Güne? enerjili sistemlerin tasarymynda global güne? y?ynymy ?iddetinin tahmini oldukça önemlidir. Güne? enerjisi, global güne? y?ynymyna ba?ly olarak ifade edilir. Bu çaly?mada, Devlet Meteoroloji Y?leri Genel Müdürlü?ü'nden temin edilen Tunceli'ye ait 2005–2009 yyllary arasynda ölçülmü?; açyk gün sayysy, nispi nem, hava basyncy, hava sycakly?y, güne?lenme süresi, rüzgâr hyzy ve güne? y?ynym ?iddeti de?erleri kullanylmy?tyr. Son yyllarda geni?çe kullanym alany bulan yapay sinir a?lary (YSA) metodu ile aylyk ortalama güne? y?ynymy ?iddeti tahmin edilmi?tir. Güne? y?ynym miktarynyn tahmini için Geriye yayylymly (GY) çok katmanly YSA kullanylmy?tyr. YSA modelinin performansy orta katmandaki nöron sayysy, giri? sayysy, ö?renme katsayysy gibi parametreler de?i?tirilerek ayarlanmy?tyr. Olu?turulan YSA modeli ile global güne? y?ynym miktary yüksek bir hassasiyetle tahmin edilmi?tir.

The prediction of global radiation which is used to obtain the potential of sun energy prediction is very important. Sun energy is expressed in terms of global radiation. In this paper; relative humidity atmospheric pressure, air temperature, sunshine duration, wind speed and amount of global solar radiation belong to Tunceli City between 2005-2009, which were taken from Turkish State Meteorological Service have been used. In this paper, back-propagation (GY) multi layer artificial neural networks were used to estimate monthly mean sun radiation in future. Performance of artificial neural networks was adjusted by changed number of neurons in hidden layer, learning rate and momentum constant. The model based on the artificial neural network (YSA) achieved solar radiation forecasting with high accuracy.