Optimization of Fuel Consumption for Rule-Based Energy Management Strategies of Hybrid Electric Vehicles: SOC Compensation Methods

To optimize the fuel consumption of hybrid electric vehicles (HEV) controlled by rule-based energy management strategies (EMS), multiple driving cycles are simulated. These driving cycles are simulated with different EMS calibrations and the optimizer compares the corresponding fuel consumptions. A drive cycle simulation usually ends with a different end state of charge (SOC) compared to the start SOC. Such an unbalanced SOC for the secondary energy source (battery) affects the consumption of the primary energy source (fuel). Therefore, it is crucial to consider the battery SOC difference when comparing fuel consumption in a drive cycle. In this paper, six different methods are presented to compensate the SOC difference or to achieve a balanced SOC, such as Multiple Sequential Drive Cycle Simulation, Variation of Start SOC, Linear Regression, Static Correction Factor, Individual Correction Factor and Linear Interpolation. These methods are compared in their applicability within a numerical optimization and, for a subset, also in their accuracy in SOC compensation using an exemplary hybrid electric vehicle model. It was determined, that Linear Interpolation requires twice as much computing time as either Static or Individual Correction Factor, but it is the most accurate method. In addition, it supports robust EMS behavior without strongly restricting the boundary conditions within the optimization.