Indirect prediction method for residual service life of lithium ion battery

Abstract

An indirect prediction method for the remaining service life of a lithium ion battery comprises the following steps: firstly, extracting an appropriate indirect health factor expressing the aging of the lithium ion battery, and verifying the correlation between the indirect health factor and the capacity; constructing an indirect health factor prediction model based on a support vector regression algorithm; then constructing a relation model between the indirect health factors and the capacity based on Gaussian process regression; and finally, inputting the indirect health factor prediction value into a Gaussian process regression model to realize lithium battery capacity prediction, and when the battery capacity reaches a set failure threshold value, outputting a residual service life prediction result. The method solves the problems that traditional support vector regression can only obtain a single-point predicted value and Gaussian process regression is poor in long-term prediction performance, the prediction performance is stable, the prediction confidence interval is obtained while an accurate prediction result is obtained, and a lithium battery system can make decision maintenance and management in advance conveniently when a battery breaks down.

Description

technical field [0001] The invention relates to the field of lithium battery electrical performance monitoring, in particular to an indirect method for predicting the remaining service life of a lithium ion battery. [0002] technical background [0003] Compared with lead-acid batteries, nickel-metal hydride batteries, and nickel-cadmium batteries, lithium-ion batteries are small in size and light in weight, greatly reducing the volume and weight of vehicle battery packs, and have high operating voltage, high energy density, and no memory effect. Small self-discharge, long cycle life and other advantages. Widely used in electronic equipment, energy storage systems, aerospace and other fields, it has become the mainstream energy source of electric vehicles and plays an important role in modern society. However, in the process of repeated use, the performance of lithium batteries will gradually degrade and become invalid, which will cause the batteries to easily leak and shor...

AI Technical Summary

Problems solved by technology

In the existing literature methods of this kind of methods, the support vector regression method is often used to predict the RUL of lithium batteries. SVR improves the generalization ability by seeking the minimum structural risk. Based on this, it can also model well when the sample size is small, but This method cannot obtain the uncertainty expression of the prediction results; the GPR model is a flexible non-parametric model based on Bayesian theory. While obtaining the prediction value point estimate, a prediction confidence interval is given to represent the prediction uncertainty. However, the long-term prediction performance is poor when using the GPR method for lithium battery RUL prediction

Moreover, these methods do not consider the impact of the entire charging and discharging process of the lithium battery and the temperature on the RUL prediction of the lithium battery

Images