Prediksi Beban Listrik Jangka Pendek Menggunakan Metode Autoregressive Integrated Moving Average (Arima)

Abstract

Intisari - Penelitian ini bertujuan untuk menjelaskan bagaimana cara menggunakan metode ARIMA (Autoregressive Integrated Moving Average) untuk memprakirakan beban konsumsi listrik jangka pendek dan mengetahui seberapa besarkah tingkat akurasi dari metode ARIMA (Autoregressive Integrated Moving Average) yang digunakan. Metode prediksi jangka pendek Autoregressive Integrated Moving Average atau ARIMA digunakan sebagai metode untuk memperhitungkan besarnya penggunaan energi listrik di Gedung H Teknik Elektro dan Teknik Mesin Fakultas Teknik Universitas Lampung pada bulan Juni dan Juli tahun 2019 dengan menggunakan data penggunaan energi listrik pada bulan April dan Mei tahun 2019. Observasi yang dilakukan adalah memperhitungkan prediksi data deret waktu berupa hubungan antara Energi listrik (kWh) terhadap waktu. Analisis prediksi menggunakan metode ARIMA (2,1,0) memberikan nilai galat rata-rata sebesar 29,59%.   Kata kunci - Prediksi, ARIMA, Energi Listrik, Galat   Abstract - Nowadays forecasting methods have been widely used for various disciplines, with no exception for electrical energy. That methods used to determine the amount of electrical energy consumtion in the future. This research will discuss short term forecasting method Autoregressive Integrated Moving Average or ARIMA for determine the amount of electrical energy consumtion in the H Building of Electrical Engineering and Mechanical Engineering Department of the Faculty of Engineering, University of Lampung in June and July 2019. This research uses data that has been stored on a server computer in the University of Lampung's ICT building by using the Electricity Measurement Smart Monitoring equipment that has been installed in the H building of the Faculty of Engineering, University of Lampung. The data used for this method is the data in April and May 2019. The observation is to forecast time series data, electrical energy consumption (kWh) againts time. Forecasting analysis using the ARIMA (2,1,0) method showed an average 29,59% of error value.   Keywords - Forecasting, ARIMA, Electrical Energy, Error