CN-122017577-A - Online detection and dynamic sorting method and system for full life cycle SOH of lithium battery

CN122017577ACN 122017577 ACN122017577 ACN 122017577ACN-122017577-A

Abstract

The application discloses an on-line detection and dynamic sorting method for a full life cycle SOH of a lithium battery, which comprises the steps of S1, connecting the battery into a system, starting normal charge and discharge, S2, generating microsecond alternating current pulses and millisecond step currents, injecting the microsecond alternating current pulses and millisecond step currents into a battery loop through a current clamp, S3, synchronously collecting voltage and current responses, operating a Kalman filtering algorithm, combining a second-order equivalent circuit model, decoupling a charge transfer resistor Rct and a lithium ion loss quantity delta QLi in real time, S4, calculating SOH according to the charge transfer resistor Rct and the lithium ion loss quantity delta QLi obtained by real-time decoupling in S3 and corresponding weights thereof, generating a JSON data packet, uploading the JSON data packet, and S5, switching the battery to a bus of a corresponding gear according to the SOH calculated in S4. By calculating SOH on line and sorting according to SOH in real time, health evaluation and sorting are synchronously completed in the battery operation process, SOH on-line detection and real-time sorting are completed within 30 seconds, and the problems of slow test, low efficiency and strong environment dependence are solved.

Inventors

LI JINYAN
ZHANG WEI
CHEN CHAOHUI
ZHAO YIQUN
LI ZEHONG
CHEN GUISHENG
XU YONGZI

Assignees

昆明理工大学
昆明冶金高等专科学校

Dates

Publication Date: 20260512
Application Date: 20260119

Claims (6)

1. The on-line detection and dynamic sorting method for the full life cycle SOH of the lithium battery is characterized by comprising the following steps of: s1, connecting a battery into a system to start normal charge and discharge; s2, generating microsecond alternating current pulses and millisecond step currents, and injecting the microsecond alternating current pulses and millisecond step currents into a battery loop through a current clamp; s3, synchronously acquiring voltage and current responses, running a Kalman filtering algorithm, and combining a second-order equivalent circuit model to decouple the charge transfer resistor Rct and the lithium ion loss delta QLi in real time; S4, calculating SOH according to the charge transfer resistor Rct and the lithium ion loss delta QLi obtained by real-time decoupling in the S3 and the corresponding weight thereof, generating and uploading a JSON data packet; And S5, switching the battery to a bus of a corresponding gear according to the SOH calculated in the S4.
2. The method for on-line detection and dynamic sorting of full life cycle SOH of lithium battery of claim 1, further comprising: And S6, matching a recommended application scene according to the SOH, the charge transfer resistor Rct and the lithium ion loss quantity delta QLi, and outputting a text report containing the recommended scene and the use suggestion.
3. The method for online detection and dynamic sorting of SOH in a lithium battery life cycle according to claim 2, wherein the SOH calculation formula in S4 is as follows: , Wherein alpha and beta are self-learning weights.
4. The method for on-line detection and dynamic sorting of full life cycle SOH of lithium battery as claimed in claim 3, wherein the JSON data packet in S4 comprises SOH, charge transfer resistor Rct and lithium ion loss delta QLi, confidence, temperature and time stamp.
5. The method for on-line detection and dynamic sorting of full life cycle SOH of lithium battery according to claim 4, further comprising: and S7, uploading the measured data to the cloud end in each preset period, and dynamically updating alpha and beta weights according to the uploaded data to realize closed-loop optimization.
6. An on-line detection and dynamic sorting system for full life cycle SOH of a lithium battery, for implementing the on-line detection and dynamic sorting method for full life cycle SOH of a lithium battery according to any of claims 1-5, comprising: The dual-scale excitation generation module is characterized by comprising a DSP or FPGA as a core, a DAC and a power amplifier, wherein the dual-scale excitation generation module is used for outputting microsecond alternating current pulses (1-10 kHz) and millisecond step currents (C/20-C/10) and injecting the microsecond alternating current pulses and the millisecond step currents into a battery loop through a current clamp; The impedance-capacity decoupling operation module is based on an ARM Cortex-M7 microcontroller and is used for running a Kalman filtering algorithm and real-time decoupling the charge transfer resistor Rct and the lithium ion loss delta QLi by combining a second-order equivalent circuit model; The SOH detection result report generation submodule integrates data packaging, diagnosis analysis and report generation functions and is used for outputting a JSON format data packet, wherein the JSON format data packet comprises SOH, rct, delta QLi, confidence level, temperature and time stamp; the balance-gating double-bit MOS module adopts two groove MOSFETs, the on resistance is less than or equal to 1mΩ, and the balance-gating double-bit MOS module is used for switching the battery to the corresponding gear bus according to the SOH value to realize dynamic sorting; The echelon utilization suggestion report generation module is used for matching a recommended application scene according to SOH, a charge transfer resistor Rct and a lithium ion loss quantity delta QLi, supporting case reasoning decisions and outputting a text report containing the recommended scene and the use suggestion; the cloud self-learning correction module is used for receiving the actual measurement capacity data, updating weights alpha and beta through Bayesian regression, and enabling the updating period to be less than or equal to 24 hours, so that model self-adaptive optimization is realized.

Description

Online detection and dynamic sorting method and system for full life cycle SOH of lithium battery Technical Field The application relates to the technical field of lithium battery detection and sorting, in particular to a method and a system for on-line detection and dynamic sorting of full life cycle SOH of a lithium battery. Background With the development of electric vehicles and energy storage systems, the echelon utilization of lithium batteries after retirement becomes a key link. The health state is a core index for measuring the performance of the battery, and the current mainstream detection method comprises a complete charge and discharge test, an incremental capacity analysis method, an alternating current impedance spectroscopy method and the like. The problems of long testing time, dependence on specific working conditions, large influence by temperature and the like generally exist in the methods, and the requirements of large-scale rapid separation of retired batteries are difficult to meet. Disclosure of Invention The application aims to provide a method and a system for on-line detection and dynamic sorting of full life cycle SOH of a lithium battery, and the specific technical scheme is as follows: A method for detecting and dynamically sorting the full life cycle of lithium battery on line includes such steps as S1, connecting the battery to system, generating microsecond AC pulse and millisecond step current, injecting them into battery loop, S3, synchronously collecting voltage and current response, running Kalman filtering algorithm, real-time decoupling the charge transfer resistor Rct and lithium ion loss delta QLi, S4, calculating SOH according to the charge transfer resistor Rct and lithium ion loss delta QLi obtained by real-time decoupling in S3, generating JSON data packet, and uploading, and S5, switching the battery to bus with corresponding gear according to SOH calculated in S4. And S6, matching a recommended application scene according to the SOH, the charge transfer resistor Rct and the lithium ion loss quantity delta QLi, and outputting a text report containing the recommended scene and the use suggestion. The calculation formula of SOH in S4 is: , Wherein alpha and beta are self-learning weights. The JSON packet in S4 includes SOH, charge transfer resistor Rct and lithium ion loss delta QLi, confidence, temperature and time stamp. And S7, uploading the measured data to the cloud end in each preset period, and dynamically updating alpha and beta weights according to the uploaded data to realize closed-loop optimization. The on-line detection and dynamic sorting system for the full life cycle SOH of the lithium battery is used for realizing the on-line detection and dynamic sorting method for the full life cycle SOH of the lithium battery and comprises a double-scale excitation generating module, a power amplifier and a power amplifier, wherein the core of the double-scale excitation generating module is a DSP (digital signal processor) or an FPGA (field programmable gate array), the double-scale excitation generating module is matched with the DAC and the power amplifier and is used for outputting microsecond alternating current pulses (1-10 kHz) and millisecond step currents (C/20-C/10), and the microsecond alternating current pulses and the millisecond step currents are injected into a battery loop through a current clamp; the system comprises an impedance-capacity decoupling operation module, a balance-gating double-bit MOS module, a gradient-utilization advice report generation module, a built-in rule base, a self-learning correction module and a Bayesian updating module, wherein the impedance-capacity decoupling operation module is based on an ARM Cortex-M7 microcontroller and is used for running a Kalman filtering algorithm, combining a second-order equivalent circuit model, decoupling a charge transfer resistor Rct and a lithium ion loss amount delta QLi in real time, the SOH detection result report generation sub-module is used for integrating data packing, diagnosis analysis and report generation functions and outputting a JSON format data packet and comprises SOH, rct, delta QLi, confidence level, temperature and time stamp, the balance-gating double-bit MOS module adopts two trench MOSFETs, on resistance is less than or equal to 1mΩ, the battery is switched to a corresponding gear bus according to SOH value, the gradient-utilization advice report generation module is used for realizing dynamic sorting, the cloud advice report generation module is used for matching recommended application scenes according to SOH, the charge transfer resistor Rct and the lithium ion loss amount delta QLi, the self-learning correction module is used for outputting a text report comprising recommendation scenes and using advice, the self-learning correction module is used for receiving the self-learning correction module to update capacity weight data, updating the self-learning co