A three-electrode-based sodium-ion battery state of health monitoring and repairing method

Abstract

The present application relates to the technical field of sodium ion battery, and particularly relates to a three-electrode-based sodium ion battery health state monitoring and repairing method. The method comprises the following steps: implanting a reference electrode in a sodium ion battery cell and performing a charge-discharge cycle test; collecting SOH values and various characteristic parameters of the positive electrode and the negative electrode in each cycle; forming a positive electrode training set and a negative electrode training set by using the SOH values and various characteristic parameters of the positive electrode and the negative electrode in a preset cycle number; training a prediction model by using the positive electrode training set and the negative electrode training set to obtain a positive electrode SOH prediction model and a negative electrode SOH prediction model; predicting a predicted positive electrode SOH value and a predicted negative electrode SOH value in each cycle by using the positive electrode SOH prediction model and the negative electrode SOH prediction model; and monitoring and repairing the health state of the sodium ion battery based on the predicted negative electrode SOH value and a grading strategy. The present application can effectively monitor and repair the health of the sodium ion battery.

Description

Technical Field

[0001] This invention relates to the field of sodium-ion battery technology, and specifically to a method for monitoring and repairing the health status of sodium-ion batteries based on a three-electrode system. Background Technology

[0002] Sodium-ion batteries, as an emerging energy storage technology, have shown broad application prospects in large-scale energy storage systems, electric vehicles, and portable electronic devices due to their advantages such as abundant resources, low cost, and environmental friendliness. Currently, the State of Health (SOH) estimation of sodium-ion batteries is mainly based on the following three technical routes: (1) The SOH estimation method based on the traditional two-electrode system mainly involves monitoring external parameters such as the total voltage, current and temperature of the battery pack, and indirectly estimating using an equivalent circuit model or a data-driven model. For example, CRRC Zhuzhou Electric Locomotive Co., Ltd. disclosed a "method, equipment and storage medium for predicting the health status of sodium-ion batteries" (publication number CN119087274 A). By setting the prediction time start point, the measured state parameters of the sodium-ion battery at the prediction time start point are obtained, and then the measured state parameters are input into the exponential decay model for prediction. In addition, there are other schemes that obtain reconstructed characterization data, construct the optimal separation hyperplane, and use a continuous scaling prediction method to separate and predict the energy feature set in order to generate the prediction relationship between the battery parameter model and the SOH prediction.

[0003] (2) Battery state monitoring methods based on three electrodes: In existing technologies, the three-electrode system has been used for battery state monitoring. For example, a patent discloses a method for estimating the health state of a lithium-ion battery, which uses the three-electrode principle to monitor the potential and potential changes between the positive electrode and the lithium sheet, the negative electrode and the lithium sheet, and the positive and negative electrodes during charging. After normalization, normalized voltage curves of the positive electrode, the negative electrode, and the whole cell are obtained, and the health state of each electrode of the battery is estimated by using the voltage curve fitting method. Another patent discloses a method for analyzing the electrode degradation of commercial lithium-ion batteries based on a three-electrode system. A three-electrode battery is fabricated based on a three-electrode device, and cycle performance and electrochemical impedance spectroscopy tests are performed, while the positive and negative electrode voltages and impedances of the battery are monitored simultaneously.

[0004] In addition, there is a method for monitoring the health status of lithium-ion batteries during fast charging cycles. This method involves placing a three-electrode lithium-ion battery at a preset constant temperature and charging and discharging it at a set charge and discharge rate. The lithium plating status of the battery is determined based on the trend of the change in the negative electrode's potential relative to the reference electrode during the charging process.

[0005] (3) Battery repair method based on three electrodes: In the field of battery repair, existing technologies have proposed the concept of three-electrode repairable lithium-ion batteries, that is, implanting a third electrode during the manufacturing process of lithium-ion batteries to achieve the effect of continuously replenishing the active lithium lost by the battery during cycling. Zhuhai Guanyu Battery Co., Ltd. has also disclosed "Reference Electrode Repair Method, Apparatus, Equipment and Storage Medium" (Publication No. CN117538395A), which determines whether the reference electrode is aged by obtaining the reaction resistance of the three-electrode lithium-ion battery. If it is aged, a cyclic lithium plating operation is performed on a single electrode.

[0006] While the aforementioned existing technologies have achieved battery health monitoring or repair to some extent, they still have the following significant shortcomings: The existing two-electrode method can only obtain data on the overall performance degradation of the battery. Although the existing three-electrode method can obtain the potential information of the positive and negative electrodes separately, it has not yet established an independent SOH prediction model for the positive and negative electrodes based on real-time three-electrode data, and therefore cannot accurately locate the cause of aging.

[0007] Prediction and regulation are disconnected, and closed-loop intelligent management is lacking: existing health status prediction methods only provide status information output and cannot automatically trigger regulation strategies based on prediction results; while existing repair methods are mostly reactive after the fact, failing to form a closed-loop management system of "monitoring → prediction → decision-making → regulation".

[0008] Difficulty in dealing with interference from battery side reactions: Due to the presence of side reactions during the charging and discharging process, the three-electrode method cannot accurately detect the amount of sodium deposited in the battery, and therefore cannot accurately predict the battery's health.

[0009] Lack of a graded adaptive control mechanism: Existing technologies lack graded and differentiated response strategies for battery degradation, and cannot take graded intervention measures from early warning prompts to active control and then to self-repair triggers according to the degree of degradation. Summary of the Invention

[0010] To address the aforementioned technical problems, the present invention aims to provide a method for monitoring and repairing the health status of sodium-ion batteries based on a three-electrode system. The specific technical solution adopted is as follows: One embodiment of the present invention provides a method for monitoring and repairing the health status of a sodium-ion battery based on a three-electrode system, the method comprising: A reference electrode was implanted inside the sodium-ion battery cell and charge-discharge cycle tests were conducted. The SOH values ​​of the positive and negative electrodes and various characteristic parameters were collected for each cycle. The SOH values ​​of the positive and negative electrodes and various characteristic parameters for a preset number of cycles were used to form a positive electrode training set and a negative electrode training set, respectively. The prediction model was trained using the positive electrode training set and the negative electrode training set respectively to obtain the positive electrode SOH prediction model and the negative electrode SOH prediction model; the positive electrode SOH prediction model and the negative electrode SOH prediction model were used to predict the predicted positive electrode SOH value and the predicted negative electrode SOH value for each cycle. Based on the predicted negative electrode SOH value, a tiered strategy is developed to monitor and repair the health status of sodium-ion batteries.

[0011] Preferably, the SOH values ​​of the positive and negative electrodes and various characteristic parameters are collected for each cycle, including: The positive and negative independent impedances after a preset number of cycles are measured using EIS technology. For other number of cycles, the positive and negative independent impedances are obtained using a linear interpolation algorithm. The positive independent impedance after one cycle is subtracted from the initial positive independent impedance and divided by the number of cycles to obtain the growth rate of the positive independent impedance in that cycle. Similarly, the growth rate of the negative independent impedance in that cycle can be obtained. The rate of change of the positive electrode potential at the end of the charging cycle is obtained by using the derivative of the potential of the positive electrode relative to the reference electrode with respect to time at the end of the charging process in one cycle; the rate of change of the negative electrode potential at the end of the discharging cycle is obtained by using the derivative of the potential of the negative electrode relative to the reference electrode with respect to time at the end of the discharging process in one cycle. The coulombic efficiency of a sodium-ion battery in one cycle is obtained by dividing the discharge capacity by the charge capacity; the discharge capacity of a sodium-ion battery in one cycle is obtained by integrating the discharge current over time. The growth rate of the positive independent impedance, the growth rate of the negative independent impedance, the rate of change of the positive electrode potential at the end of charging, the rate of change of the negative electrode potential at the end of discharging, the coulombic efficiency, and the discharge capacity are the various characteristic parameters of this cycle.

[0012] Preferably, the SOH values ​​of the positive and negative electrodes for a preset number of cycles, along with various characteristic parameters, are used to form the positive electrode training set and the negative electrode training set, respectively, including: The various feature parameters are combined into a matrix after 10 consecutive cycles, and this matrix is ​​used as a sample. For a sample, the SOH value of the positive electrode after a preset number of cycles after 10 consecutive cycles is used as the positive electrode label of the sample, and the SOH value of the negative electrode after a preset number of cycles after 10 consecutive cycles is used as the negative electrode label of the sample. The positive electrode training set is formed using samples with positive electrode labels, and the negative electrode training set is formed using samples with negative electrode labels.

[0013] Preferably, the predicted positive electrode SOH value and the predicted negative electrode SOH value for each cycle are predicted using the positive electrode SOH prediction model and the negative electrode SOH prediction model, including: Using a window of preset length, the model slides along the loop count with a step size of one loop. The feature parameters of each loop within a window are combined into a matrix, which serves as the input for that window. The input corresponding to the current window is then input into the positive electrode SOH prediction model and the negative electrode SOH prediction model, respectively, and the predicted positive electrode SOH value and the predicted negative electrode SOH value corresponding to the current window are output.

[0014] Preferably, the predicted positive electrode SOH value and the predicted negative electrode SOH value for each cycle are predicted using the positive electrode SOH prediction model and the negative electrode SOH prediction model, including: Using a window of preset length, the model slides along the loop count with a step size of one loop. The feature parameters of each loop within a window are combined into a matrix, which serves as the input for that window. The input corresponding to the current window is then input into the positive electrode SOH prediction model and the negative electrode SOH prediction model, respectively, and the predicted positive electrode SOH value and the predicted negative electrode SOH value corresponding to the current window are output.

[0015] Preferably, a graded strategy is formulated based on the predicted negative electrode SOH value to monitor and repair the health status of the sodium-ion battery, including: The first-level response is that if the predicted positive electrode SOH value and the predicted negative electrode SOH value corresponding to the current window are less than the first preset threshold, the BMS will issue a health status warning signal. The secondary response is as follows: if the predicted negative electrode SOH value corresponding to the next window of the current window is less than the second preset threshold after the primary response is triggered, and the negative electrode potential at the end of the charging measured by the BMS in the current actual cycle is lower than the preset safe potential threshold, then the BMS will automatically switch the charging strategy from 1C constant current charging to 1C pulse charging. After the Level 2 response is executed, the BMS will continuously monitor and process the data according to the following logic: Diversion ①: If the predicted negative electrode SOH value corresponding to the window obtained by subsequent prediction is greater than the second preset threshold, and the actual measured negative electrode potential at the end of charging is no longer lower than the safe potential threshold, then the BMS will automatically revert to the original 1C constant current charging strategy, exit the secondary response state, and return to the normal monitoring mode. Diversion ②: If the predicted negative electrode SOH value is within the stalemate range after 100 consecutive cycles, the BMS will automatically adjust the pulse charging parameters. If the predicted negative electrode SOH value is within the stalemate range after another 50 cycles, it will be judged as a control failure and forced to enter the third-level response. Diversion ③: If, during the execution of the secondary response, the predicted negative electrode SOH value continues to decrease, and the sum of the decreases of the three consecutive predicted negative electrode SOH values ​​is greater than or equal to 1%, then the tertiary response is triggered, and the BMS immediately enters the tertiary response. The Level 3 response is as follows: when the control failure condition in shunting ③ or shunting ② is met, the BMS applies a square wave AC current to the third electrode and the negative electrode for a preset time to perform repair. If the predicted negative electrode SOH values ​​obtained after repair are greater than or equal to the first preset threshold, the repair is successful. If there is a predicted negative electrode SOH value less than the first preset threshold among the predicted negative electrode SOH values ​​obtained after repair, the Level 3 response is repeated, up to a maximum of 2 times. If it is still ineffective after 2 repetitions, an alarm is issued indicating self-repair failure and suggesting replacement of the battery cell.

[0016] The embodiments of the present invention have at least the following beneficial effects: This application implants a reference electrode inside a sodium-ion battery cell and performs charge-discharge cycle tests, collecting the SOH values ​​of the positive and negative electrodes and various characteristic parameters for each cycle. The prediction model is trained using the positive and negative electrode training sets respectively to obtain the positive electrode SOH prediction model and the negative electrode SOH prediction model. By separating the aging of the positive and negative electrodes and modeling them separately, the SOH estimation accuracy is much higher than that of the traditional two-electrode method, which can accurately locate the main cause of performance degradation (positive electrode material degradation or negative electrode sodium deposition), providing a basis for targeted optimization; Based on the predicted negative electrode SOH value, a graded strategy is formulated to monitor and repair the health status of sodium-ion batteries. Through graded regulation and self-repair strategies, battery aging can be effectively delayed and its effective service life extended. A closed-loop management system of "monitoring → prediction → decision-making → regulation" is constructed, providing core algorithms and hardware foundation for the next generation of intelligent and autonomous battery management systems. Attached Figure Description

[0017] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart illustrating a method for monitoring and repairing the health status of a sodium-ion battery based on a three-electrode system, as provided in this application embodiment; Figure 2 This is a schematic diagram of the three electrodes of a sodium-ion battery, which is provided as an embodiment of the present application for a method for monitoring and repairing the health status of a sodium-ion battery based on a three-electrode system. Detailed Implementation

[0019] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a three-electrode-based sodium-ion battery health status monitoring and repair method proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0021] The following description, in conjunction with the accompanying drawings, details a specific scheme for a three-electrode-based sodium-ion battery health status monitoring and repair method provided by the present invention.

[0022] Example: The main application scenario of this invention is to monitor the health status of sodium-ion batteries and repair them according to their status.

[0023] Please see Figure 1 The diagram illustrates a flowchart of a method for monitoring and repairing the health status of a sodium-ion battery based on a three-electrode system, provided by an embodiment of the present invention. The method includes the following steps: Step S1: Implant a reference electrode inside the sodium-ion battery cell and perform charge-discharge cycle tests; collect the SOH values ​​of the positive and negative electrodes and various characteristic parameters for each cycle; and form a positive electrode training set and a negative electrode training set by combining the SOH values ​​of the positive and negative electrodes and various characteristic parameters for a preset number of cycles.

[0024] First, a reference electrode needs to be implanted inside the sodium-ion battery cell to construct a three-electrode system. Specifically, during the cell stacking process, a 50μm diameter copper wire is implanted between the separators as the reference electrode. One end of the copper wire is wrapped with the separator material to ensure that it does not contact the positive and negative electrode plates; the other end of the copper wire extends outside the battery as the third tab. The positive tab, negative tab, and third tab are connected to a dedicated sampling board of the BMS for data acquisition. The reference electrode is made of copper wire, gold wire, or sodium-plated copper wire. Before use, the reference electrode is soaked in concentrated sulfuric acid for 1 hour, ultrasonically cleaned for 30 minutes, and baked in a 50°C vacuum oven for 30 minutes to remove surface oxides and impurities. Figure 2 As shown in the figure, the copper wire is the reference electrode.

[0025] Furthermore, the assembled three-electrode sodium-ion battery underwent charge-discharge cycle testing. Specifically, the assembled three-electrode sodium-ion battery was subjected to charge-discharge cycle testing at a rate of 1C / 1C under a constant temperature of 25℃. One complete cycle included charging and discharging. During the test, the BMS collected the following data in real time at a sampling frequency of 1Hz: the potential V+ of the positive electrode relative to the reference electrode (recorded separately during the charging and discharging phases), the potential V- of the negative electrode relative to the reference electrode (recorded separately during the charging and discharging phases), the full cell voltage Vcell, the battery surface temperature T, and the charge-discharge current I (recorded as +I during the charging phase and -I during the discharging phase).

[0026] Furthermore, it is necessary to acquire various characteristic parameters, filter and normalize the collected raw data, and extract a set of characteristic parameters for each cycle (corresponding to one complete charge and discharge cycle), including the rate of change of positive electrode potential at the end of charging, the rate of change of negative electrode potential at the end of discharging, the growth rate of positive independent impedance R+ and negative independent impedance R- under each cycle number, coulombic efficiency, and discharge capacity.

[0027] Specifically, the positive independent impedance R+ and negative independent impedance R- of a preset number of cycles are measured using EIS technology. For other number of cycles, the positive and negative independent impedances are obtained using a linear interpolation algorithm. Then, the positive independent impedance after one cycle is subtracted from the initial positive independent impedance and divided by the number of cycles to obtain the growth rate of the positive independent impedance in that cycle. Similarly, the growth rate of the negative independent impedance in that cycle can be obtained. The rate of change of the positive electrode potential at the end of the charging cycle is obtained by using the derivative of the potential of the positive electrode relative to the reference electrode with respect to time at the end of the charging process in one cycle; the rate of change of the negative electrode potential at the end of the discharging cycle is obtained by using the derivative of the potential of the negative electrode relative to the reference electrode with respect to time at the end of the discharging process in one cycle. The coulombic efficiency of a sodium-ion battery in one cycle is obtained by dividing the discharge capacity by the charge capacity; the discharge capacity of a sodium-ion battery in one cycle is obtained by integrating the discharge current over time.

[0028] Therefore, the growth rate of the positive independent impedance, the growth rate of the negative independent impedance, the rate of change of the positive electrode potential at the end of charging, the rate of change of the negative electrode potential at the end of discharging, the coulombic efficiency, and the discharge capacity are the various characteristic parameters of this cycle.

[0029] Simultaneously, the SOH value needs to be measured for each cycle. It should be noted that the positive independent impedance R+ and the negative independent impedance R- are measured once using EIS technology after every 50 cycles. For other cycles, R+ and R- are calculated using linear interpolation. That is, the preset number of cycles in this application is 50, and the implementer can adjust the number of cycles according to the actual situation. In this application, the end of charging is the range of SOC from 80% to 100% during charging, and the end of discharging is the range of SOC from 20% to 0% during discharging.

[0030] Each iteration of the above-obtained feature parameters yields a complete set of various feature parameters.

[0031] Furthermore, the various feature parameters from 10 consecutive cycles are combined into a matrix, which is used as a sample. For a sample, the SOH value of the positive electrode after a preset number of cycles following 10 consecutive cycles is used as the positive electrode label of the sample, and the SOH value of the negative electrode after a preset number of cycles following 10 consecutive cycles is used as the negative electrode label of the sample. The positive electrode training set is formed using samples with positive electrode labels, and the negative electrode training set is formed using samples with negative electrode labels.

[0032] The samples in the positive and negative training sets are the same; the difference lies in the sample labels. The matrix composed of various feature parameters from 10 consecutive cycles is 10 cycles × 6 feature parameters, forming a 10 × 6 dimensional matrix. The preset number of cycles is 50, meaning that the data for the next 50 cycles is predicted using 10 consecutive cycles. For example, the positive and negative labels of the samples corresponding to cycles 1 to 10 correspond to the SOH values ​​of the positive and negative electrodes after the 60th cycle. The positive and negative labels of the samples composed of feature parameters from cycles 2 to 11 correspond to the SOH values ​​of the positive and negative electrodes after the 61st cycle. In other words, cycles 1-10 predict the SOH values ​​of the 60th cycle, and cycles 2-11 predict the SOH values ​​of the 61st cycle. SOH is defined as the ratio (expressed as a percentage) of the discharge capacity of the current cycle to the discharge capacity of the first cycle. In this application, the training set is constructed using data obtained from the first two hundred cycles. If the number of samples is insufficient, the number of samples can be increased or supplemented with data from other historical sodium-ion battery cycle tests.

[0033] Step S2: Train the prediction model using the positive electrode training set and the negative electrode training set respectively to obtain the positive electrode SOH prediction model and the negative electrode SOH prediction model; use the positive electrode SOH prediction model and the negative electrode SOH prediction model to predict the predicted positive electrode SOH value and the predicted negative electrode SOH value for each cycle.

[0034] The above steps construct positive and negative electrode training sets. Further training is needed using these sets to obtain positive and negative electrode SOH prediction models, respectively. In this embodiment, the prediction model is a Long Short-Term Memory (LSTM) network. The input to the prediction model is a matrix composed of various feature parameters for 10 consecutive iterations. This involves sliding a window of a preset length across the iterations. The matrix of various parameters corresponding to each iteration within each window corresponds to the matrix for each window. The preset length is 10 iterations. The output is the predicted positive and negative electrode SOH values ​​for the next 50 iterations after 10 consecutive iterations. These predicted positive and negative SOH values ​​are denoted as the predicted positive SOH value and the predicted negative SOH value, respectively. Each window corresponds to one predicted positive SOH value and one predicted negative SOH value.

[0035] After the model training is completed, the positive electrode SOH prediction model and the negative electrode SOH prediction model are used to predict the predicted positive electrode SOH value and the predicted negative electrode SOH value for each cycle.

[0036] Specifically, a window of preset length is used to slide along the loop count with a step size of one loop count. The feature parameters of each loop within a window are combined into a matrix, which serves as the input for that window. The input corresponding to the current window is then input into the positive electrode SOH prediction model and the negative electrode SOH prediction model, respectively, and the predicted positive electrode SOH value and the predicted negative electrode SOH value corresponding to the current window are output.

[0037] This allows us to obtain the predicted positive electrode SOH value and the predicted negative electrode SOH value for each window, or in other words, the predicted positive electrode SOH value and the predicted negative electrode SOH value for each cycle.

[0038] Step S3: Based on the predicted negative electrode SOH value, a graded strategy is formulated to monitor and repair the health status of the sodium-ion battery.

[0039] The above process yields the predicted SOH value of the negative electrode. Furthermore, the predicted SOH value of the negative electrode is used to formulate a graded strategy for graded regulation.

[0040] Specifically, the first-level response is that if the predicted positive electrode SOH value and the predicted negative electrode SOH value corresponding to the current window are less than the first preset threshold (90%), the BMS will issue a "health status warning" signal to prompt the operation and maintenance personnel to pay attention.

[0041] The secondary response is triggered after the primary response. If the predicted negative electrode SOH value corresponding to the next window of the current window is less than the second preset threshold (80%), and the negative electrode potential at the charging end measured by the BMS in the current actual cycle is lower than the preset safety potential threshold (e.g., ≤0.05V vs. Na), then... +If the risk of sodium precipitation is present (e.g., / Na), the BMS will automatically switch the charging strategy from 1C constant current charging to "1C pulse charging" (charging for 5 seconds, then resting for 1 second) until subsequent predictions show that the risk has been eliminated.

[0042] After the Level 2 response is executed, the BMS will continuously monitor and process the data according to the following logic: Diversion ①—Risk Relief: If subsequent predictions show that the predicted SOH value of the negative electrode in the 50th cycle rises to above the second threshold (80%), that is, the predicted negative electrode SOH value corresponding to the window obtained by subsequent prediction is greater than the second preset threshold, and the negative electrode potential at the end of charging is no longer lower than the safe potential threshold, then the BMS will automatically revert to the original 1C constant current charging strategy, exit the secondary response state, and return to the normal monitoring mode. Diversion ②—Continuous Stagnation: If, after 100 consecutive cycles, the predicted negative electrode SOH value is within the stagnation range (i.e., within the stagnation range of approximately 80% ± 2%), the BMS will automatically adjust the pulse charging parameters (e.g., adjust the charging 5 seconds / resting 1 second to charging 4 seconds / resting 2 seconds or charging 3 seconds / resting 3 seconds). If, after another 50 cycles, the predicted negative electrode SOH value is within the stagnation range, it is determined that the control has failed and is forced to enter the third-level response. Diversion ③—Continued deterioration: If, during the execution of the secondary response, the predicted negative electrode SOH value continues to decrease, and the sum of the decreases of the three consecutive predicted negative electrode SOH values ​​is greater than or equal to 1%, then the tertiary response is triggered, and the BMS immediately enters the tertiary response.

[0043] c. Level 3 Response - Self-Repair Trigger: When the "regulation failure" condition in the above-mentioned shunting ③ or shunting ② is met, the BMS applies a square wave AC current (e.g., a square wave AC current with an amplitude of 0.2C and a frequency of 10Hz, which can be adjusted according to the actual situation) to the third electrode and the negative electrode, and continues for a preset time (e.g., 30 seconds) to attempt to repair the performance of the negative electrode.

[0044] After a Level 3 response is executed, the process and termination criteria are as follows: After a Level 3 response is executed, the process re-enters step (II) for a new round of data collection and evaluation. Repair successful: If, in 10 consecutive cycles after repair, the predicted negative electrode SOH values ​​show a stable or rising trend (the 10 predicted negative electrode SOH values ​​no longer accelerate down to below 90% of the first preset threshold, and are all greater than or equal to the first preset threshold), then the self-repair is determined to be successful, the system exits the level 3 response, and returns to the normal monitoring mode; Repair Failure: If the SOH continues to decrease after repair, the Level 3 response (applying a square wave AC current) will be repeated, up to a maximum of 2 times. If it is still ineffective after 2 repetitions, an alarm "Self-repair failed, cell replacement recommended" will be issued, terminating the active control process for that cell.

[0045] Additionally, it should be noted that during the aforementioned cyclic testing process, if the reference electrode used is a sodium-plated copper wire, when the BMS detects a drift in the potential response of the reference electrode (i.e., the reference potential deviation of V+ or V- exceeds ±5mV of the initial value), it indicates that the sodium on the surface of the reference electrode has been completely consumed. The BMS will then pause the cyclic testing and automatically perform a reference electrode repair operation: short-circuit the reference electrode with the negative electrode, apply a small current of 0.01C-0.05C for a brief charging (lasting 5-10 minutes), so that the surface of the reference electrode is re-plated with a layer of sodium metal, restoring its reference function. After the repair is completed, once the potential response of the reference electrode is confirmed to have returned to normal, the cyclic testing and control process will continue.

[0046] Meanwhile, the cycle test for sodium-ion batteries is not an infinite cycle; the entire test is terminated under any of the following conditions: Normal Termination—SOH Retirement Threshold Achieved: When the actual SOH value of either the positive or negative electrode drops below 80%, it indicates that the battery has reached the end of its service life, and the test is terminated. Normal Termination - Target Loop Count Achieved: The cumulative number of loops reaches the preset target value (e.g., 2000 times), the full lifecycle verification is completed, and the test terminates; Abnormal Termination - Safety Failure Trigger: In the event of a safety anomaly such as short circuit, leakage, or bulging, the test will be terminated immediately and safety protection will be activated. Abnormal Termination - Repair Failure Trigger: After the Level 3 response is repeated twice, the predicted negative electrode SOH value continues to decrease, and an alarm "Self-repair failed, cell replacement recommended" is issued before termination.

[0047] In addition, firstly, the BMS obtains the predicted positive and negative SOH values ​​by running the trained positive and negative SOH prediction models, respectively. When both the predicted positive and negative SOH values ​​are below 85% (the third prediction threshold), and the R+ and R- values ​​measured by EIS during the current actual cycle test both exceed 150% of the initial values ​​(where the initial values ​​refer to the initial R+ and initial R- values ​​obtained by EIS measurement after battery assembly and before the first charge-discharge cycle), the BMS determines that the increase in polarization resistance is the main cause of performance degradation. At this time, the BMS skips the first and second level responses and directly executes the repair strategy: Apply a square wave AC current (e.g., a square wave AC current with an amplitude of 0.3C and a frequency of 15Hz) to the third tab and the positive (or negative) tab for a preset time (e.g., 45 seconds). After processing, measure R+ and R- again to confirm whether the polarization internal resistance has decreased. If R+ and R- decrease to within 120% of their initial values, the "self-repair" is considered successful; if not, repeat the application 1-2 times. Each time, the interval should be 30 seconds. If the requirement is still not met after 3 repetitions, the self-repair is considered a failure, a maintenance alarm is issued, and the active control process of the cell is terminated.

[0048] In summary, the core of this application lies in automatically triggering a three-level closed-loop intervention mechanism based on the three-electrode prediction results, from "early warning prompts → active regulation → self-repair triggers".

[0049] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0050] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0051] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for monitoring and repairing the health status of a sodium-ion battery based on a three-electrode system, characterized in that, The method includes: A reference electrode was implanted inside the sodium-ion battery cell and charge-discharge cycle tests were conducted. The SOH values ​​of the positive and negative electrodes and various characteristic parameters were collected for each cycle. The SOH values ​​of the positive and negative electrodes and various characteristic parameters for a preset number of cycles were used to form a positive electrode training set and a negative electrode training set, respectively. The prediction model was trained using the positive electrode training set and the negative electrode training set respectively to obtain the positive electrode SOH prediction model and the negative electrode SOH prediction model; the positive electrode SOH prediction model and the negative electrode SOH prediction model were used to predict the predicted positive electrode SOH value and the predicted negative electrode SOH value for each cycle. Based on the predicted negative electrode SOH value, a tiered strategy is developed to monitor and repair the health status of sodium-ion batteries.

2. The method for monitoring and repairing the health status of a sodium-ion battery based on a three-electrode system according to claim 1, characterized in that, The collection of SOH values ​​and various characteristic parameters of the positive and negative electrodes for each cycle includes: The positive and negative independent impedances after a preset number of cycles are measured using EIS technology. For other number of cycles, the positive and negative independent impedances are obtained using a linear interpolation algorithm. The positive independent impedance after one cycle is subtracted from the initial positive independent impedance and divided by the number of cycles to obtain the growth rate of the positive independent impedance in that cycle. Similarly, the growth rate of the negative independent impedance in that cycle can be obtained. The rate of change of the positive electrode potential at the end of the charging cycle is obtained by using the derivative of the potential of the positive electrode relative to the reference electrode with respect to time at the end of the charging process in one cycle; the rate of change of the negative electrode potential at the end of the discharging cycle is obtained by using the derivative of the potential of the negative electrode relative to the reference electrode with respect to time at the end of the discharging process in one cycle. The coulombic efficiency of a sodium-ion battery in one cycle is obtained by dividing the discharge capacity by the charge capacity; the discharge capacity of a sodium-ion battery in one cycle is obtained by integrating the discharge current over time. The growth rate of the positive independent impedance, the growth rate of the negative independent impedance, the rate of change of the positive electrode potential at the end of charging, the rate of change of the negative electrode potential at the end of discharging, the coulombic efficiency, and the discharge capacity are the various characteristic parameters of this cycle.

3. The method for monitoring and repairing the health status of a sodium-ion battery based on a three-electrode system according to claim 1, characterized in that, The step of forming positive and negative electrode training sets and negative electrode training sets respectively by using the SOH values ​​of the positive and negative electrodes for a preset number of cycles and various characteristic parameters includes: The various feature parameters are combined into a matrix after 10 consecutive cycles, and this matrix is ​​used as a sample. For a sample, the SOH value of the positive electrode after a preset number of cycles after 10 consecutive cycles is used as the positive electrode label of the sample, and the SOH value of the negative electrode after a preset number of cycles after 10 consecutive cycles is used as the negative electrode label of the sample. The positive electrode training set is formed using samples with positive electrode labels, and the negative electrode training set is formed using samples with negative electrode labels.

4. The method for monitoring and repairing the health status of a sodium-ion battery based on a three-electrode system according to claim 1, characterized in that, The method of predicting the predicted positive and negative SOH values ​​for each cycle using the positive and negative electrode SOH prediction models includes: Using a window of preset length, the model slides along the loop count with a step size of one loop. The feature parameters of each loop within a window are combined into a matrix, which serves as the input for that window. The input corresponding to the current window is then input into the positive electrode SOH prediction model and the negative electrode SOH prediction model, respectively, and the predicted positive electrode SOH value and the predicted negative electrode SOH value corresponding to the current window are output.

5. The method for monitoring and repairing the health status of a sodium-ion battery based on a three-electrode system according to claim 1, characterized in that, The method of predicting the predicted positive and negative SOH values ​​for each cycle using the positive and negative electrode SOH prediction models includes: Using a window of preset length, the model slides along the loop count with a step size of one loop. The feature parameters of each loop within a window are combined into a matrix, which serves as the input for that window. The input corresponding to the current window is then input into the positive electrode SOH prediction model and the negative electrode SOH prediction model, respectively, and the predicted positive electrode SOH value and the predicted negative electrode SOH value corresponding to the current window are output.

6. The method for monitoring and repairing the health status of a sodium-ion battery based on a three-electrode system according to claim 1, characterized in that, The step-by-step strategy for monitoring and repairing the health status of sodium-ion batteries based on the predicted negative electrode SOH value includes: The first-level response is that if the predicted positive electrode SOH value and the predicted negative electrode SOH value corresponding to the current window are less than the first preset threshold, the BMS will issue a health status warning signal. The secondary response is as follows: if the predicted negative electrode SOH value corresponding to the next window of the current window is less than the second preset threshold after the primary response is triggered, and the negative electrode potential at the end of the charging measured by the BMS in the current actual cycle is lower than the preset safe potential threshold, then the BMS will automatically switch the charging strategy from 1C constant current charging to 1C pulse charging. After the Level 2 response is executed, the BMS continues to monitor and handles the situation according to the following logic: Diversion ①: If the predicted negative electrode SOH value corresponding to the window obtained by subsequent prediction is greater than the second preset threshold, and the actual measured negative electrode potential at the end of charging is no longer lower than the safe potential threshold, then the BMS will automatically revert to the original 1C constant current charging strategy, exit the secondary response state, and return to the normal monitoring mode. Diversion ②: If the predicted negative electrode SOH value is within the stalemate range after 100 consecutive cycles, the BMS will automatically adjust the pulse charging parameters. If the predicted negative electrode SOH value is within the stalemate range after another 50 cycles, it will be judged as a control failure and forced to enter the third-level response. Diversion ③: If, during the execution of the secondary response, the predicted negative electrode SOH value continues to decrease, and the sum of the decreases of the three consecutive predicted negative electrode SOH values ​​is greater than or equal to 1%, then the tertiary response is triggered, and the BMS immediately enters the tertiary response. The Level 3 response is as follows: when the control failure condition in shunting ③ or shunting ② is met, the BMS applies a square wave AC current to the third electrode and the negative electrode for a preset time to perform repair. If the predicted negative electrode SOH values ​​obtained after repair are greater than or equal to the first preset threshold, the repair is successful. If there is a predicted negative electrode SOH value less than the first preset threshold among the predicted negative electrode SOH values ​​obtained after repair, the Level 3 response is repeated, up to a maximum of 2 times. If it is still ineffective after 2 repetitions, an alarm is issued indicating self-repair failure and suggesting replacement of the battery cell.

Images