full2011_inter.pdf - page 348

2011 International Conference on Alternative Energy in Developing Countries and Emerging Economies
- 348 -
The last part of this section provides a general analysis
on the forecasting performance of two models in (5) and
(6) by comparing the forecast values with the actual of
wind speed data in the second set. The MSE in (4) of
both prediction models are presented in Table 4. The
result indicates that the SARIMA model outperforms
than the additive decomposition model because this
model has the MSE less than the other one. The
comparison of the actual values of 3-h wind speed in
August 2010 and the forecast values by two statistical
models are shown in Fig. 10. In order to compare the
results clearly, we have presented the results of
comparative forecast range in terms of three intervals that
is the day at 1 to 10, at 11 to 20 and at 21 to 31 August,
as shown in Fig. 11 – 13.
TABLE IV.
MSE VALUES OF THE SARIMAAND ADDITIVE DECOMPOSITION
MODELS FOR THE SECOND WIND SPEED DATA SET
SARIMA Model Additive Decomposition Model
MSE 0.6629 0.7745
IV. CONCLUSIONS
The forecasting models of 3-h wind speed at 30 m
height along the coast of southern Thailand at Chana
district in Songkhla province are presented in this paper
using two statistical methods, the first method is Box-
Jenkins and the second one is decomposition method.
BIC and Ljung-Box Q statistics were two main criteria to
select the best tentative Box-Jenkins models. Finally,
based on observed wind speed at Chana district in
Songkhla province during May until July 2010, the best
model based upon Box-Jenkins method of 3-h wind speed
at 30 m height is SARIMA(1, 0, 1)(1, 1, 1)8, that shown
in (5). The appropriate decomposition model for 3-h wind
speed at 30 m height is the additive model which can be
decomposed to constant plus seasonal indices. The
additive decomposition model in order to forecast the
wind speed for each period is shown in (6). Results from
MSE analysis shows that the SARIMA(1, 0, 1)(1, 1, 1)8 is
better than the additive decomposition model. Therefore,
this model could be applied not only for short-term wind
speed prediction but also for assisting in wind farm yield
prediction and operation. However, the forecast of an
hourly wind speed in a year period should be studied
further for either investigating the model quality or
creating the new model for higher accuracy in
forecasting. Finally, based on the statistical model, the
novel technique such as Artificial Neural Networks
(ANN) modeling or wavelet modeling should be applied
and compared the quality for short to long terms
forecasting of the wind speed.
ACKNOWLEDGMENT
This research is (partially) supported by the Centre of
Excellence in Mathematics, the Commission of Higher
Education, Thailand. The authors also gratefully
acknowledge the National Research Council of Thailand
(NRCT) for research funding in wind resource
assessment in southern, Thailand.
REFERENCES
[1] Sfetsos, A., 2002, A novel approach for the forecasting of
mean hourly wind speed time series, Renewable Energy,
Vol. 27, pp. 163-174.
[2] Huang, Z. and Chalabi, Z. S., 1995, Use of time-series
analysis to model and forecast wind speed, Journal of
Wind Engineering and Industrial Aerodynamics , Vol. 56,
pp. 311-322.
[3] Box, G. E. P., Jenkins G. M. and Reinsel, G. C., 1994,
Time series analysis: forecasting and control, 3 rd ed.,
Englewood Cliffs, Prentice Hall, New Jersey.
[4] Lei, M., Shiyan, L., Chuanwen, J., Hongling, L. and Yan,
Z., 2009, A review on the forecasting of wind speed and
generated power, Renewable & Sustainable Energy
Reviews , Vol. 13, pp. 915-920.
[5] Giorgi, M. G., Ficarella, A. and Tarantino, M., 2011, Error
analysis of short term wind power prediction models,
Applied Energy, Vol. 88, pp. 1298-1311.
[6] Torres, J. L., Garcia, A., Blas M. De. and Francisco, A.
De., 2005, Forecast of hourly average wind speed with
ARMA models in Navarre (Spain), Solar Energy, Vol. 79,
pp. 65-77.
[7] Cadenas, E. and Rivera, W., 2007, Wind speed forecasting
in the south coast of Oaxaca, Mexico, Renewable Energy ,
Vol. 32, pp. 2116-2128.
[8] Costa, A., Crespo, A., Navarro, J., Lizcano, G., Madsen, H.
and Feitosa, E., 2008, A review on the young history of the
wind power short-term prediction, Renewable &
Sustainable Energy Reviews, Vol. 12, pp. 1725-1744.
[9] Erdem, E. and Shi, J., 2011, ARMA based approaches for
forecasting the tuple of wind speed and direction, Applied
Energy, Vol. 88, pp. 1405-1414.
[10] Sfetsos, A., 2000, A comparison of various forecasting
technique applied to mean hourly wind speed time series,
Renewable Energy, Vol. 21, pp. 23-35.
[11] Sankar, T. J.; Prabakaran, R.; Kannan, S.; and Suresh, S.,
2010, Stochastic modeling for cattle production
forecasting. Journal of Modern Mathematics and Statistics,
Vol. 4, pp. 53-57.
[12] Schwarz, G., 1978, Estimating the dimension of a model.
The Annals of Statistics, Vol. 6, pp. 461-464.
[13] Bowerman, B. L., O’Connell, R. T., 2000, Forecasting and
time series: an application approach, 3rd ed., Wiley, New
York.
[14] Makridakis, S., Wheelwright, S. C., and McGee, V. E.,
1997, Forecasting: methods and applications, 3rd ed.,
Wiley, New York.