Design and optimization of fuel cell/battery/supercapacitor hybrid power sources for electric vehicles.

摘要

Fuel Cell powered Hybrid electric Vehicles (FCHVs) are considered to be the most promising alternatives of Internal Combustion Engines (ICE) vehicles. One of the most important research aspects of FCHVs is their hybrid power sources study; however, current research is insufficient and there are many open problems to be further studied, such as hybrid power sources topology analysis, vehicle power sources design and optimization, and related dynamic models construction. Also a powerful simulation package of FCHVs is indispensable to evaluate and support the related research. Addressing these problems, this dissertation carries out a series of studies step by step which includes the following three parts.; The first part is focused on modeling electrochemical components. A novel fully dynamic lithium-ion battery model is developed, which accounts for battery nonlinear equilibrium potentials, rate- and temperature-dependencies, thermal effects and response to transient power demand. A multi-stage resistor capacitor ladder supercapacitor model is constructed with unique characteristics including automatic order selection and capacity scaling. In the modeling of a Proton Exchange Membrane (PEM) fuel cell system, each component is first modeled separately, and then these modules are coupled together into an integrative unit.; The second part concentrates on the study of hybrid power source topology analysis. Firstly, the concepts of passive hybrid and active hybrid are defined. Secondly, an active hybrid is constructed and explored through both experiments and simulations. Finally, the comparison of passive and active hybrids, along with battery alone, is undertaken to study the performance extension of power sources using power converters. The study process is further extended and applied in hybrid power source design for FCHVs.; In the third part, a complete simulation package for FCHVs is constructed in the Virtual Test Bed (VTB) computational environment. The modeling approach is forward looking (causal) and the system setup is modular, thus the package recognizes the dynamic interaction among different vehicle components and provides a powerful capability to simulate different topologies. After that, a novel 8-step energy component size determination and optimization method is proposed, which is based on the operation of the decreasing rearrangement distribution function of drive cycles.