Methods, systems, devices, and media for estimating lithium-ion battery state of health

Abstract

This invention relates to a method, system, device, and medium for estimating the state of health (SQU) of lithium-ion batteries. The method establishes a database of voltage data for lithium-ion batteries under constant power charge-discharge conditions throughout their entire lifespan. The battery management unit collects and preprocesses voltage data from actual use. A dynamic time programming algorithm is used to calculate the similarity between the preprocessed data and the data in the database, obtaining the minimum similarity distance, defined as the optimal similar health state. By combining the health states during charge-discharge processes and assigning weights using an importance function, the comprehensive health state of the battery cell is calculated. This method achieves a more comprehensive and accurate estimation of the health state of lithium-ion batteries under full capacity and all operating conditions, thereby improving the stability and sustainability of energy storage systems.

Description

Technical Field

[0001] This invention relates to battery management system technology and the field of renewable energy storage, specifically to a method, system, device, and medium for estimating the health status of lithium-ion batteries. Background Technology

[0002] With the rapid development of renewable energy and the construction of smart grids, the demand for stability and sustainability of energy storage systems has further increased. Among various energy storage technologies, lithium-ion batteries have attracted much attention due to their high energy density and long cycle life. However, the difficulty in measuring the internal electrochemical characteristics of lithium-ion batteries increases the difficulty in estimating their State of Health (SOH), affecting the application of battery systems over long time scales and restricting the further development of energy storage systems.

[0003] Accurate State of Health (SOH) estimation for lithium-ion batteries is one of the core objectives of the entire battery control algorithm. Existing active computing schemes mainly focus on local features of battery external characteristics, employing big data models such as neural networks as the basic algorithm, and training reliable and accurate SOH estimation models through a large amount of experimental data. Dynamic Time Warping (DTW) algorithms are widely used in the field of speech recognition. They can perform comparative matching of long and short speech segments, complete feature similarity calculation of asymmetric time data sequences, reduce the number of samples in the data analysis process, and reduce the computational power requirements.

[0004] Existing technologies have applied the DTW algorithm to SOH estimation of lithium-ion batteries. These schemes target on-board power lithium-ion batteries, applying the scenarios of dynamic discharge and constant-current / constant-voltage standard charging. The collected data primarily consists of voltage data during the stable constant-current / constant-voltage charging phase, with some battery discharge voltage data also collected. Considering that on-board batteries in practical applications cannot cover the entire battery capacity, and that using only partial data for similarity calculation cannot guarantee an accurate representation of the battery's health status, it is necessary to extract Pearson coefficients, Spearman coefficients, and Kendall coefficients as SOH reference values ​​in addition to similarity calculations. This increases the computational and time complexity of the algorithm, and its suitability for SOH estimation of energy storage platform batteries operating primarily under constant-power charging and discharging conditions within the full battery capacity needs improvement.

[0005] However, simply introducing the relevant design scheme into the energy storage system will face the following problems:

[0006] 1. The existing solution only applies to a portion of the data. If the existing solution is continued, the full potential of the data cannot be explored, and the system complexity of calculating the battery health status will be increased.

[0007] 2. Existing solutions use the calculated health status value of a cell under a specific working condition as the actual health status value, while ignoring the impact of the cell's resting period on the battery's health status. Summary of the Invention

[0008] This invention provides a method, system, device, and medium for estimating the health status of lithium-ion batteries. Its purpose is to overcome the limitations of existing technologies that rely only on partial data and specific operating conditions, and to achieve a more comprehensive and accurate estimation of the full capacity and health status of lithium-ion batteries under all operating conditions, thereby improving the stability and sustainability of energy storage systems.

[0009] To achieve the above objectives, the first aspect of the present invention provides a method for estimating the state of health of a lithium-ion battery, comprising the following steps:

[0010] Acquire voltage data of lithium-ion batteries under constant power charge and discharge conditions throughout their entire life cycle, and establish a database to form a one-to-one correspondence between voltage data and battery life.

[0011] The battery management unit collects voltage data of the battery cells during constant power charging in actual use;

[0012] The voltage data collected during the charging process is preprocessed;

[0013] The similarity between the preprocessed voltage data and the voltage data sequence in the database is calculated using a dynamic time planning algorithm to obtain the minimum similarity distance. The health state corresponding to the minimum similarity distance is defined as the best similar health state during charging.

[0014] The voltage data of the same cell during the next adjacent constant power discharge process is collected in the battery management unit;

[0015] The voltage data collected during the discharge process is preprocessed;

[0016] The similarity between the preprocessed voltage data and the voltage data sequence in the database is calculated using a dynamic time programming algorithm to obtain the minimum similarity distance. The health state corresponding to the minimum similarity distance is defined as the best similar health state during discharge.

[0017] Combining the best similar health state during charging and the best similar health state during discharging, an importance function is used to assign importance coefficients to the best similar health states during charging and discharging, respectively, and the sum of these two coefficients is 1.

[0018] Based on the allocation results, the overall health status of the battery cell is calculated, and the overall health status is used as the current health status assessment of the battery cell.

[0019] Furthermore, the preprocessing of the voltage data collected during the charging process and the voltage data collected during the discharging process includes filtering the voltage data to remove noise and adding time stamps.

[0020] Furthermore, methods for calculating the similarity of voltage data sequences using dynamic time programming algorithms include:

[0021] Initialize a similarity matrix with time and category dimensions. Set the initial value of each element in the similarity matrix to infinity and set the starting element of the similarity matrix to 0 to begin similarity calculation.

[0022] Furthermore, the specific formula for calculating each element in the similarity matrix is ​​as follows:

[0023]

[0024] For the remaining elements:

[0025] D(i,j)=|V(i)-V m (j)|+min(D(i-1,j),D(i,j-1),D(i-1,j-1))

[0026] Where V(i) represents the voltage value at the i-th point of the preprocessed voltage data, V m (j) represents the voltage value of the j-th point in the voltage data in the database; D represents the similarity matrix, which is a two-dimensional array used to store intermediate similarity values ​​during the calculation process. The size of the similarity matrix is ​​m×n, where m and n are the control sequence V and the voltage value of the control sequence V, respectively. m The length of the input sequence V, and D(i,j) represent the element in the i-th row and j-th column of the matrix, representing the first i elements of the input sequence V and the comparison with the control sequence V. m The cumulative minimum similarity distance between the first j elements;

[0027] By finding the similarity matrix D m×n The minimum value in the equation is used to determine the best similarity match between two time series, thereby assessing the health status of the battery.

[0028] Furthermore, methods for calculating the overall health status of the battery cell include:

[0029] Provide an importance function that assigns importance coefficients to the health status during charging and discharging processes;

[0030] The importance coefficient needs to be determined by combining the actual engineering application scenario and battery characteristics, while ensuring that their sum is 1.

[0031] The best similar health state during charging and the best similar health state during discharging are weighted using an importance coefficient, as shown in the formula:

[0032]

[0033] Where, α ch α represents the importance coefficient of the health status during the charging process. dis SOH is the importance coefficient of the health status during the discharge process. ch For optimal similar health status during charging, SOH dis This represents the optimal similar health state during discharge.

[0034] The calculated SOH value is output as the current comprehensive health status assessment value of the battery cell.

[0035] Furthermore, data acquisition involves real-time monitoring of battery status via sensors.

[0036] Furthermore, the established database includes standard charge and discharge voltage data under 25-degree Celsius temperature conditions.

[0037] To achieve the above objectives, a second aspect of the present invention provides a system for estimating the state of health of a lithium-ion battery, comprising the following modules:

[0038] The data acquisition module is used to acquire the voltage data of lithium-ion batteries under constant power charge and discharge conditions throughout their entire life cycle, and to establish a database to record the one-to-one correspondence between voltage data and battery health status.

[0039] The voltage acquisition module is located in the battery management unit and is used to collect voltage data of the battery cell during constant power charging and discharging in actual use.

[0040] The data preprocessing module is used to preprocess the voltage data collected during the charging and discharging process to ensure that the data is suitable for subsequent analysis.

[0041] The similarity calculation module, based on the dynamic time planning algorithm, calculates the similarity between the preprocessed voltage data and the voltage data sequence in the database, obtains the minimum similarity distance, and defines the optimal similar health state during charging and discharging based on the battery health state corresponding to the minimum similarity distance.

[0042] The health status assessment module combines the best similar health status during charging and the best similar health status during discharging; it uses an importance function to assign importance coefficients to the best similar health status during charging and discharging, while ensuring that the sum of these two coefficients is 1.

[0043] The comprehensive health status calculation module calculates the comprehensive health status of the battery cell based on the assigned importance coefficients and outputs it as the current health status assessment result of the battery cell.

[0044] To achieve the above objectives, a third aspect of the present invention provides an electronic device, the electronic device comprising: a processor coupled to a memory;

[0045] The memory is used to store computer programs;

[0046] The processor is configured to execute the computer program stored in the memory, so that the electronic device performs the method.

[0047] To achieve the above objectives, a fourth aspect of the present invention provides a computer-readable storage medium including instructions that, when executed on a computer, cause the computer to perform the method described thereon.

[0048] The beneficial effects of this invention are:

[0049] Compared with existing technologies, this invention provides a method, system, device, and medium for estimating the state of health (SOH) of lithium-ion batteries. By employing a Dynamic Time Programming (DTW) algorithm to accurately estimate the SOH from full voltage data, and comprehensively considering the impact of lithium-ion batteries during resting and cyclic aging, it overcomes the limitations of existing technologies that rely only on partial data and specific operating conditions. This technical solution can comprehensively collect and analyze complete voltage data of batteries in energy storage systems under constant power charge and discharge conditions, utilize the DTW algorithm to evaluate the similarity between data points, and thus accurately match the actual health state of the battery. Furthermore, through an importance function, this invention can also rationally allocate the influence weights of charging and discharging states, optimizing the health state assessment of batteries under unconventional operating environments. This method not only improves the accuracy of SOH estimation but also reduces the complexity of the algorithm and computational requirements, making it more suitable for real-time monitoring and management of battery health status in large-scale energy storage systems. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below.

[0051] Figure 1 This is a flowchart of a method for estimating the health status of a lithium-ion battery, as disclosed in an embodiment of the present invention.

[0052] Figure 2 This is a flowchart of a DTW algorithm disclosed in an embodiment of the present invention.

[0053] Figure 3 This is a pseudocode algorithm structure diagram of DTW disclosed in an embodiment of the present invention.

[0054] Figure 4 This is a graph showing the results of 800 cycles of DTW calculation disclosed in an embodiment of the present invention. Detailed Implementation

[0055] The technical solutions of the embodiments of this application will now be described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0056] like Figure 1 As shown, the present invention provides a method for estimating the state of health of a lithium-ion battery, comprising the following steps:

[0057] Step S100: Obtain the voltage data of the lithium-ion battery under constant power charge and discharge conditions throughout its entire life cycle, and establish a database to form a one-to-one correspondence between voltage data and battery life.

[0058] Step S200: Collect voltage data of the battery cell during constant power charging in actual use in the battery management unit;

[0059] Step S300: Preprocess the voltage data collected during the charging process;

[0060] Step S400: Use a dynamic time planning algorithm to calculate the similarity between the preprocessed voltage data and the voltage data sequence in the database, thereby obtaining the minimum similarity distance, and define the health state corresponding to the minimum similarity distance as the best similar health state during charging.

[0061] Step S500: Collect voltage data of the same cell during the next adjacent constant power discharge process in the battery management unit;

[0062] Step S600: Preprocess the voltage data collected during the discharge process;

[0063] Step S700: Use a dynamic time programming algorithm to calculate the similarity between the preprocessed voltage data and the voltage data sequence in the database, thereby obtaining the minimum similarity distance, and define the health state corresponding to the minimum similarity distance as the best similar health state during discharge.

[0064] Step S800: Combining the best similar health state during charging and the best similar health state during discharging, use an importance function to assign importance coefficients to the best similar health states during charging and discharging, and make the sum of the two coefficients equal to 1.

[0065] Step S900: Based on the allocation results, calculate the overall health status of the battery cell and use the overall health status as the current health status assessment of the battery cell.

[0066] In this embodiment, as described in steps S100-S700 above, the similarity between the collected voltage data and the voltage data in the database is calculated to obtain the similarity distance between the collected data and the data under different health states. The health state represented by the smallest similarity distance is the best similar health state. The specific process is as follows:

[0067] Firstly, during the cell testing phase, it's necessary to acquire voltage data from constant-power charging and discharging throughout the entire battery lifecycle and establish a database to establish a one-to-one correspondence between different battery health states and voltage data. Subsequently, the health state data in this database will be used as standard data for matching with data collected during actual use. In practical applications, the Battery Management Unit (BMU) can obtain all voltage data of the cell during constant-power charging. This data is collected, stored, and preprocessed to remove outliers and generate time tags. The similarity between the data sequence and the charging data sequence in the database is calculated using the DTW algorithm. The minimum similarity distance is obtained, and the battery health state (SOH) corresponding to this similarity is the optimal similarity health state (SOH) for that cell considering the charging data at that moment. ch .

[0068] Similarly, constant power discharge voltage data within the same charge-discharge cycle is collected, stored, and preprocessed to remove outliers and generate time tags. The similarity between the data sequence and the discharge data sequence in the database is calculated using the DTW algorithm. The minimum similarity distance is then obtained, and the battery health state corresponding to this similarity is the optimal similar health state (SOH) for that cell considering the discharge data at that time. dis .

[0069] In this embodiment, as described in steps S800 and S900 above, the optimal similar health state of the battery is obtained by calculating and processing the constant power charge and discharge data of one cycle. Since there is a resting condition between the charging and discharging conditions, i.e., there is a resting aging condition between two cycle aging, in order to comprehensively consider the impact of cycle aging and resting aging on the battery health state, importance values ​​are assigned to the two optimal similar health states. This value is related to the material characteristics and discharge characteristics of the battery cell and needs to be analyzed according to different engineering applications. The specific allocation formula is as follows:

[0070]

[0071] Where, α ch α represents the importance coefficient of the health status during the charging process. disSOH is the importance coefficient of the health status during the discharge process. ch For optimal similar health status during charging, SOH dis This represents the optimal similar health state during discharge. To ensure the reasonableness of the upper and lower limits of battery health state, the sum of the two coefficients must be 1. However, by calculating the importance function allocation, a more accurate overall battery health state that reflects actual operating characteristics can be obtained.

[0072] It should be noted that the DTW algorithm is a dynamic time programming algorithm used to calculate the similarity between two sets of asymmetric sequence data. The data matrix to be calculated is defined as V = [V(1)V(2)...V(n)], and the comparison data matrix is ​​defined as V... m =[V m (1)V m (1)...V m (m)], the specific process is as follows Figure 2 As shown.

[0073] First, the entire similarity matrix is ​​initialized as follows:

[0074]

[0075] In the matrix, m and n represent the time dimension and category dimension of the data, respectively. Then, local similarity is calculated as follows:

[0076]

[0077] The results for the remaining positions of the similarity matrix are then calculated as follows:

[0078] D(i,j)=|V(i)-V m (j)|+min(D(i-1,j),D(i,j-1),D(i-1,j-1))

[0079] Where V(i) represents the voltage value at the i-th point of the preprocessed voltage data, V m (j) represents the voltage value of the j-th point in the voltage data in the database; D represents the similarity matrix, which is a two-dimensional array used to store intermediate similarity values ​​during the calculation process. The size of the similarity matrix is ​​m×n, where m and n are the control sequence V and the voltage value of the control sequence V, respectively. m The length of the input sequence V, and D(i,j) represent the element in the i-th row and j-th column of the matrix, representing the first i elements of the input sequence V and the comparison with the control sequence V. m The cumulative minimum similarity distance between the first j elements.

[0080] After calculating the complete similarity matrix, the similarity distance between the two sets of sequence data can be obtained. This distance can be used to evaluate the similarity between the two sets of data. The specific algorithm structure is as follows: Figure 3 As shown.

[0081] This invention employs the DTW algorithm to effectively match the actual cycle aging number of a battery, i.e., its health status, based on constant power charge-discharge full voltage data. Specifically, it calculates the minimum DTW distance between battery discharge voltage data from 800 cycles and a pre-built discharge cycle voltage database. The results are as follows: Figure 4 As shown.

[0082] It can be observed that the DTW algorithm can accurately locate the actual aging state of the battery. The minimum DTW distance occurs at 800 cycles. At the same time, the DTW distance increases with the distance from the actual aging cycle number, thus verifying the robustness of the algorithm.

[0083] The DTW positioning results of the charging and discharging processes are assigned using an importance formula, which can optimize the impact of unconventional operating conditions on the battery health status, such as the charging and discharging process at high and low temperatures, the resting process, and the overcurrent charging and discharging process. In particular, the resting process can be combined with the positioning results after resting based on the resting aging and cyclic working aging phenomena to obtain the comprehensive health status of the battery.

[0084] In summary, in this embodiment, by using full voltage data as the basic data for DTW, the battery voltage characteristic information is fully explored, and the simple secondary processing of health status reduces the computational complexity of the system. In addition, the impact of static aging and cyclic aging under energy storage conditions is fully considered, and the combination of the two yields a comprehensive health status that better reflects the health status of the battery cell.

[0085] The method provided in this application can be applied to various electronic devices. These electronic devices may include, but are not limited to, smartphones, televisions, tablet computers, tablet PCs, laptop computers (also known as notebook computers or handheld computers), and personal computers (PCs). Of course, the specific form of the electronic device is not limited in the following embodiments.

[0086] For example, the electronic device includes: a processor coupled to a memory; the memory for storing a computer program; and the processor for executing the computer program stored in the memory to cause the electronic device to perform the method.

[0087] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware, or it can be implemented by special-purpose hardware including application-specific integrated circuits, special-purpose CPUs, special-purpose memory, special-purpose components, etc. Generally, any function performed by a computer program can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can also be diverse, such as analog circuits, digital circuits, or special-purpose circuits. However, for this application, software program implementation is more often the preferred implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium, such as a computer floppy disk, USB flash drive, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0088] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. The computer-readable storage medium can be any available medium that a computer can store, or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., a solid-state disk (SSD)).

Claims

1. A method for estimating the state of health of a lithium-ion battery, characterized in that, Includes the following steps: Acquire voltage data of lithium-ion batteries under constant power charge and discharge conditions throughout their entire life cycle, and establish a database to form a one-to-one correspondence between voltage data and battery life. The battery management unit collects voltage data of the battery cells during constant power charging in actual use; The voltage data collected during the charging process is preprocessed; The similarity between the preprocessed voltage data and the voltage data sequence in the database is calculated using a dynamic time planning algorithm to obtain the minimum similarity distance. The health state corresponding to the minimum similarity distance is defined as the best similar health state during charging. The voltage data of the same cell during the next adjacent constant power discharge process is collected in the battery management unit; The voltage data collected during the discharge process is preprocessed; The similarity between the preprocessed voltage data and the voltage data sequence in the database is calculated using a dynamic time programming algorithm to obtain the minimum similarity distance. The health state corresponding to the minimum similarity distance is defined as the best similar health state during discharge. Combining the best similar health state during charging and the best similar health state during discharging, an importance function is used to assign importance coefficients to the best similar health states during charging and discharging, respectively, and the sum of these two coefficients is 1. Based on the allocation results, the overall health status of the battery cell is calculated, and the overall health status is used as the current health status assessment of the battery cell.

2. The method for estimating the state of health of a lithium-ion battery as described in claim 1, characterized in that, Preprocessing the voltage data collected during the charging process and the voltage data collected during the discharging process includes filtering the voltage data to remove noise and adding time stamps.

3. The method for estimating the state of health of a lithium-ion battery as described in claim 1, characterized in that, Methods for calculating the similarity of voltage data sequences using dynamic time programming algorithms include: Initialize a similarity matrix with time and category dimensions. Set the initial value of each element in the similarity matrix to infinity and set the starting element of the similarity matrix to 0 to begin similarity calculation.

4. The method for estimating the state of health of a lithium-ion battery as described in claim 3, characterized in that, The steps for calculating each element in the similarity matrix, and the specific calculation formula, are as follows: For the remaining elements: D(i,j)=|V(i)-V m (j)+min(D(i-1,j),D(i,j-1),D(i-1,j-1)) Where V(i) represents the voltage value at the i-th point of the preprocessed voltage data, V m (j) represents the voltage value of the j-th point in the voltage data in the database; D represents the similarity matrix, which is a two-dimensional array used to store intermediate similarity values ​​during the calculation process. The size of the similarity matrix is ​​m×n, where m and n are the control sequence V and the voltage value of the control sequence V, respectively. m The length of the input sequence V, and D(i,j) represent the element in the i-th row and j-th column of the matrix, representing the first i elements of the input sequence V and the comparison with the control sequence V. m The cumulative minimum similarity distance between the first j elements; By finding the similarity matrix D m×n The minimum value in the equation is used to determine the best similarity match between two time series, thereby assessing the health status of the battery.

5. The method for estimating the state of health of a lithium-ion battery as described in claim 1, characterized in that, Methods for calculating the overall health status of battery cells include: Provide an importance function that assigns importance coefficients to the health status during charging and discharging processes; The importance coefficient needs to be determined by combining the actual engineering application scenario and battery characteristics, while ensuring that their sum is 1. The best similar health state during charging and the best similar health state during discharging are weighted using an importance coefficient, as shown in the formula: Where, α ch α represents the importance coefficient of the health status during the charging process. dis SOH is the importance coefficient of the health status during the discharge process. ch For optimal similar health status during charging, SOH dis This represents the optimal similar health state during discharge. The calculated SOH value is output as the current comprehensive health status assessment value of the battery cell.

6. The method for estimating the state of health of a lithium-ion battery as described in claim 1, characterized in that, Data acquisition involves real-time monitoring of battery status using sensors.

7. The method for estimating the state of health of a lithium-ion battery as described in claim 1, characterized in that, The established database includes standard charge and discharge voltage data under 25-degree Celsius temperature conditions.

8. A system for estimating the state of health of a lithium-ion battery, characterized in that, Includes the following modules: The data acquisition module is used to acquire the voltage data of lithium-ion batteries under constant power charge and discharge conditions throughout their entire life cycle, and to establish a database to record the one-to-one correspondence between voltage data and battery health status. The voltage acquisition module is located in the battery management unit and is used to collect voltage data of the battery cell during constant power charging and discharging in actual use. The data preprocessing module is used to preprocess the voltage data collected during the charging and discharging process to ensure that the data is suitable for subsequent analysis. The similarity calculation module, based on the dynamic time planning algorithm, calculates the similarity between the preprocessed voltage data and the voltage data sequence in the database, obtains the minimum similarity distance, and defines the optimal similar health state during charging and discharging based on the battery health state corresponding to the minimum similarity distance. The health status assessment module combines the best similar health status during charging and the best similar health status during discharging; it uses an importance function to assign importance coefficients to the best similar health status during charging and discharging, while ensuring that the sum of these two coefficients is 1. The comprehensive health status calculation module calculates the comprehensive health status of the battery cell based on the assigned importance coefficients and outputs it as the current health status assessment result of the battery cell.

9. An electronic device, characterized in that, The electronic device includes: a processor coupled to a memory; The memory is used to store computer programs; The processor is configured to execute the computer program stored in the memory, so that the electronic device performs the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, Includes instructions that, when executed on a computer, cause the computer to perform the method as described in any one of claims 1 to 7.

Images