Least Square Support Vector Machine (LSSVM) is proposed to improve short-term wind speed forecasting precision. The EEMD is firstly utilized to decompose the original wind speed time series into a set of subseries. Then the LSSVM models are established to forecast these subseries. Partial autocorrelation function is adopted to analyze the inner relationships between the historical wind speed series in order to determine input variables of LSSVM models for prediction of every subseries. Finally, the superposition principle is employed to sum the predicted values of every subseries as the final wind speed prediction. The performance of hybrid model is evaluated based on six metrics. Compared with LSSVM, Back Propagation Neural Networks (BP), Auto-Regressive Integrated Moving Average (ARIMA), combination of Empirical Mode Decomposition (EMD) with LSSVM, and hybrid EEMD with ARIMA models, the wind speed forecasting results show that the proposed hybrid model outperforms these models in terms of six metrics. Furthermore, the scatter diagrams of predicted versus actual wind speed and histograms of prediction errors are presented to verify the superiority of the hybrid model in short-term wind speed prediction.

Wind speed parameters sensitivity analysis based on fractals and neuro-fuzzy selection technique

Fluctuation of wind speed affects wind energy systems since the potential wind power is proportional the cube of wind speed. Hence precise prediction of wind speed is very important to improve the performances of the systems. Due to unstable behavior of the wind speed above different terrains, in this study fractal characteristics of the wind speed series were analyzed. According to the self-similarity characteristic and the scale invariance, the fractal extrapolate interpolation prediction can be performed by extending the fractal characteristic from internal interval to external interval. Afterward neuro-fuzzy technique was applied to the fractal data because of high nonlinearity of the data. The neuro-fuzzy approach was used to detect the most important variables which affect the wind speed according to the fractal dimensions. The main goal was to investigate the influence of terrain roughness length and different heights of the wind speed on the wind speed prediction.

Two combined forecasting models based on singular spectrum analysis and intelligent optimized algorithm for short-term wind speed

Short-term wind speed forecasting has become an important technology to utilize sustainable energy, reduce the impact of wind power grid and improve the control of wind turbines. Current forecasting models based on individual algorithm and hybrid optimization algorithm could be applied to a variety of wind speed forecast. However, these algorithms ignored the limitation of optimizing parameter and the use of recent data, which may result in forecasting poor accuracy. In this paper, integrated approaches, combining singular spectrum analysis technique, cuckoo search algorithm (CSA) and harmony search algorithm (HSA), and back propagation neural network (BPNN), were introduced for conducting short-term wind speed forecasting. Firstly, the SSA technique is used for identifying and extracting instable components from raw wind speed signals. Then, the parameters of BPNN are employed to be optimized by the CSA/HSA, which improve precision of forecast results. Finally, the BPNN is utilized to deal with the wind speed series. The proposed hybrid model is conducted by the wind speed at three stations located in Penglai, China. Our experiment reveals that the proposed hybrid model can generate a more accurate, reliable and robust result and SSA technique is found to be superior technique to preprocess wind speed series. Furthermore, we compare to the forecast results with SSA and non SSA denoising procedure and found that hybrid model with SSA technique outperforms other prediction model.

Characteristics and forecast of short-term wind speed series in the Donghai Bridge wind farm

As a sort of worldwide renewable energy, wind energy can reduce environmental pollution and relieve the energy shortage. In order to reduce the adverse effect with the integration of wind energy into electricity grids and the operating cost of power supply system, it is becoming increasingly significant to acquire accurate short-term wind speed forecasts. In this paper, based on the analysis of the measured wind speed data in the Donghai Bridge wind farm, we suggest that the short-term wind speed series has volatility clustering effect and asymmetric effect, and the volatility feed-back effect is not significant. And then the possible causes for this phenomenon are elucidated in detail from the viewpoint of physics. In addition, in order to select the forecast model which is appropriate to the Donghai Bridge wind farm, we use these indexes of error: mean absolute error (MAE), mean absolute percentage error (MAPE) and root mean square error (RMSE) as the criterion, and compare the prediction accuracy of the five models by quantitative analysis. The results show that ARMA-EGARCH model and ARMA-EGARCH-M model are very close to each other in both single-step and multi-step forecasting, and they are superior to other models. What’s more, with the increase of the number of advance forecasting steps, error growth rate of these two models is low.

Spectral structure of 5 year time series of horizontal wind speed at the Boulder Atmospheric Observatory

AbstractWe investigate the spectral structures of 5 year, 1 min time series of horizontal wind speeds at 100 and 10 m heights at the Boulder Atmospheric Observatory tower located in the eastern slope of the Rocky Mountains, USA. In the full‐scale spectra, the diurnal spectral peak, which is usually insignificant at a coastal or offshore site, is the most significant at both heights. The spectrum is enhanced on the low‐frequency side of the diurnal peak during winter, but on the high‐frequency side during summer, which indicates frequent synoptic weather events during winter supplanted by mesoscale events during summer. In terms of the spectral density in the spectral gap of Van der Hoven (1957), separating boundary layer turbulence from the synoptic‐scale fluctuations, at a frequency between 10−4 and 10−3 Hz, we rank the daily time series at 100 m height and sample the summer top and winter bottom 10 percentile cases. The winter cases of the reduced spectral density in the gap region present the f−3 spectrum (f is frequency) and negatively skewed velocity increment distributions, which are the signatures of enstrophy (the integral of squared vorticity) cascade of turbulent two‐dimensional (2‐D) flows. In contrast, the summer cases of the enhanced spectral density present the f−5/3 spectrum and positively skewed velocity increment distributions, which are the signatures of upscale energy cascade of 2‐D flows. In these mesoscale events that fill up the gap, the turbulence intensity‐wind speed relationship is very sensitive to the choice of the averaging period.

Homogenization and Assessment of Observed Near-Surface Wind Speed Trends across Sweden, 1956–2013

Abstract Multidecadal variability of observed near-surface wind speed from 24 stations across Sweden has been analyzed for 1956–2013, with a focus on 1979–2008 (incorporating an additional 9 stations) for comparison with previous studies. Wind speed data have been subjected to a robust data processing protocol, consisting of quality control, reconstruction, and homogenization, by using geostrophic wind speed series as reference. The homogenized dataset displays a significant (at p < 0.05) downward trend for 1956–2013 (−0.06 m s−1 decade−1) and an even larger decreasing trend for 1979–2008 (−0.14 m s−1 decade−1). However, differences have been observed seasonally, with significant decreasing values in spring, summer, and autumn and a small downward trend in winter for 1956–2013. Most interestingly, a nonsignificant wind speed increase has been detected in winter for 1979–2008, which contrasts with the marked “stilling” reported for this season in much of midlatitude regions. The decreasing rate in wind speed is larger for coastal stations and in the southern part of Sweden. Decreasing trends were found at 91.7% of the stations during summer, whereas 58.3% of the stations displayed decreasing trends in winter. On the contrary, increasing trends occurred in 41.7% of the stations for winter and in only 8.3% for summer. The possible impact of the North Atlantic Oscillation (NAO) index has also been investigated, showing evidence that the small increasing trend in winter for 1979–2008 is hypothetically associated with the positive tendency of the NAO index during the last decades. These results reveal the influence of large-scale atmospheric circulation on wind speed variability across Sweden.

A Gaussian process regression based hybrid approach for short-term wind speed prediction

This paper proposes a hybrid model based on autoregressive (AR) model and Gaussian process regression (GPR) for probabilistic wind speed forecasting. In the proposed approach, the AR model is employed to capture the overall structure from wind speed series, and the GPR is adopted to extract the local structure. Additionally, automatic relevance determination (ARD) is used to take into account the relative importance of different inputs, and different types of covariance functions are combined to capture the characteristics of the data. The proposed hybrid model is compared with the persistence model, artificial neural network (ANN), and support vector machine (SVM) for one-step ahead forecasting, using wind speed data collected from three wind farms in China. The forecasting results indicate that the proposed method can not only improve point forecasts compared with other methods, but also generate satisfactory prediction intervals.

The climate-archive dune: Sedimentary record of annual wind intensity

Abstract Active dunes contain an unread high-resolution archive of past wind strength. Variations in the grain-size spectrum of an active dune through time are compared with historical time series of wind speed. Annually averaged, sediment-derived wind intensity and instrumental records of wind speed show a correlation as high as 0.75. The potential of eolian dunes to gain long-term data series of wind intensity in areas and for time periods lacking an instrumental record is demonstrated by reconstructing 20th century wind-intensity variations in the southern North Sea area. The approach can be used in both recent and fossil dune systems. Potential applications include the validation of climate models, the reconstruction of supraregional wind systems, and the monitoring of future shifts in the climate system.

Deep belief network based deterministic and probabilistic wind speed forecasting approach

With the rapid growth of wind power penetration into modern power grids, wind speed forecasting (WSF) plays an increasingly significant role in the planning and operation of electric power and energy systems. However, the wind speed time series always exhibits nonlinear and non-stationary features, making it very difficult to be predicted accurately. Recognizing this challenge, a novel deep learning based approach is proposed for deterministic and probabilistic WSF. The approach is a hybrid of wavelet transform (WT), deep belief network (DBN) and spine quantile regression (QR). WT is employed to decompose raw wind speed data into different frequency series with better behaviors. The nonlinear features and invariant structures of each frequency are completely extracted by layer-wise pre-training based DBN. Then, the uncertainties in wind speed are statistically synthesized via the QR method. Case studies using real wind farm data from China and Australia have been presented. The comparative results demonstrate that the high-level nonlinear and non-stationary feature in the wind speed series can be learned better, and competitive performance can thus be obtained. It is therefore convinced that the proposed method has a high potential for practical applications in electric power and energy systems.

A Hybrid Multi-Step Rolling Forecasting Model Based on SSA and Simulated Annealing—Adaptive Particle Swarm Optimization for Wind Speed

With the limitations of conventional energy becoming increasing distinct, wind energy is emerging as a promising renewable energy source that plays a critical role in the modern electric and economic fields. However, how to select optimization algorithms to forecast wind speed series and improve prediction performance is still a highly challenging problem. Traditional single algorithms are widely utilized to select and optimize parameters of neural network algorithms, but these algorithms usually ignore the significance of parameter optimization, precise searching, and the application of accurate data, which results in poor forecasting performance. With the aim of overcoming the weaknesses of individual algorithms, a novel hybrid algorithm was created, which can not only easily obtain the real and effective wind speed series by using singular spectrum analysis, but also possesses stronger adaptive search and optimization capabilities than the other algorithms: it is faster, has fewer parameters, and is less expensive. For the purpose of estimating the forecasting ability of the proposed combined model, 10-min wind speed series from three wind farms in Shandong Province, eastern China, are employed as a case study. The experimental results were considerably more accurately predicted by the presented algorithm than the comparison algorithms.