Hybrid Electric Vehicles (HEVs) are regarded as an important (transition) element of sustainable transportation. Exploiting the full potential of HEVs requires (i) a suitable route selection and (ii) suitable power management, i.e., deciding on the split between combustion engine and electric motor usage as well as the mode of the electric motor, i.e., driving or charging the battery (recuperation). The coupling of the two subproblems motivates to formulate the routing and power management as an integrated optimization problem, i.e., optimizing simultaneously the route selection and the split between combustion engine and electric motor over the entire route selection. We present an eco-routing approach that embeds a hybrid (mechanistic/data-driven) model of the HEV powertrain in an integrated routing and power management optimization problem. The hybrid model uses mechanistic equations for those parts of the powertrain that can easily be modeled with mechanistic equations and data-driven surrogates for those parts of the powertrain that are difficult to model mechanistically and to solve. Formulating the integrated routing problem with the hybrid model yields a mixed-integer bilinear program which we reformulate and solve a mixed-integer linear program using a state-of-the-art solver. The results show the validity of the developed hybrid powertrain model and demonstrate that the ecorouting approach with powertrain model embedded can be applied on large-scale problems. We consider optimization for minimal travel time and minimum fuel consumption. The latter results in fuel demand reductions up to 70 %. Alternatively we minimize the fuel consumption while constraining the travel time to a maximum value resulting in up to 50 % fuel demand reductions. The highest fuel demand reductions are achieved in urban environments. The entire framework is written in python and provided as an opensource version (MIT License) under https://git.rwth-aachen.de/avt-svt/public/optimal-routing that can readily be applied.
manuscript under revision.