Optimasi Performa Kendaraan Listrik Konversi Honda Beat FI 110 CC Menggunakan Response Surface Method

Abstract

The transportation sector's reliance on fossil fuels has driven the development of converted electric vehicles as a more economical alternative. However, the performance of these converted vehicles—specifically the torque of the Brushless DC (BLDC) motor—remains suboptimal due to a lack of systematic controller parameter settings. This study aims to analyze the effects of power limit and load on the electric motor torque of a converted Honda Beat FI 110 cc, develop a mathematical model using Response Surface Methodology (RSM), and determine the optimal parameter combination. A quantitative experimental approach was employed using a face-centered Central Composite Design (CCD), resulting in 13 experimental combinations. Independent variables included power limit (50%, 80%, and 100%) and load (60, 70, and 80 kg), while the dependent variable was torque, measured using a chassis dynamometer. Data were analyzed via second-order polynomial regression, ANOVA, a lack-of-fit test, and desirability optimization. The results showed torque values ​​ranging from 42.87 to 51.81 N·m, with the highest value achieved at a combination of 80% power limit and 70 kg load. The RSM model was statistically significant (F = 17.72; p = 0.001) and showed no lack of fit (p = 0.973). Optimization identified the best conditions at a power limit of 87.88% and a load of 70.10 kg, yielding a predicted torque of 49.66 N·m and a composite desirability value of 0.8197. This study demonstrates that the RSM-CCD approach is effective for modeling and optimizing the torque of converted electric vehicles, thereby serving as a reference for more systematic controller parameter settings.