Battery health degradation and optimal life management

In this paper, we discuss the battery health degradation and optimal battery life management. As the first contribution of the paper, we discuss how battery health degradation manifests itself as the battery capacity fading due to repeated charging and discharging cycles. We present two models, namely, LAR-αβγ and CVD (control variable-dependent) regression models, for battery capacity fade modeling, which are characterized as functions of the number of cycles and two charge control parameters, viz., maximum terminal voltage of the battery and maximum charge current. The development of these models is based on curve-fitting of data from copious aging experiments performed on Samsung GS4 batteries. These models are compared to a bi-exponential capacity model and the superiority of the proposed capacity models over the bi-exponential capacity model is demonstrated using the experimental data. The CVD battery capacity fade model is integral to a battery life management scheme, which is the second contribution of this paper. The battery life management scheme consists of a two-level strategy which provides fast charging, while minimizing long term health degradation effects. At level-I, the CVD battery capacity fade model is used for developing an optimal charging parameter selection method, which provides the best control variables, viz., the maximum terminal voltage and the maximum charge current to achieve a pre-specified desired “useful cycle life”, while attaining the fastest possible time-to-charge (TTC). The selection of the best control variables utilizes information about the present condition of the battery, such as the number of cycles that the battery has been exposed to, the normalized capacity of the battery, and the battery resistance at the present cycle. Practically, estimates of the required parameters may be obtained by a battery fuel gauge. At level-II, the maximum terminal voltage and the maximum charge current calculated at level-I are used as limiting conditions for the optimal charging strategy. The proposed optimal charging parameter selection method is illustrated via numerical results.