Battery intelligent classification and performance evaluation device and method

Abstract

The invention discloses a battery intelligent classification and performance evaluation device and method, and the device comprises a battery data testing module which is used for carrying out the charging and discharging testing of a decommissioned battery under a set control condition, and obtaining the battery parameters of the decommissioned battery; the data generation module is used for generating capacity data and internal resistance data conforming to Weibull distribution based on the obtained battery parameters; the data screening module is configured to perform preliminary screening on the generated capacity data and internal resistance data based on normal distribution to complete abnormal point elimination; the clustering classification module is used for clustering the screened capacity data and internal resistance data by using a K-means + + algorithm; and the performance evaluation module is used for evaluating the consistency of the clustered and classified batteries, and judging the number of clusters selected when the performance of the classification system is optimal in combination with system operation time verification. According to the method, the batteries with similar consistency can be quickly clustered and sorted in a high-quality manner, and the safety and efficiency of echelon utilization of the retired batteries are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of battery intelligent classification, and specifically relates to a battery intelligent classification and performance evaluation device and method. Background Art

[0002] The statements herein merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] With the intensification of environmental pollution and the increasing depletion of non-renewable energy, the world faces the dual challenges of energy transition and environmental protection. This presents new opportunities for the development of lithium-ion batteries (LIBs) in the new energy industry. Lithium-ion batteries, with their high energy density, low self-discharge rate, and long lifespan, have become the primary power source for electric vehicles and energy storage systems. However, battery parameters gradually degrade with long-term use. When capacity drops to 80%, the battery's endurance and performance decline significantly, posing a potential retirement challenge. Repurposing retired batteries not only maximizes their residual value but also reduces environmental impact and resource waste.

[0004] Battery usage directly impacts their residual value and safety. Therefore, efficient recycling requires classification and screening based on their aging state. The inventors discovered that classification methods based on a single parameter (such as capacity) fail to fully reflect battery performance. Furthermore, clustering algorithms, which divide samples into clusters with similar characteristics, are a core step in battery sorting. Based on their principles and implementation, existing clustering algorithms can be categorized into four types: hierarchical, density-based, model-based, and partitioning. Hierarchical clustering methods are complex and require strict data set morphology; density-based clustering methods produce poor clustering quality when the density of the dataset is uneven; and model-based clustering methods are sensitive to initial parameter settings, resulting in poor interpretability of clustering results. Partitioning clustering methods, such as K-means, offer simple principles, fast clustering speed, and high clustering flexibility, making them widely used in the sorting of retired batteries. However, they still have inherent drawbacks, such as sensitivity to noise and abnormal battery data and difficulty in selecting initial cluster centers. Therefore, there is an urgent need for an intelligent classification method in the field of retired battery recycling to achieve efficient sorting and performance evaluation of retired batteries. Summary of the Invention

[0005] The purpose of the present invention is to provide a battery intelligent classification and performance evaluation device and method to solve the problem of differences in parameters such as voltage, resistance and capacity of retired batteries caused by manufacturing processes, operating conditions and poor battery management systems, overcome the shortcomings of the existing technology, quickly and high-quality cluster sorting of retired batteries with similar consistency, and improve the safety and efficiency of the cascade utilization of retired batteries.

[0006] In order to achieve the above object, the present invention is implemented through the following technical solutions: In one aspect, the technical solution of the present invention provides a battery intelligent classification and performance evaluation device, comprising: The battery data testing module is configured to: perform charge and discharge tests on retired batteries under set control conditions to obtain battery parameters of the retired batteries; A data generation module is configured to: generate capacity data and internal resistance data that conform to Weibull distribution based on the obtained battery parameters; The data screening module is configured to: perform preliminary screening of the generated capacity data and internal resistance data based on normal distribution to eliminate abnormal points; The clustering classification module is configured to: cluster the filtered capacity data and internal resistance data using the K-means++ algorithm; The performance evaluation module is configured to: evaluate the consistency of the batteries after cluster classification, and determine the number of clusters selected when the classification system performs best in combination with system running time verification.

[0007] In at least one embodiment, the battery parameters include battery capacity, battery internal resistance, and open circuit voltage.

[0008] In at least one embodiment, the battery data testing module includes a battery testing unit and a test data acquisition and recording unit; the battery testing unit is used to perform charge and discharge tests on retired batteries under set control conditions, specifically including: Place several batteries in a temperature-controlled box with a set temperature; charging the retired battery at a constant current of a first set rate until the charging cut-off voltage is reached; Switch to constant voltage charging until the current drops to the second set rate and then stand for the set time; Discharging the retired battery at a constant current of a first set rate until the discharge cut-off voltage is reached; Repeat the above charge-discharge process to determine the internal resistance, capacity and open circuit voltage data of the retired battery in the fully charged state and record them through the test data recording and acquisition recording unit.

[0009] In at least one embodiment, the data generation module includes a capacity data generation unit and an internal resistance data generation unit; the capacity data generation unit is used to generate capacity data that conforms to the Weibull distribution based on the obtained battery parameters, and the capacity data that conforms to the Weibull distribution is specifically: ; The internal resistance data generating unit is used to generate internal resistance data that conforms to the Weibull distribution based on the obtained battery parameters. The internal resistance data that conforms to the Weibull distribution is specifically: ; Where, is the proportional parameter of capacity and internal resistance distribution, are the shape parameters of the capacity and internal resistance distribution, is the mathematical expectation of the internal resistance distribution.

[0010] In at least one embodiment, the data screening module screens the generated capacity data and internal resistance data based on the normal distribution, specifically: screening the generated capacity data and internal resistance data based on the normal distribution, retaining The battery data within the range is considered as abnormal data and the data points outside this range are removed.

[0011] In at least one embodiment, the cluster classification module is used to implement a cluster classification process, specifically including: A data normalization unit is configured to perform normalization processing on the filtered capacity data and internal resistance data as input to the K-means++ clustering algorithm; an initial centroid selection unit, configured to randomly select a first centroid from the data set and calculate the distance between each data point in the data set and the selected nearest centroid; The subsequent centroids are selected based on the principle that the probability of a data point being selected as a new centroid is proportional to the square of its distance from the nearest centroid; A centroid position updating unit is configured to iteratively assign data points to the nearest centroid, calculate the distance of each point to all centroids, assign the data points to the cluster corresponding to the nearest centroid, update the centroid position, and recalculate the centroid position of each cluster; repeat the assignment and updating process until the centroid position tends to be stable or a preset maximum number of iterations is reached; The clustering result output unit is configured to output the clustering classification result and restore the original scale of the data using reverse normalization.

[0012] In at least one embodiment, the performance evaluation module includes a discrete coefficient evaluation unit and a system operating time evaluation unit; the discrete coefficient evaluation unit is configured to use the discrete coefficient to evaluate the consistency of retired batteries after clustering classification; the system operating time evaluation unit is configured to comprehensively consider the sorting system operating time as one of the performance evaluation criteria.

[0013] On the other hand, the technical solution of the present invention also provides a battery intelligent classification and performance evaluation method, including: Perform charge and discharge tests on retired batteries under set control conditions to obtain battery parameters of the retired batteries; Generate capacity data and internal resistance data that conform to Weibull distribution based on the obtained battery parameters; Perform preliminary screening on the generated capacity data and internal resistance data based on normal distribution to eliminate abnormal points; Use K-means++ algorithm to cluster and classify the filtered capacity data and internal resistance data; The consistency of the battery after cluster classification is evaluated, and combined with the system running time verification, the number of clusters selected when the classification system performs best is determined.

[0014] In at least one embodiment, the consistency of the retired battery cluster after cluster classification is evaluated using the discrete coefficient, and the superiority of the method is judged based on the system operation time; the optimal number of categories for the retired battery cluster sorting is determined by combining the classification performance indicators of consistency and operation time through the discrete coefficient evaluation unit and the system operation time evaluation unit. In at least one embodiment, the coefficient of variation is the ratio of the standard deviation of the battery parameter to its mean value, specifically expressed as: ; Where, is the standard deviation of the battery parameters, is the average value of battery parameters.

[0015] The beneficial effects of the technical solution of the present invention are as follows: 1) The battery intelligent classification and performance evaluation device and method of the present invention can respond to the needs of the comprehensive performance evaluation system, quickly screen large quantities of retired batteries, and return the number of clusters and the cluster number of each cell at the time of optimal classification to guide the rapid and accurate grouping and cascade utilization of retired batteries, thereby improving the safety and efficiency of the cascade utilization of retired batteries and further improving the system performance of retired battery classification.

[0016] 2) The present invention utilizes the K-means++ algorithm to not only accelerate the convergence speed of the algorithm but also improve the final clustering quality. Finally, consistency and runtime evaluation are performed to achieve efficient and reliable classification and screening of retired batteries.

[0017] 3) The present invention selects parameters such as the battery's open circuit voltage, discharge capacity, and DC internal resistance as health characteristics, which solves the defect of the existing classification method based on a single parameter that cannot fully reflect battery performance issues. It is also time-saving and simple to operate.

[0018] 4) The present invention proposes a normal distribution A multi-level screening strategy is used to preliminarily remove abnormal data, and the K-means++ clustering algorithm is used to improve the initial cluster center points, which solves the problems of the existing clustering algorithm being sensitive to noise and abnormal battery data and the difficulty in selecting the initial cluster center. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0020] Figure 1 Schematic diagram of the battery intelligent classification and performance evaluation device disclosed in Example 1 of the present invention; Figure 2 Schematic diagram of the implementation process of the battery intelligent classification and performance evaluation method disclosed in Example 2 of the present invention; Figure 3 This is a schematic diagram of the embedded computer disclosed in Example 3 of the present invention. DETAILED DESCRIPTION

[0021] It should be noted that the following detailed description is illustrative and is intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used in the present invention have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0022] As introduced in the background technology, the purpose of the present invention is to provide a battery intelligent classification and performance evaluation device and method to solve the problem of differences in parameters such as voltage, resistance and capacity of retired batteries caused by manufacturing processes, operating conditions and poor battery management systems, overcome the shortcomings of the existing technology, and quickly and high-quality cluster sorting of retired batteries with similar consistency, thereby improving the safety and efficiency of the cascade utilization of retired batteries.

[0023] Example 1 In a typical embodiment of the present invention, Figure 1 As shown, this embodiment discloses a battery intelligent classification and performance evaluation device, including: The battery data testing module is configured to: perform charge and discharge tests on retired batteries under set control conditions to obtain battery parameters of the retired batteries; A data generation module is configured to: generate capacity data and internal resistance data that conform to Weibull distribution based on the obtained battery parameters; The data screening module is configured to: perform preliminary screening of the generated capacity data and internal resistance data based on normal distribution to eliminate abnormal points; The clustering classification module is configured to: cluster the filtered capacity data and internal resistance data using the K-means++ algorithm; The performance evaluation module is configured to: evaluate the consistency of the batteries after cluster classification, and determine the number of clusters selected when the classification system performs best in combination with system running time verification.

[0024] The above-mentioned battery intelligent classification and performance evaluation device is described in detail below with reference to specific embodiments.

[0025] In this embodiment, the battery data testing module includes a battery testing unit and a test data collection and recording unit, which are used to perform charge and discharge tests on retired batteries under set control conditions, obtain battery parameters of the retired batteries, and collect and record test data.

[0026] Among them, the battery testing unit uses lithium iron phosphate batteries as the test object to conduct full life cycle cycle aging and performance experimental tests to obtain battery-related parameters such as battery capacity, battery internal resistance and open circuit voltage.

[0027] Specifically, several lithium iron phosphate batteries are first placed in a temperature-controlled box at 25°C, and a constant current charging mode is adopted to charge the batteries at a constant current of 1 / 3C until the upper cut-off voltage is reached; then, the charging mode is switched to a constant voltage charging mode until the current drops to 1 / 20C and the batteries are left to stand for two hours; after the standing time is completed, the charging mode is switched to a constant current discharge mode to discharge the batteries at a constant current of 1 / 3C until the lower cut-off voltage is reached; the above charging-discharging process is repeated three times, and the voltage, current, discharge capacity and other data are recorded in real time by the test data recording and acquisition recording unit, and the battery internal resistance is calculated based on the known data to determine the ohmic resistance, capacity and open circuit voltage of the retired battery in a fully charged state.

[0028] Among them, C in 1 / 3C and 1 / 20C is C-rate , i.e. the charge and discharge rate, 1C corresponds to the rated capacity of the battery. In this embodiment, the charge cut-off voltage is 4.2V and the discharge cut-off voltage is 3.0V.

[0029] Due to the limitations of experimental conditions, it is difficult to continue to repeat the battery charge and discharge test to obtain a large amount of battery data. Therefore, in this embodiment, the data generation module includes a capacity data generation unit and an internal resistance data generation unit, which are respectively used to randomly generate a large amount of capacity and internal resistance data that conform to the Weibull distribution based on the actual battery test sample data obtained in the battery data test module.

[0030] First, the cumulative distribution function is used to describe the probability that a random variable is less than or equal to the set parameter, which is defined as: (1); in, x are random variables of capacity and internal resistance data; and They are shape parameters and scale parameters, which are obtained through test data and further applied.

[0031] The mathematical characteristics of Weibull distribution mainly include mathematical expectation and variance , where the mathematical expectation It reflects the size of the mean value of the random variable and the variance It is the average value of mathematical expectation, which mainly reflects the degree of dispersion of distribution data, as shown in formula (2).

[0032] (2); Where, is the gamma function.

[0033] Capacity and resistance do not have a direct linear relationship with battery aging and performance evaluation. The relationship between them needs to be described by the Weibull distribution. Therefore, in order to obtain battery capacity and internal resistance data that conform to the Weibull distribution, it is necessary to solve the capacity probability density function and internal resistance probability density function of the retired battery respectively, as shown below: The probability density function of capacity data that conforms to the Weibull distribution is specifically: (3); Where, is the proportional parameter of capacity distribution, is the shape parameter of the capacity distribution.

[0034] The probability density function of the internal resistance data that conforms to the Weibull distribution is specifically: (4); Where, is the proportional parameter of the internal resistance distribution, is the shape parameter of the internal resistance distribution, is the mathematical expectation of the internal resistance distribution.

[0035] Based on the data obtained from the experiment, the capacity data generation unit and the internal resistance data generation unit are used to solve formula (3) and formula (4) respectively to obtain a large number of effective battery data that conforms to the Weibull distribution of lithium-ion battery capacity and internal resistance.

[0036] After obtaining the Weibull distribution of lithium-ion batteries, because the low-capacity data deviates too much from the actual working capacity range of the battery, and the high-resistance data deviates too much from the actual resistance range of the battery, the part in the data set should be further refined and subsequently processed.

[0037] In this embodiment, the data screening module screens the generated capacity data and internal resistance data based on normal distribution, and retains About 99.7% of the capacity data points and internal resistance data points fall within the range. Data points outside this range are usually regarded as abnormal data and need to be eliminated.

[0038] In the context of battery performance testing, this principle helps to exclude retired batteries that exhibit abnormal characteristics due to manufacturing defects, damage or other non-random factors, thereby reducing the possibility of failure and improving the safety and service life of sorted and reassembled battery packs.

[0039] After the data screening module completes the removal of abnormal points in the capacity data and internal resistance data, this embodiment implements the clustering classification process based on the clustering classification module, using K-means ++ The algorithm clusters and classifies the filtered capacity data and internal resistance data, and outputs the cluster label of each data point and the mass points of each cluster, including: The data normalization unit is configured to: normalize the data as input to the K-means++ clustering algorithm.

[0040] Since the dimensions of the extracted features are not uniform, the data needs to be normalized before clustering. This is done by scaling all features to the same scale to eliminate the influence of different dimensions and make their range between 0 and 1, thereby accelerating the convergence of the algorithm. The normalization equation is: (5); The K-means++ algorithm is used to cluster and classify the filtered data. In the K-means algorithm, the center of each cluster is all the representative points in the cluster. The algorithm optimizes the cluster division by minimizing the sum of the squares of the cluster distances. Its objective function is: (6); The K-means clustering algorithm usually uses the Euclidean distance to measure the distance from a point to the cluster center, and its equation is: (7); The K-means++ clustering algorithm used in this embodiment includes a more intelligent initialization process. To ensure that the initial cluster centers are evenly distributed, in this embodiment, the initial centroids are selected by an initial centroid selection unit using a probabilistic method. Specifically, the initial centroid selection unit is configured as follows: Randomly select the first centroid from the dataset , calculate for each data point in the dataset Distance to the nearest selected centroid: (8); The subsequent centroid selection unit is configured to select the subsequent centroid based on the principle that the probability of a data point being selected as a new centroid is proportional to the square of its distance from the nearest centroid: When performing K-means ++During the algorithm, the first centroid is randomly selected, and subsequent centroids are selected according to a probability distribution of squared Euclidean distances from existing centroids. This ensures the dispersion of the initial centroids and helps avoid local optimal solutions: (9); The centroid position update unit is configured to: assign data points to the nearest centroid, calculate the distance of each point to all centroids, assign them to the cluster corresponding to the nearest centroid, update the centroid position, and recalculate the intra-cluster mean (i.e., centroid) of each cluster; repeat the assignment and update process until the centroid position tends to stabilize or reaches the preset maximum number of iterations, so as to achieve the purpose of optimizing the quality and stability of the clustering results.

[0041] The clustering result output unit is configured to: after the above clustering process is completed, output the cluster classification result, obtain the cluster label to which each data point belongs and the center coordinates of each cluster, and use reverse normalization to restore the original scale of the data.

[0042] In order to facilitate the verification and presentation of clustering results, inverse normalization is usually used to help restore the original scale of the data, so as to more intuitively understand the distribution and characteristics of the data and compare the differences between different data sets. The inverse normalization equation is: (10); Based on the retired battery data after preliminary screening, the K-means++ algorithm was further used to cluster the screened retired battery data. The batteries within the same cluster were more similar in performance parameters, and the separation between different clusters showed that the algorithm effectively identified the inherent groupings in the data.

[0043] S5. Evaluate the consistency of the batteries after clustering and classification, and determine the best performance based on the system running time.

[0044] In this embodiment, the performance evaluation module includes a discrete coefficient evaluation unit and a system operating time evaluation unit, wherein the discrete coefficient evaluation unit is configured to use the discrete coefficient to evaluate the consistency of retired batteries after clustering classification, and the system operating time evaluation unit is configured to comprehensively consider the operating time of the sorting system as one of the performance evaluation criteria.

[0045] Specifically, the dispersion coefficient evaluation unit uses the dispersion coefficient to evaluate the consistency of retired batteries after cluster classification. The dispersion coefficient is the ratio of the standard deviation of battery parameters (i.e., capacity and internal resistance) to their average values, usually expressed as a percentage, as follows: (11); in, is the standard deviation of the battery parameters, is the average value of the battery parameter. The coefficient of dispersion (C.D.) reflects the relative distance of the battery parameter fluctuation from the average value. A smaller C.D. indicates greater consistency in the battery parameters and more stable battery performance. A larger C.D. indicates greater fluctuation in the battery parameters and less consistent battery performance. Generally speaking, a C.D. below 3% indicates low data variability and high consistency; below 5% indicates moderate dispersion and good stability; and above 8% indicates high dispersion and data fluctuation.

[0046] Furthermore, due to the large number of retired electric vehicle battery samples, the retired battery sorting and grouping algorithm is a daunting task, so the clustering algorithm runtime is also a key consideration in practical applications. Therefore, this embodiment uses the dispersion coefficient to evaluate the consistency of retired battery clusters after clustering and classification. Simultaneously, a system runtime evaluation unit is used to determine the superiority of the method based on the system runtime. The dispersion coefficient evaluation unit and the system runtime evaluation unit combine the classification performance indicators of consistency and runtime to determine the optimal number of categories for retired battery clustering and sorting, thereby guiding the rapid and accurate grouping and cascade utilization of retired batteries.

[0047] Example 2 In a typical embodiment of the present invention, Figure 2 As shown, this embodiment discloses a battery intelligent classification and performance evaluation method, including: S1. Perform charge and discharge tests on retired batteries under set control conditions to obtain battery parameters of the retired batteries; S2. generating capacity data and internal resistance data conforming to the Weibull distribution based on the obtained battery parameters; S3. Perform preliminary screening on the generated capacity data and internal resistance data based on normal distribution to eliminate abnormal points; S4. Use K-means++ algorithm to cluster and classify the filtered capacity data and internal resistance data; S5. Evaluate the consistency of the batteries after cluster classification, and combine it with the system running time verification to determine the number of clusters selected when the classification system performs best.

[0048] In this embodiment, the performance evaluation is a comprehensive evaluation based on the discrete coefficient evaluation and the system running time.

[0049] The dispersion coefficient is used to evaluate the consistency of retired batteries after cluster classification. The dispersion coefficient is the ratio of the standard deviation of battery parameters (i.e., capacity and internal resistance) to their average values, usually expressed as a percentage, as follows: (11); in, is the standard deviation of the battery parameters, is the average value of the battery parameter. The coefficient of dispersion (C.D.) reflects the relative distance of the battery parameter fluctuation from the average value. A smaller C.D. indicates greater consistency in the battery parameters and more stable battery performance. A larger C.D. indicates greater fluctuation in the battery parameters and less consistent battery performance. Generally speaking, a C.D. below 3% indicates low data variability and high consistency; below 5% indicates moderate dispersion and good stability; and above 8% indicates high dispersion and data fluctuation.

[0050] Due to the large number of retired electric vehicle battery samples, the retired battery sorting and grouping algorithm is a daunting task. Therefore, the clustering algorithm's runtime is also a key consideration in practical applications. Therefore, this embodiment uses the dispersion coefficient to evaluate the consistency of retired battery clusters after clustering and classification, and judges the superiority of the method based on the system runtime. The dispersion coefficient evaluation unit and the system runtime evaluation unit combine the classification performance indicators of consistency and runtime to determine the optimal number of categories for retired battery clustering and sorting, thereby guiding the rapid and accurate grouping and cascade utilization of retired batteries.

[0051] Example 3 In this embodiment, the apparatus of embodiment 1 and the method of embodiment 2 are deployed in an embedded computer system. The overall architecture is as follows: Figure 3 As shown, the specific introduction is: Embedded computer systems are computing devices designed specifically for specific tasks, primarily consisting of hardware, middleware, and software layers. They can be deployed locally for edge computing, offering advantages such as strong real-time responsiveness, low power consumption, and high reliability. They serve as the physical carrier for this example.

[0052] The hardware layer of an embedded system is its physical foundation, responsible for performing calculations, storing data, connecting to external devices, and implementing specific functions. Among these layers are the embedded microprocessor, its "brain," which performs its core computing functions; storage systems that balance speed, capacity, and reliability, including ROM and SDRAM; peripheral interfaces for connecting sensors, actuators, and communication modules, including DACs, ADCs, and SPIOs; general interfaces for external communication, including serial ports, Ethernet, and USB; and human-computer interaction interfaces that respond to user operations, including touch screens, mice, and keyboards.

[0053] The driver layer of an embedded system is a key module that connects hardware devices to the operating system / middleware. It is responsible for directly controlling hardware peripherals and providing standardized operational interfaces. Its core task is to transform physical hardware behavior into logical services that can be called by software.

[0054] The software layer of an embedded system is the top-level architecture directly oriented toward business logic and user interaction, implementing high-level tasks such as functional integration, data processing, and decision-making control. It includes a real-time operating system (RTOS), various libraries, protocol stacks, and middleware such as communication protocols and file systems. The embedded software layer is the ultimate vehicle for delivering business value, and its design requires a deep integration of domain knowledge, hardware features, and algorithmic capabilities, such as the algorithm design in this example.

[0055] This embodiment operates on a lithium-ion battery system, comprising hundreds or even thousands of lithium-ion battery cells, as well as auxiliary subsystems and supporting components. The battery management system's core tasks include balancing management, safety protection, and communication interaction. The power conversion system is responsible for converting AC and DC power, regulating voltage and frequency, and controlling bidirectional energy flow. Auxiliary equipment includes air conditioning and fire protection systems. Sensors used for battery status monitoring include current sensors, voltage sensors, and temperature sensors.

[0056] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. Battery intelligent classification and performance evaluation device, characterized in that: include: The battery data testing module is configured to: perform charge and discharge tests on retired batteries under set control conditions to obtain battery parameters of the retired batteries; A data generation module is configured to: generate capacity data and internal resistance data that conform to Weibull distribution based on the obtained battery parameters; The data screening module is configured to: perform preliminary screening of the generated capacity data and internal resistance data based on normal distribution to eliminate abnormal points; The clustering classification module is configured to: cluster the filtered capacity data and internal resistance data using the K-means++ algorithm; The performance evaluation module is configured to: evaluate the consistency of the batteries after cluster classification, and determine the number of clusters selected when the classification system performs best in combination with system running time verification.

2. The battery intelligent classification and performance evaluation device according to claim 1, characterized in that: The battery parameters include battery capacity, battery internal resistance and open circuit voltage.

3. The battery intelligent classification and performance evaluation device according to claim 1, characterized in that: The battery data testing module includes a battery testing unit and a test data acquisition and recording unit; the battery testing unit is used to perform charge and discharge tests on retired batteries under set control conditions, specifically including: Place several batteries in a temperature-controlled box with a set temperature; charging the retired battery at a constant current of a first set rate until the charging cut-off voltage is reached; Switch to constant voltage charging until the current drops to the second set rate and then stand for the set time; Discharging the retired battery at a constant current of a first set rate until the discharge cut-off voltage is reached; Repeat the above charge-discharge process to determine the internal resistance, capacity and open circuit voltage data of the retired battery in the fully charged state and record them through the test data recording and acquisition recording unit.

4. The battery intelligent classification and performance evaluation device according to claim 1, characterized in that: The data generation module includes a capacity data generation unit and an internal resistance data generation unit; the capacity data generation unit is used to generate capacity data that conforms to the Weibull distribution based on the obtained battery parameters, and the capacity data that conforms to the Weibull distribution is specifically: ； The internal resistance data generating unit is used to generate internal resistance data that conforms to the Weibull distribution based on the obtained battery parameters. The internal resistance data that conforms to the Weibull distribution is specifically: ； Where, is the proportional parameter of capacity and internal resistance distribution, are the shape parameters of the capacity and internal resistance distribution, is the mathematical expectation of the internal resistance distribution.

5. The battery intelligent classification and performance evaluation device according to claim 1, characterized in that: The data screening module screens the generated capacity data and internal resistance data based on the normal distribution, specifically: screening the generated capacity data and internal resistance data based on the normal distribution, retaining The battery data within the range is considered as abnormal data and the data points outside this range are removed.

6. The battery intelligent classification and performance evaluation device according to claim 1, characterized in that: The cluster classification module is used to implement the cluster classification process, which specifically includes: A data normalization unit is configured to perform normalization processing on the filtered capacity data and internal resistance data as input to the K-means++ clustering algorithm; an initial centroid selection unit, configured to randomly select a first centroid from the data set and calculate the distance between each data point in the data set and the selected nearest centroid; The subsequent centroids are selected based on the principle that the probability of a data point being selected as a new centroid is proportional to the square of its distance from the nearest centroid; A centroid position updating unit is configured to iteratively assign data points to the nearest centroid, calculate the distance of each point to all centroids, assign the data points to the cluster corresponding to the nearest centroid, update the centroid position, and recalculate the centroid position of each cluster; repeat the assignment and updating process until the centroid position tends to be stable or a preset maximum number of iterations is reached; The clustering result output unit is configured to output the clustering classification result and restore the original scale of the data using reverse normalization.

7. The battery intelligent classification and performance evaluation device according to claim 1, characterized in that: The performance evaluation module includes a discrete coefficient evaluation unit and a system operation time evaluation unit; the discrete coefficient evaluation unit is configured to evaluate the consistency of retired batteries after cluster classification using the discrete coefficient; The system operation time evaluation unit is configured to comprehensively consider the operation time of the sorting system as one of the performance evaluation criteria.

8. A battery intelligent classification and performance evaluation method, characterized in that: include: Perform charge and discharge tests on retired batteries under set control conditions to obtain battery parameters of the retired batteries; Generate capacity data and internal resistance data that conform to Weibull distribution based on the obtained battery parameters; Perform preliminary screening on the generated capacity data and internal resistance data based on normal distribution to eliminate abnormal points; Use K-means++ algorithm to cluster and classify the filtered capacity data and internal resistance data; The consistency of the battery after cluster classification is evaluated, and combined with the system running time verification, the number of clusters selected when the classification system performs best is determined.

9. The battery intelligent classification and performance evaluation method according to claim 8, characterized in that: The consistency of retired battery clusters after cluster classification is evaluated using the discrete coefficient, and the superiority of the method is judged based on the system running time. The classification performance indicators of consistency and running time are comprehensively evaluated by the discrete coefficient evaluation unit and the system running time evaluation unit to determine the optimal number of categories for cluster sorting of retired batteries.

10. The battery intelligent classification and performance evaluation method according to claim 9, characterized in that: The coefficient of variation is the ratio of the standard deviation of a battery parameter to its mean value, specifically expressed as: ； Where, is the standard deviation of the battery parameters, is the average value of battery parameters.

Images