Prediction method for cyclic diving of lithium ion battery
By adding a constant voltage charging stage at regular intervals during the lithium-ion battery cycling process, recording the current change curve and identifying characteristic peaks, and combining this with the battery capacity retention rate or cycle count, the problem of predicting battery degradation was solved, improving prediction accuracy and efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 安徽国轩新能源汽车科技有限公司
- Filing Date
- 2026-01-14
- Publication Date
- 2026-04-24
AI Technical Summary
Existing technologies struggle to effectively predict the water drop phenomenon in lithium-ion batteries during cycling, affecting battery lifespan and safety.
During battery cycling, a constant voltage charging stage is added at regular intervals, and the current change curve is recorded. The characteristic peak is used to determine whether the battery will experience a sudden drop in voltage. The battery capacity retention rate or the number of cycles is used as the termination condition.
It improves the accuracy and efficiency of battery price drop prediction, reduces redundant testing time, and increases the testing efficiency of good batteries.
Smart Images
Figure CN121918006A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of battery manufacturing and health management technology, and in particular to a method for predicting the cycle degradation of lithium-ion batteries. Background Technology
[0002] Lithium-ion batteries play a vital role in modern society, and their widespread use has profoundly impacted people's lifestyles. However, some batteries experience a "water drop" phenomenon during cycling, where their performance deteriorates rapidly, significantly affecting their lifespan and, in severe cases, even causing battery safety issues. Therefore, predicting this water drop phenomenon during battery cycling is crucial for extending battery life and improving battery safety.
[0003] Currently, there are many methods for predicting battery cycle degradation. Patent CN118244119A describes a machine learning-based method in a system for online joint prediction of lithium-ion battery capacity degradation. This method extracts feature values during battery cycling and learns the mapping relationship between these feature values and the current charge-discharge cycle's lithium-ion battery health status to predict battery cycle degradation. The accuracy of battery cycle degradation prediction is crucial in how effectively useful feature values for predicting battery cycle degradation are extracted. Summary of the Invention
[0004] To overcome the shortcomings of existing technologies in that it is difficult to lock the associated features of battery cyclic cascading, this invention proposes a method for predicting cyclic cascading of lithium-ion batteries. This method can quickly extract the feature values associated with battery cascading and effectively predict the phenomenon of battery cyclic cascading.
[0005] The present invention proposes a method for predicting the cycle drop of lithium-ion batteries, which involves cyclically charging and discharging the battery under constant temperature conditions, with a constant voltage charging stage added at the end of the battery charging process at a set interval of M cycles. Record the current change curves for each constant voltage charging stage and identify characteristic peaks; If a characteristic peak appears in the current change curve before the detection cutoff, it is predicted that the battery cycle will experience a drop in capacity; otherwise, it is predicted that the battery cycle will not experience a drop in capacity before the detection cutoff. The detection cutoff condition is set as follows: the battery capacity retention rate is lower than the set threshold C0 or the number of cycles reaches the preset threshold H.
[0006] Preferably, the criteria for determining the characteristic peak are: the current change curve satisfies d(△I) / dt=0 and d²(△I) / dt²<0, where d represents the derivative, t represents time, and △I represents the current change.
[0007] Preferably, if a characteristic peak appears in N consecutive current change curves before the detection cutoff, it is predicted that the battery cycle will experience a drop in voltage; otherwise, it is predicted that the battery cycle will not experience a drop in voltage before the detection cutoff; N>2.
[0008] Preferably, it includes the following steps: S1. Place the battery under test in a constant temperature environment; S2. Perform cyclic charging and discharging of the battery and monitor its capacity; S3. Determine whether the number of battery charge-discharge cycles has reached a multiple of M; M≥100; If yes, then add a constant voltage charging stage at the end of the charging period, record the current change curve of the constant voltage charging node, and then execute step S4. No, proceed to step S6; S4. Determine whether the latest current change curve shows a characteristic peak; No, then clear the characteristic peak register value to zero, and then execute step S6; the initial value of the characteristic peak register value is 0; If yes, increment the characteristic peak register value by 1 and store it, then execute step S5; S5. Determine whether the characteristic peak register value is greater than or equal to N; No, proceed to step S6; If yes, then it is determined that the battery will experience a significant drop in water pressure. S6. Determine whether the battery capacity retention rate is lower than C0 or the number of cycles reaches the preset threshold H; If yes, then it is determined that the battery will not experience a sudden drop in water level. If not, return to step S2.
[0009] Preferably, in step S3, when the number of battery charge-discharge cycles reaches a multiple of M, it is first determined whether the end of the charging period is constant voltage charging. If yes, the current change curve of the current constant voltage charging stage is recorded; otherwise, a constant voltage charging stage is added at the end of the charging period, and the current change curve of the constant voltage charging node is recorded.
[0010] Preferably, the cutoff current during the constant voltage charging stage is ≤0.05C, and the sampling time is ≤30s.
[0011] Preferably, C0 ≥ 60% and H ≥ 200.
[0012] Preferably, the charging and discharging system can adopt constant current and constant voltage charging, stepped charging, constant power charging, or power mapping.
[0013] The present invention proposes a prediction system for lithium-ion battery cycle degradation, comprising a memory and a processor. The memory stores a computer program, and the processor is connected to the memory. The processor is used to execute the computer program to implement the prediction method for lithium-ion battery cycle degradation.
[0014] The present invention proposes a storage medium storing a computer program, which, when executed, is used to implement the method for predicting the cycle failure of a lithium-ion battery.
[0015] The advantages of this invention are: This invention proposes a method for predicting battery cycle degradation during lithium-ion battery cycling. By adding a constant-voltage charging phase at regular intervals during battery cycling testing, the method predicts whether the battery will experience a degradation event by analyzing the current change trend during the constant-voltage charging phase. This invention effectively captures the characteristic values of battery cycle degradation by adding a constant-voltage charging phase at regular intervals during battery cycling. Experimental verification has shown that this method can effectively predict battery cycle degradation. The method is simple to operate, non-destructive, and highly accurate, making it suitable for actual battery testing and thus possessing significant application value.
[0016] In this invention, the continuous occurrence of characteristic peaks is used as the termination condition for judgment, which is beneficial to predict the water drop phenomenon as early as possible, improve the prediction efficiency, and avoid redundant test time. This invention uses the battery capacity retention rate (or the number of cycle tests) as the test termination condition, which improves the testing efficiency of good batteries while ensuring the reliability of judgment. Attached Figure Description
[0017] Figure 1 Example 1: Curve of current change during constant voltage charging in battery cycling process; Figure 2 Example 1: Capacity decay curve during battery cycling; Figure 3 Example 2: Curve of current change during constant voltage charging in battery cycling process; Figure 4 Example 2: Capacity decay curve during battery cycling; Figure 5 Example 3: Curve of current change during constant voltage charging in battery cycling process; Figure 6 Example 3: Capacity decay curve during battery cycling; Figure 7 Example 4: Curve of current change during constant voltage charging in battery cycling process; Figure 8 Example 4: Capacity decay curve during battery cycling; Figure 9 This is a flowchart of a method for predicting the cycling failure of a lithium-ion battery proposed in this invention. Figure 10 This is a flowchart of another method for predicting the cycling failure of lithium-ion batteries proposed in this invention. Detailed Implementation
[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.
[0019] This embodiment proposes a method for predicting the battery's capacity drop during cyclic charging and discharging. The battery under test is placed in a constant-temperature environment for cyclic charging and discharging until the battery capacity retention rate is lower than a set value C0 or the number of cycles reaches a preset threshold H; 60%≤C0, specifically 80%, H≥200; every M charge-discharge cycles, a constant-voltage charging stage is added at the end of the charging process, M≥100; during the cyclic charging and discharging process, the current change curve of the constant-voltage charging stage is recorded; if a characteristic peak appears in the current change curve, it is predicted that the battery will experience a capacity drop during cyclic charging and discharging.
[0020] Based on experience, if a characteristic peak is detected, it can be predicted that the battery will experience a significant drop in battery capacity. If no characteristic peak is detected before the battery capacity retention rate is lower than the set value C0 or the number of cycles reaches the preset threshold H, it means that the battery will not experience a significant drop in battery capacity before the battery capacity retention rate is lower than the set value C0 or the number of cycles reaches the preset threshold H.
[0021] It's worth noting that if M is set too small, such as 50, the current change curve may show continuous characteristic peaks. In this case, you can set the current change curve to only show characteristic peaks after N consecutive occurrences (N>2, for example, N=3) to determine if the battery is about to experience a sudden drop in voltage. Conversely, if M is set too large, such as 600, the battery will be considered to experience a sudden drop in voltage as soon as a characteristic peak appears in the current change curve.
[0022] In this embodiment, the constant voltage charging cutoff current is ≤0.05C, and the constant voltage charging sampling time is ≤30s. The temperature value T of the constant temperature environment can be taken in the range [-20℃, 50℃].
[0023] Reference Figure 9 The method includes the following steps: S1. Place the battery under test in a constant temperature environment; S2. Cycle charge and discharge the battery and monitor its capacity; the charge and discharge regime can be constant current and constant voltage charging, stepped charging, constant power charging or power map; S3. Determine whether the number of battery charge / discharge cycles has reached a multiple of M; If yes, then add a constant voltage charging stage at the end of the charging period, record the current change curve of the constant voltage charging node, and then execute step S4. No, proceed to step S6; S4. Determine whether the latest current change curve has a characteristic peak; the condition for determining the characteristic peak is: the current change curve satisfies d(△I) / dt=0 and d²(△I) / dt²<0, where d represents the derivative, t represents time; △I represents the current change, that is, the difference in charging current at two adjacent sampling time points t.
[0024] No, then clear the characteristic peak register value to zero, and then execute step S6; the initial value of the characteristic peak register value is 0; If yes, increment the characteristic peak register value by 1 and store it, then execute step S5; S5. Determine whether the characteristic peak register value is greater than or equal to N; No, proceed to step S6; If yes, then it is determined that the battery will experience a significant drop in water pressure. S6. Determine whether the detection cutoff condition has been met; The detection cutoff condition can be set as needed: whether the battery capacity retention rate is lower than C0 or whether the number of cycles reaches a preset threshold H; If yes, then it is determined that the battery will not experience a sudden drop in water level before the detection cutoff condition. If not, return to step S2.
[0025] Reference Figure 10 In practice, if constant voltage charging is used at the end of the battery cycle, there is no need to add an extra constant voltage charging stage. The current change curve of the constant voltage charging stage can be recorded directly in step S3.
[0026] The method proposed in this invention will be verified by referring to several specific embodiments below.
[0027] Example 1
[0028] The specific steps for predicting a battery voltage drop during constant current and constant voltage cycling tests are as follows: S001. Place two lithium iron phosphate batteries in a constant temperature chamber at 25°C and perform constant current and constant voltage cycle tests. The detailed test steps are as follows: charge at 1C constant current to 3.65V and then switch to 3.65V constant voltage charging. The cutoff current for constant voltage charging is 0.05C. Let it rest for 30 minutes. Discharge at 1C to 2.5V and let it rest for 30 minutes. Repeat the above cycle 1300 times. S002. Record the capacity changes and current and voltage changes during the battery cycle, with a sampling time of 1 second.
[0029] S003. The capacity decay curve during battery cycling and the current change curve during the constant voltage stage at 100-cycle intervals are calculated: S004. Predict the risk of battery degradation during cycling: Depend on Figure 1 It can be seen that during the battery cycle of 600 cycles, characteristic peaks appear on the current curves of the constant voltage stage at 300 / 400 / 500 / 600 cycles, indicating a risk of a sudden drop in current as the battery continues to cycle. During the battery cycle of 1300 cycles, [the following text appears to be unrelated and possibly a separate sentence fragment: "from..."] Figure 2 It can be seen that the battery experienced a water drop phenomenon around 1000 cycles. The above phenomenon indicates that this method can predict whether a water drop phenomenon will occur during battery cycling.
[0030] Example 2
[0031] The specific steps for predicting battery voltage drops during constant power cycling tests are as follows: S001. Place two lithium iron phosphate batteries in a 35°C constant temperature chamber for constant power cycle testing. During the cycle, a constant voltage charging step is added at the end of the charging process every 100 cycles. The detailed test steps are as follows: 1) Charge 1P to 3.65V, let it rest for 30 minutes, discharge 1P to 2.5V, let it rest for 30 minutes, repeat the cycle 100 times; 2) Charge 1P to 3.65V and switch to 3.65V constant voltage charging. The constant voltage charging is cut off when the current is 0.05C. Let it rest for 30 minutes. Discharge 1P to 2.5V and let it rest for 30 minutes.
[0032] 3) Repeat steps 1) to 2) 15 times.
[0033] S002. Record the capacity changes and current and voltage changes during the battery cycle, with a sampling time of 30 seconds. S003. The capacity decay curve during battery cycling and the current change curve during the constant voltage stage at 100-cycle intervals are calculated: S004. Predict the risk of battery degradation during cycling: Depend on Figure 3 It can be seen that during the battery cycle of 700 times, characteristic peaks appear on the current curves of the constant voltage stage at 300 / 400 / 500 / 600 / 700 cycles, indicating a risk of a sudden drop in current as the battery continues to cycle. During the battery cycle of 1700 times, combined with... Figure 4 It can be seen that the battery experienced a water drop phenomenon around 1100 cycles. The above phenomenon indicates that this method can effectively predict whether a water drop phenomenon will occur during battery cycling.
[0034] Example 3
[0035] The specific steps for predicting a drop in battery performance during a stepped charging cycle test are as follows: S001. Place two ternary lithium batteries in a 25°C constant temperature chamber for stepped charging cycle testing. During the cycle, a constant voltage charging step is added at the end of the charging process every 100 cycles. The detailed test steps are as follows: 1) Charge at 3C constant current to 3.5V, charge at 1.5C constant current to 3.8V, charge at 0.5C constant current to 4.2V, let stand for 30 minutes, discharge at 0.5C to 3V, let stand for 30 minutes, cycle 100 times; 2) Charge at 3C constant current to 3.5V, charge at 1.5C constant current to 3.8V, charge at 0.5C constant current to 4.2V and then switch to 4.2V constant voltage charging. Cut off constant voltage charging when the current is 0.05C, let stand for 30 minutes, discharge at 0.5C to 3V, and let stand for 30 minutes. 3) Repeat steps 1) to 2) 18 times; S002. Record the capacity changes and current and voltage changes during the battery cycle, with a sampling time of 30 seconds. S003. The capacity decay curve during battery cycling and the current change curve during the constant voltage stage at 100-cycle intervals are calculated: S004. Predict the risk of battery degradation during cycling: Depend on Figure 5 It can be seen that during the battery cycle to 900 cycles, characteristic peaks appear on the current curves of the constant voltage stage at 500 / 600 / 700 / 800 / 900 cycles, indicating a risk of a sudden drop in current as the battery continues to cycle. During the battery cycle to 1800 cycles, combined with... Figure 6 It can be seen that the battery experienced a water drop around 1300 cycles. The above phenomenon indicates that this method can effectively predict whether a water drop will occur during battery cycling.
[0036] Example 4
[0037] The specific steps for predicting a battery voltage drop during constant current and constant voltage cycling are as follows: S001. Place two lithium iron phosphate batteries in a constant temperature chamber at 25°C and perform a cycle test. The detailed test steps are as follows: charge at 1C constant current to 3.65V, then switch to 3.65V constant voltage charging. The cutoff current for constant voltage charging is 0.05C. Let it rest for 30 minutes. Discharge at 1C to 2.5V. Let it rest for 30 minutes. Repeat the above cycle until the capacity retention rate is 84%. S002. Record the capacity changes and current and voltage changes during the battery cycle, with a sampling time of 30 seconds. S003. The capacity decay curve during battery cycling and the current change curve during the constant voltage stage at 100-cycle intervals are calculated: S004. Predict the risk of battery degradation during cycling: Depend on Figure 7 It can be seen that during the battery cycle to 84% capacity retention, no characteristic peaks appeared on the current curve during the constant voltage phase of every 100 cycles. Figure 8 It can be seen that the battery did not experience a water drop during the cycle. The above phenomenon indicates that this method can predict whether a water drop will occur during the battery cycle.
[0038] The test results of Examples 1-4 above are statistically summarized in the table below.
[0039] Table 1: Statistical analysis of experimental results in Examples 1-4
[0040] from Figure 2 , Figure 4 and Figure 6 It can be seen that the number of charge-discharge cycles during battery cycling is much greater than the number of cycles in which the characteristic peak appears, and the characteristic peaks all appear continuously during the constant voltage charging phase. This demonstrates the reliability of setting the N value and threshold C0 (or threshold H) in this invention.
[0041] Of course, those skilled in the art will recognize that the present invention is not limited to the details of the exemplary embodiments described above, but also includes the same or similar structures that can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.
[0042] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.
[0043] The technologies, shapes, and structures not described in detail in this invention are all known technologies.
Claims
1. A method for predicting the cycle degradation of lithium-ion batteries, characterized in that, The battery is cycled in a constant temperature environment, with a set interval of M cycles, and a constant voltage charging stage is added at the end of the battery charging process. Record the current change curves for each constant voltage charging stage and identify characteristic peaks; If a characteristic peak appears in the current change curve before the detection cutoff, it is predicted that the battery cycle will experience a drop in capacity; otherwise, it is predicted that the battery cycle will not experience a drop in capacity before the detection cutoff. The detection cutoff condition is set as follows: the battery capacity retention rate is lower than the set threshold C0 or the number of cycles reaches the preset threshold H.
2. The method for predicting the cycle degradation of lithium-ion batteries as described in claim 1, characterized in that, The criteria for identifying characteristic peaks are: the current change curve satisfies d(△I) / dt=0 and d²(△I) / dt²<0, where d represents the derivative, t represents time, and △I represents the current change.
3. The method for predicting the cycle degradation of lithium-ion batteries as described in claim 1, characterized in that, If a characteristic peak appears in N consecutive current change curves before the detection deadline, it is predicted that the battery cycle will experience a drop in current. Conversely, it is predicted that the battery cycle will not experience a drop in current before the detection deadline; N>2.
4. The method for predicting the cycle degradation of lithium-ion batteries as described in claim 3, characterized in that, Includes the following steps: S1. Place the battery under test in a constant temperature environment; S2. Perform cyclic charging and discharging of the battery and monitor its capacity; S3. Determine whether the number of battery charge / discharge cycles has reached a multiple of M; M≥100； If yes, then add a constant voltage charging stage at the end of the charging period, record the current change curve of the constant voltage charging node, and then execute step S4. No, proceed to step S6; S4. Determine whether the latest current change curve shows a characteristic peak; No, then clear the characteristic peak register value to zero, and then execute step S6; the initial value of the characteristic peak register value is 0; If yes, increment the characteristic peak register value by 1 and store it, then execute step S5; S5. Determine whether the characteristic peak register value is greater than or equal to N; No, proceed to step S6; If yes, then it is determined that the battery will experience a significant drop in water pressure. S6. Determine whether the battery capacity retention rate is lower than C0 or the number of cycles reaches the preset threshold H; If yes, then it is determined that the battery will not experience a sudden drop in water level. If not, return to step S2.
5. The method for predicting the cycle degradation of lithium-ion batteries as described in claim 4, characterized in that, In step S3, when the number of battery charge-discharge cycles reaches a multiple of M, first determine whether the end of the charging period is constant voltage charging. If so, record the current change curve of the current constant voltage charging stage. If not, add a constant voltage charging stage at the end of the charging process and record the current change curve at the constant voltage charging node.
6. The method for predicting the cycle failure of lithium-ion batteries as described in claim 1, characterized in that, The cutoff current during the constant voltage charging stage is ≤0.05C, and the sampling time is ≤30s.
7. The method for predicting the cycle failure of lithium-ion batteries as described in claim 1, characterized in that, C0≥60%, H≥200.
8. The method for predicting the cycle degradation of lithium-ion batteries as described in any one of claims 1-7, characterized in that, The charging and discharging system can adopt constant current and constant voltage charging, stepped charging, constant power charging, or power map.
9. A prediction system for the cycle failure of a lithium-ion battery, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program, the processor is connected to the memory, and the processor is used to execute the computer program to implement the method for predicting the cycle drop of lithium-ion batteries as described in any one of claims 1-8.
10. A storage medium, characterized in that, The device contains a computer program that, when executed, is used to implement the method for predicting the cycle failure of a lithium-ion battery as described in any one of claims 1-8.