Toward accurate online state of power estimation for lithium-ion batteries in electric vehicles

State of power (SOP) is a key index for the evaluation of lithium-ion battery peak power performance in electric vehicles. Battery SOP estimation highly relies on accurate battery modelling. However, most existing approaches employed a first-order resistor-capacitance model due to its simple structure, which will lead to biased SOP estimations for a relatively lengthy prediction window. In this paper, an online SOP estimation method is proposed based on a partial-adaptive fractional-order model (P A-FOM). First, this P A-FOM separately considers the time constants for battery transient responses under loads and during relaxations. Second, the parameterization of the P A-FOM combines both offline and online techniques to accommodate fast dynamics of a battery under changeable load conditions and relatively slow dynamics in a lengthy prediction window. Third, the proposed SOP estimation method linearizes the P A-FOM only at the beginning of a prediction window, thus retaining its nonlinear characteristic to the greatest extent. Experimental validations have demonstrated the promising performance of the proposed method under an EV driving profile.