Statistic Method for Extraction of Synthetic Load Cycles

Paper i proceeding, 2009

Despite rapid development, the battery is still the single most expensive component in a HEV drivetrain. Consequently, its durability is critical to the overall feasibility of the vehicle. The battery ageing mechanisms and the resulting cycle life of HEV-optimised batteries are highly non-linear and difficult to test. In addition, the selection of load cycle profile is of great significance when battery cycle life is to be verified experimentally. This paper presents a statistic method for evaluation and simplification of dynamic load profiles based on measured load profiles from heavy-duty HEV applications. The presented method has been used to extract simplified synthetic load cycles with configurable energy throughput as well as different strategies for state-of-charge control. These cycles were also compared with reference cycles and evaluated regarding power distribution, energy distribution, energy window and energy throughput. The presented method was found to be a usable tool for creating new battery load cycles for cyclelife tests. In addition, it may be a useful to evaluate and compare statistical properties of measured cycles before initiating laboratory battery tests.