Journal of Power and Energy Engineering, 2014, 2, 161-168
Published Online April 2014 in SciRes. http://www.scirp.org/journal/jpee
http://dx.doi.org/10.4236/jpee.2014.24023
How to cite this paper: Chang, W.-Y. (2014) A Literature Review of Wind Forecasting Methods. Journal of Power and
Energy Engineering, 2, 161-168. http://dx.doi.org/10.4236/jp ee.2014.24023
A Literature Review of Wind Forecasting
Methods
Wen-Yeau Chang
Department of Electrical Engineering, St. John’s University, New Taipei City, Taiwan
Email: changwy@mail.sju .edu.tw
Received December 2013
Abstract
In this paper, an overview of new and current developments in wind forecasting is given where
the focus lies upon principles and practical implementations. High penetration of wind power in
the electricity system provides many challenges to the power system operators, mainly due to the
unpredictability and variability of wind power generation. Although wind energy may not be dis-
patched, an accurate forecasting method of wind speed and power generation can help the power
system operators reduce the risk of unreliability of electricity supply. This paper gives a literature
survey on the categories and major methods of wind forecasting. Based on the assessment of wind
speed and power forecasting methods, the future development direction of wind forecasting is
proposed.
Keywords
Literature Survey; Wind Forecasting Categories; Wind Speed and Power Forecasting Methods
1. Introduction
Rising crude oil prices highlights the exploitation of renewable energy applications. The wind power is one of
the most attractive renewable energy technologies because of its high efficiency and low pollution [1]. However,
since the power produced by the wind energy conversion systems (WECS) is varied with the atmosphere mete-
orology and wind speed [2-3], unexpected variations of the WECS power generation may increase operating
costs for the electricity system because the requirements of primary reserves will be increased, and place poten-
tial risks to the reliability of electricity supply [4].
The power system operators have to predict changes of the wind power production in order to schedule the
spinning reserve capacity and to manage the grid operations [4]. To reduce the reserve capacity and increase the
wind power penetration, the accurate forecasting of wind speed is needed [5]. In addition, wind power forecast-
ing plays an important role in the allocation of balancing power. Besides, wind power forecasting is used for the
day-ahead scheduling of conventional power plants and trading of electricity on the spot market [6].
Although the prediction accuracy of wind power forecasting is lower than the prediction accuracy of load fo-
recasting. Wind power forecasts s till play a key role to address the operation challenges in electricity supply.
Recently, several methods have been employed for the wind power forecasting. Many literatures have been de-
voted to the improvements of wind power forecasting approaches by researchers with wide experience in the
W.-Y. Chang
162
field tests. A number of wind power forecasting methods have been developed and launched on wind sites. The
wind power forecasting methods can be generally categorized into six groups, persistence method, physical me-
thod, statistical method, spatial correlation method, artificial intelligence method, and hybrid approach.
This paper presents a detailed review on existing tools used in wind speed and wind power prediction over
time-scales, and further identifies possible developments in the future.
2. Classification of Wind Power Forecasting According to Time-Scales
Various methods classified according to time-scales or methodology, are available for wind power forecasting.
Time-scale classification of wind power forecasting methods is different in various literature descriptions.
However, combined with some literatures [7-9], according to the time-scales wind power forecasting methods
can be divided into 4 categories:
Ultra-shor t -te rm forecasting: From few minutes to 1 hour ahead.
Short-term forecasting: From 1 hour to several hours ahead.
Medium-term forecasting: From several hours to 1 week ahead.
Long-term forecasting: From 1 week to 1 year or more ahead.
Table 1 also presents the specific time-scale in view of the operation of electricity systems. The applications
of specific time-scale in electricity systems are consequential different.
3. Overall of the Wind Power Forecasting Method
3.1. Persistence Method
Persistence method uses a simple assumption that the wind speed or wind power at a certain future time will be
the same as it is when the forecast is made [9]. If the measured wind speed and wind power at t are v(t) and P(t),
then the forecasting wind speed and wind power at t+t can be formulated as the following term:
() ()vtt vt+∆ =
(1)
() ()Ptt Pt+∆ =
(2)
The persistence method is somehow more accurate than other wind forecasting methods in ultra-short-term
forecasting. But the accuracy of persistence method will degrade rapidly when with the time-scale of forecasting
is increasing [10].
The persistence method is not only the simplest but is also the most economical wind speed or power fore-
casting method. Electrical utility uses the persistence method for ultra-short-term forecasts. Hence, any forecast
method that is developed should, first, be tested against classical benchmark of persistence method to check how
Table 1. Time-scale classification for wind forecasting.
Time-scale Range Applications
Ultra-short-term few minutes to 1 hour ahead
Electricity market clearing
Real-time grid operations
Regulation actions
Short-term 1 hour to several hours ahead
Economic load dispatch planning
Load reasonable decisions
Operational security in electricity market
Medium-term several hours to 1 week ahead
Unit commitment decisions
Reserve requirement decisions
Generator online/offline decisions
Long-term 1 week to 1 year or more ahead
Maintenance planning
Operation management
Optimal operating cost
Feasibility study for design of the wind farm
W.-Y. Chang
163
much it can improve over the persistence derived forecasts [11].
3.2. Physical Methods
Physical method is based on lower atmosphere or numerical weather prediction (NWP) using weather forecast
data like temperature, pressure, surface roughness and obstacles. NWP model is developed by meteorologists for
large scale area weather prediction [5]. In general, wind speed obtained from the local meteorological service
and transformed to the wind turbines at the wind farm is converted to wind power [7]. Physical methods are to
increase the real resolution of NWP model in order to achieving accurate prediction of the weather [9]. The
physical methods are rendered on supercomputers as they need lots of computations.
Physical systems use parameterizations based on a detailed physical description of the atmosphere, to reach
the best prediction precision. Usually, wind speed given by the weather service on a coarse grid is transformed
to the onsite conditions at the location of the wind farm [6].
Existing commercial wind power forecasting methods use NWP wind forecasts as the input data. Physical
systems, using the input data from NWP, carry out the necessary refinement of these output data (wind speed
forecast) to the on-site conditions by methods that are based on the physics of the lower atmospheric boundary
layer [12].
3.3. Statistical Methods
Statistical methods aim at finding the relationship of the on-line measured power data. For a statistical model,
the historical data of the WECS may be used. Statistical models are easy to model and cheaper to develop com-
pared to other models. Basically, statistical method is good for short time periods. The disadvantage with this
method is that the prediction error increases as the prediction time increases.
Statistical methods include the auto regressive (AR), auto regressive moving average (ARMA), auto regres-
sive integrated moving average (ARIMA), Bayesian approach, and gray predictions. Statistical methods can be
used to solve the problems in engineering, economics and natural sciences that have a great deal of data where
the observations are interdependent.
Firat et al. [13] presented a new statistical method based on AR model and independent component analysis.
Based on the obtained results, the proposed method obviously gives higher accuracy compared with direct fore-
casting.
Since the wind speed has very good succession and randomness, it is quite appropriate to use ARMA model
of times series to forecast the wind speed. Erdem and Shi [14] proposed four approaches based on ARMA me-
thod for the forecasting of the tuple of wind speed and direction. The results showed that the component model
is better at predicting the wind direction than the traditional-linked ARMA model, whereas the opposite is ob-
served for wind speed forecasting.
Li et al. [15] presented an ARMA model combines with wavelet transform for wind speed prediction. The
wavelet transform is used to pick up the low frequency parts of the whole wind speed. ARMA model is used to
forecast the wind speed on the gentled data. The combination model can effectively improve the prediction ac-
curacy.
Palomares-Salas et al. [16] used an ARIMA model for time-series forecast involving wind speed measure-
ments. The paper presents the process of model validation, along with a regression analysis, based in real-life
data. Results show that ARIMA model is better than back propagation neural network for short time-intervals to
forecast.
Miranda and Dunn [17] focused on one-hour-ahead wind speed prediction using the Bayesian approach to
characterize the wind resource. This paper presented the development and simulation results of an AR model
using the Bayesian approach. The simulation results indicated that the Bayesian approach can be a useful tool in
both wind speed and wind power predictions.
3.4. Spatial Correlation Models
The spatial correlation models take the spatial relationship of different sites’ wind speed into account. In spatial
correlation models, the wind speed time-series of the predicted point and its neighboring points are employed to
predict the wind speed [5]. A spatial correlation model is u sed for predicting wind speed at one site based on
W.-Y. Chang
164
measurements at another site. Its behavior has been tested with satisfactory verification using data collected over
seven years [18].
Alexiadis et al. [19] illustrated a technique for forecasting wind speed and power output up to several hours
ahead, based on cross correlation at neighboring sites. Based on spatial correlation models, this paper developed
an ANN approach that significantly improves forecasting accuracy comparing to the persistence forecasting
model.
Barbounis and Theocharis [20] suggested a locally feedback dynamic fuzzy neural network (LF-DFNN) with
application to the wind speed prediction using spatial correlation. According to the position of the base site, re-
mote meteorological stations are installed at two reference sites, such that the positions of the three sites are
lined along the direction of the prevailing winds. In this paper, the LF-DFNN is employed for predicting mul-
ti-step ahead wind speed in base site, using spatial information from remote meteorological stations. Simulation
results revealed that the LF-DFNN exhibits superior performance compared to other network models evaluated
on this application.
3.5. Artificial Intelligence Methods
Recently, with the development of artificial in tellige nce (AI) , various new AI methods for wind speed and pow-
er prediction have been developed. The new developed methods include artificial neural network (ANN), adap-
tive neuro-fuzzy inference system (ANFIS), fuzzy logic methods, support vector machine (SVM), neuro-fuzzy
network , and evolutionary optimization algorithms.
ANN could deal with non-linear and complex problems in terms of classification or forecasting. The ANN
models can represent a complex nonlinear relationship and extract the dependence between variables through
the training process [21]. ANN based methods include back propagation neural networks, recurrent neural net-
works, radial basis function (RBF) neural networks, ridgelet neural network, and adaptive linear element neural
network. ANN based method is an appropriate method to apply into the problem to forecast the wind power.
Sfetsos [22] presented an ANN method for the forecasting of mean hourly wind speed data using time series
analysis. The proposed methodology has an additional benefit for utility that have significant wind penetration
level and use hourly interval for power system operational procedures such as economic dispatch and unit com-
mitment.
Chang [23] described a wind power forecasting methodologies based on back propagation neural network.
The developed model for short-term wind forecasting showed a very good accuracy to be used by a 2400kW
WECS in Taichung coast for the energy supply.
More and Deo [24] presented two wind forecasting methodologies based on back propagation neural network
and recurrent neural network. The neural networks forecasting are found to be more accurate than traditional
statistical time series analysis.
Chang [25] gave a method to do time series prediction forecast of wind power generation using RBF neural
network. The good agreements between the realistic values and forecasting values are obtained; the numerical
results show that the proposed forecasting method is accurate and reliable.
Guo et al. [26] investigated a modified empirical mode decomposition (EMD) based feed-forward neural
network (FNN) wind forecasting method. The proposed method shows the best accuracy comparing with basic
FNN and unmodified EMD-based FNN through multi-step forecasting the mean monthly and daily wind speed
in Zhangye of China.
Li and Shi [27] investigated three types of typical ANN, namely, adaptive linear element, back propagation,
and radial basis function, to forecast the wind speed. The comparing results of three types of ANN show that
even for the same wind dataset, no single ANN model outperforms others universally in terms of all evaluation
metrics. Moreover, the selection of the type of ANN for best performance is also dependent upon the data
sources.
Yang et al. [28] introduced an ANFIS method to interpolate the missing and invalid wind data. The p erfor-
mance tests are given, 12 measured wind data sets from a wind farm in North China are interpolated and ana-
lyzed, respectively. The test results proved the effectiveness of ANFIS method.
Zeng and Qiao [29] described a SVM-based method for wind power forecasting. Simulation studies are car-
ried out by using real wind speed and wind power data obtained from the National Renewable Energy Labora-
tory. Results show that the proposed SVM method outperforms the persistence model and the RBF neural net-
W.-Y. Chang
165
work-based model.
Zhou et al. [30] described a systematic study on fine tuning of least-squares support vector machines
(LSSVM) model parameters for one-step ahead wind speed forecasting. Three SVM kernels, namely linear,
Gaussian, and polynomial kernels, are implemented. It is found that LSSVM methods can outperform the per-
sistence model in the majority of cases.
Xia et al. [31] presented a neuro-fuzzy network method for short-term wind power forecasting. The forecast-
ing approach is applied for the wind power forecasting of a real wind farm located in China. The test results
showed that the trained neuro-fuzzy networks are powerful for modeling the wind farm and forecasting the wind
power.
Jursa and Rohrig [32] presented a new short-term prediction method based on the application of evolutionary
optimization algorithms for the automated specification of neural networks and the nearest neighbour search.
The test results showed that the wind power prediction error can be reduced by using the proposed automated
specification method.
3.6. Hybrid Methods
The object of hybrid models is to benefit from the advantages of each model and obtain a globally optimal fore-
casting performance [10]. Since the information contained in the individual forecasting method is limited, hybr-
id method can maximize the available information, integrate individual model information and make the best
use of the advantages of multiple forecasting methods thus improving the prediction accuracy [8]. The hybrid
methods combine different approaches such as mixing physical and statistical approaches or combine short-term
and medium-term models [11].
Many types of hybrid models were utilized to predict wind power. The types of combinations can be:
Combination of physical and artificial intelligence approaches
Combination of statistical and artificial intelligence approaches
Combination of alternative artificial intelligence models
Zhao et al. [33] investigated a hybrid wind forecasting method consists of a NWP model and ANN models.
The NWP model is established by coupling the Global Forecasting system (GFS) with the Weather Research
and Forecasting (WRF) system together to predict meteorological parameters. This hybrid forecasting system is
profitable for increasing the wind energy penetration level in China.
Shi et al. [34] presented two hybrid models, namely, ARIMA-ANN and ARIMA-SVM, for wind speed and
power forecasting. This paper systematically and comprehensively investigates the applicability of the proposed
hybrid models based on two case studies on wind speed and wind power generation, respectively. The results
show that the hybrid approaches are viable options for forecasting both wind speed and wind power generation
time series, but they do not always produce superior forecasting performance for all the forecasting time hori-
zons investigated.
Guo et al. [35] described a new hybrid wind speed forecasting method based on a back propagation neural
network and the idea of eliminating seasonal effects from actual wind speed datasets using seasonal exponential
adjustment. The test results showed that the proposed method performed better than the single back propagation
neural network.
Catalão et al. [36] proposed a hybrid approach, based on the combination of ANN with wavelet transform, for
short-term wind power forecasting in Portugal. The wavelet transform is used to decompose the wind power se-
ries into a set of better-behaved constitutive series. The test results presented confirm the considerable value of
the proposed hybrid approach in forecasting wind power.
4. The Future of Wind Forecasting
Due to the high penetration of wind power in the electricity system, the forecast accuracy of wind power predic-
tion systems becomes increasingly important. In recent years, many scholars have done a lot of research on wind
power prediction. The forecast accuracy has improved constantly, and it can be expected that intense research
and development efforts are already on track. In order to further improve wind power forecasts, combined with
some literatures [5,9,37,38], anticipated improvements for the further research include the following areas:
Study on more novel artificial intelligence methods and improve their training algorithms to achieve the
forecast accuracy. In addition, new methods on complex terrain are the focus of future research.
W.-Y. Chang
166
Do further research on the hybrid methods combine different approaches such as mixing physical and statis-
tical approaches to achieve good results both in long-term and short-term prediction.
The existing prediction method should put into use in actual WECS. Do further research on the practical ap-
plication of the methods, not only in theoretical.
Establish th e more accurate evaluation system and the standard for measurement of performance of methods.
Further improvements in the NWP models and more frequent updates of the weather predictions will im-
prove the input data for wind power forecasting.
Deepen further research on the use of online wind measurement data, especially for short-term wind fore-
casting.
Do further research on the adaptive parameter estimation. The models have ability to automatically adapt to
the changes of the farms and the surroundings.
Deepen further research on the NWP models designed for offshore environment. Improve the availability of
meteorological data to validate NWP outputs for the offshore locations.
5. Conclusion
This paper presented a review on forecasting of wind speed and power under different time-scales. Six catego-
ries of forecasting methods, which have their own characteristics, were discussed. Papers were selected to em-
phasize the diversity of forecasting methods and the time-scales of forecasting methods. Some of these methods
have good performances at short-term prediction while others perform better in different time-scale prediction. It
is difficult to evaluate the performance of various methods, as the existing applications were in different
time-scale and different way. But various wind forecast method s are available in the power system, which will
help the wind farm owners to identify their wind forecasting method as per their needs. Based on the develop-
ment history of wind speed and power prediction, the future development directions of wind speed and power
forecasting are proposed in the end.
Acknowledgements
The author would like to express his acknowledgements to the National Science Council of ROC for the finan-
cial support under Grant NSC 102-2221-E-129-008.
References
[1] Chang, W.Y. (2013) Short-Term Wind Power Forecasting Using EPSO Based Hybrid Method. Energies, 6, 4879-4896.
http://dx.doi.org/10.3390/en6094879
[2] Chang, W.Y. (2013) Comparison of Three Short Term Wind Power Forecasting Systems. Advanced Materials Re-
search, 684, 671-675. http://dx.doi.org/10.4028/www.scientific.net/AMR.684.671
[3] Chang, W.Y. (2013) An RBF Neural Network Combined with OLS Algorithm and Genetic Algorithm for Short-Term
Wind Power Forecasting. Journal of Applied Mathematics, 2013, Article ID: 971389, 9 p.
[4] Sideratos, G. and Hatziargyriou, N.D. (2007) An Advanced Statistical Method for Wind Power Forecasting. IEEE
Transactions on Power Systems, 22, 258-265. http://dx.doi.org/10.1109/TPWRS.2006.889078
[5] Ma, L., Luan, S.Y., Jiang, C.W., Liu, H L. and Zhang, Y. (2009) A Review on the Forecasting of Wind Speed and
Generated Power. Renewable and Sustainable Energy Reviews, 13, 915-920.
http://dx.doi.org/10.1016/j.rser.2008.02.002
[6] Lange, M. and Focken, U. (2008) New Developments in Wind Energy Forecasting. Proceedings of the 2008 IEEE
Power and Energy Society General Meeting, Pittsburgh, 20-24 July 2008, 1-8.
[7] Wang, X.C. , Guo, P. and Huang, X.B. (2011) A Review of Wind Power Forecasting Models. Energy Procedia, 12,
770-778. http://dx.doi.org/10.1016/j.egypro.2011.10.103
[8] Zhao, D.M., Zhu, Y.C. and Zhang, X. (2011) Research on Wind Power Forecasting in Wind Farms. Proceedings of the
2011 IEEE Power Engineering and Automation Conference, Wuhan, 8-9 September 2011, 175-178.
http://dx.doi.org/10.1109/PEAM.2011.6134829
[9] Zhao, X., Wang, S.X. and Li, T. (2011) Review of Evaluation Criteria and Main Methods of Wind Power Forecasting.
Energy Procedia, 12, 761-769. http://dx.doi.org/10.1016/j.egypro.2011.10.102
[10] Wu, Y. K. and Hon, J. S. (2007) A Literature Review of Wind Forecasting Technology in the World. Proceedings of the
W.-Y. Chang
167
IEEE Conference on Power Tech, Lausanne, 1-5 July 2007, 504-509.
[11] Soman, S.S., Zareipour, H., Malik, O. and Mandal, P. (2010) A Review of Wind Power and Wind Speed Forecasting
Methods with Different Time Horizons. Proceedings of the 2010 North American Power Symposium, Arlington, 26-28
September 2010, 1-8. http://dx.doi.org/10.1109/NAPS.2010.5619586
[12] Bhaskar, K. and Singh, S.N. (2012) AWNN-Assisted Wind Power Forecasting Using Feed-Forward Neural Network.
IEEE Transactions on Sustainable Energy, 3, 306-315. http://dx.doi.org/10.1109/TSTE.2011.2182215
[13] Firat, U., Engin, S.N., Saraclar, M. and Ertuzun, A.B. (2010) Wind Speed Forecasting Based on Second Order Blind
Identification and Autoregressive Model. Proceedings of the 9th International Conference on Machine Learning and
Applications, Washington, 12-14 December 2010, 618-621.
[14] Erdem, E. and Shi, J. (2011) ARMA Based Approaches for Forecasting the Tuple of Wind Speed and Direction. Ap-
plied Energy, 88, 1405-1414. http://dx.doi.org/10.1016/j.apenergy.2010.10.031
[15] Li, L.L., Li, J.H., He, P.J. and Wang, C.S. (2011) The Use of Wavelet Theory and ARMA Model in Wind Speed Pre-
diction. Proceedings of the 1st International Conference on Electric Power Equipment-Switching Technology, Xi’an,
23-27 October 2011, 395-398.
[16] Palomares-Salas, J.C., de la Rosa, J.J.G. , Ramiro, J.G., Melgar, J., Aguera, A. and Moreno, A. (2009) ARIMA vs.
Neural Networks for Wind Speed Forecasting. Proceedings of the IEEE International Conference on Computational
Intelligence for Measurement Systems and Applications, Hong Kong, 11-13 May 2009, 129-133.
[17] Miranda, M.S. and Dunn, R. W. (2006) One-Hour-Ahead Wind Speed Prediction Using a Bayesian Methodology. Pro-
ceedings of the 2006 IEEE Power Engineering Society General Meeting, Montreal, 18-22 June 2006, 1-6.
[18] Alexiadis, M.C., Dokopoulos, P.S., Sahsamanoglou, H. S. and Manousaridis, I.M. (1998) Short-Term Forecasting of
Wind Speed and Related Electrical Power . Solar Energy, 63, 61-68. http://dx.doi.org/10.1016/S0038-092X(98)00032-2
[19] Alexiadis, M.C., Dokopoulos, P.S. and Sahsamanoglou, H.S. (1999) Wind Speed and Power Forecasting Based on
Spatial Correlation Models. IEEE Transactions on Energy Conversion, 14, 836-842.
http://dx.doi.org/10.1109/60.790962
[20] Barbounis, T.G. and Theocharis, J.B. (2007) A Locally Recurrent Fuzzy Neural Network with Application to the Wind
Speed Prediction Using Spatial Correlation. Neurocomputing, 70, 1525-1542.
http://dx.doi.org/10.1016/j.neucom.2006.01.032
[21] Wu, Y. K., Lee, C.Y., Tsai, S.H. and Yu, S.N. (2010) Actual Experience on the Short-Term Wind Power Forecasting at
Penghu-From an Island Pe rspectiv e. Proceedings of the 2010 International Conference on Power System Technology,
Hangzhou, 24-28 October 2010, 1-8. http://dx.doi.org/10.1109/POWERCON.2010.5666092
[22] Sfetsos, A. (2002) A Novel Approach for the Forecasting of Mean Hourly Wind Speed Time Series. Renewable Energy,
27, 163-174. http://dx.doi.org/10.1016/S0960-1481(01)00193-8
[23] Chang, W.Y. (2013) Application of Back Propagation Neural Network for Wind Power Generation Forecasting. Inter-
national Journal of Digital Content Technology and its Application, 7, 502-509.
[24] More, A. and Deo, M.C. (2003) Forecasting Wind with Neural Networks. Marine Structures, 16, 35-49.
http://dx.doi.org/10.1016/S0951-8339(02)00053-9
[25] Chang, W. Y. (2013) Wind Energy Conversion System Power Forecasting Using Radial Basis Function Neural Net-
work. Applied Mechanics and Materials, 284-287, 1067-1071.
http://dx.doi.org/10.4028/www.scientific.net/AMM.284-287.1067
[26] Guo, Z.H., Zhao, W.G., Lu, H.Y. and Wa ng, J.Z. (2012) Multi-Step Forecasting for Wind Speed Using a Modified
EMD-Based Artificial Neural Network Model. Renewable Energy, 37, 241-249.
http://dx.doi.org/10.1016/j.renene.2011.06.023
[27] Li, G. and Shi, J. (2010) On Comparing Three Artificial Nneural Networks for Wind Speed Forecasting. Applied
Energy, 87, 2313-2320. http://dx.doi.org/10.1016/j.apenergy.2009.12.013
[28] Yang, Z.L., Liu, Y.Q. and Li, C.R. (2011) Interpolation of Missing Wind Data Based on ANFIS. Renewable Energy,
36, 993-998. http://dx.doi.org/10.1016/j.renene.2010.08.033
[29] Zeng, J.W. and Qiao, W. (2011) Support Vector Machine-Based Short-Term Wind Power Forecasting. Proceedings of
the IEEE/PES Power Systems Conference and Exposition, Phoenix, 20-23 March 2011, 1-8.
[30] Zhou, J.Y., Shi, J. and Li , G. (2011) Fine Tuning Support Vector Machines for Short-Term Wind Speed Forecasting.
Energy Conversion and Management, 52, 1990-1998. http://dx.doi.org/10.1016/j.enconman.2010.11.007
[31] Xia, J.R., Zhao, P. and Dai, Y.P. (2010) Neuro-Fuzzy Networks for Short-Term Wind Power Forecasting. Proceedings
of the International Conference on Power System Technology, Hangzhou, 24-28 October 2010, 1-5.
http://dx.doi.org/10.1115/1.859612
[32] Jursa, R. and Rohrig, K. (2008) Short-Term Wind Power Forecasting Using Evolutionary Algorithms for the Auto-
W.-Y. Chang
168
mated Specificatio n of Artificial Intelligence Models. International Journal of Forecasting, 24, 694-709.
http://dx.doi.org/10.1016/j.ijforecast.2008.08.007
[33] Zhao, P., Wang, J.F. , Xia, J.R., Dai, Y.P., Sheng, Y.X. and Yue, J. (2012) Performance Evaluation and Accuracy En-
hancement of a Day-Ahead Wind Power Forecasting System in China. Renewable Energy, 43, 234-241.
http://dx.doi.org/10.1016/j.renene.2011.11.051
[34] Shi, J., Guo, J.M. and Zheng, S.T. (2012) Evaluation of Hybrid Forecasting Approaches for Wind Speed and Power
Generation Time Series. Renewable and Sustainable Energy Reviews, 16, 3471-3480.
http://dx.doi.org/10.1016/j.rser.2012.02.044
[35] Guo, Z.H., Wu, J., Lu, H.Y. and Wang , J.Z. (2011) A Case Study on a Hybrid Wind Speed Forecasting Method Using
BP Neural Network. Knowledge-Based Systems, 24, 1048-1056. http://dx.doi.org/10.1016/j.knosys.2011.04.019
[36] Catalão, J.P.S., Pousinho, H.M.I. and Mendes, V.M.F. (2011) Short-Term Wind Power Forecasting in Portugal by
Neural Networks and Wavelet Transform. Renewable Energy, 36, 1245-1251.
http://dx.doi.org/10.1016/j.renene.2010.09.016
[37] Ernst, B., Oakleaf, B., Ahlstrom, M.L., Lange, M., Moehrlen, C., Lange, B., Focken, U. and Rohrig, K. (2007) Pre-
dicting the Wind. IEEE Power and Energy Magazine, 5, 78-89. http://dx.doi.org/10.1109/MPE.2007.906306
[38] Foley, A.M., Leahy, P.G. , Marvuglia, A. and McKeogh, E.J. (2012) Current Methods and Advances in Forecasting of
Wind Power Generation. Renewable Energy, 37, 1-8. http://dx.doi.org/10.1016/j.renene.2011.05.033