Energy Optimal Routing of Vehicle Fleet with Heterogeneous Powertrains

This dissertation examines the benefit of energy optimization in the operation of a vehicle system at an individual vehicle level and the fleet level. For energy optimization in an individual vehicle, a hybridized Class 6 Pickup and Delivery truck with a Range Extended Electric Vehicle configuration is considered. The truck's components were chosen for minimal energy consumption while meeting all the performance requirements of a conventional, diesel-powered vehicle of that class and application. Dynamic Programming is used to determine the best possible energy consumption performance over the course of a working day for the hybrid truck. Energy consumption is then determined using a causal energy management controller on a forward simulator that is compatible with implementation in real-time, where this dissertation introduces the use of a distance-based driver that accurately matches the distance traveled by the vehicle from every start-to-stop in the drive cycle even if the performance constraints of the components prevent the exact matching of the drive cycle speed. The energy consumption results with the forward simulator demonstrate that with increasing levels of information of the expected duty cycle of the day, the onboard energy management can be easily adapted to obtain better fuel consumption performance. For energy optimization in a vehicle fleet, a delivery vehicle fleet is considered that consists of Battery Electric Vehicles (BEVs), Hybrid Electric Vehicles (HEVs) and conventional Internal Combustion Engine Vehicles (ICEVs) operating over the same service area, from a shared depot. This dissertation develops a methodology for route optimization of such a heterogeneous delivery vehicle fleet while taking into account information related to static parameters of the service area (such as topography, payload and driving distance) and dynamic driving conditions (such as traffic incidents and traffic lights). The benefit of route optimization of the fleet for energy consumption and time is then demonstrated for multiple sets of customer locations and depot locations on a real-world road map.