COMPARATIVE ANALYSIS OF FUZZY TIME SERIES CHEN AND MARKOV CHAIN METHODS FOR FORECASTING ELECTRICITY CONSUMPTION IN MATARAM CITY

Nirwanto, Nirwanto
Bahri, Syamsul
Harsyiah, Lisa

BAREKENG: Jurnal Ilmu Matematika dan Terapan
Vol 18 No 4 (2024): BAREKENG: Journal of Mathematics and Its Application

The consumption of electrical energy continues to experience fluctuations every month, and these fluctuations cannot be accurately predicted. This uncertainty can become a problem if not projected and planned effectively. Therefore, PT PLN (Persero) needs to be able to provide and distribute electricity supply in an appropriate amount. This research aims to forecast electricity consumption based on historical data from January 2016 to April 2023 using the Fuzzy Time Series Chen (FTSC) method and the Fuzzy Time Series Markov Chain (FTSMC) method. The results of this research show that the forecast for May 2023 using the FTSC and FTSMC methods are 136.878.489 kWh and 143.498.523 kWh, respectively, with MAPE values of 11.61739% and 4.85428%, respectively. Therefore, forecasting in May 2023 using the FTSMC method is better than the FTSC method because the MAPE value is smaller.