A Wavelet-Based Battery State of Health Estimation Method

Abstract

The invention discloses a wavelet-based battery health state estimation method. The method comprises steps that firstly, multi-resolution analysis of wavelet transform and electrochemical impedance spectrum band analysis of batteries are combined to determine the wavelet decomposition layer number, secondly, voltage signals of the batteries with different aging degrees responding to the same simulation operation condition at the same environment temperature and the same SOC are obtained, DWT-based MRA specific layer number decomposition of the voltage signals is performed, the SOH of the batteries is estimated through utilizing statistical characteristics of standard deviation of an approximate signal with required-high frequency resolution obtained through decomposition and a detail signal. Compared with a method based on data driving, the method is advantaged in that no extensive experiments are required, SOH prediction of the batteries can be performed online, prediction accuracy ishigh, the experimental life of the batteries is not influenced, compared with a model-based SOH estimation method, the algorithm is simple, and calculation efficiency is high.

Description

technical field [0001] The invention belongs to the technical field of battery health state estimation, in particular to a wavelet-based battery health state estimation method. Background technique [0002] Due to the deterioration of the environment and the shortage of energy, electric vehicles driven by motors have the advantages of saving energy, reducing waste emissions, and low noise compared with traditional fuel vehicles, and have broad development prospects. The electric vehicle battery management system is an important link connecting the on-board power battery and the electric vehicle. It can monitor the parameters of the battery in real time and estimate the state of charge (SOC) and state of health (SOH) of the battery. provide valid vehicle information. A suitable battery management system can not only give full play to the advantages of the battery, but also give the battery the best protection, thereby prolonging the service life of the battery. [0003] As ...

AI Technical Summary

Problems solved by technology

At present, there are mainly two methods to predict battery SOH at home and abroad. The first one is based on the data-driven method. Through a large number of experimental data, the relationship between battery internal characteristic parameters such as ohmic internal resistance and polarization internal resistance and SOH is fitted or passed. Black box methods such as artificial intelligence and machine learning use a large amount of experimental data for training to establish the relationship between external input parameters such as temperature, depth of discharge, and battery open circuit voltage and SOH. This type of method consumes a lot of time in experiments, and the established corresponding relationship is not universal. The second is the model-based SOH estimation method. First, the rated capacity is used instead of the actual capacity to establish an observer to estimate the SOC, and then the SOC is used to estimate the actual capacity, and this cycle is estimated. The error cumulative effect is obvious, and the algorithm is complex and the calculation efficiency is low.

[0004] To sum up, the data-driven method consumes a lot of time in experiments, and the corresponding relationship established is not universal. The error accumulation effect of the model-based SOH estimation method is obvious, and the algorithm is complex and the calculation efficiency is low.

Images