SOC estimation for lithium‐ion batteries based on a novel model

Abstract

Lithium‐ion batteries (LIBs) are widely used in electric vehicles because of their high energy density and less pollution. As an important parameter of the battery management system, accurate estimation of the state of charge (SOC) of the battery can ensure the energy distribution and safe use of the battery. This paper obtains better estimation accuracy from four aspects. First, the battery model is established via Thevenin equivalent circuit model, and the parameters are identified by the forgetting factor recursive least squares. Second, the influence of dual extended Kalman filter on SOC estimation is analysed, a novel algorithm‐based improved dual Kalman filter is proposed. Besides, to reduce the influence of the system noise on the estimation results, an adaptive intelligent algorithm is applied to promote the accuracy of SOC estimation. Finally, compared with the estimated SOC results of the traditional algorithm, the experimental results show the effectiveness of the algorithm.

A novel adaptive H‐infinity filtering method for the accurate SOC estimation of lithium‐ion batteries based on optimal forgetting factor selection

“…Combining Equation (11) and Equation ( 12), the expression of the cost function J 1 can be obtained, as shown in Equation (13).…”

“…To solve the above problems, battery management system (BMS) came into being. [12][13][14] BMS can detect the physical parameters of the lithium-ion battery and estimate state of charge (SOC). Lithium-ion battery SOC is the core parameter of BMS; it can characterize the remaining power of the battery.…”

“…With the in‐depth application of lithium‐ion batteries in the field of new energy vehicles, its safety and reliability have been severely tested, due to individual differences in batteries, overcharging, overdischarging, and overheating often occur during use. To solve the above problems, battery management system (BMS) came into being 12–14 . BMS can detect the physical parameters of the lithium‐ion battery and estimate state of charge (SOC).…”

online state of charge estimation of lithium-ion batteries using gated recurrent unit neural network and a least mean square adaptive filter

“…Model-based estimation methods like Kalman Filter produce a highly reliable estimation as they can mitigate the inaccuracies caused by measurement errors or changing work conditions because this method implements closed-loop feedback control [8,9]. Nevertheless, this method depends on the characteristics of battery models, and any error in calculating battery parameters _which is predictable in operational conditions_ will affect this method's accuracy [6,10].data-driven estimation method is independent of any battery model because it assumes the battery as a black box, nonetheless the training data and method is very important in this method.…”

“…Adaptive filters have been widely used in various applications like prediction, noise canceling, and system identification due to their simplicity and robustness [24]. Authors in [10] tried to specify the conditions for implementing real-time adaptive prediction filters. Huang et al [7] provide a predictor which combines LMS and LSTM-RNN in Wireless Sensor Network.…”