Estimasi Parameter Model ARIMA untuk Peramalan Debit Air Sungai Menggunakan Least Square dan Goal Programming

Abstract

Forecasting is a technique to make a desicion in the future considered by data from the past and present. This forecasting is in hydrology sector which is river flow forecasting. River flow forecasting is one way to anticipate the instability of the river flow. The aim of this research was to determine the best ARIMA model based on analysis of the river flow of Karang Mumus, Samarinda. This research will explain the procedure of ARIMA model building using the Least Square and Goal Programming to predict the river flow of Karang Mumus, Samarinda. The data used montly from January until December. The model of ARIMA (2,1,2)to predict the river flow of Karang Mumus using Goal Programming is : Z_t=μ-0,0492Z_t-1-0,0523Z_t-2-0,9969Z_t-3+0,9247a_t-1+0,9339a_t-2+a_t

ARIMA (2,1,2) for river flow forecasting using Goal Programming is : Z_t=1,17Z_t-1-0,17Z_t-2+a_t+0,31a_t-1

The best ARIMA model for river flow forecasting of Karang Mumus is ARIMA (2,1,2) using Least Square method. Result for river flow forecasting of Karang Mumus river in Samarinda from January until Desember 2015 are 1.733 m³, 1.729 m³, 1.730 m³, 1.730 m³, 1.729 m³, 1.730 m³, 1.732 m³, 1.729 m³, 1.730 m³, 1.732 m³, 1.729 m³, dan 1.730 m³.