Remaining useful life prediction of the lithium-ion battery based on CNN-LSTM fusion model and grey relational analysis

There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

Abstract

<abstract> <p>The performance of lithium-ion batteries will decline dramatically with the increase in usage time, which will cause anxiety in using lithium-ion batteries. Some data-driven models have been employed to predict the remaining useful life (RUL) model of lithium-ion batteries. However, there are limitations to the accuracy and applicability of traditional machine learning models or just a single deep learning model. This paper presents a fusion model based on convolutional neural network (CNN) and long short-term memory network (LSTM), named CNN-LSTM, to measure the RUL of lithium-ion batteries. Firstly, this model uses the grey relational analysis to extract the main features affecting the RUL as the health index (HI) of the battery. In addition, the fusion model can capture the non-linear characteristics and time-space relationships well, which helps find the capacity decay and failure threshold of lithium-ion batteries. The experimental results show that: 1) Traditional machine learning is less effective than LSTM. 2) The CNN-LSTM fusion model is superior to the single LSTM model in predicting performance. 3) The proposed model is superior to other comparable models in error indexes, which could reach 0.36% and 0.38e-4 in mean absolute percentage error (MAPE) and mean square error (MSE), respectively. 4) The proposed model can accurately find the failure threshold and the decay fluctuation for the lithium-ion battery.</p> </abstract>

Related collections

Most cited references 52

Record: found
Abstract: found
Article: not found

The Li-ion rechargeable battery: a perspective.

John B Goodenough, Kyu-Sung Park (2013)

Each cell of a battery stores electrical energy as chemical energy in two electrodes, a reductant (anode) and an oxidant (cathode), separated by an electrolyte that transfers the ionic component of the chemical reaction inside the cell and forces the electronic component outside the battery. The output on discharge is an external electronic current I at a voltage V for a time Δt. The chemical reaction of a rechargeable battery must be reversible on the application of a charging I and V. Critical parameters of a rechargeable battery are safety, density of energy that can be stored at a specific power input and retrieved at a specific power output, cycle and shelf life, storage efficiency, and cost of fabrication. Conventional ambient-temperature rechargeable batteries have solid electrodes and a liquid electrolyte. The positive electrode (cathode) consists of a host framework into which the mobile (working) cation is inserted reversibly over a finite solid-solution range. The solid-solution range, which is reduced at higher current by the rate of transfer of the working ion across electrode/electrolyte interfaces and within a host, limits the amount of charge per electrode formula unit that can be transferred over the time Δt = Δt(I). Moreover, the difference between energies of the LUMO and the HOMO of the electrolyte, i.e., electrolyte window, determines the maximum voltage for a long shelf and cycle life. The maximum stable voltage with an aqueous electrolyte is 1.5 V; the Li-ion rechargeable battery uses an organic electrolyte with a larger window, which increase the density of stored energy for a given Δt. Anode or cathode electrochemical potentials outside the electrolyte window can increase V, but they require formation of a passivating surface layer that must be permeable to Li(+) and capable of adapting rapidly to the changing electrode surface area as the electrode changes volume during cycling. A passivating surface layer adds to the impedance of the Li(+) transfer across the electrode/electrolyte interface and lowers the cycle life of a battery cell. Moreover, formation of a passivation layer on the anode robs Li from the cathode irreversibly on an initial charge, further lowering the reversible Δt. These problems plus the cost of quality control of manufacturing plague development of Li-ion rechargeable batteries that can compete with the internal combustion engine for powering electric cars and that can provide the needed low-cost storage of electrical energy generated by renewable wind and/or solar energy. Chemists are contributing to incremental improvements of the conventional strategy by investigating and controlling electrode passivation layers, improving the rate of Li(+) transfer across electrode/electrolyte interfaces, identifying electrolytes with larger windows while retaining a Li(+) conductivity σ(Li) > 10(-3) S cm(-1), synthesizing electrode morphologies that reduce the size of the active particles while pinning them on current collectors of large surface area accessible by the electrolyte, lowering the cost of cell fabrication, designing displacement-reaction anodes of higher capacity that allow a safe, fast charge, and designing alternative cathode hosts. However, new strategies are needed for batteries that go beyond powering hand-held devices, such as using electrode hosts with two-electron redox centers; replacing the cathode hosts by materials that undergo displacement reactions (e.g. sulfur) by liquid cathodes that may contain flow-through redox molecules, or by catalysts for air cathodes; and developing a Li(+) solid electrolyte separator membrane that allows an organic and aqueous liquid electrolyte on the anode and cathode sides, respectively. Opportunities exist for the chemist to bring together oxide and polymer or graphene chemistry in imaginative morphologies.