Global Sensitivity Analysis of Battery Single Particle Model Parameters

Parameters sensitivity analysis plays a vital role in the system identification domain. This study presents a time-varying global sensitivity analysis of 14 parameters in a single particle model (SPM) for the graphite LiCoO2 battery. The sensitivity ranking of each parameter is achieved by the elementary effect test coupled with a latin hypercubes sampling method. Simulations at 1/2C, 1C rate discharging and dynamic tests are sufficiently studied to investigate the correlation between working conditions and parameters sensitivity ranking. Moreover, the convergence analysis is assessed to address whether the obtained sensitivity ranking is independent of the sampling size.