Sains Malaysiana 31: 149-158 (2002) Pengajian Kuantitatif / Quantitative Studies
Peramalan Sinaran Suria Global Melalui Kaedah Berstatistik (Forecasting Global Solar Radiation Using Statistical Method)
Mohd. Anuar bin Abdul Karim Perbadanan Produktiviti Negara (NPC) Lorong Produktiviti Off Jalan Sultan 46200 Petaling Jaya, Selangor D.E., Malaysia
Ahmad Mahir bin Razali Program Statistik Pusat Pengajian Sains Matematik Fakulti Sains Teknologi Universiti Kebangsaan Malaysia 43600 UKM Bangi, Selangor D.E., Malaysia
ABSTRAK
Kajian ini bermatlamat untuk meramal purata bulanan sinaran suria harian berdasarkan kepada faktor-faktor meteorologi seperti jumlah jam penerimaan matahari, kelembapan relatif, jumlah hujan serta kelajuan angin. Dua kaedah pemodelan berstatistik digunakan iaitu kaedah pemodelan regresi linear berganda dan kaedah siri masa yang berdasarkan model gabungan auto regresi purata bergerak (ARPB). Berdasarkan analisis yang telah dijalankan didapati model regresi linear berganda yang mempunyai dua pemboleh ubah tidak bersandar menghasilkan nilai min kuasadua ralat yang paling minimum berbanding dengan model-model peramalan yang lain. Hasil peramalan daripada model tersebut bagi data selama sebulan iaitu bulan Disember 1997 mendapati kecekapan peramalan yang dibuat adalah sebanyak 91.93%. Oleh itu, adalah dicadangkan bahawa model yang diperolehi digunakan sebagai model peramalan jumlah sinaran suria harian di Lapangan Terbang Sultan Abdul Aziz Shah Subang dan amnya di Malaysia.
ABSTRACT
The purpose of this study is to forecast the monthly average daily solar radiation in Malaysia based on the meteorology factors such as sunshine hours, relative humidity, total rainfall and wind speed. Two types of statistical modeling have been used. There were multiple linear regression and the time series auto regression moving average model. Analysis showed that the multiple linear regressions with two independence variables provide the minimum mean square error compared to the other forecasting models. The efficiency of forecasting was 91.93 %. The percentages were gathered from the differences from the actual value of solar radiation data (December 1997) and the forecasting value. It is suggested that the model will be implemented as a forecasting model for gathering the daily solar radiation data in Lapangan Terbang Sultan Abdul Aziz Shah Subang and generally in Malaysia.
RUJUKAN/REFERENCES
Angstrom, A. 1924. Solar and terrestrial radiation. Quart. J.R. Met. Soc. 50: 121-125. Box, G.E. & Jenkins, G.M. 1970. Time Series Analyses Forecasting and control. San Francisco: Holden-Day. Brinkworth, B.J. 1976. Autocorrelation and Stochastic modeling of insolation sequences. Solar Energy 19: 343-347. Carlson, W.L. & Thorne, B. 1997. Applied Statistical Methods. New Jersey: Prentice-Hall. Chuah, D.G.S. & Lee, S.L. 1981. Solar Radiation estimate in Malaysia. Solar Energy 26: 33-40 Goh, T.N. & Tan, KJ. 1976. Stochastic modelling and forecasting of solar radiation data. Solar Energy 19: 755-757 Jennrich, R.I. 1995. An Introduction to Computational Statistics. New Jersey : Prentice-Hall: 97-143 Kamaruzzaman, S. & Mohd. Yusof, O. 1992. Estimates of monthly average daily global solar radiation in Malaysia. Renewable Energy 2: 319-325 Mokhtar Abdullah. 1994. Analisis Regresi. Kuala Lumpur: Dewan Bahasa dan Pustaka: 109-115. Reddy, S.J. 1971. An empirical method for the estimation of the total solar radiation. Solar Energy 13: 291 Sabbagh, J.A. , Sayigh, A.M. & El-Salam, M.A. 1975. Estimation of the total solar radiation from the meteorological data. Solar Energy 19: 307-311 Yusof Mohd. Taib. 1999. Siri Masa. Bangi: Universiti Kebangsaan Malaysia, 54-125
|
|||
|
