The rising popularity of electric motorcycles (EMs) has drawn significant attention from manufacturers and researchers. However, the limitation of small battery packs restricts riding distance and overall performance. Accurately predicting remaining range is crucial for developing advanced range management system (RMS) to optimize battery usage and ease rider anxiety. This paper introduces a machine learning (ML) approach for precise remaining range estimation (RRE) in EMs. To select the most appropriate method, a comparative analysis of various ML methods is firstly conducted using consistent data derived from a standard riding profile. An EM model is then used to assess the proposed strategy against traditional approach. Experimentation is conducted using a hardware-in-the-loop (HiL) simulation system under real-world conditions to evaluate the performance and applicability of the developed model in real-time conditions. The training accuracy achieves 96.9% while the testing attains 95.8% of accuracy score with the representative riding profile. Experimental results show the developed approach significantly improves prediction accuracy of 62.6% and reduces average computational load of 74.6% compared to the conventional method.
Real-Time Prediction of Remaining Driving Range for Electric Motorcycle Applications
IEEE Transactions on Intelligent Transportation Systems ; 25 , 11 ; 18171-18184
01.11.2024
16124527 byte
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Remaining driving range estimation of electric vehicle
IEEE | 2012
|Digital Twin-Based Remaining Driving Range Prediction for Connected Electric Vehicles
SAE Technical Papers | 2023
|Driving structure for electric motorcycle and electric motorcycle
Europäisches Patentamt | 2024
|Electric vehicle remaining driving range prediction method and device considering road environment
Europäisches Patentamt | 2024
|