Peramalan Penjualan kWh pada Kajian Kelayakan Finansial Penyambungan Pelanggan Industri dengan Integrated Learning Time-Series dan Cross-Sectional Variabel Ekonomi Makro

Andika, Wahab (2024) Peramalan Penjualan kWh pada Kajian Kelayakan Finansial Penyambungan Pelanggan Industri dengan Integrated Learning Time-Series dan Cross-Sectional Variabel Ekonomi Makro. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6010221011-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (14MB) | Request a copy

Abstract

Proses peramalan dan perencanaan yang berbasis risiko dengan metode yang lebih menyeluruh dan kompleks akan memperkecil tingkat kesalahan atau peningkatan akurasi ramalan yang dihasilkan. Analisa pemakain energi listrik pelanggan industri menunjukkan pola pemakaian yang seragam tanpa melihat lama berlangganan. Variabel consumer price index, world governance indicator, dan suku bunga melalui metode forward stepwise regression menunjukkan pengaruh terhadap konsumsi pemakain energi listrik industri di Indonesia. Metode ANN menunjukkan performa terbaik dalam peramalan time-series dibandingkan dengan metode lainnya, meskipun belum bisa menunjukkan pola seasonal yang baik seperti pada metode SARIMAX. Kombinasi peramalan time-series ANN dan causal SVR menunjukkan performa terbaik pada peramalan cross-sectional. Peramalan secara cross-sectional belum mampu menunjukkan peramalan yang baik terhadap adanya disrupsi dan juga belum menunjukkan pola seasonal, namun dapat menyediakan hasil peramalan dari variabel-variabel bebas penyusun sebagai informasi tambahan dalam mitigasi risiko. Penggunaaan hasil peramalan dalam simulasi kajian kelayakan finansial pada 59 calon pelanggan menunjukkan peningkatan potensi penjualan sebesar 63% lebih tinggi dari penggunaan pemakaian minimum.
============================================================
Risk-based planning and forecasting with comprehensive and complex methods will decrease the error in forecasts or increase the forecasted benefit value. Electricity usage analysis on different industrial customer groups shows the same pattern despite how long they have been customers. The consumer price index, world governance indicator, and interest rate through the forward stepwise regression method show a relationship with Indonesia’s industrial electricity usage. The ANN method shows better performance in time-series forecasting than other simulated method, although it fails to provide a decent seasonal pattern as SARIMAX. ANN time-series forecasting combined with SVR causal forecasting also displays the best performance in the cross-sectional method. Cross-sectional forecasting haven’t able to show disruption effect and seasonal pattern in forecasting results, but is capable of presenting related independent variable forecasting as risk mitigation additional information. The forecasted results which being used in the financial feasibility study simulation on 59 potential customers, show an increasing electricity sales opportunity by 63% higher than the study using minimum electricity usage.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	cross-sectional time-series, kajian kelayakan, machine learning, peramalan penjualan kWh, risiko ekonomi, electricity sales forecasting, feasibility study, macroeconomic risk
Subjects:	H Social Sciences > HD Industries. Land use. Labor > HD30.27 Business forecasting T Technology > T Technology (General) > T57.5 Data Processing
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Industrial Engineering > 26101-(S2) Master Thesis
Depositing User:	Wahab Andika
Date Deposited:	17 Jul 2024 05:21
Last Modified:	17 Jul 2024 05:21
URI:	http://repository.its.ac.id/id/eprint/108373

Actions (login required)

View Item