摘要:AbstractA data-driven methodology is proposed to identify a multi-input single-output (MISO) polynomial nonlinear state-space model (PNLSS) for battery’s short-term electrical response, starting from the best linear approximation of the electrical impedance at multiple operating conditions. Furthermore, a transient handling methodology is utilised to minimise the effect of transients arising due to concatenating data from multiple experiments during the identification of the PNLSS model. Experimental validation is done using the data acquired from the operating points lying in the nonlinear regime of an almost depleted battery in terms of the state of charge (SoC) i.e. between 2%-10% at temperatures between 5°C-40°C.
关键词:KeywordsNonlinear state-space modelsFrequency domain system identificationBattery