Online early warning method and system for NMC battery diving point based on hybrid mechanism model

Abstract

The present invention discloses an online warning method and system for NMC battery capacity degradation based on a hybrid mechanism model. This method belongs to the field of battery storage technology. The method includes: establishing a hybrid mechanism model by considering the effects of SEI film growth, lithium plating, electrolyte oxidation, and manganese decomposition on NMC battery capacity degradation; fitting the hybrid mechanism model using an early NMC battery capacity loss signal consisting of the total capacity loss values ​​of the NMC battery over the first several cycles to determine model parameters; measuring the actual capacity degradation vector of the NMC battery within a time window, calculating the theoretical capacity degradation vector of the NMC battery within the time window using the hybrid mechanism model, and obtaining a linear correlation equation between the theoretical capacity degradation vector and the actual capacity degradation vector; and using a sliding window to obtain the linear correlation equation within different time windows. When the slope of the linear correlation equation reaches a minimum value, an NMC battery capacity degradation warning is issued. This method ensures the safe and reliable operation of lithium batteries.

Description

Technical Field

[0001] The present invention belongs to the field of battery storage technology, and more specifically, relates to an online early warning method and system for NMC battery diving points based on a hybrid mechanism model. Background Art

[0002] The increasing popularity of electric vehicles is driving significant growth in the global lithium-ion battery industry. Lithium-ion batteries, renowned for their high energy density and efficiency, have become the cornerstone of electric vehicle development. However, due to irreversible electrochemical reactions, lithium batteries experience capacity degradation over time. When battery capacity declines to a certain point, it can suddenly accelerate, reaching the end of its lifespan within a relatively short period of time. This phenomenon is known as capacity drop. Sudden battery lifespan degradation can disrupt battery usage plans and even cause large equipment to cease operation, resulting in unforeseen losses. Therefore, online early warning of lithium battery capacity drop points is essential.

[0003] With the development of artificial intelligence (AI) technology, numerous machine learning methods have been applied to predicting the potential drop point of lithium batteries, achieving promising results. However, these machine learning algorithms use black-box models to predict the moment of the potential drop point, lacking physical interpretability. Furthermore, the models' robustness and generalizability are low, significantly limiting their application. Furthermore, existing lithium battery health systems, which collect real-time signals, are ineffective in predicting potential drop points. Therefore, there is an urgent need to develop an online early warning framework for nickel-manganese-cobalt oxide (NMC) battery potential drop points based on a hybrid mechanism model. Summary of the Invention

[0004] In response to the defects of the existing technology and the need for improvement, the present invention provides an NMC battery diving point online early warning method and system based on a hybrid mechanism model, which aims to solve the technical problems of low performance of the existing online early warning battery diving point and poor interpretability of the online early warning signal.

[0005] To achieve the above objectives, according to one aspect of the present invention, an online warning method for an NMC battery diving point based on a hybrid mechanism model is provided, comprising a model establishment stage S1-S2 and an online warning stage S3-S4; S1, considering the effects of SEI film growth, lithium plating, electrolyte oxidation, and manganese decomposition on the capacity decay of the NMC battery in the NMC battery, a hybrid mechanism model is established to describe the change of the total capacity loss of the NMC battery over time; S2, using an early NMC battery capacity loss signal composed of the total capacity loss values ​​of the NMC battery in the initial several cycles, the hybrid mechanism model is fitted to determine the parameters of the hybrid mechanism model; S3, measuring the actual capacity decay vector of the NMC battery in a time window, calculating the theoretical capacity decay vector of the NMC battery in the time window using the hybrid mechanism model determined by the parameters, and obtaining a linear correlation equation between the theoretical capacity decay vector and the actual capacity decay vector in the time window; S4, using a sliding window to obtain the linear correlation equations in different time windows, and issuing an NMC battery diving point warning when the slope of the linear correlation equation reaches a minimum value.

[0006] Furthermore, the effect of SEI film growth on NMC battery capacity decay is expressed as:

[0007]

[0008] Among them, Q sei The capacity decay of NMC battery caused by SEI film growth, A sei is the first coefficient, B sei is the second coefficient, and t is the time.

[0009] Furthermore, the effect of lithium plating on the capacity decay of NMC batteries is expressed as:

[0010] Q lp =B lp t 1.5

[0011] Among them, Q lp NMC battery capacity degradation caused by lithium plating, B lp is the third coefficient, and t is time. Furthermore, the effect of electrolyte oxidation on NMC battery capacity degradation is expressed as:

[0012] Q eox =C eox t

[0013] Among them, Q eox The capacity decay of NMC battery caused by electrolyte oxidation, C eox is the fourth coefficient, and t is the time.

[0014] Furthermore, the effect of manganese decomposition on the capacity decay of NMC batteries is expressed as:

[0015] Q Mn =C Mn t+B Mn t 1.5

[0016] Among them, Q Mn The capacity decay of NMC battery caused by manganese decomposition, C Mn is the fifth coefficient, B Mn is the sixth coefficient, and t is the time.

[0017] Furthermore, the hybrid mechanism model is:

[0018]

[0019] Among them, Q tot is the total capacity loss of NMC battery, Q sei The capacity decay of NMC battery caused by SEI film growth, Q lp The capacity of NMC battery is reduced due to lithium plating. eox The capacity of NMC battery is reduced due to electrolyte oxidation. Mn is the capacity decay of NMC battery caused by manganese decomposition, A, B, and C are the three parameters of the mixed mechanism model, and t is time.

[0020] According to another aspect of the present invention, an online early warning system for an NMC battery diving point based on a hybrid mechanism model is provided, comprising: a processor; and a memory storing a computer-executable program, wherein when the program is executed by the processor, the processor executes the online early warning method for an NMC battery diving point based on the hybrid mechanism model as described above.

[0021] According to another aspect of the present invention, a computer-readable storage medium is provided, on which a computer program is stored. When the program is executed by a processor, the online early warning method for the NMC battery diving point based on the hybrid mechanism model as described above is implemented.

[0022] In general, the above technical solutions conceived by the present invention can achieve the following beneficial effects:

[0023] (1) An online warning method for the NMC battery diving point based on a hybrid mechanism model is provided. The effects of SEI film growth, lithium plating, electrolyte oxidation, and manganese decomposition on the capacity decay of NMC batteries are comprehensively considered, and a hybrid mechanism model describing the capacity decay of NMC batteries is obtained. The model is physically interpretable, simple in form, and easy to apply, and can well fit the capacity decay trend of NMC batteries.

[0024] It points out that there is a linear correlation between the output of the hybrid mechanism model and the actual capacity, and extracts the slope signal of the linear correlation equation as an online early warning signal for the diving point. This overcomes the limitation of the poor interpretability of the online early warning signal and can warn of the diving point in advance before the diving point occurs.

[0025] (2) Using the sliding window interception method, the lengths of the output vector of the hybrid mechanism model and the true capacity vector are fixed, and the slope of the linear correlation equation corresponding to the output of the hybrid mechanism model and the true discharge capacity data in each window is continuously obtained through the sliding window to form a slope signal, thereby improving the calculation efficiency;

[0026] The slope signal is regarded as an online warning signal for the diving point, and whether the slope signal reaches the minimum value is used as the warning standard. When the slope signal reaches the minimum value, a warning alarm is issued, realizing online warning of the diving point of the NMC battery, solving the problem of poor performance of online warning of the diving point in the existing technology;

[0027] (3) This method can be applied to the lithium battery health management system. By using a small amount of early cycle data to obtain the parameters of the hybrid mechanism model, the online warning signal of the NMC lithium battery diving point can be quickly obtained, thereby realizing the online warning of the diving point of the NMC battery and ensuring the safe and reliable operation of the lithium battery. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 Flowchart of the online early warning method for NMC battery water drop point based on the hybrid mechanism model provided by an embodiment of the present invention;

[0029] Figure 2 This is a flow chart of the hybrid mechanism model provided by an embodiment of the present invention;

[0030] Figure 3 is a diagram of the fitting results of the hybrid mechanism model provided by an embodiment of the present invention;

[0031] Figure 4 This is a result diagram of the online early warning diving point provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0032] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely for the purpose of explaining the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.

[0033] In the present invention, the terms "first", "second", etc. (if any) in the present invention and the drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0034] Example 1

[0035] An online early warning method for NMC battery diving point based on a hybrid mechanism model. Figure 1 , combined with Figure 2-Figure 4 The online warning method for NMC battery water drop point based on the hybrid mechanism model in this embodiment is described in detail. The method includes a model establishment phase (operation S1-operation S2) and an online warning phase (operation S3-operation S4).

[0036] Operation S1 considers the effects of SEI film growth, lithium plating, electrolyte oxidation, and manganese decomposition on the capacity decay of NMC batteries, and establishes a hybrid mechanism model that describes the change of total capacity loss of NMC batteries over time.

[0037] The derivation process of the hybrid mechanism model is as follows Figure 2 Specifically, the effect of solid electrolyte interface (SEI) film growth on NMC battery capacity decay is expressed as:

[0038]

[0039] Among them, Q sei The capacity decay of NMC battery caused by SEI film growth, A sei is the first coefficient, B sei is the second coefficient, and t is the time.

[0040] The effect of lithium plating on the capacity decay of NMC batteries is expressed as:

[0041] Q lp =B lp t 1.5

[0042] Among them, Q lp NMC battery capacity degradation caused by lithium plating, B lp is the third coefficient.

[0043] The effect of electrolyte oxidation on NMC battery capacity decay is expressed as:

[0044] Q eox =C eox t

[0045] Among them, Q eox The capacity decay of NMC battery caused by electrolyte oxidation, C eox is the fourth coefficient. The effect of manganese decomposition on the capacity decay of NMC batteries is expressed as:

[0046] Q Mn =C Mn t+B Mn t 1.5

[0047] Among them, Q Mn The capacity decay of NMC battery caused by manganese decomposition, C Mn is the fifth coefficient, B Mn is the sixth coefficient.

[0048] The mixing mechanism model is:

[0049]

[0050] The mixing mechanism model can be simplified as:

[0051]

[0052] Among them, Q tot is the total capacity loss of NMC battery, and A, B, and C are the three parameters of the mixed mechanism model.

[0053] Operation S2 is to fit the hybrid mechanism model using an early NMC battery capacity loss signal consisting of total capacity loss values ​​of the NMC battery in the first several cycles to determine parameters of the hybrid mechanism model.

[0054] Preferably, in operation S2, the hybrid mechanism model is fitted using an early NMC battery capacity loss signal consisting of the total capacity loss value of the NMC battery in the first 40 cycles to determine the parameters of the hybrid mechanism model.

[0055] Operation S3 measures the capacity decay real vector of the NMC battery within the time window, calculates the capacity decay theoretical vector of the NMC battery within the time window using the hybrid mechanism model determined by the parameters, and obtains a linear correlation equation between the capacity decay theoretical vector and the capacity decay real vector within the time window.

[0056] In this embodiment, it is found that the capacity decay theoretical value calculated by the hybrid mechanism model obtained in operation S2 is linearly correlated with the actual capacity decay value. The evaluation index of the hybrid mechanism model obtained in operation S2 is R 2 The result is expressed as a fraction. Figure 3 As shown, it can be seen that the average fitting effect of the mixed mechanism model can reach 98%, which can well fit the battery total capacity discharge data.

[0057] Furthermore, it is theoretically proven that the hybrid mechanism model has a linear correlation with the capacity loss model caused by SEI growth, as shown below:

[0058]

[0059] Since SEI film growth is the main factor causing capacity decay in the early cycles of NMC batteries, it is believed that the output of the mixed mechanism model after parameter fitting is the theoretical capacity loss caused by SEI film growth.

[0060] Operation S4: Sliding window to obtain linear correlation equations in different time windows. When the slope of the linear correlation equation reaches the minimum value, an NMC battery diving point warning is issued.

[0061] Specifically, the lengths of the theoretical capacity decay vector and the actual capacity decay vector are fixed, and through the sliding window interception method, the slope and intercept of the linear correlation equation corresponding to the theoretical capacity decay value and the actual capacity decay value in each window are continuously obtained to form a slope signal and an intercept signal. The slope signal is regarded as an online warning signal for the diving point. Whether the slope signal reaches the minimum value is used as the warning standard. When the slope signal reaches the minimum value, a warning alarm is issued, realizing an online warning of the diving point of the NMC battery.

[0062] In order to verify the effectiveness of the NMC battery diving point online warning method based on the hybrid mechanism model in the embodiment of the present invention, the diving point warning performance of 47 Sanyo / Panasonic 18650 lithium manganese cobalt oxide (NMC) / Gr batteries in the Aachen University of Technology dataset was verified. Among them, the battery has a nominal capacity of 1.85Ah and a cut-off voltage of 3V and 4.1V, respectively, and was tested under the same aging conditions at 25°C. In the cycle test, the battery was charged and discharged at a current of 4A between 3.5V and 3.9V for 30 minutes under the constant current constant voltage (CC-CV) state.

[0063] The final warning effect on the data set is as follows Figure 4 See Figure 4 As can be seen, this method has a high success rate in online early warning of NMC lithium battery diving points, and the warning deviation is also within a reasonable range. In summary, this method can be used to online warn of the occurrence of NMC lithium battery diving points, and the warning signal is characterized by real-time acquisition and strong physical interpretability. This fills the gap in online early warning of NMC battery diving points and can be widely used in battery health management systems to ensure the safe and reliable operation of lithium batteries.

[0064] Example 2

[0065] An online warning system for NMC battery water-diving points based on a hybrid mechanism model includes a processor and a memory storing a computer-executable program. When executed by the processor, the program causes the processor to execute the online warning method for NMC battery water-diving points based on the hybrid mechanism model. The related technical solutions are the same as those in Example 1 and are not further described here.

[0066] Example 3

[0067] A computer-readable storage medium stores a computer program that, when executed by a processor, implements the aforementioned online warning method for NMC battery water-diving points based on a hybrid mechanism model. The related technical solutions are the same as those in Example 1 and will not be further elaborated here.

[0068] It will be easily understood by those skilled in the art that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An online early warning method for NMC battery diving point based on a hybrid mechanism model, characterized in that: It includes the model building phase S1-S2 and the online warning phase S3-S4; S1, considering the effects of SEI film growth, lithium plating, electrolyte oxidation, and manganese decomposition on the capacity decay of NMC batteries, a hybrid mechanism model is established to describe the total capacity loss of NMC batteries over time; S2, fitting the hybrid mechanism model using an early NMC battery capacity loss signal consisting of a total capacity loss value of the NMC battery over a number of initial cycles to determine parameters of the hybrid mechanism model; S3, measuring the capacity decay real vector of the NMC battery within the time window, calculating the capacity decay theoretical vector of the NMC battery within the time window using a hybrid mechanism model determined by parameters, and obtaining a linear correlation equation between the capacity decay theoretical vector and the capacity decay real vector within the time window; S4, the sliding window obtains the linear correlation equation in different time windows. When the slope of the linear correlation equation reaches the minimum value, an NMC battery diving point warning is issued.

2. The NMC battery diving point online early warning method based on the hybrid mechanism model according to claim 1 is characterized in that: The effect of SEI film growth on NMC battery capacity decay is expressed as: Among them, Q sei The capacity decay of NMC battery caused by SEI film growth, A sei is the first coefficient, B sei is the second coefficient, and t is the time.

3. The NMC battery diving point online early warning method based on the hybrid mechanism model according to claim 1 is characterized in that: The effect of lithium plating on the capacity decay of NMC batteries is expressed as: Q lp =B lp t 1.5 Among them, Q lp NMC battery capacity decay caused by lithium plating, B lp is the third coefficient, and t is the time.

4. The online early warning method for NMC battery diving point based on a hybrid mechanism model according to claim 1 is characterized in that: The effect of electrolyte oxidation on NMC battery capacity decay is expressed as: Q eox =C eox t Among them, Q eox The capacity of NMC battery is reduced due to electrolyte oxidation, C eox is the fourth coefficient, and t is the time.

5. The online early warning method for NMC battery diving point based on a hybrid mechanism model according to claim 1 is characterized in that: The effect of manganese decomposition on the capacity decay of NMC batteries is expressed as: Q Mn =C Mn t+B Mn t 1.5 Among them, Q Mn The capacity decay of NMC battery caused by manganese decomposition, C Mn is the fifth coefficient, B Mn is the sixth coefficient, and t is the time.

6. The online early warning method for NMC battery diving point based on a hybrid mechanism model according to any one of claims 1 to 5, characterized in that: The mixing mechanism model is: Among them, Q tot is the total capacity loss of NMC battery, Q sei The capacity decay of NMC battery caused by SEI film growth, Q lp The capacity of NMC battery is reduced due to lithium plating. eox The capacity of NMC battery is reduced due to electrolyte oxidation. Mn is the capacity decay of NMC battery caused by manganese decomposition, A, B, and C are the three parameters of the mixed mechanism model, and t is time.

7. An online warning system for NMC battery diving point based on a hybrid mechanism model, characterized by: include: processor; A memory storing a computer executable program, wherein when the program is executed by the processor, the processor executes the online early warning method for the NMC battery diving point based on the hybrid mechanism model according to any one of claims 1 to 6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, an online early warning method for an NMC battery diving point based on a hybrid mechanism model according to any one of claims 1 to 6 is implemented.

Images