Liquid metal battery capacity sudden drop early warning method, equipment and medium

Abstract

The invention relates to the field of battery capacity monitoring, and discloses a liquid metal battery capacity sudden drop early warning method and device and a medium, and the method comprises the steps: collecting the data of a liquid metal battery, and obtaining a battery capacity attenuation track; decomposing the battery capacity data by using an empirical model decomposition method to obtain components of local feature signals containing different time scales, and extracting aging features; for different aging characteristics, a Gaussian mixture model is used for describing probability density distribution of the aging characteristics, and the probability density distribution is in a hump shape; jS divergence is used for quantitatively describing hump evolution rules of different aging degrees so as to determine the capacity sudden drop characteristic of the battery, and early warning is carried out according to the characteristic; the problem that a traditional slope method is difficult to apply to the field of liquid metal batteries is solved. And meanwhile, sudden reduction of the capacity of the liquid metal battery is warned in advance, further management and regulation of the battery can be facilitated, and a foundation is laid for large-scale application of the battery.

Description

Technical Field

[0001] The present invention relates to the field of battery capacity monitoring, and in particular to a method, device and medium for early warning of sudden capacity drop of a liquid metal battery. Background Art

[0002] Liquid metal battery is a new type of energy storage technology used in large-scale energy storage. Compared with traditional solid-state electrode energy storage batteries, it has the advantages of high safety, long life and low price. At the working temperature, the negative electrode, alloy positive electrode and molten salt electrolyte of the liquid metal battery are all liquid, and they are automatically stratified due to density differences. This structure that abandons solid-state electrodes saves the liquid metal battery from aging processes such as lithium precipitation and changes in electrode structure, ensuring the long life of the liquid metal battery, while being easy to scale up and having stable performance.

[0003] However, although liquid metal batteries have an ultra-long cycle life, their aging trajectory will have a sudden capacity drop during the aging process. This sudden capacity drop indicates that the aging process of liquid metal batteries is nonlinear and complex, which poses a major challenge to the management and regulation of liquid metal batteries. Although there have been a series of studies on the capacity drop process in the field of lithium-ion batteries, the slope characteristic method commonly used in the field of lithium-ion batteries is not suitable for studying the capacity drop process of liquid metal batteries. On the one hand, the capacity drop mechanism and data characteristics of liquid metal batteries are significantly different from those of lithium-ion batteries. Unlike lithium-ion batteries, where the capacity drop is caused by the gradual accumulation of aging mechanisms such as electrolyte decomposition and electrode changes, the capacity drop of liquid metal batteries is due to the formation of irreversible solid substances, which consumes active materials and causes a significant drop in capacity, so the capacity drop process of liquid metal batteries is more sudden. On the other hand, the impact of capacity drop on battery life is different. After the capacity drop, the liquid metal battery can still continue to operate, so its slope will not continue to increase after the capacity drop. Therefore, the traditional slope analysis method is difficult to apply in the field of liquid metal batteries. Summary of the invention

[0004] The purpose of the present invention is to propose a method for early warning of sudden capacity drop of a liquid metal battery, so as to solve the technical problem that the prior art lacks capacity monitoring for liquid metal batteries.

[0005] Specifically, the present invention provides a liquid metal battery capacity sudden drop warning method, comprising the following steps:

[0006] S1. Collect liquid metal battery data and obtain battery capacity decay trajectory;

[0007] S2: Decompose the battery capacity data using the empirical model decomposition method to obtain components containing local characteristic signals at different time scales and extract aging characteristics;

[0008] S3: For different aging characteristics, a mixed Gaussian model is used to describe their probability density distribution, which is hump-shaped;

[0009] S4: Use JS divergence to quantitatively describe the hump evolution law of different aging degrees to determine the capacity drop characteristics of the battery and issue an early warning based on this characteristic.

[0010] A storage medium stores instructions and data for implementing a liquid metal battery capacity sudden drop warning method.

[0011] A liquid metal battery capacity sudden drop warning device comprises: a processor and the storage medium; the processor loads and executes instructions and data in the storage medium to implement a liquid metal battery capacity sudden drop warning method.

[0012] The beneficial effects provided by the present invention are: a data-driven early warning method for the sudden drop in the capacity of a liquid metal battery, such as first using the empirical mode decomposition method (EMD) to decompose the capacity data of the liquid metal battery, and then using the Gaussian mixture model (GMM) to construct the distribution of each aging feature; finally, using the Jensen-Shannon (JS) divergence method to quantify the evolution of the aging characteristics to give an early warning of the sudden drop in capacity, ultimately solving the problem that the traditional slope method is difficult to apply to the field of liquid metal batteries. At the same time, early warning of the sudden drop in the capacity of the liquid metal battery can help further management and regulation of the battery, paving the way for its large-scale application. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a schematic flow chart of the method of the present invention;

[0014] Figure 2 Capacity trajectory diagrams of 20Ah and 50Ah Li||Sb4Sn6 provided by the present invention;

[0015] Figure 3 The 20Ah and 50Ah battery capacity sudden drop early warning diagram provided by the present invention;

[0016] Figure 4 It is a schematic diagram of the working of the hardware device of an embodiment of the present invention. DETAILED DESCRIPTION

[0017] To make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0018] Before formally describing the present invention, the scheme of the present invention is first generally described for easy understanding.

[0019] Please refer to Figure 1 The present invention provides a liquid metal battery capacity sudden drop warning method, comprising the following steps:

[0020] S1. Collect liquid metal battery data and obtain battery capacity decay trajectory;

[0021] As an example, in S1, the aging data used in the present invention is collected from 20Ah Li||Sb4Sn6 and 50Ah Li||Sb4Sn6 liquid metal batteries. Since liquid metal batteries have an ultra-long cycle life, the present invention adopts an accelerated aging test scheme that quantifies the number of cycles of accelerated aging and then calibrates the actual capacity of the battery. Preferably, the present invention uses a high-rate charge and discharge current to cycle 100 cycles, and then low-rate charge and discharge to obtain the actual capacity of the battery in its current state.

[0022] S2: Decompose the battery capacity data using the empirical model decomposition method to obtain components containing local characteristic signals at different time scales and extract aging characteristics;

[0023] As an example, in S2, EMD is used to decompose the battery capacity attenuation trajectory. The EMD method can decompose a complex signal into a finite number of intrinsic mode functions (IMFs) and a residual component (Residual, Res). At the same time, combined with the aging characteristics of liquid metal batteries, the sum of all IMFs is used as the first feature, and Res is used as the second feature.

[0024] S3: For different aging characteristics, a mixed Gaussian model is used to describe their probability density distribution, which is hump-shaped;

[0025] As an embodiment, in S3, a Gaussian mixture model GMM is used to obtain the probability density distribution between the first feature and the second feature.

[0026] In the mixed Gaussian model, it is assumed that each data point in the dataset is generated by a Gaussian distribution, and these Gaussian distributions are mixed together with a certain probability. In other words, the mixed Gaussian model assumes that the data is a mixture of multiple Gaussian distributions.

[0027] Gaussian distribution is the most common continuous probability distribution in statistics and probability theory, and its PDF is described as follows:

[0028]

[0029] Among them, μ, σ, σ 2represent mean, standard deviation and variance respectively.

[0030] GMM consists of several Gaussian distributions, each of which is called a component and has its own unique mean and covariance. The PDF of the Gaussian mixture model can be expressed by formula (2):

[0031]

[0032] Where n represents the number of Gaussian distributions, π i , μ i ,∑ i is the weight, mean, and covariance matrix of the i-th Gaussian distribution, and the sum of the weights of all components is 1.

[0033] S4: Use JS divergence to quantitatively describe the hump evolution law of different aging degrees to determine the capacity drop characteristics of the battery and issue an early warning based on this characteristic.

[0034] It should be noted that in S4, the use of JS divergence is an effective method for measuring the difference or similarity between two probability distributions. JS divergence combines the advantages of Kullback-Leibler (KL) divergence and provides a better measurement method to measure the similarity between two probability distributions through the improvement of symmetry and non-negativity. Its formula is as follows:

[0035]

[0036] Where M is the mean of P and Q, and JS divergence is calculated by first calculating the KL divergence between P and Q and their mean distributions, and then taking the average of these two KL divergences. JS divergence is symmetrical, and the range of JS divergence is between 0 and 1. The closer the value is to 0, the more similar the two probability distributions are, and the closer the value is to 1, the greater the difference between the two probability distributions.

[0037] In order to observe the change of JS divergence more clearly, the overlap rate O is used rate To describe the similarity between these humps, it is defined as follows:

[0038] O rate =1-D JS (4)

[0039] In step S4, when the overlap ratio O rate When the preset value is reached, it indicates that the capacity of the liquid metal battery has dropped suddenly.

[0040] Example:

[0041] The present invention preferably uses a 20Ah and 50Ah liquid metal battery;

[0042] Please refer to Figure 2 , Figure 2 The capacity attenuation trajectory diagram of the 20Ah and 50Ah liquid metal batteries provided by the present invention shows that under most aging conditions, the slope curve maintains a relatively stable fluctuation. Only at the capacity drop point does the slope drop sharply, indicating the importance of early warning of the capacity drop process.

[0043] Please refer to Figure 3 , Figure 3 This is the early warning result of the capacity drop process based on the data-driven method provided by the present invention. As can be seen from the figure, whether it is a 20Ah or 50Ah liquid metal battery, under most aging conditions, the slope curve remains relatively stable, and only drops sharply at the capacity drop point. This shows that slope analysis can only detect data changes at the moment of capacity drop. In contrast, the data-driven method proposed in the present invention captures the changing characteristics of the data before the capacity drops suddenly, proving the rationality of using data decomposition to analyze the battery capacity drop process.

[0044] In summary, the data method provided by the present invention can effectively quantify the capacity change of the liquid metal battery before the capacity suddenly drops, thereby warning of the sudden drop in capacity, which is helpful for accurate prediction of battery life and active regulation in practical applications.

[0045] See also Figure 4 , Figure 4 4 is a schematic diagram of the working of the hardware device of an embodiment of the present invention, wherein the hardware device specifically comprises: a liquid metal battery capacity sudden drop warning device 401, a processor 402 and a storage medium 403.

[0046] A liquid metal battery capacity sudden drop warning device 401: the liquid metal battery capacity sudden drop warning device 401 implements the liquid metal battery capacity sudden drop warning method.

[0047] Processor 402: The processor 402 loads and executes the instructions and data in the storage medium 403 to implement the liquid metal battery capacity sudden drop warning method.

[0048] Storage medium 403: The storage medium 403 stores instructions and data; the storage medium 403 is used to implement the liquid metal battery capacity sudden drop warning method.

[0049] The beneficial effects of the present invention are: a data-driven early warning method for the sudden drop in the capacity of a liquid metal battery, such as first using the empirical mode decomposition method (EMD) to decompose the capacity data of the liquid metal battery, and then using the Gaussian mixture model (GMM) to construct the distribution of each aging feature; finally, using the Jensen-Shannon (JS) divergence method to quantify the evolution of the aging characteristics to give an early warning of the sudden drop in capacity, ultimately solving the problem that the traditional slope method is difficult to apply to the field of liquid metal batteries. At the same time, early warning of the sudden drop in the capacity of the liquid metal battery can help further management and regulation of the battery, paving the way for its large-scale application.

[0050] 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 spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A liquid metal battery capacity sudden drop warning method, characterized by: The method comprises the following steps: S1. Collect liquid metal battery data and obtain battery capacity decay trajectory; S2: Decompose the battery capacity data using the empirical model decomposition method to obtain components containing local characteristic signals at different time scales and extract aging characteristics; S3: For different aging characteristics, a mixed Gaussian model is used to describe their probability density distribution, which is hump-shaped; S4: Use JS divergence to quantitatively describe the hump evolution law of different aging degrees to determine the capacity drop characteristics of the battery and issue an early warning based on this characteristic.

2. A liquid metal battery capacity sudden drop warning method as claimed in claim 1, characterized in that: In step S1, the liquid metal battery is charged and discharged at a high rate for a preset number of cycles, and then charged and discharged at a low rate to obtain the actual capacity of the battery in the current state, thereby obtaining the liquid metal battery data and attenuation trajectory.

3. A liquid metal battery capacity sudden drop warning method as claimed in claim 1, characterized in that: Step S2 is specifically as follows: S21, using the empirical model decomposition (EMD) method to decompose the liquid metal battery data into a finite number of intrinsic mode functions (IMFs) and a residual component (Res); S22. The aging feature includes a first feature and a second feature, the first feature is the sum of the intrinsic mode functions, and the second feature is the residual component Res.

4. A liquid metal battery capacity sudden drop warning method as claimed in claim 1, characterized in that: In step S3, a Gaussian mixture model GMM is used to obtain the probability density distribution between the first feature and the second feature.

5. A liquid metal battery capacity sudden drop warning method as claimed in claim 1, characterized in that: The sudden drop characteristics in step S4 are: overlap rate O rate , the calculation formula is as follows: rate =1-D JS ; where D JS is the JS divergence.

6. A liquid metal battery capacity sudden drop warning method as claimed in claim 5, characterized in that: In step S4, when the overlap ratio O rate When the preset value is reached, it indicates that the capacity of the liquid metal battery has dropped suddenly.

7. A storage medium, characterized in that: The storage medium stores instructions and data for implementing a liquid metal battery capacity sudden drop warning method as described in any one of claims 1 to 6.

8. A liquid metal battery capacity sudden drop warning device, characterized by: include: A processor and a storage medium; the processor loads and executes instructions and data in the storage medium to implement a liquid metal battery capacity sudden drop warning method as described in any one of claims 1 to 6.

Images