Long Short-Term Memory Recurrent Neural Network for Remaining Useful Life Prediction of Lithium-Ion Batteries

Zhang, Yongzhi; Xiong, Rui; He, Haibo; Pecht, Michael

doi:10.1109/tvt.2018.2805189

Cited by 877 publications

(355 citation statements)

References 26 publications

Supporting

Mentioning

299

Contrasting

Unclassified

Order By: Relevance

“…A lot of work is focusing on recurrent neural networks (RNNs). In References 10–14, LSTM (long short‐term memory) algorithms are used to predict the RUL and SOH of the battery. Hybrid versions of LSTM are also presented in References 9 and 15.…”

Section: Introductionmentioning

confidence: 99%

Online state of health prediction method for lithium‐ion batteries, based on gated recurrent unit neural networks

2020

View full text Add to dashboard Cite

Summary Online state of health (SOH) prediction of lithium‐ion batteries remains a very important problem in assessing the safety and reliability of battery‐powered systems. Deep learning techniques based on recurrent neural networks with memory, such as the long short‐term memory (LSTM) and gated recurrent unit (GRU), have very promising advantages, when compared to other SOH estimation algorithms. This work addresses the battery SOH prediction based on GRU. A complete BMS is presented along with the internal structure and configuration parameters. The neural network was highly optimized by adaptive moment estimation (Adam) algorithm. Experimental data show very good estimation results for different temperature values, not only at room value. Comparisons performed against other relevant estimation methods highlight the performance of the recursive neural network algorithms such as GRU and LSTM, with the exception of the battery regeneration points. Compared to LSTM, the GRU algorithm gives slightly higher estimation errors, but within similar prediction error range, while needing significantly fewer parameters (about 25% fewer), thus making it a very suitable candidate for embedded implementations.

show abstract

Section: Introductionmentioning

confidence: 99%

Online state of health prediction method for lithium‐ion batteries, based on gated recurrent unit neural networks

2020

View full text Add to dashboard Cite

show abstract

“…Alternatively, machine learning techniques have been introduced to extract battery health-related features from data and estimate the battery health. Exemplary works include exploration and exploitation of artificial neural network [20], long-short-term memory network [21], supporting vector machine [22], and random forest regression [23] for battery applications.…”

Section: Introductionmentioning

confidence: 99%

Aging trajectory prediction for lithium-ion batteries via model migration and Bayesian Monte Carlo method

et al. 2019

View full text Add to dashboard Cite

This paper develops a new prediction method for the aging trajectory of lithium-ion batteries with significantly reduced experimental tests. This method is driven by data collected from two types of battery operation modes. The first type is accelerated aging tests that are performed under stress factors, such as overcharging, over-discharging and large current rates, and cover most of the battery lifespan. In the second operation mode, the same kinds of cells are aged at normal speeds to generate a partial aging profile. An accelerated aging model is developed based on the first type of data and is then migrated as a new model to describe the normalspeed aging behavior. Under the framework of Bayesian Monte Carlo algorithms, the new model is parameterized based on the second type of data and is used for prediction of the remaining battery aging trajectory. The proposed prediction method is validated on three types of commercial batteries and also compared with two benchmark algorithms. The sensitivity of results to the number of cycles is investigated for both modes. Illustrative results demonstrate that based on the normal-speed aging data collected in the first 30 cycles, the proposed method can predict the entire aging trajectories (up to 500 cycles) at a root-mean-square error of less than 2.5% for all considered scenarios. When only using the first five-cycle data for model training, such a prediction error is bounded by 5% for aging trajectories of all the tested batteries.

show abstract

“…Data-driven models, which are free of assuming any mechanism a priori, are also gaining increasing attention in the battery state-of-health (SOH) estimation and remaining useful life diagnosis [14]. Different intelligent techniques such as support vector regression [15], [16], Bayesian prediction [17], [18], and artificial neural network [19]- [21] have been successfully applied to build data-driven models for battery cyclic aging prediction. On the one hand, some review papers have summarised these state-of-the-art applications [22], [23], concluding that several limitations still exist as: 1) data-driven approaches are mainly used to capture battery cyclic aging states but very few attempts have been done for calendar aging diagnosis.…”

mentioning

confidence: 99%

Gaussian Process Regression With Automatic Relevance Determination Kernel for Calendar Aging Prediction of Lithium-Ion Batteries

Liu

et al. 2020

IEEE Trans. Ind. Inf.

276

119

View full text Add to dashboard Cite

Battery calendar aging prediction is of extreme importance for developing durable electric vehicles. This paper derives machine learning-enabled calendar aging prediction for lithium-ion batteries. Specifically, the Gaussian process regression (GPR) technique is employed to capture the underlying mapping among capacity, storage temperature, and SOC. By modifying the isotropic kernel function with an automatic relevance determination (ARD) structure, high relevant input features can be effectively extracted to improve prediction accuracy and robustness. Experimental battery calendar aging data from nine storage cases are utilized for model training, validation, and comparison, which is more meaningful and practical than using the data from a single condition. Illustrative results demonstrate that the proposed GPR model with ARD Matern32 (M32) kernel outperforms other counterparts and can achieve reliable prediction results for all storage cases. Even for the partial-data training test, multi-step prediction test and accelerated aging training test, the proposed ARDbased GPR model is still capable of excavating the useful features, therefore offering good generalization ability and accurate prediction results for calendar aging under various storage conditions. This is the first known data-driven application that utilizes the GPR with ARD kernel to perform battery calendar aging prognosis.Index Terms-Lithium-ion batteries, calendar aging prediction, battery health, Gaussian process regression, datadriven model.

show abstract

Long Short-Term Memory Recurrent Neural Network for Remaining Useful Life Prediction of Lithium-Ion Batteries

Cited by 877 publications

References 26 publications

Online state of health prediction method for lithium‐ion batteries, based on gated recurrent unit neural networks

Online state of health prediction method for lithium‐ion batteries, based on gated recurrent unit neural networks

Aging trajectory prediction for lithium-ion batteries via model migration and Bayesian Monte Carlo method

Gaussian Process Regression With Automatic Relevance Determination Kernel for Calendar Aging Prediction of Lithium-Ion Batteries

Contact Info

Product

Resources

About