Graphical Interpretation of the Extended Kalman Filter: Estimating the State-of-Charge of a Lithium Iron Phosphate Cell

Abstract

Electric vehicles (EVs) fall in line with a new ideology of less waste and more conscious usage of resources, slowly picking up speed. In this context, energy storage is of paramount importance, making batteries a key element in the architecture of the electric vehicles. The state of the battery pack must be thoroughly monitored to prolong lifetime and extend vehicle range. For this, measurable physical quantities (i.e. terminal voltage, charge/discharge current, temperature) must be monitored and processed, while the inferred parameters (e.g. state-of-charge (SoC), state-of-health (SoH)) are computed and continuously updated. Whether we are talking about control of a noisy system, ill-defined decision-making processes or data analysis, estimation theory comes into play on a regular basis. The estimation algorithm is critical for appropriate usage of all available power, therefore, research effort is required to allow development of an optimum for a given application, by exploring design alternatives and their effects. This paper evaluates graphically an extended Kalman filter (EKF) for determining the SoC of lithium-ion batteries (LIBs) considering various cell models, initial conditions and charge/discharge profiles. The results are qualitatively and quantitatively assessed by extracting and visualizing the dynamics of the internal variables of the filter during operation.

Keywords

• battery management systems
• electric vehicles
• Kalman filters
• Lithium batteries
• parameter estimation