Deep neural network‐driven in‐situ detection and quantification of lithium plating on anodes in commercial lithium‐ion batteries

Abstract

Lithium plating seriously threatens the life of lithium-ion batteries at low temperatures charging conditions, but the onboard detection and quantification of lithium plating are severely hampered by the limited available signals and volatile operating conditions in real scenarios. Herein, we propose a detection method to predict the occurrence and quantification of lithium plating under uncertain conditions by only using constant-current curves during charging based on deep learning. A deep neural network (DNN) is developed to extract data-driven features induced by lithium plating from the charge curves, avoiding the challenge of manual feature selection. Only using the most common voltage and current signals as inputs, the network exhibits superior adaptability and accuracy. The detection accuracy of the proposed method is 98.64%, while the quantity of the lithium plating can be accurately predicted with a root-mean-square error <4.1712 mg. Moreover, the generalization ability of the proposed method is verified by its reliable detection accuracy under conditions that are not used in the training dataset. The detection accuracy is 92.39% for brand new charging conditions and 95.53% for brand new aging states. This method shortens the detection time that currently takes more than several hours (the widely used differential curve analysis) to milliseconds and eliminates the need for a rigorous testing environment, showing great potential for onboard application in future battery management systems.

Building Smarter Aqueous Batteries

“…Therefore, it can be anticipated that more AI applications in ABs will be involved in materials, manufacturing, characterization, and prognosis/diagnosis. [121,122] On one hand, machine learning/deep learning methods (combined with simulation [152] or advanced characterizations [125,153,154] ) can assist the investigation of electrode structure evolution, such as ion plating/dendrite growth [152,155] and crack formation. [156] On the other hand, these AI methods can also promote battery states and performance prediction, including capacity, [157,158] lifetime, [124,159] and cycling protocols.…”

Feature–target pairing in machine learning for battery health diagnosis and prognosis: A critical review

“…130,131 In recent years, ML has also facilitated the noninvasive diagnosis of such degradation principles for indicating battery internal health. For instance, Tian et al 132 first developed a deep neural network (DNN) as an insitu detection tool for quantifying the Li-plating in LIBs, which was only fed the voltage and current signals as features and the Li-plating content manually labeled by DVA as targets. By disassembling cells for cross-validation, Chen et al 133,32 continuingly established ML frameworks for early differentiating the Li-plating from SEI formation as dominating aging mechanisms of LIBs, trained by physically meaningful electrochemical signatures as key features.…”

Design and Implementation of a Battery Big Data Platform Through Intelligent Connected Electric Vehicles

“…With the change of the environment and the aging of the battery, the model structure and parameters need to be adjusted accordingly [11], which adds difficulties to developing accurate battery model laboratory test data; The machine learning methods extract the complex nonlinear relationship between battery state and various variables through huge battery operation data. They include support vector regression (SVR) [12], random forest (RF) [13], neural network (NN) [14,15], etc. However, the accuracy of the estimation results depends on the quality and quantity of the data used for training.…”