Incorporating Drivability Metrics into Optimal Energy Management Strategies for Hybrid Vehicles.

Hybrid Vehicle fuel economy performance is highly sensitive to the "Energy Management" strategy used to regulate power flow among the various energy sources and sinks. Optimal solutions are easy to specify if the drive cycle is known a priori. It is very challenging to compute controllers that yield good fuel economy for a class of drive cycles representative of typical driver behavior. Additional challenges come in the form of constraints on powertrain activity, like shifting and starting the engine, which are commonly called "drivability" metrics. These constraints can adversely affect fuel economy. In this dissertation, drivability restrictions are included in a Shortest Path Stochastic Dynamic Programming (SPSDP) formulation of the energy management problem to directly address this tradeoff and generate optimal, causal controllers. The controllers are evaluated on Ford Motor Company's highly accurate proprietary vehicle model and compared to a controller developed by Ford for a prototype vehicle. The SPSDP-based controllers improve fuel economy more than 15% compared to the industrial controller on government test cycles. In addition, the SPSDP-based controllers can directly quantify tradeoffs between fuel economy and drivability. Hundreds of thousands of simulations are conducted using real-world drive cycles to evaluate performance and robustness in the real world, demonstrating 10% improvement compared to the baseline. Finally, the controllers are tested in a real vehicle.PhDMechanical EngineeringUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/77818/1/dopila_1.pd