Lithium-ion batteries (LiB) are nowadays in the center of the green revolution and are considered as one of the most important players to transform our energy sector towards a more sustainable system. Among all LiB, lithium iron phosphate (LFP) cells are again gaining significant visibility thanks to the cheap and diffuse raw materials and their acceptable lifetime features. However, their characteristics in terms of performance make still their modelling and onboard diagnostic rather complicated. In fact, the presence of hysteresis in the open circuit voltage and flatness of the voltage curves are still big concerns. Additionally, when modeling the charge behavior, voltage curve differs highly from OCV at high values of State-of-Charge (SoC), whereby current and temperature have a significant impact on this (not linear).
Therefore, the aim of this work is to model accurately the charge behavior of LFP cells under different operating conditions. The proposed model is a mix of equivalent electrical circuit model and physicochemical approaches. Parameterization tests are carried out on a small 2.5 Ah LFP cell. Model is validated by means of different validation profiles, considering temperature and current (constant and variable) as important parameters. By means of the proposed approach, errors in modeling charging behavior towards high SoC values are reduced in respect to employing simple lookup tables in all the tested conditions of temperature and current (RMSE error reduction up to approx. 80% for moderate temperature and up to approx. 30% for low temperatures).
Proposed model approach can be employed as simulation tools for designing and validating algorithms for onboard battery state detection for LFP or for designing operating strategies at system level (charging protocols).