CN-121978571-A - Lithium battery health state estimation method based on electrochemical impedance spectrum characteristic peak tracking

CN121978571ACN 121978571 ACN121978571 ACN 121978571ACN-121978571-A

Abstract

The invention relates to the technical field of lithium battery health management, in particular to a lithium battery health state estimation method based on electrochemical impedance spectrum characteristic peak tracking, which comprises the following steps: collecting battery voltage, current, temperature and SOC data, screening the working conditions to obtain local analysis data, constructing local electrochemical impedance spectrum through short-time frequency domain conversion, obtaining characteristic peak distribution through self-adaptive peak enhancement, constructing a limited search interval by combining temperature, SOC and historical peak positions, identifying and tracking characteristic peaks in the interval, correcting parameter sequences under abnormal working conditions such as drift, overlapping, disappearance and the like, extracting health characteristics such as peak positions, amplitude values, peak widths and the like, and calculating the SOH of the lithium battery according to a preset mapping relation. According to the invention, the local electrochemical impedance spectrum is constructed through working condition screening, the health characteristics are extracted and SOH is estimated through self-adaptive peak enhancement and characteristic peak identification tracking, and the problems of single parameter dependence, poor working condition adaptability and insufficient estimation stability of the traditional method are solved.

Inventors

MA MINGHUI
LOU TAISHAN
SHI LEI
QI RENLONG
MA YONGLI
WANG FEI
ZHU XIAOHUI

Assignees

精规智(河南)智能装备有限公司
郑州科技学院

Dates

Publication Date: 20260505
Application Date: 20260330

Claims (10)

1. The lithium battery health state estimation method based on electrochemical impedance spectrum characteristic peak tracking is characterized by comprising the following steps of: S1, acquiring operation data including voltage data, current data, temperature data and state of charge (SOC) data in the operation process of a lithium battery, and screening the acquired operation data according to a preset working condition screening rule to obtain local analysis data meeting screening conditions; S2, performing short-time frequency domain conversion on the voltage data and the current data in the local analysis data, calculating frequency domain impedance and constructing local electrochemical impedance spectrum; s3, performing self-adaptive peak enhancement conversion on the local electrochemical impedance spectrum to obtain characteristic peak distribution data; s4, combining the temperature data, the SOC data and the historical characteristic peak position data, determining target characteristic peak position prediction data and constructing a limited search interval of a target characteristic peak; s5, identifying a target characteristic peak in a limited search interval, extracting a target characteristic peak parameter and constructing a target characteristic peak parameter time sequence, and aiming at parameter jump or deletion of the target characteristic peak parameter time sequence under the abnormal working conditions of drift, overlap and disappearance, performing continuous tracking and correction processing based on time sequence correlation on the target characteristic peak parameter time sequence; s6, extracting health features from the corrected target feature peak parameter time sequence, and calculating an SOH estimation result of the health state of the lithium battery according to a preset mapping relation between the health features and the SOH.
2. The method for estimating the health state of a lithium battery based on electrochemical impedance spectrum characteristic peak tracking according to claim 1, wherein in S1, the step of screening the collected operation data according to a preset working condition screening rule includes: Dividing the operation data into a plurality of time windows according to a preset time length; Calculating the change rate of the current data, the change rate of the voltage data, the change amount of the temperature data and the change amount of the state of charge (SOC) data in each time window respectively; Comparing the current data change rate with a preset current change rate threshold, comparing the voltage data change rate with a preset voltage change rate threshold, comparing the temperature data change amount with a preset temperature change amount threshold, and comparing the state of charge (SOC) data change amount with a preset SOC change amount threshold; and when the current data change rate is smaller than a preset current change rate threshold value, the voltage data change rate is smaller than a preset voltage change rate threshold value, the temperature data change amount is smaller than a preset temperature change amount threshold value and the state of charge (SOC) data change amount is smaller than a preset SOC change amount threshold value, determining the operation data in the corresponding time window as local analysis data.
3. The method for estimating a state of health of a lithium battery based on peak tracking of electrochemical impedance spectrum characteristics according to claim 1, wherein in S2, the step of performing short-time frequency domain conversion on voltage data and current data in the local analysis data comprises: Dividing voltage data and current data in the local analysis data into a plurality of time windows according to a preset time length; Performing window function weighting processing on the voltage data and the current data in each time window respectively; Performing frequency domain transformation on the weighted voltage data and the weighted current data respectively to obtain a voltage data frequency domain representation and a current data frequency domain representation of a corresponding time window; and arranging the voltage data frequency domain representation and the current data frequency domain representation obtained by each time window according to a time sequence to form a short-time frequency domain conversion result of the voltage data and the current data.
4. The method for estimating the state of health of a lithium battery based on peak tracking of electrochemical impedance spectrum characteristics according to claim 3, wherein in S2, the step of calculating the frequency domain impedance and constructing the local electrochemical impedance spectrum comprises: calculating impedance values of corresponding frequency points according to the voltage data frequency domain representation and the current data frequency domain representation; decomposing the impedance value to obtain an impedance real part and an impedance imaginary part; combining the real part and the imaginary part of the impedance with the corresponding frequency values to form an impedance frequency sequence; A local electrochemical impedance spectrum is constructed from the impedance frequency sequence.
5. The method of estimating a state of health of a lithium battery based on peak tracking of electrochemical impedance spectra as set forth in claim 4, wherein in S3 the step of performing adaptive peak enhancement transformation on the local electrochemical impedance spectra comprises: extracting an impedance amplitude sequence from an impedance frequency sequence of the local electrochemical impedance spectrum, and performing smoothing on the impedance amplitude sequence; Calculating the change rate and the change rate difference between adjacent frequency points according to the smoothed impedance amplitude sequence to obtain a peak response sequence; Calculating weight coefficients corresponding to all frequency points according to the impedance amplitude sequences; And carrying out combination calculation according to the peak response sequence and the weight coefficient to obtain an enhanced peak response sequence, and generating characteristic peak distribution data according to the enhanced peak response sequence.
6. The method for estimating a state of health of a lithium battery based on characteristic peak tracking of electrochemical impedance spectrum according to claim 1, wherein in S3, the adaptive peak enhancement transform enhances characteristic peak response intensities related to charge transfer process and diffusion process in electrochemical impedance spectrum by a second-order differential response based on local curvature weighting.
7. The method for estimating the state of health of a lithium battery based on electrochemical impedance spectrum characteristic peak tracking according to claim 1, wherein in S4, the step of determining target characteristic peak position prediction data and constructing a limited search interval of target characteristic peaks comprises: The historical characteristic peak position data are arranged according to the time sequence to obtain a historical characteristic peak position sequence; Determining a historical characteristic peak position data subset of a corresponding working condition interval in a historical characteristic peak position sequence according to the temperature data and the state of charge (SOC) data; Calculating a peak position change sequence among historical characteristic peak positions according to the historical characteristic peak position data subset, and calculating target characteristic peak position prediction data by combining temperature data and state of charge (SOC) data; and determining a peak position deviation range according to the historical characteristic peak position data and the target characteristic peak position prediction data, and determining a limited search interval of the target characteristic peak according to the target characteristic peak position prediction data and the peak position deviation range.
8. The method for estimating a state of health of a lithium battery based on electrochemical impedance spectrum feature peak tracking according to claim 7, wherein in S5, the step of identifying a target feature peak within a limited search interval comprises: Extracting characteristic peak distribution data of a corresponding frequency interval from the characteristic peak distribution data according to the limited search interval of the target characteristic peak and forming a characteristic peak distribution data sequence; Traversing each data point in the characteristic peak distribution data sequence according to the frequency sequence, and determining the current data point as a candidate characteristic peak when the value of the current data point is larger than that of the adjacent data point; extracting and screening candidate characteristic peak parameters according to frequency values and peak amplitudes corresponding to the candidate characteristic peaks, and determining target characteristic peak parameters; And arranging the target characteristic peak parameters obtained at each time point according to a time sequence, and constructing a target characteristic peak parameter time sequence.
9. The method for estimating the state of health of a lithium battery based on electrochemical impedance spectrum characteristic peak tracking according to claim 8, wherein in S5, the step of performing continuous tracking processing and correction on the target characteristic peak parameter time series comprises: Traversing adjacent time points in the time sequence of the target characteristic peak parameters according to the time sequence, and calculating the peak position change rate in the target characteristic peak parameters corresponding to the adjacent time points; when the peak position change rate exceeds a preset change range, marking the corresponding time point as a peak position drift state; When the target characteristic peak parameter is not detected in the limited search interval of the target characteristic peak, marking the corresponding time point as a peak vanishing state; when a plurality of candidate characteristic peak parameters are detected at the same time point, screening and determining target characteristic peak parameters according to the time sequence correlation of each candidate characteristic peak parameter and the historical characteristic peak position, and marking the corresponding time point as a peak overlapping state; And according to the target characteristic peak parameters of the adjacent time points, performing interpolation correction or neighborhood smoothing processing on the time points marked as peak position drift states or peak disappearance states.
10. The method for estimating a state of health of a lithium battery based on peak tracking of electrochemical impedance spectrum characteristics according to claim 1, wherein in S6, the health characteristics include at least one of a target characteristic peak-to-peak position variation, a target characteristic peak-to-peak amplitude variation rate, and a target characteristic peak-to-peak width variation.

Description

Lithium battery health state estimation method based on electrochemical impedance spectrum characteristic peak tracking Technical Field The invention relates to the technical field of lithium battery health management, in particular to a lithium battery health state estimation method based on electrochemical impedance spectrum characteristic peak tracking. Background With the rapid development of new energy automobiles, energy storage systems and portable electronic devices, lithium ion batteries are widely used as the main electrochemical energy storage units. The aging phenomena such as capacity decay and internal resistance increase of the battery can occur in the long-term charge and discharge process, so that the accurate assessment of the state of health of the battery becomes one of key technologies in a battery management system. The traditional lithium battery health state estimation method mostly adopts a method for carrying out state estimation based on a single operation parameter or static impedance characteristic. Such methods typically assume that the battery is operating in a steady state condition, and the state of health is inferred through an offline calibrated impedance spectrum or empirical model. However, in actual dynamic operating conditions, the current, temperature and state of charge of the battery continuously change, resulting in drift, overlap and even transient disappearance of characteristic peaks of the electrochemical impedance spectrum with operating conditions. The traditional full frequency domain peak detection method is difficult to distinguish a noise pseudo peak from a real characteristic peak due to the fact that the peak is searched in a full frequency range, and a fixed search interval cannot adapt to dynamic changes of the characteristic peak, so that time sequence consistency of characteristic extraction is poor, an SOH estimation result jumps, and the requirement of a battery management system on estimation stability cannot be met. Therefore, how to realize stable tracking of impedance spectrum characteristic peaks under complex dynamic working conditions becomes a technical problem to be solved in lithium battery health state estimation. Disclosure of Invention In order to overcome the defects, the invention provides a lithium battery health state estimation method based on electrochemical impedance spectrum characteristic peak tracking, which solves the problems of unstable characteristic peak identification and low health state estimation precision of the traditional method under dynamic working conditions through working condition screening, self-adaptive peak enhancement, limited search interval construction and time sequence tracking correction. The invention provides a lithium battery health state estimation method based on electrochemical impedance spectrum characteristic peak tracking, which comprises the following steps: S1, acquiring operation data including voltage data, current data, temperature data and State of Charge (SOC) data in the operation process of a lithium battery, and screening the acquired operation data according to a preset working condition screening rule to obtain local analysis data meeting screening conditions; S2, performing short-time frequency domain conversion on the voltage data and the current data in the local analysis data, calculating frequency domain impedance and constructing local electrochemical impedance spectrum; s3, performing self-adaptive peak enhancement conversion on the local electrochemical impedance spectrum to obtain characteristic peak distribution data; s4, combining the temperature data, the SOC data and the historical characteristic peak position data, determining target characteristic peak position prediction data and constructing a limited search interval of a target characteristic peak; s5, identifying a target characteristic peak in a limited search interval, extracting a target characteristic peak parameter and constructing a target characteristic peak parameter time sequence, and aiming at parameter jump or deletion of the target characteristic peak parameter time sequence under the abnormal working conditions of drift, overlap and disappearance, performing continuous tracking and correction processing based on time sequence correlation on the target characteristic peak parameter time sequence; S6, extracting Health features from the corrected target feature peak parameter time sequence, and calculating an SOH estimation result of the Health State of the lithium battery according to a mapping relation between the preset Health features and SOH (State of Health), namely the Health State. According to the technical scheme, the working condition screening is carried out on the lithium battery operation data, the local electrochemical impedance spectrum is constructed, the self-adaptive peak enhancement conversion is carried out on the local electrochemical impedance spectrum, the target characteris