(35a) Power System Design for a Hybrid Fuel Cell Vehicle: A Globally Optimal Approach

Ahmed, S. K. - Presenter, Illinois Institute of Technology
Chmielewski, D. J. - Presenter, Illinois Institute of Technology

The notion of a fuel cell powered vehicle has captured the imagination of many as a clean/efficient alternative to the internal combustion engine. However, the capital cost of a fuel cell power unit is significant, and threatens to be a show-stopping hurtle. As such, many have looked to hybridize the fuel cell with lower cost energy storage devices; a rechargeable battery and / or super-capacitor. These low energy / high power density devices will provide supplemental power during periods of large demand and thus allow for the selection of a smaller / lower cost fuel cell unit. The basic design challenge is to size of the fuel cell and supplemental energy storage units so as to minimize overall cost.

Due to the drive cycle expected during vehicle operation, the proposed design scheme should include power demand variability as an underlying element. It is further recognized that a high level controller is needed to coordinate device interactions as well as ensure that demanded power is sent to the electric motor driving the vehicle. Clearly, the characteristics of this high level controller will influence the required size of each device. As such, controller characteristics have been included as design variables. Additional design criteria include; max power and energy density for each type of device, losses associated with energy conversion, friction losses, the impact of device size on vehicle mass and overall power demand.

Concerning computational issues, we will show that nearly all of the nonlinear constraints discussed above can be converted to a convex form. The exception is a limited set of scalar reverse-convex inequalities, used to define the relation between variance and standard deviation for key process variables. Given this convex / reverse-convex form a computationally efficient branch-and-bound algorithm has been developed to obtain globally optimal solutions to the proposed hybrid vehicle design problem.