A, ANALISIS AKURASI RUNGE-KUTTA SEBAGAI NILAI AWAL ADAMS-BASHFORTH-MOULTON PADA PREDIKSI PRODUKSI KELAPA SAWIT PROVINSI KALIMANTAN TIMUR

Abstract

Palm oil production is one of the important sectors in the Indonesian economy, including in East Kalimantan Province. However, in recent years, production has shown slow growth and fluctuations, so an accurate prediction model is needed to support planning and decision making. This study aims to form a logistic differential equation model and analyze the accuracy level of the Adams–Bashforth–Moulton method of order 5 with the initial value method of the Classical Runge–Kutta order 5 and Runge–Kutta Fehlberg 45. The data used are palm oil production data of East Kalimantan Province for 2015–2024 obtained from the Central Bureau of Statistics of Indonesia. The obtained model has a growth rate parameter of 0.4 and a capacity of 4,100,900 tons. The model solution is carried out using ABM order 5 with RK5 and RKF45 as the initial value method, where in RKF45 it is limited by using the order 5 approach. The accuracy level is evaluated using the Mean Absolute Percentage Error (MAPE). The results showed that RK5 produced a MAPE value of 9.88841685%, while RKF45 was 9.88841880%. RK5 was chosen as the best method because it produced a smaller MAPE value. Based on these methods, the predicted palm oil production in 2025–2028 is 3,981,149.3717 tons, 4,0019,602.672 tons, 4,045,869.3641 tons, and 4,063,720.3084 tons, respectively.

Keywords: Logistic Growth Model, Runge-Kutta, Adams-Bashfort-Moulton, Palm Oil Production, East Kalimantan.