Data-Driven Identification of Combustion Operating Regimes and Fuel Consumption Sensitivity in Gasoline Plug-In Hybrid Electric Vehicles Under Real-Driving Conditions

Abstract

Abstract

Plug-in hybrid electric vehicles (PHEVs) combine an internal combustion engine with an electric propulsion system, creating highly variable engine operating conditions that differ substantially from those of conventional gasoline vehicles. The frequent transitions between engine-off, engine-on, electric-assisted, charge-sustaining, and transient operating states make the identification of combustion-related operating regimes particularly important for understanding real-world fuel consumption. This study develops a data-driven framework for identifying distinct combustion operating regimes and quantifying the sensitivity of fuel consumption to vehicle operating conditions in gasoline PHEVs under real-driving conditions. The proposed framework integrates time-resolved vehicle operating parameters, including vehicle speed, acceleration, engine operating state, battery state of charge, engine restart events, propulsion mode, and fuel consumption, to characterize the relationship between driving dynamics and engine fuel demand. Operating data are classified into physically meaningful regimes according to the combined behavior of engine activation, vehicle demand, and battery state. Statistical sensitivity analysis is subsequently applied to quantify the contribution of individual operating variables and their interactions to variations in fuel consumption. Particular attention is given to transient engine operation and repeated engine activation, which can produce operating conditions substantially different from steady-state engine behavior. The analysis is structured to distinguish the effects associated with electric-energy availability from those directly related to gasoline engine operation. This approach enables the identification of operating regions in which fuel consumption becomes particularly sensitive to changes in vehicle speed, acceleration, battery state of charge, and engine utilization. The resulting framework provides a quantitative basis for interpreting real-world fuel-consumption variability in gasoline PHEVs and for identifying combustion operating conditions that are associated with inefficient fuel utilization. The study also establishes a reproducible analytical pathway for transforming real-world vehicle measurements into combustion-oriented operating regimes, thereby supporting more precise assessment of PHEV energy performance and providing a methodological basis for future optimization of engine operating strategies and powertrain control.