• Energy Research

Among the many rechargeable lithium batteries, lithium-titanate, or lithium-titanium oxide cells are characterized by the highest thermal stability and operational safety levels, which makes them particularly well suited for highly demanding applications. This paper presents the results of experimental characterization of a lithium-titanate battery cell for the purpose of building a control-oriented battery equivalent electrical circuit model suitable for designing a state-of-charge estimator. Battery cell model identification has been carried out in three stages: (i) Recording of the open-circuit voltage vs. state-of-charge characteristic by means of low-current charging interspersed with battery voltage recuperation intervals over the full-range of battery voltages, (ii) Determination of the effective battery charge capacity, and (iii) Identification of battery cell series resistance and polarization resistance and lag time constant for different charging current values using recursive parameter estimation based on pseudo-random binary sequence test signal. Based on the experimentally identified battery cell equivalent circuit model (the so-called Thevenin model) an extended Kalman filter has been designed to serve as the state-of-charge estimator. The effectiveness of the proposed estimator is verified by means of simulations and experimentally on the battery test bench.

Among the many rechargeable lithium batteries, lithium-titanate, or lithium-titanium oxide cells are characterized by the highest thermal stability and operational safety levels, which makes them particularly well suited for highly demanding applications. This paper presents the results of experimental characterization of a lithium-titanate battery cell for the purpose of building a control-oriented battery equivalent electrical circuit model suitable for designing a state-of-charge estimator. Battery cell model identification has been carried out in three stages: (i) Recording of the open-circuit voltage vs. state-of-charge characteristic by means of low-current charging interspersed with battery voltage recuperation intervals over the full-range of battery voltages, (ii) Determination of the effective battery charge capacity, and (iii) Identification of battery cell series resistance and polarization resistance and lag time constant for different charging current values using recursive parameter estimation based on pseudo-random binary sequence test signal. Based on the experimentally identified battery cell equivalent circuit model (the so-called Thevenin model) an extended Kalman filter has been designed to serve as the state-of-charge estimator. The effectiveness of the proposed estimator is verified by means of simulations and experimentally on the battery test bench.

This paper presents the design of an extended Kalman filter (EKF) as a state-of-charge (SoC) estimator for a lithium–titanate (LTO) battery cell. The design is based on a Thevenin model of the equivalent battery electrical circuit with two parallel resistive-capacitive (RC) elements for the emulation of electrolyte polarization effects. The effectiveness of the proposed SoC estimator is verified through computer simulations on the comprehensive nonlinear battery simulation model featuring battery parameter vs. SoC variations, which is subject to highly dynamic loading regimes.

This paper presents the design of an extended Kalman filter (EKF) as a state-of-charge (SoC) estimator for a lithium–titanate (LTO) battery cell. The design is based on a Thevenin model of the equivalent battery electrical circuit with two parallel resistive-capacitive (RC) elements for the emulation of electrolyte polarization effects. The effectiveness of the proposed SoC estimator is verified through computer simulations on the comprehensive nonlinear battery simulation model featuring battery parameter vs. SoC variations, which is subject to highly dynamic loading regimes.