A hybrid method for online cycle life prediction of lithium‐ion batteries

Abstract

Many industrial applications use lithium-ion batteries, but lack of maintenance, harsh use environments, and poor charging operations accelerate their degradation. Therefore, online remaining useful lifetime (RUL) prediction is a hot research topic. The RUL estimation analysis of a battery can be based on the normalized capacity as the state of health of its cycle life. We propose a hybrid method based on a bidirectional long short-term memory model with an attention mechanism (BiLSTM-AM) model and a support vector regression (SVR) model for online cycle life prediction. Once the sensor collects temperature readings, it uses SVR to update the initial data online to obtain a multistep advance prediction of the temperature and then uses BiLSTM-AM to predict cycle life.The proposed model is verified using 12 lithium-ion phosphate/graphite cells, and the results show that the average RUL estimation error is 3.72%.

Research on the remaining useful life prediction method of lithium‐ion batteries based on aging feature extraction and multi‐kernel relevance vector machine optimization model

“…Among them, f(x) is the original objective function, δ(t)H(x) is the penalty term, δ(t)is the penalty intensity, and H(x) is the penalty factor. Also, the penalty term needs to vary with the number of iterations t. Drawing on the non-fixed multi-segment mapping penalty function method in References [39], there are the following update Equations (8)(9)(10)(11)(12):…”

“…There are currently three types of RUL prediction methods 8,9 for lithium-ion batteries, namely mechanistic model, data-driven model, and hybrid methods. The mechanistic model method builds the RUL prediction model by analyzing the internal mechanism of lithiumion batteries, 10,11 but the physical and chemical properties of different batteries are different.…”

A data‐driven approach with error compensation and uncertainty quantification for remaining useful life prediction of lithium‐ion battery

“…Therefore, it is important to establish the simple and effective health indicator. Some studies have shown that trend of degeneration is closely related to interval of charging saturation voltage 2,36,37 . Nevertheless, it is difficult to obtain the curve of complete charging saturation voltage.…”

“…Some studies have shown that trend of degeneration is closely related to interval of charging saturation voltage. 2,36,37 Nevertheless, it is difficult to obtain the curve of complete charging saturation voltage. Therefore, the ECVT is established for extraction health indicator.…”

Life Prediction under Charging Process of Lithium-Ion Batteries Based on AutoML

“…Zraibi et al [13] pointed out a CNN-LSTM-DNN algorithm for RUL prediction, in which the three hybrid methods respectively play a critical role. Wang et al [14] proposed a hybrid method combined with a BiLSTM-AM model and a support vector regression (SVR) model for online life prediction, and the collected initial data are updated by SVR, and BiLSTM-AM is used to predict cycle life. Tang et al [15] decomposed the original data into high-and low-frequency parts precisely through an ensemble empirical mode decomposition, and the parts separately are predicted by DNN and a self-designed LSTM network, named IRes2Net-BiGRU-FC, which showed a high robustness of RUL prediction in both the CC and CV stages.…”