TY - JOUR
T1 - Short-Term Peak Load Forecasting Using Interval Type-2 Fuzzy Logic - Horse Herd Optimization Algorithm in Sulbagsel Electricity, System
AU - Syafaruddin,
AU - Robandi, Imam
AU - Hasanah, Rini Nur
AU - Akil, Yusri Syam
AU - Guntur, Harus Laksana
AU - Lystianingrum, Vita
AU - Djalal, Muhammad Ruswandi
AU - Prakasa, Mohamad Almas
N1 - Publisher Copyright:
© (2025), (Intelligent Network and Systems Society). All rights reserved.
PY - 2025
Y1 - 2025
N2 - This paper discusses short-term peak load forecasting for the South Sulawesi system (Sulbagsel), Indonesia. The peak load is forecasted using interval type-2 fuzzy logic (IT2FL) combined with the horse herd optimization algorithm (HHOA). The HHOA method is employed to optimize the footprint of uncertainty (FOU) in fuzzy logic, including both the antecedent (X, Y) and the consequent (Z). This approach is applied using daily peak load data from the four days prior to the forecasted day (d-4) and the forecasted day itself (d). To compare the HHOA method, similar swarm intelligence techniques, the cuckoo search algorithm (CSA) and Bat Algorithm (BA), are also used. The test results show that IT2FL-HHOA provides more accurate forecasting, as indicated by a significantly lower mean absolute percentage error (MAPE). The MAPE for IT2FL-HHOA is 1.5567%, while for IT2FL-CSA, it is 1.6289%, and for IT2FL-BA, it is 1.6386%. For the Type-1 fuzzy logic (IT1FL-HHOA) method, the MAPE is 1.6604%, for IT1FL-CSA it is 1.6730%, and for IT1FL-BA it is 1.6704%.
AB - This paper discusses short-term peak load forecasting for the South Sulawesi system (Sulbagsel), Indonesia. The peak load is forecasted using interval type-2 fuzzy logic (IT2FL) combined with the horse herd optimization algorithm (HHOA). The HHOA method is employed to optimize the footprint of uncertainty (FOU) in fuzzy logic, including both the antecedent (X, Y) and the consequent (Z). This approach is applied using daily peak load data from the four days prior to the forecasted day (d-4) and the forecasted day itself (d). To compare the HHOA method, similar swarm intelligence techniques, the cuckoo search algorithm (CSA) and Bat Algorithm (BA), are also used. The test results show that IT2FL-HHOA provides more accurate forecasting, as indicated by a significantly lower mean absolute percentage error (MAPE). The MAPE for IT2FL-HHOA is 1.5567%, while for IT2FL-CSA, it is 1.6289%, and for IT2FL-BA, it is 1.6386%. For the Type-1 fuzzy logic (IT1FL-HHOA) method, the MAPE is 1.6604%, for IT1FL-CSA it is 1.6730%, and for IT1FL-BA it is 1.6704%.
KW - HHOA
KW - Load forecasting
KW - MAPE
KW - Short-term
KW - Sulbagsel electricity system
UR - http://www.scopus.com/inward/record.url?scp=85214201734&partnerID=8YFLogxK
U2 - 10.22266/ijies2025.0229.19
DO - 10.22266/ijies2025.0229.19
M3 - Article
AN - SCOPUS:85214201734
SN - 2185-310X
VL - 18
SP - 268
EP - 278
JO - International Journal of Intelligent Engineering and Systems
JF - International Journal of Intelligent Engineering and Systems
IS - 1
ER -