Lithium ion battery health state estimation method based on codec model

Abstract

The invention discloses a lithium ion battery health state estimation method based on a codec model, which comprises the following steps of: (1) acquiring battery charging and discharging period dataincluding terminal voltage and current data and the maximum discharging capacity in each charging and discharging period; and (2) constructing a codec model of an attention mechanism according to thecharacteristics of the acquired data, wherein the codec model comprises an encoder and a decoder, voltage and current values are used as encoder input, an SOH estimated value of the battery is obtained and used as decoder output, and the number of nodes of each layer is determined; (3) preprocessing and normalizing the data acquired in the step (1), inputting the data into the codec model with randomly initialized weight, and minimizing the output error of the codec model through Adam algorithm training; and (4) inputting a new test sample into the codec model trained in the step (3), and calculating a prediction error so as to evaluate the accuracy of model prediction.

Description

technical field [0001] The invention belongs to the field of lithium battery PHM, and relates to a lithium battery SOH estimation method based on an attention mechanism codec model. Background technique [0002] Lithium-ion batteries have broader application prospects due to their high energy density, long cycle life, and low self-discharge rate, including consumer electronics, power storage, and transportation. However, the performance of lithium batteries is affected by factors such as charge and discharge cycles, temperature, and aging. Among them, the state of health (SOH) and remaining service life (RUL) are important factors to evaluate whether the battery can operate. Therefore, real-time and accurate SOH and RUL estimation methods have become an urgent requirement to ensure the safe and stable operation of lithium batteries. The existing SOH estimation methods are mainly divided into the following three categories: direct measurement method, modeling method and data...

AI Technical Summary

Problems solved by technology

Although traditional data-driven algorithms can effectively use prior information, they are also limited by the selected features, while deep learning algorithms have more powerful data processing capabilities and nonlinear mapping capabilities. Some SOH and RUL estimation methods based on deep learning are only The original sampling data is required, and the characteristics of the model are used to mine the data characteristics, which also means that the estimation is more comprehensive and the process is simpler

Images