AdaBoosting Neural Network for Short-Term Wind Speed Forecasting Based on Seasonal Characteristics Analysis and Lag Space Estimation

High accurary in wind speed forcasting remains hard to achieve due to wind’s random distribution nature and its seasonal characteristics.Randomness,intermittent and nonstationary usually cause the portion problem of the wind speed forecasting.Seasonal characteristics of wind speed means that its feature distribution is inconsistent.This typically results that the persistence of excitation for modeling can not be guaranteed,and may severely reduce the possibilities of high precise forecasting model.In this paper,we proposed two effective solutions to solve the problems caused by the randomness and seasonal characteristics of the wind speed.(1)Wavelet analysis is used to extract the robust components of time series and reduce the influence of randomness.(2)Based on the energy distribution about the extracted amplitude and associated frequency,seasonal characteristics of wind speed are analyzed based on self-similarity in periodogram under scales range generated by wavelet transformation.Thus,the original dataset is reasonably divided into subsest which can effectively reflect the seasonal distribution characteristics of wind speed.In addition,two strategies are given to optimal model structure and improve the forecasting accuracy:(1)The forecasting model’s lag space is approximately estimated by the Lipschitz quotient to improve the generality ability of the feedforward neural network.(2)The forecasting accuracy and model robustness are further improved by the wavelet decomposition combined with AdaBoosting neural network.Finally,experimental evaluation based on the dataset from National Renewable Energy Laboratory(NREL)is given to demonstrate the performance of the proposed approach.

Evaluation on monthly sea surface wind speed of four reanalysis data sets over the China seas after 1988

This study investigates the long-term changes of monthly sea surface wind speeds over the China seas from 1988 to 2015. The 10-meter wind speeds products from four major global reanalysis datasets with high resolution are used: Cross-Calibrated Multi-Platform data set(CCMP), NCEP climate forecast system reanalysis data set(CFSR),ERA-interim reanalysis data set(ERA-int) and Japanese 55-year reanalysis data set(JRA55). The monthly sea surface wind speeds of four major reanalysis data sets have been investigated through comparisons with the longterm and homogeneous observation wind speeds data recorded at ten stations. The results reveal that(1) the wind speeds bias of CCMP, CFSR, ERA-int and JRA55 are 0.91 m/s, 1.22 m/s, 0.62 m/s and 0.22 m/s, respectively.The wind speeds RMSE of CCMP, CFSR, ERA-int and JRA55 are 1.38 m/s, 1.59 m/s, 1.01 m/s and 0.96 m/s,respectively;(2) JRA55 and ERA-int provides a realistic representation of monthly wind speeds, while CCMP and CFSR tend to overestimate observed wind speeds. And all the four data sets tend to underestimate observed wind speeds in Bohai Sea and Yellow Sea;(3) Comparing the annual wind speeds trends between observation and the four data sets at ten stations for 1988-1997, 1988–2007 and 1988–2015, the result show that ERA-int is superior to represent homogeneity monthly wind speeds over the China seaes.

Temporal-spatial cross-correlation analysis of non-stationary near-surface wind speed time series

Temporal-spatial cross-correlation analysis of non-stationary wind speed time series plays a crucial role in wind field reconstruction as well as in wind pattern recognition.Firstly,the near-surface wind speed time series recorded at different locations are studied using the detrended fluctuation analysis(DFA),and the corresponding scaling exponents are larger than 1.This indicates that all these wind speed time series have non-stationary characteristics.Secondly,concerning this special feature( i.e.,non-stationarity)of wind signals,a cross-correlation analysis method,namely detrended cross-correlation analysis(DCCA) coefficient,is employed to evaluate the temporal-spatial cross-correlations between non-stationary time series of different anemometer pairs.Finally,experiments on ten wind speed data synchronously collected by the ten anemometers with equidistant arrangement illustrate that the method of DCCA cross-correlation coefficient can accurately analyze full-scale temporal-spatial cross-correlation between non-stationary time series and also can easily identify the seasonal component,while three traditional cross-correlation techniques(i.e.,Pearson coefficient,cross-correlation function,and DCCA method) cannot give us these information directly.

Performance analysis of 20 Pole 1.5 KW Three Phase Permanent Magnet Synchronous Generator for low Speed Vertical Axis Wind Turbine

This paper gives performance analysis of a three phase Permanent Magnet Synchronous Generator (PMSG) connected to a Vertical Axis Wind Turbine (VAWT). Low speed wind condition (less than 5 m/s) is taken in consideration and the entire simulation is carried in Matlab/Simulink environment. The rated power for the generator is fixed at 1.5 KW and number of pole at 20. It is observed under low wind speed of6 m/s, a turbine having approximately1 mof radius and2.6 mof height develops 150 Nm mechanical torque that can generate power up to 1.5 KW. The generator is designed using modeling tool and is fabricated. The fabricated generator is tested in the laboratory with the simulation result for the error analysis. The range of error is about 5%-27% for the same output power value. The limitations and possible causes for error are presented and discussed.

Urban Wind Speed Analysis in Global Climate Change Perspective: <i>Karachi as a Case Study</i>

It is now well known that coastal urban local climate has been showing changing pattern due to global climate change. This communication attempts to explore fluctuating pattern of urban average monthly wind speed during past 50 years (1961-2010). It shows peculiar results taking Karachi (24?53'N, 67?00'E), a coastal mega-city of Pakistan, as a case study. Mann-Kendall trend test shows that March, April and October and both summer and winter seasons show positive trends for the average monthly wind speed during the whole study period (1961-2010). For the earlier 25 years data, it has been found that January, March, May, August, November and December and annual wind speed data have shown the negative trends. Only summer season has shown the positive trend for the wind speed. Similarly, for the most recent 25 years data it has been found that January, February, March, April, May, June, October, November and December and annual and both summer and winter wind speed data have shown the positive trends showing some degree of change in wind speed pattern. Probabilistic analysis reveals that average monthly wind speed data sets follow lognormal, logistic, largest extreme value, and Weibull (two-and three-parameters) probability distributions. Change point analysis has also confirmed the change in the pattern of observed average monthly wind speed data near 1992. The analysis performed reveals the effect of global warming on the local urban wind speed which appears to be temporal non-stationary.

An Extreme Value Analysis of Wind Speed over the European and Siberian Parts of Arctic Region

Multiyear observed time series of wind speed for selected points of the Arctic region (data of station network from the Kola Peninsula to the Chukotka Peninsula) are used to highlight the important peculiarities of wind speed extreme statistics. How largest extremes could be simulated by climate model (the INM-CM4 model data from the Historical experiment of the CMIP5) is also discussed. Extreme value analysis yielded that a volume of observed samples of wind speeds are strictly divided into two sets of variables. Statistical properties of one population are sharply different from another. Because the common statistical conditions are the sign of identity of extreme events we therefore hypothesize that two groups of extreme wind events adhere to different circulation processes. A very important message is that the procedure of selection can be realized easily based on analysis of the cumulative distribution function. The authors estimate the properties of the modelled extremes and conclude that they consist of only the samples, adhering to one group. This evidence provides a clue that atmospheric model with a coarse spatial resolution does not simulate special mechanism responsible for appearance of largest wind speed extremes. Therefore, the tasks where extreme wind is needed cannot be explicitly solved using the output of climate model. The finding that global models are unable to capture the wind extremes is already well known, but information that they are members of group with the specific statistical conditions provides new knowledge. Generally, the implemented analytical approach allows us to detect that the extreme wind speed events adhere to different statistical models. Events located above the threshold value are much more pronounced than representatives of another group (located below the threshold value) predicted by the extrapolation of law distributions in their tail. The same situation is found in different areas of science where the data referring to the same nomenclature are adhering to different statistical models. This result motivates our interest on our ability to detect, analyze, and understand such different extremes.

CFD-Based Performance Analysis and Experimental Investigation of Design Factors of Vertical Axis Wind Turbines under Low Wind Speed Conditions in Thailand

This paper presents effects of design factors on mechanical performance of Vertical Axis Wind Turbines (VAWTs), and an experimental investigation of optimal VAWT performance under low wind speed conditions in Thailand. Design factors include types of wind turbines, number of blades, types of materials, height-to-radius ratios, and design modifications. Potential VAWT models with different design factors are numerically analyzed within a virtual wind tunnel at various wind speeds by utilizing XflowTM?Computational Fluid Dynamics (CFD) software. The performance curves of each VAWT are obtained as plots of power coefficients against tip speed ratios. It is found that the type of wind turbine, number of blades, and height-to-radius ratio have significant effects on mechanical performance whereas types of materials result in shifts of operating speeds of VAWTs. Accordingly, an optimal VAWT prototype is developed to operate under actual low speed wind conditions. The performance curve from experimental results agrees with the CFD results. The proposed methodology can be used in the computer design of VAWTs to improve mechanical performance before physical fabrication.

Analysis of typhoon wind hazard in Shenzhen City by Monte-Carlo Simulation

As one of the most serious natural disasters,many typhoons affect southeastern China every year.Taking Shenzhen,a coastal city in southeast China as an example,we employed a Monte-Carlo simulation to generate a large number of virtual typhoons for wind hazard analysis.By analyzing 67-year historical typhoons data from 1949 to 2015 using the Best Track Dataset for Tropical Cyclones over the Western North Pacific recorded by the Shanghai Typhoon Institute,China Meteorological Administration(CMASTI),typhoon characteristic parameters were extracted and optimal statistical distributions established for the parameters in relation to Shenzhen.We employed the Monte-Carlo method to sample each distribution to generate the characteristic parameters of virtual typhoons.In addition,the Yah Meng(YM)wind field model was introduced,and the sensitivity of the YM model to several parameters discussed.Using the YM wind field model,extreme wind speeds were extracted from the virtual typhoons.The extreme wind speeds for different return periods were predicted and compared with the current structural code to provide improved wind load information for wind-resistant structural design.

Relationships between Terrain Features and Forecasting Errors of Surface Wind Speeds in a Mesoscale Numerical Weather Prediction Model

Numerical weather prediction(NWP)models have always presented large forecasting errors of surface wind speeds over regions with complex terrain.In this study,surface wind forecasts from an operational NWP model,the SMS-WARR(Shanghai Meteorological Service-WRF ADAS Rapid Refresh System),are analyzed to quantitatively reveal the relationships between the forecasted surface wind speed errors and terrain features,with the intent of providing clues to better apply the NWP model to complex terrain regions.The terrain features are described by three parameters:the standard deviation of the model grid-scale orography,terrain height error of the model,and slope angle.The results show that the forecast bias has a unimodal distribution with a change in the standard deviation of orography.The minimum ME(the mean value of bias)is 1.2 m s^(-1) when the standard deviation is between 60 and 70 m.A positive correlation exists between bias and terrain height error,with the ME increasing by 10%−30%for every 200 m increase in terrain height error.The ME decreases by 65.6%when slope angle increases from(0.5°−1.5°)to larger than 3.5°for uphill winds but increases by 35.4%when the absolute value of slope angle increases from(0.5°−1.5°)to(2.5°−3.5°)for downhill winds.Several sensitivity experiments are carried out with a model output statistical(MOS)calibration model for surface wind speeds and ME(RMSE)has been reduced by 90%(30%)by introducing terrain parameters,demonstrating the value of this study.

Analysis of Observed and Modelled Near-Surface Wind Extremes over the Sub-Arctic Northeast Pacific

Wind speed extremes in the sub-Arctic realm of the North-East Pacific region were investigated through extreme value analysis of wind speed obtained from wind simulations of the COSMO-CLM (Consortium for Small-scale Modelling, climate version) mesoscale model, as well as using observed data. The analysis showed that the set of wind speed extremes obtained from observations is a mixture of two different subsets each neatly described by the Weibull distribution. Using special metaphoric terminology, they are labelled as “Black Swans” and “Dragons”. The “Dragons” are responsible for strongest extremes. It has been shown that both reanalysis and GCM (general circulation model) data have no “Dragons”. This means that such models underestimate wind speed maxima, and the important circulation process generating the anomalies is not simulated. The COSMO-CLM data have both “Black Swans” and “Dragons”. This evidence provides a clue that an atmospheric model with a detailed spatial resolution (we used in this work the data from domain with 13.2 km spatial resolution) does reproduce the special mechanism responsible for the generation of the largest wind speed extremes. However, a more thorough analysis shows that the differences in the parameters of the cumulative distribution functions are still significant. The ratio between the modelled Dragons and Black Swans can reach up to only 10%. It is much less than 30%, which was the level established for observations.

Very Short-Term Generating Power Forecasting for Wind Power Generators Based on Time Series Analysis

In recent years, there has been introduction of alternative energy sources such as wind energy. However, wind speed is not constant and wind power output is proportional to the cube of the wind speed. In order to control the power output for wind power generators as accurately as possible, a method of wind speed estimation is required. In this paper, a technique considers that wind speed in the order of 1 - 30 seconds is investigated in confirming the validity of the Auto Regressive model (AR), Kalman Filter (KF) and Neural Network (NN) to forecast wind speed. This paper compares the simulation results of the forecast wind speed for the power output forecast of wind power generator by using AR, KF and NN.

Short-term forecasting optimization algorithms for wind speed along Qinghai-Tibet railway based on different intelligent modeling theories

To protect trains against strong cross-wind along Qinghai-Tibet railway, a strong wind speed monitoring and warning system was developed. And to obtain high-precision wind speed short-term forecasting values for the system to make more accurate scheduling decision, two optimization algorithms were proposed. Using them to make calculative examples for actual wind speed time series from the 18th meteorological station, the results show that: the optimization algorithm based on wavelet analysis method and improved time series analysis method can attain high-precision multi-step forecasting values, the mean relative errors of one-step, three-step, five-step and ten-step forecasting are only 0.30%, 0.75%, 1.15% and 1.65%, respectively. The optimization algorithm based on wavelet analysis method and Kalman time series analysis method can obtain high-precision one-step forecasting values, the mean relative error of one-step forecasting is reduced by 61.67% to 0.115%. The two optimization algorithms both maintain the modeling simple character, and can attain prediction explicit equations after modeling calculation.

Joint Occurrence Period of Wind Speed and Wave Height Based on Both Service Term and Risk Probability

Return periods calculated for different environmental conditions are key parameters for ocean platform design.Many codes for offshore structure design give no consideration about the correlativity among multi-loads and over-estimate design values.This frequently leads to not only higher investment but also distortion of structural reliability analysis.The definition of design return period in existing codes and industry criteria in China are summarized.Then joint return periods of different ocean environmental parameters are determined from the view of service term and danger risk.Based on a bivariate equivalent maximum entropy distribution,joint design parameters are estimated for the concomitant wave height and wind speed at a site in the Bohai Sea.The calculated results show that even if the return period of each environmental factor,such as wave height or wind speed,is small,their combinations can lead to larger joint return periods.Proper design criteria for joint return period associated with concomitant environmental conditions will reduce structural size and lead to lower investment of ocean platforms for the exploitation of marginal oil field.

Long-term Correlations and Extreme Wind Speed Estimations

In this paper, we use fluctuation analysis to study statistical correlations in wind speed time series. Each time series used here was recorded hourly over 40 years. The fluctuation functions of wind speed time series were found to scale with a universal exponent approximating to 0.7, which means that the wind speed time series are long-term correlated. In the classical method of extreme estimations, data are commonly assumed to be independent (without correlations). This assumption will lead to an overestimation if data are long-term correlated. We thus propose a simple method to improve extreme wind speed estimations based on correlation analysis. In our method, extreme wind speeds are obtained by simply scaling the mean return period in the classical method. The scaling ratio is an analytic function of the scaling exponent in the fluctuation analysis.

Correlation between sea surface temperature and wind speed in Greenland Sea and their relationships with NAO variability

The North Atlantic Oscillation （NAO） is one of the major causes of many recent changes in the Arctic Ocean. Generally, it is related to wind speed, sea surface temperature （SST）, and sea ice cover. In this study, we analyzed the distributions of and correlations between SST, wind speed, NAO, and sea ice cover from 2003 to 2009 in the Greenland Sea at 10°W to 10°E, 65°N to 80°N. SST reached its peak in July, while wind speed reached its minimum in July. Seasonal variability of SST and wind speed was different for different regions. SST and wind speed mainly had negative correlations. Detailed correlation research was focused on the 75~N to 80~N band. Regression analysis shows that in this band, the variation of SST lagged three months behind that of wind speed Ice cover and NAO had a positive correlation, and the correlation coefficient between ice cover and NAO in the year 2007 was 0.61 SST and NAO also had a positive correlation, and SST influenced NAO one month in advance. The correlation coefficients between SST and NAO reached 0.944 for the year 2005, 0.7 for the year 2008, and 0.74 for the year 2009 after shifting SST one month later. NAO also had a positive correlation with wind speed, and it also influenced wind speed one month in advance. The correlation coefficients between NAO and wind speed reached 0.783, 0.813, and 0.818 for the years 2004, 2005, and 2008, respectively, after shifting wind speed one month earlier.

Change Point Detection and Trend Analysis for Time Series

Trend analysis and change point detection in a time series are frequent analysis tools.Change point detection is the identification of abrupt variation in the process behaviour due to natural or artificial changes,whereas trend can be defined as estimation of gradual departure from past norms.We analyze the time series data in the presence of trend,using Cox-Stuart methods together with the change point algorithms.We applied the methods to the nearsurface wind speed time series for Australia as an example.The trends in near-surface wind speeds for Australia have been investigated based upon our newly developed wind speed datasets,which were constructed by blending observational data collected at various heights using local surface roughness information.The trend in wind speed at 10 m is generally increasing while at 2 m it tends to be decreasing.Significance testing,change point analysis and manual inspection of records indicate several factors may be contributing to the discrepancy,such as systematic biases accompanying instrument changes,random data errors(e.g.accumulation day error)and data sampling issues.Homogenization technique and multiple-period trend analysis based upon change point detections have thus been employed to clarify the source of the inconsistencies in wind speed trends.

Advanced aerostatic analysis of long-span suspension bridges

As the span length of suspension bridges increases, the diameter of cables and thus the wind load acting on them, the nonlinear wind-structure interaction and the wind speed spatial non-uniformity all increase consequently, which may have unnegligible influence on the aerostatic behavior of long-span suspension bridges. In this work, a method of advanced aerostatic analysis is presented firstly by considering the geometric nonlinearity, the nonlinear wind-structures and wind speed spatial non-uniformity. By taking the Runyang Bridge over the Yangtze River as example, effects of the nonlinear wind-structttre interaction, wind speed spatial non-uniformity, and the cable＇s wind load on the aerostatic behavior of the bridge are investigated analytically. The results showed that these factors all have important influence on the aerostatic behavior, and should be considered in the aerostatic analysis of long and particularly super long-span suspension bridges.

Wind speed inversion and in-orbit assessment of the imaging altimeter on Tiangong-2 space station

Imaging altimeter(IALT)is a new type of radar altimeter system.In contrast to the conventional nadir-looking altimeters,such as HY-2 A altimeter,Jason-1/2,and TOPEX/Poseidon,IALT observes the earth surface at low incident angles(2.5°–8°),so its swath is much wider and its spatial resolution is much higher than the previous altimeters.This paper presents a wind speed inversion method for the recently launched IALT onboard Tiangong-2 space station.Since the current calibration results of IALT do not agree well with the well-known wind geophysical model function at low incidence angles,a neural network is used to retrieve the ocean surface wind speed in this study.The wind speed inversion accuracy is evaluated by comparing with the ECMWF reanalysis wind speed,buoy wind speed,and in-situ ship measurements.The results show that the retrieved wind speed bias is about–0.21 m/s,and the root-mean-square(RMS)error is about 1.85 m/s.The wind speed accuracy of IALT meets the performance requirement.

Restoration of Wind Speed in Qinzhou, Guangxi during Typhoon Rammasun

In 2014,Typhoon Rammasun invaded Qinzhou,Guangxi,causing damage to the wind tower sensor at 80 m in Qinzhou.In order to restore the wind speed at 80 m at that time,this paper was based on the hourly average wind speed data of the wind tower and meteorological station from 2017–2019,and constructed the wind speed related model of Meteorological Station and the wind measuring tower in Qinzhou,Moreover,this paper Based on the hourly average wind speed data of Qinzhou Meteorological Station in 2014,Restored the hourly average wind speed of the anemometer tower during Rammasun landfalled.The results showed it is significant correlation that the hourly mean wind speed of the wind tower at 80 m and the hourly mean wind speed of meteorological station at 100 m(R2=0.9632),and speed of the wind measuring tower and speed of meteorological station constitutes an equation,This equation is Y=0.7834X.The hourly average wind speed of the wind tower at 80 m during the 2014 Rammasun Landing was restored using this model.See the results in Schedule 4.

Short-term local prediction of wind speed and wind power based on singular spectrum analysis and locality-sensitive hashing

With the growing penetration of wind power in power systems, more accurate prediction of wind speed and wind power is required for real-time scheduling and operation. In this paper, a novel forecast model for shortterm prediction of wind speed and wind power is proposed,which is based on singular spectrum analysis(SSA) and locality-sensitive hashing(LSH). To deal with the impact of high volatility of the original time series, SSA is applied to decompose it into two components: the mean trend,which represents the mean tendency of the original time series, and the fluctuation component, which reveals the stochastic characteristics. Both components are reconstructed in a phase space to obtain mean trend segments and fluctuation component segments. After that, LSH is utilized to select similar segments of the mean trend segments, which are then employed in local forecasting, so that the accuracy and efficiency of prediction can be enhanced. Finally, support vector regression is adopted forprediction, where the training input is the synthesis of the similar mean trend segments and the corresponding fluctuation component segments. Simulation studies are conducted on wind speed and wind power time series from four databases, and the final results demonstrate that the proposed model is more accurate and stable in comparison with other models.