Pada penelitian ini, membahas hasil peramalan beban listrik jangka pendek. Metode peramalan yang digunakan adalah Feed Forward Backpropagation Neural Network, Cascade Forward Backpropagation Neural Network dan Generalized Regression Neural Network. Peramalan akan kebutuhan energi listrik jangka pendek mengacu pada perhitungan beban harian, suhu udara harian. Peramalan ini akan meramalkan beban listrik pada tanggal 7-13 April 2007. Pada penelitian ini, akan membandingkan hasil peramalan beban listrik menggunakan metode Feed Forward Backpropagation Neural Network, Cascade Forward Backpropagation Neural Network dan Generalized Regression Neural Network. Hasil peramalan beban yang hasilnya mendekati beban aktual terdapat pada metode Cascade Forward Backpropagation Neural Network pada tanggal 8 April 2007 dengan nilai MAPE sebesar 7,6% sedangkan hasil peramalan beban yang perbedaan hasilnya sangat besar dengan beban aktual terdapat pada metode Cascade Forward Backpropagation Neural Network dengan nilai MAPE sebesar 39,7%. Sehingga pada penelitian ini metode peeramalan yang paling baik menggunakan metode Cascade Forward Backpropagation Neural Network..

Kata Kunci: Peramalan, Jangka Pendek, Feed Forward Backpropagation, Cascade Forward Backpropagation, Generalized Regression, Neural Network.

In this study, discussing the results of short-term electricity load forecasting. Forecasting methods used are Feed Forward Backpropagation Neural Network, Cascade Forward Backpropagation Neural Network and Generalized Regression Neural Network. Forecasting short-term electrical energy needs refers to the calculation of daily load, daily air temperature. This forecast will predict the electricity load on 7-13 April 2007. In this study, will compare the results of electricity load forecasting using the method of Feed Forward Backpropagation Neural Network, Cascade Forward Backpropagation Neural Network and Generalized Regression Neural Network. Load forecasting results which results close to the actual load contained in the Cascade Forward Backpropagation Neural Network method on April 8, 2007 with a MAPE value of 7.6% while the results of load forecasting results that are very large difference with the actual load contained in the Cascade Forward Backpropagation Neural Network method with MAPE value of 39.7%. So that in this study the best method of forecasting uses the Cascade Forward Backpropagation Neural Network method.

Keywords: Forecasting, Short-term, Feed Forward Backpropagation, Cascade Forward Backpropagation, Generalized Regression, Neural Network.