Wind Speed Time Series Data Research Articles

The Use of Odd and Even Class Wind Speed Time Series of Distribution Histogram to Estimate Weibull Parameters

In this article, we introduced a new approach based on graphical method (GPM), maximum likelihood method (MLM), energy pattern factor method (EPFM), empirical method of Justus (EMJ), empirical method of Lysen (EML) and moment method (MOM) using the even or odd classes of wind speed series distribution histogram with 1 m/s as bin size to estimate the Weibull parameters. This new approach is compared on the basis of the resulting mean wind speed and its standard deviation using seven reliable statistical indicators (RPE, RMSE, MAPE, MABE, R2, RRMSE and IA). The results indicate that this new approach is adequate to estimate Weibull parameters and can outperform GPM, MLM, EPF, EMJ, EML and MOM which uses all wind speed time series data collected for one period. The study has also found a linear relationship between the Weibull parameters K and C estimated by MLM, EPFM, EMJ, EML and MOM using odd or even class wind speed time series and those obtained by applying these methods to all class (both even and odd bins) wind speed time series. Another interesting feature of this approach is the data size reduction which eventually leads to a reduced processing time.Article History: Received February 16th 2018; Received in revised form May 5th 2018; Accepted May 27th 2018; Available onlineHow to Cite This Article: Salami, A.A., Ajavon, A.S.A., Kodjo, M.K. , Ouedraogo, S. and Bédja, K. (2018) The Use of Odd and Even Class Wind Speed Time Series of Distribution Histogram to Estimate Weibull Parameters. Int. Journal of Renewable Energy Development 7(2), 139-150.https://doi.org/10.14710/ijred.7.2.139-150

Computational Fluid Dynamics Approach to Predict the Actual Wind Speed over Complex Terrain

This paper proposes a procedure for predicting the actual wind speed for flow over complex terrain with CFD. It converts a time-series of wind speed data acquired from field observations into a time-series data of actual scalar wind speed by using non-dimensional wind speed parameters, which are determined beforehand with the use of CFD output. The accuracy and reproducibility of the prediction procedure were investigated by simulating the flow with CFD with the use of high spatial resolution (5 m) surface elevation data for the Noma Wind Park in Kagoshima Prefecture, Japan. The errors of the predicted average monthly wind speeds relative to the observed values were less than approximately 20%.

Time-series prediction of wind speed using machine learning algorithms: A case study Osorio wind farm, Brazil

Machine learning algorithms (MLAs) are applied to predict wind speed data for Osorio wind farm that is located in the south of Brazil, near the Osorio city. Forecasting wind speed in wind farm regions is valuable in order to obtain an intelligent management of the generated power and to promote the utilization of wind energy in grid-connected and isolated power systems. In this study, multilayer feed-forward neural network (MLFFNN), support vector regression (SVR), fuzzy inference system (FIS), adaptive neuro-fuzzy inference system (ANFIS), group method of data handling (GMDH) type neural network, ANFIS optimized with particle swarm optimization algorithm (ANFIS-PSO) and ANFIS optimized with genetic algorithm (ANFIS-GA) are developed to predict the time-series wind speed data. The Time-series prediction describes a model that predicts the future values of the system only using the past values. Past data is entered as input and future data to be used for represents MLA output. The developed models are examined on 5-min, 10-min, 15-min and 30-min intervals of wind speed data. The results demonstrated that the GMDH model for all time intervals can successfully predict the time-series wind speed data with a high accuracy. Also, the combination of ANFIS models with PSO and GA algorithms can increase the prediction accuracy of the ANFIS model for all time intervals.

Short-term wind speed and wind power prediction using hybrid empirical mode decomposition and kernel ridge regression

This paper presents an efficient non-iterative hybrid empirical mode decomposition (EMD) and kernel ridge regression (KRR) for significantly accurate short-term wind speed and wind power prediction. The original non-linear and non-stationary wind speed and wind power time series data are decomposed using EMD to isolate the mutual effects between different components. The proposed EMD-KRR model is tested for predicting wind speed and wind power time series data over a time horizon spanning intervals of 10 min, 30 min, 1 h and 3 h ahead, respectively. Further the performance of EMD-KRR prediction model is compared with two other widely used non-iterative prediction models like EMD based Random vector functional link network (RVFL), and EMD based Extreme Learning Machine (ELM). Also an iterative Mutated Firefly Algorithm with Global Optima concept (MFAGO) optimized RVFL network is used for comparison to prove the advantages of non-iterative models over the iterative ones. The performance metrics of the proposed EMD-KRR and EMD-RVFL confirm the effectiveness and precision in producing an accurate forecast of both wind speed and wind power in comparison to all other prediction models using the wind power data from three real world wind farms. Further a fast reduced version of the EMD-KRR is presented in the paper to reduce the computational overhead substantially using randomly selected support vectors from the data set while resulting in a reasonably accurate forecast.

Autoencoder for wind power prediction

Successful integration of renewable energy sources like wind power into smart grids largely depends on accurate prediction of power from these intermittent sources. Production of wind power cannot be controlled as the wind speed can vary based on weather conditions. Accurate prediction of wind power can assist smart grid that intelligently decides on the usage of alternative power sources based on demand forecast. Time series wind speed data are normally used for wind power prediction. In this paper, we have investigated the usage of a set of secondary features obtained using deep learning for wind power prediction. Deep learning is a special form on neural network that is capable of capturing the structural properties of time series data in terms of a set of numeric features. More precisely, we have designed a two-stage autoencoder (a particular type of deep learning) and incorporated the structural features into a prediction framework. Using the structural features, we have achieved as high as 12.63% better prediction accuracy than traditionally used statistical features.

The Relationships between Wind Speed and Temperature Time Series in Bangkok, Thailand.

In this research we investigate the relationships between wind speed and temperature time series data in Bangkok, Thailand, from the time interval of January 2009 to December 2011 using wavelet transform (WT), cross wavelet transform (XWT) and wavelet coherence (WTC). The results from all three wavelet analysis show the strong periodicity around period 1 day (hourly data) and period band 256-450 days (daily data) variations that are exhibited in both wind speed and temperature data across the entire power spectrum from 2009 to 2011. These two oscillations are connected with the natural day time effects and the annual natural season cycle. Although the daily periodic for the temperature is appeared nearly uniform all year but it is not the case for wind speed. In 2009 this wind speed oscillations appear only from mid-February to mid-April in summer and from the fourth week of May to the third week of August in rainy season. XWT also detects strong high common power between the wind speed and temperature at a period band of 14-25 days in summer 2009, a period band of 4-8 days in summer 2009, July 2009, summer 2010 and summer 2011. WTC shows the coherence period band around 10-30 days appeared in summer and rainy season and 32-50 days in summer 2009 and rainy season in 2010. From these three wavelet analysis, the wind speed and temperature time series data show the strong correlation especially at 1 day and 256-450 days period band and also at several different scales. This studied will be helpful in predicting the wind speed and temperature for the future used.

Comparative analysis of Weibull parameters for wind data measured from met-mast and remote sensing techniques

The remote sensing techniques are gaining attention worldwide for the comprehensive assessment of wind resource in flat, complex, and mountainous terrain. This paper presents an attempt to increase the confidence on remote sensing technique to compute Weibull parameters at higher heights for assessment of wind energy resource. The measurement campaign was conducted at Kayathar, Tamil Nadu, India. The 10 min average time series wind speed data for the period of one month (September 2014) were recorded simultaneously at 80 m and 100 m using cup anemometer installed in the proximity of 120 m meteorological mast, Second Wind Triton SODAR (Sound Detection and Ranging) and Continuous-wave wind LIDAR (Light Detection and Ranging). The data obtained from this measurement were analyzed and Weibull parameters were evaluated using nine different methods. The applicability of nine methods for calculation of Weibull parameters is analyzed based on statistical analysis i.e. Root Mean Square Error Test, Coefficient of Determination, Mean Absolute Percentage Error, and Chi-square Test for both measuring techniques. This study also shows the performance of remote sensing technology by comparing with result obtained from met mast. The outcome of this study is expected to encourage the deployment of remote sensing techniques at Indian sites.

A generalized dynamic fuzzy neural network based on singular spectrum analysis optimized by brain storm optimization for short-term wind speed forecasting

Wind speed forecasting plays a pivotal role in power dispatching and normal operations of power grids. However, it is both a difficult and challenging problem to achieve high-precision forecasting for the wind speed because the original sequence includes many nonlinear stochastic signals. The current conventional forecasting methods are more suitable for capturing linear trends, and artificial neural networks easily fall into a local optimum. This paper proposes a model that combines a denoising method with a dynamic fuzzy neural network to address the problems above. Singular spectrum analysis optimized by brain storm optimization is applied to preprocess the original wind speed data to obtain a smoother sequence, and a generalized dynamic fuzzy neural network is utilized to perform the forecasting. With a smaller and simpler structure of the neural network, the model can effectively achieve a rapid learning rate and accurate forecasting. Three experimental results, which cover 10-min, 30-min and 60-min interval wind speed time series data, demonstrate that the model can both satisfactorily approximates the actual value and be used as an effective and simple tool for the planning of smart grids.

サポートベクトル回帰による多地点時系列データを用いた風速予測におけるk-NNとVARによる学習データ選択の検討

The output of the wind power is dependent on natural phenomena, the influence on the power system connection is of particular concern. Therefore, to predict the electricity generated by the wind power, efficient utilization of the wind energy can be expected. In this study, the wind speed prediction method of wind power with support vector regression SVR, k-NN and VAR model using the time-series wind speed data of multiple neighboring sites is proposed. The wind speed prediction performance with wind data of multiple sites including meteorological towers in Tohoku district is evaluated.

The impact of the North Atlantic Oscillation on electricity markets: A case study on Ireland

The North Atlantic Oscillation (NAO) is a large-scale atmospheric circulation pattern driving climate variability in north-western Europe. As the deployment of wind-powered generation expands on electricity networks across Europe, the impacts of the NAO on the electricity system will be amplified. This study assesses the impact of the NAO, via wind-power generation, on the electricity market considering thermal generation costs, wholesale electricity prices and wind generation subsidies. A Monte Carlo approach is used to model NAO phases and generate hourly wind speed time-series data, electricity demand and fuel input data. A least-cost unit commitment and economic dispatch model is used to simulate an island electricity system, modelled on the all-island Irish electricity system. The impact of the NAO obviously depends on the level of wind capacity within an electricity system. Our results indicate that on average a switch from negative to positive NAO phase can reduce thermal generation costs by up to 8%, reduce wholesale electricity prices by as much as €1.5/MWh, and increase wind power generators' revenue by 12%.

A Novel Combined Model Based on an Artificial Intelligence Algorithm—A Case Study on Wind Speed Forecasting in Penglai, China

Wind speed forecasting plays a key role in wind-related engineering studies and is important in the management of wind farms. Current forecasting models based on different optimization algorithms can be adapted to various wind speed time series data. However, these methodologies cannot aggregate different hybrid forecasting methods and take advantage of the component models. To avoid these limitations, we propose a novel combined forecasting model called SSA-PSO-DWCM, i.e., particle swarm optimization (PSO) determined weight coefficients model. This model consisted of three main steps: one is the decomposition of the original wind speed signals to discard the noise, the second is the parameter optimization of the forecasting method, and the last is the combination of different models in a nonlinear way. The proposed combined model is examined by forecasting the wind speed (10-min intervals) of wind turbine 5 located in the Penglai region of China. The simulations reveal that the proposed combined model demonstrates a more reliable forecast than the component forecasting engines and the traditional combined method, which is based on a linear method.

An Approach to Determine the Weibull Parameters and Wind Power Analysis of Saint Martin’s Island, Bangladesh

This paper explores wind speed distribution using Weibull probability distribution and Rayleigh distribution methods that are proven to provide accurate and efficient estimation of energy output in terms of wind energy conversion systems. Two parameters of Weibull (shape and scale parameters k and c respectively) and scale parameter of Rayleigh distribution have been determined based on hourly time-series wind speed data recorded from October 2014 to October 2015 at Saint Martin’s island, Bangladesh. This research has been carried out to examine three numerical methods namely Graphical Method (GM), Empirical Method (EM), Energy Pattern Factor method (EPF) to estimate Weibull parameters. Also, Rayleigh distribution method has been analyzed throughout the study. The results in the research revealed that the Graphical method followed by Empirical method and Energy Pattern Factor method were the most accurate and efficient way for determining the value of k and c to approximate wind speed distribution in terms of estimating power error. Rayleigh distribution gives the most power error in the research. Potential for wind energy development in Saint Martin’s island, Bangladesh as found from the data analysis has been explained in this paper.

Carbon dioxide (CO2) emissions from electricity: The influence of the North Atlantic Oscillation

The North Atlantic Oscillation (NAO) is a large-scale circulation pattern driving climate variability in north-western Europe. In recent years there has been an increasing deployment of wind-powered generation technology, i.e. wind farms, on electricity networks across Europe. As this deployment increases it is important to understand how climate variability will affect both wind-powered and non-renewable power generation. This study extends the literature by assessing the impact of NAO, via wind-power generation, on carbon dioxide emissions from the wider electricity system. A Monte Carlo approach is used to model NAO phases, generate hourly wind speed time-series data, electricity demand and fuel input data. A unit commitment, least-cost economic dispatch model is used to simulate an entire electricity system, modelled on the all-island Irish electricity system. Our results confirm that the NAO has a significant impact on monthly mean wind speeds, wind power output, and carbon dioxide emissions from the entire electricity system. The impact of NAO on emissions obviously depends on the level of wind penetration within an electricity system but our results indicate that emissions intensity within the Irish electricity system could vary by as much as 10% depending on the NAO phase within the next few years.

Investigating Potential of Wind Energy in Mahshahr, Iran

In this paper, measured time-series wind speed data in two sites in Mahshahr (Iran) were analyzed to find wind energy potential of this region. The objective was to evaluate the most important characteristics of wind energy in the studied sites using means of meteorological and Weibull distribution functions. The statistical approach was used to estimate mean wind speed, wind speed distribution function, mean wind power density and mean wind energy density of the sites at heights of 10 and 40 m. Annual mean wind speeds at heights of 10 and 40 m above the ground were 4.11 and 6.08 m/s, respectively. The results revealed wide variation in the values of monthly power and energy densities in the region. The lowest and highest values of mean wind energy densities at height of 10 m were 30.8 and 111.8 kWh/m2/month in July and January, respectively. The winds had mean energy density of 1210 kWh/m2/year at 40 m height. It can be concluded that Mahshahr region has a relatively good potential for harnessing winds.

Chaos and periodicity in solar wind speed: cycle 23

The solar wind speed time series data from 1st January, 1997 to 28th October, 2003 has been pre-processed using simple exponential smoothing, discrete wavelet transform for denoising to investigate the underneath dynamics of it. Recurrence plot and recurrence quantification analysis has revealed that the time series is non-stationary one with deterministic chaotic behavior. The Hilbert-Huang Transform has been used in search of the underlying periods of the data series. Present investigation has revealed the periods of 21 days, 32.5 days, 43.6 days, 148.86 days, 180.7 days, 355.5 days, 403.2 days, 413.6 days, 490.72 days, 729.6 days, 1086.76 days, 1599.4 days and 1892.6 days.

Hardware-in-the-Loop Wind Turbine Simulation Platform for a Laboratory Feeder Model

As part of a larger project to build a laboratory-based smart grid feeder, there was a need to incorporate renewable energy sources into the feeder model. The goal in this project was to create a hardware-in-the-loop (HIL) wind turbine simulation platform. Software is used to calculate a dynamic optimal power point from time-series wind speed data and to generate control signals for the platform’s hardware. The hardware consists of a dc motor providing mechanical power to a coupled generator. To validate the HIL simulation platform, time-series wind speed and output power data were used from an installed and operating utility-scale turbine. The steady-state performance of the HIL simulator correlates well with the provided manufacturer’s curve, but the dynamic performance of the system is more varied with periods of excellent correlation as well as poor. A richer data set from the utility containing more than wind speed and power output (such as wind direction and turbine yaw position) could provide substantial improvements to the dynamic performance.

Statistical Analysis of Wind Power Density Based on the Weibull and Rayleigh Models of Selected Site in Malaysia

The demand for electricity in Malaysia is growing in tandem with its Gross Domestic Product (GDP) growth. Malaysia is going to need even more energy as it strives to grow towards a high-income economy. Malaysia has taken steps to exploring the renewable energy (RE) including wind energy as an alternative source for generating electricity. In the present study, the wind energy potential of the site is statistically analyzed based on 1-year measured hourly time-series wind speed data. Wind data were obtained from the Malaysian Meteorological Department (MMD) weather stations at nine selected sites in Malaysia. The data were calculated by using the MATLAB programming to determine and generate the Weibull and Rayleigh distribution functions. Both Weibull and Rayleigh models are fitted and compared to the Field data probability distributions of year 2011. From the analysis, it was shown that the Weibull distribution is fitting the Field data better than the Rayleigh distribution for the whole year 2011. The wind power density of every site has been studied based on the Weibull and Rayleigh functions. The Weibull distribution shows a good approximation for estimation of wind power density in Malaysia.

Spatial and temporal patterns of global onshore wind speed distribution

Wind power, a renewable energy source, can play an important role in electrical energy generation. Information regarding wind energy potential is important both for energy related modeling and for decision-making in the policy community. While wind speed datasets with high spatial and temporal resolution are often ultimately used for detailed planning, simpler assumptions are often used in analysis work. An accurate representation of the wind speed frequency distribution is needed in order to properly characterize wind energy potential. Using a power density method, this study estimated global variation in wind parameters as fitted to a Weibull density function using NCEP/climate forecast system reanalysis (CFSR) data over land areas. The Weibull distribution performs well in fitting the time series wind speed data at most locations according to R2, root mean square error, and power density error. The wind speed frequency distribution, as represented by the Weibull k parameter, exhibits a large amount of spatial variation, a regionally varying amount of seasonal variation, and relatively low decadal variation. We also analyzed the potential error in wind power estimation when a commonly assumed Rayleigh distribution (Weibull k = 2) is used. We find that the assumption of the same Weibull parameter across large regions can result in non-negligible errors. While large-scale wind speed data are often presented in the form of mean wind speeds, these results highlight the need to also provide information on the wind speed frequency distribution.

Mining Markov chain transition matrix from wind speed time series data

Extracting important statistical patterns from wind speed time series at different time scales is of interest to wind energy industry in terms of wind turbine optimal control, wind energy dispatch/scheduling, wind energy project design and assessment, and so on. In this paper, a systematic way is presented to estimate the first order (one step) Markov chain transition matrix from wind speed time series by two steps. Wind speed time series data is used first to generate basic estimators of transition matrices (i.e. first order, second order, third order, etc.) based on counting techniques. Then an evolutionary algorithm (EA), specifically double-objective evolutionary strategy algorithm (ES), is proposed to search for the first order Markov chain transition matrix which can best match these basic estimators after transforming the first order transition matrix into its higher order counterparts. The evolutionary search for the first order transition matrix is guided by a predefined cost function which measures the difference between the basic estimators and the first order transition matrix, and its high order transformations. To deal with the potential high dimensional optimization problem (i.e. large transition matrices), an enhanced offspring generation procedure is proposed to help the ES algorithm converge efficiently and find better Pareto frontiers through generations. The proposed method is illustrated with wind speed time series data collected from individual 1.5 MW wind turbines at different time scales.

Influence of the level of fit of a density probability function to wind-speed data on the WECS mean power output estimation

Static methods which are based on statistical techniques to estimate the mean power output of a WECS (wind energy conversion system) have been widely employed in the scientific literature related to wind energy. In the static method which we use in this paper, for a given wind regime probability distribution function and a known WECS power curve, the mean power output of a WECS is obtained by resolving the integral, usually using numerical evaluation techniques, of the product of these two functions. In this paper an analysis is made of the influence of the level of fit between an empirical probability density function of a sample of wind speeds and the probability density function of the adjusted theoretical model on the relative error ε made in the estimation of the mean annual power output of a WECS. The mean power output calculated through the use of a quasi-dynamic or chronological method, that is to say using time-series of wind speed data and the power versus wind speed characteristic of the wind turbine, serves as the reference. The suitability of the distributions is judged from the adjusted R 2 statistic ( R a 2 ) . Hourly mean wind speeds recorded at 16 weather stations located in the Canarian Archipelago, an extensive catalogue of wind-speed probability models and two wind turbines of 330 and 800 kW rated power are used in this paper. Among the general conclusions obtained, the following can be pointed out: (a) that the R a 2 statistic might be useful as an initial gross indicator of the relative error made in the mean annual power output estimation of a WECS when a probabilistic method is employed; (b) the relative errors tend to decrease, in accordance with a trend line defined by a second-order polynomial, as R a 2 increases.