Optimisation of electrochemical energy storage based on deep learning algorithm

Currently, machine learning is widely used in the field of energy storage batteries, and its superiority has been proved in terms of time efficiency and prediction accuracy. Among them, various algorithms have different characteristics and adaptive ranges, and choosing a suitable machine-learning algorithm is a key step in constructing a machine-learning model, which greatly affects its prediction accuracy and generalization ability. The key to energy storage batteries is materials. Data-driven materials science makes use of the large amount of data accumulated by traditional experimental, theoretical, and computational simulation methods to model and analyze the performance-driving mechanisms of electrochemical energy storage materials with the help of data-driven artificial intelligence methods in order to accelerate the research, development, and design of new high-performance electrochemical energy storage materials. Currently, as one of the typical representatives of data-driven artificial intelligence methods, machine learning has been widely applied to materials performance prediction and new materials discovery. The application of machine learning in the field of batteries can be traced back to the use of fuzzy logic methods by Salkind et al. in 1999 to determine the state of charge and health of batteries. This paper is based on the combination of deep learning big data algorithms and electrochemical energy storage, which provides a breakthrough and analysis in the field of convergence.