Lithium battery life cycle SOH online detection and dynamic sorting method and system

Abstract

The invention discloses a lithium battery life cycle SOH online detection and dynamic sorting method, which comprises the following steps: S1, connecting a battery into a system, and starting normal charging and discharging; s2, generating a microsecond alternating current pulse and a millisecond step current, and injecting the microsecond alternating current pulse and the millisecond step current into a battery loop through a current clamp; s3, synchronously acquiring voltage and current responses, running a Kalman filtering algorithm, and decoupling charge transfer resistance Rct and lithium ion loss amount delta QLi in real time in combination with a second-order equivalent circuit model; s4, calculating SOH according to the charge transfer resistance Rct and the lithium ion loss amount delta QLi obtained through real-time decoupling in the S3 and corresponding weights thereof, generating a JSON data packet and uploading the JSON data packet; and S5, switching the battery to the bus of the corresponding gear according to the SOH calculated in the step S4. The SOH is calculated on line, and real-time sorting is carried out according to the SOH, so that health assessment 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 environmental dependence are solved.

Description

Technical Field

[0001] This application relates to the field of lithium battery testing and sorting technology, specifically to a method and system for online detection and dynamic sorting of SOH (State of Health) throughout the entire life cycle of lithium batteries. Background Technology

[0002] With the development of electric vehicles and energy storage systems, the reuse of retired lithium batteries has become a crucial aspect. Battery health is a core indicator of battery performance, and current mainstream testing methods include complete charge-discharge testing, incremental capacity analysis, and AC impedance spectroscopy. However, these methods generally suffer from long testing times, reliance on specific operating conditions, and significant temperature sensitivity, making them unsuitable for the large-scale, rapid sorting needs of retired batteries. Summary of the Invention

[0003] The purpose of this application is to provide a method and system for online detection and dynamic sorting of state of harm (SOH) throughout the entire life cycle of lithium batteries. The specific technical solution is as follows:

[0004] A method for online detection and dynamic sorting of state of harm (SOH) throughout the life cycle of a lithium battery includes: S1, connecting the battery to the system and starting normal charging and discharging; S2, generating microsecond-level AC pulses and millisecond-level step currents, and injecting them into the battery circuit through current clamping; S3, synchronously acquiring voltage and current responses, running a Kalman filter algorithm, and combining a second-order equivalent circuit model to decouple the charge transfer resistance Rct and lithium-ion loss ΔQLi in real time; S4, calculating the SOH based on the charge transfer resistance Rct and lithium-ion loss ΔQLi obtained in real time decoupling in S3 and their corresponding weights, generating a JSON data packet and uploading it; S5, switching the battery to the corresponding bus based on the SOH calculated in S4.

[0005] S6. Based on SOH, charge transfer resistance Rct, and lithium ion loss ΔQLi, match recommended application scenarios and output a text report containing recommended scenarios and usage suggestions.

[0006] The formula for calculating SOH in S4 is:

[0007] ,

[0008] Where α and β are self-learning weights.

[0009] The JSON data packet in S4 includes: SOH, charge transfer resistance Rct and lithium-ion loss ΔQLi, confidence level, temperature and timestamp.

[0010] S7. The measured data is uploaded to the cloud in each preset period, and the α and β weights are dynamically updated based on the uploaded data to achieve closed-loop optimization.

[0011] A lithium battery lifecycle SOH online detection and dynamic sorting system is provided to implement the aforementioned lithium battery lifecycle SOH online detection and dynamic sorting method. The system includes: a dual-scale excitation generation module, whose core is a DSP or FPGA, coupled with a DAC and power amplifier, to output microsecond-level AC pulses (1–10 kHz) and millisecond-level step currents (C / 20–C / 10), which are injected into the battery circuit via current clamping; and an impedance-capacity decoupling calculation module, which is based on ARM. A Cortex-M7 microcontroller is used to run the Kalman filter algorithm, combined with a second-order equivalent circuit model, to decouple the charge transfer resistance Rct and lithium-ion loss ΔQLi in real time. A SOH detection result report generation submodule integrates data packaging, diagnostic analysis, and report generation functions, outputting JSON format data packets containing SOH, Rct, ΔQLi, confidence level, temperature, and timestamps. A balance-gated dual-MOSFET module uses two trench MOSFETs with an on-resistance ≤1mΩ to switch batteries to the corresponding bus based on SOH values, achieving dynamic sorting. A tiered utilization suggestion report generation module has a built-in rule base, matching recommended application scenarios based on SOH, charge transfer resistance Rct, and lithium-ion loss ΔQLi, while also supporting case-based reasoning decisions, outputting text reports containing recommended scenarios and usage suggestions. A cloud-based self-learning correction module receives measured capacity data and updates weights α and β through Bayesian regression with an update cycle ≤24 hours, achieving adaptive model optimization.

[0012] The beneficial effect of this application is that by calculating the State of Health (SOH) online and sorting the batteries in real time based on the SOH, the health assessment and sorting can be completed simultaneously during battery operation. The system can complete online SOH detection, real-time sorting, and generate tiered utilization suggestions within 30 seconds, solving the problems of slow testing, low efficiency, and strong environmental dependence. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the method flow of this application;

[0014] Figure 2 This is a schematic diagram of the system structure of this application;

[0015] Figure 3 This is a schematic diagram of the second-order equivalent circuit based on Kalman filtering in this application;

[0016] Figure 4 A schematic diagram of a balanced-gated dual-bit MOS module circuit. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to specific embodiments and accompanying drawings. It should be understood that these descriptions are merely exemplary and not intended to limit the scope of this application. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0018] like Figure 1 As shown, a method for online detection and dynamic sorting of state of harm (SOH) throughout the entire life cycle of lithium batteries includes:

[0019] S1. Connect the battery to the system and begin normal charging and discharging.

[0020] S2 generates microsecond-level AC pulses and millisecond-level step currents, which are then injected into the battery circuit via current clamping.

[0021] S3. Synchronously acquire voltage and current responses, run the Kalman filter algorithm, and combine it with a second-order equivalent circuit model, such as Figure 3 As shown, the real-time decoupling charge transfer resistance Rct and lithium-ion loss ΔQLi are represented.

[0022] S4. Calculate the State of Health (SOH) based on the charge transfer resistance Rct and lithium-ion loss ΔQLi obtained from real-time decoupling in S3, along with their corresponding weights. Generate a JSON data packet and upload it. The JSON data packet includes: SOH, charge transfer resistance Rct and lithium-ion loss ΔQLi, confidence level, temperature, and timestamp.

[0023] S5. Based on the SOH calculated in S4, switch the battery to the corresponding bus position. For example... Figure 4 As shown, in practical applications, it is divided into three levels: A, B, and C. When SOH is greater than 85%, level A is activated; when SOH is greater than 80% but less than or equal to 85%, level B is activated; and when SOH is less than 80%, level C is activated.

[0024] S6. Based on SOH, charge transfer resistance Rct, and lithium ion loss ΔQLi, match recommended application scenarios and output a text report containing recommended scenarios and usage suggestions.

[0025] S7. The measured data is uploaded to the cloud in each preset period, and the α and β weights are dynamically updated based on the uploaded data to achieve closed-loop optimization.

[0026] like Figure 2As shown, a lithium battery full-lifecycle SOH online detection and dynamic sorting system is used to implement the above-mentioned lithium battery full-lifecycle SOH online detection and dynamic sorting method. It includes: a dual-scale excitation generation module, the core of which is a DSP or FPGA, coupled with a DAC and power amplifier, to output microsecond-level AC pulses (1–10 kHz) and millisecond-level step currents (C / 20–C / 10), which are injected into the battery circuit through current clamping; and an impedance-capacity decoupling calculation module, which is based on ARM. A Cortex-M7 microcontroller is used to run the Kalman filter algorithm, combined with a second-order equivalent circuit model, to decouple the charge transfer resistance Rct and lithium-ion loss ΔQLi in real time. A SOH detection result report generation submodule integrates data packaging, diagnostic analysis, and report generation functions, outputting JSON format data packets containing SOH, Rct, ΔQLi, confidence level, temperature, and timestamps. A balance-gated dual-MOSFET module uses two trench MOSFETs with an on-resistance ≤1mΩ to switch batteries to the corresponding bus based on SOH values, achieving dynamic sorting. A tiered utilization suggestion report generation module has a built-in rule base, matching recommended application scenarios based on SOH, charge transfer resistance Rct, and lithium-ion loss ΔQLi, while also supporting case-based reasoning decisions, outputting text reports containing recommended scenarios and usage suggestions. A cloud-based self-learning correction module receives measured capacity data and updates weights α and β through Bayesian regression with an update cycle ≤24 hours, achieving adaptive model optimization.

[0027] Based on the above method and system, in actual testing, 200 retired NMC ternary lithium batteries were tested. Under the environment of 25±5℃, the root mean square error of SOH estimation and standard capacity test was 1.5%-1.8%, the maximum error was <2%, and the single cell processing time was ≤35 seconds.

Claims

1. A method for online detection and dynamic sorting of state of harm (SOH) throughout the entire life cycle of lithium batteries, characterized in that, include: S1. Connect the battery to the system and begin normal charging and discharging; S2. Generate microsecond-level AC pulses and millisecond-level step currents, and inject them into the battery circuit through current clamping; S3. Synchronously acquire voltage and current responses, run Kalman filtering algorithm, and combine second-order equivalent circuit model to decouple charge transfer resistance Rct and lithium ion loss ΔQLi in real time. S4. Calculate SOH based on the charge transfer resistance Rct and lithium ion loss ΔQLi obtained from real-time decoupling in S3 and their corresponding weights, generate a JSON data packet and upload it. S5. Based on the SOH calculated in S4, switch the battery to the bus of the corresponding gear.

2. The method for online detection and dynamic sorting of state of harm (SOH) throughout the entire life cycle of lithium batteries as described in claim 1, characterized in that, Also includes: S6. Based on SOH, charge transfer resistance Rct, and lithium ion loss ΔQLi, match recommended application scenarios and output a text report containing recommended scenarios and usage suggestions.

3. The online detection and dynamic sorting method for SOH (State of Harm) throughout the entire life cycle of lithium batteries as described in claim 2, characterized in that, The formula for calculating SOH in S4 is: ， Where α and β are self-learning weights.

4. The lithium battery full-lifecycle SOH online detection and dynamic sorting method as described in claim 3, characterized in that, The JSON data packet in S4 includes: SOH, charge transfer resistance Rct and lithium ion loss ΔQLi, confidence level, temperature and timestamp.

5. The online detection and dynamic sorting method for SOH throughout the entire life cycle of lithium batteries as described in claim 4, characterized in that, Also includes: S7. The measured data is uploaded to the cloud in each preset period, and the α and β weights are dynamically updated based on the uploaded data to achieve closed-loop optimization.

6. A lithium battery full-lifecycle SOH online detection and dynamic sorting system, used to implement the lithium battery full-lifecycle SOH online detection and dynamic sorting method as described in any one of claims 1-5, characterized in that, include: A dual-scale excitation generation module, the core of which is a DSP or FPGA, in conjunction with a DAC and a power amplifier, is used to output microsecond-level AC pulses (1–10 kHz) and millisecond-level step currents (C / 20–C / 10), which are injected into the battery circuit through current clamping; Impedance-capacitance decoupling operation module, which is based on ARM Cortex-M7 microcontroller, is used to run Kalman filter algorithm, combined with second-order equivalent circuit model, to decouple charge transfer resistance Rct and lithium ion loss ΔQLi in real time; The SOH test result report generation submodule integrates data packaging, diagnostic analysis, and report generation functions, and is used to output JSON format data packets containing SOH, Rct, ΔQLi, confidence level, temperature, and timestamp. The balanced-gating dual-position MOS module uses two trench MOSFETs with an on-resistance of ≤1mΩ. It is used to switch the battery to the corresponding bus based on the SOH value to achieve dynamic sorting. The tiered utilization suggestion report generation module, with its built-in rule base, matches recommended application scenarios based on SOH, charge transfer resistance Rct, and lithium ion loss ΔQLi, and also supports case-based reasoning decision-making, to output a text report containing recommended scenarios and usage suggestions; The cloud-based self-learning correction module receives measured capacity data and updates the weights α and β through Bayesian regression with an update cycle of ≤24 hours to achieve adaptive optimization of the model.

Images