Optimization of blended battery packs

Abstract

This thesis reviews the traditional battery pack design process for hybrid and electric vehicles, and presents a dynamic programming (DP) based algorithm that eases the process of cell selection and pack design, especially for blended battery packs (those containing two or more different energy sources). The proposed algorithm simultaneously optimizes the size of the battery pack while determining the ideal control strategy for the power split between the two sources. To test the algorithm, a simulation experiment is presented that compares the results of the DP based algorithm with single energy source options and a peak shaving heuristic strategy. The results of this experiment show that the algorithm reliably picks the lowest cost solution, and illustrates that blended battery packs have great potential for cost reduction in hybrid vehicles.