A prediction method of battery capacity decay trajectory based on transplanted neural network

Abstract

The invention relates to a method for predicting battery capacity decay trajectory based on a transplanted neural network, and belongs to the technical field of battery management. The method includes: S1 selecting the power battery to be tested, formulating two cycle conditions of different working conditions, carrying out battery aging experiments respectively for the cycle conditions of the two different working conditions, and collecting battery test data; S2 according to the collected battery Test data, calculate the capacity attenuation of the battery under the corresponding working conditions, generate the benchmark model training database and transplant the neural network training database; S3 select the type of the base model, and use all the data in the benchmark model training database to identify the parameters of the base model; S4 The collected transplanted neural network training database is used for transplanted neural network training, and the transplanted neural network model is established; S5 predicts the future capacity trajectory of the battery with slower decay based on the transplanted neural network model. The invention has the advantages of low cost, low complexity, good portability and the like.

Description

technical field [0001] The invention belongs to the technical field of battery management, and relates to a method for predicting battery capacity decay trajectory based on a transplanted neural network. Background technique [0002] In the practical application of electric vehicles and energy storage systems, accurate estimation of the state of health of the battery is crucial to the safety and efficiency of the battery. In battery state of health estimates, future time battery capacity information can help ensure that the battery operates under reliable and safe conditions and reduce user concerns about battery life. Therefore, it is crucial to develop effective methods for predicting battery capacity fade trajectories. [0003] At present, there are mainly the following types of predictions on the battery capacity decay trajectory of the battery in the future: [0004] 1) Use a complex nonlinear model such as a battery electrochemical model to generate electrochemical a...

AI Technical Summary

Problems solved by technology

[0004] 1) Using a complex nonlinear model such as a battery electrochemical model to generate electrochemical aging mechanism or capacity decay information inside the battery in the future, this method is complex and computationally intensive, and is not suitable for online use

[0005] 2) An empirical model with a specific mathematical expression is used to predict the battery capacity decay curve online after identifying parameters offline. This method has poor adaptive ability and is sensitive to noise, which is prone to huge errors under different working conditions.

[0006] 3) Use machine learning techniques such as neural networks, support vector machines, etc. to derive data-driven models for future capacity attenuation estimation, but traditional data-driven models require a large amount of experimental data for training, thus consuming a large amount of experimental cost and time cost

Images