<![CDATA[The development of electric vehicles (EVs) in Indonesia is accelerating following government policies aimed at reducing greenhouse gas emissions. Despite their benefits, the adoption of electric motorcycles remains limited due to concerns about battery life and charging station availability. This study proposes a machine learning-based model to predict distance coverage (DC) based on the state of charge of the battery (SoC) for electric motorcycles, specifically under a full throttle dominant usage pattern. The research employs multiple regression and classification algorithms, including Linear Regression, Random Forest Regression, and Support Vector Regression (SVR) for prediction, along with Random Forest Classifier, Logistic Regression, and K-Nearest Neighbors (KNN) Classifier for travel classification. The results demonstrate that Linear Regression outperforms other models for DC prediction, achieving an R2 value of 0.9818, while the Random Forest Classifier achieves 98% accuracy in classifying travel distances. A graphics user interface (GUI)-based software was developed to integrate these models, enabling real-time prediction and travel classification for users. The findings indicate that ML-based DC prediction can enhance user confidence and optimize battery usage in electric motorcycles.]]>
Copyrights © 2025
