Real-world validated MVEM-based modeling and SoC-dependent driving scenario definition for a parallel hybrid vehicle

This study presents a real-world data–validated framework for driving scenario definition and robustness analysis of a parallel hybrid vehicle using meta-heuristic optimization. A holistic MVEM-based ICE is developed in MATLAB/Simulink and validated using CAN-bus data, achieving trend correlation values above 0.91 and a fuel consumption error of less than 1%. Hybrid operation is formulated as a SoC-aware parametric driving scenario definition problem, in which the hybrid mode speed threshold governs electric-to-combustion transitions. Across initial SoC levels of 25%, 50%, 75%, and 100%, GA, JS, and WOA converge to consistent SoC-dependent optimal mode speed thresholds, ranging from approximately 15.8 km/h at low SoC to 42.4 km/h at full SoC. Implementation of these optimized scenarios under real-world driving conditions yields fuel consumption reductions of 6.96%–18.62% compared to a pure ICE configuration. The results demonstrate that SoC-aware scenario formulation exerts a stronger influence on solution characteristics than algorithm selection.