Ex-service battery echelon sorting method and system based on multi-dimensional parameters

Abstract

The invention discloses a multi-dimensional parameter-based decommissioned battery echelon sorting method and system, relates to the technical field of battery recovery, and can fully excavate the internal relation between multi-dimensional operating parameters and battery faults by constructing and training a battery fault analysis model, comprehensively analyze decommissioned batteries from multiple dimensions, and improve the efficiency of battery recovery. The accuracy and reliability of a fault analysis result are improved, and a solid foundation is provided for subsequent performance rating and echelon determination; the multi-dimensional operation parameters of the target retired battery are read and input into the trained model, so that a fault analysis result can be quickly and accurately output, the sorting efficiency is improved, and the sorting cost is reduced; performance rating is performed according to a fault analysis result, an application echelon is determined, batteries with different performances are distributed to a proper application scene, and the utilization efficiency of the whole battery echelon is improved.

Description

Technical Field

[0001] This invention relates to the field of battery recycling technology, specifically a method and system for sorting retired batteries based on multi-dimensional parameters. Background Technology

[0002] With the booming development of the new energy vehicle industry, a large number of power batteries are facing retirement. Directly recycling these retired batteries would result in a huge waste of resources and increase the environmental burden. Therefore, the secondary use of retired batteries has significant economic and environmental implications; however, current methods for sorting retired batteries have many shortcomings. Most existing sorting methods rely on only a single parameter, such as remaining capacity. This method is difficult to comprehensively and accurately assess battery performance, and some retired batteries often have faults, the differences of which can affect the sorting results.

[0003] How to select reusable fields based on the performance score and fault type of retired batteries, so that retired batteries can be applied to more suitable application areas and maximize their utilization value, is a problem we need to solve. To this end, we now provide a retired battery tiered sorting method and system based on multi-dimensional parameters. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for sorting retired batteries based on multi-dimensional parameters.

[0005] The objective of this invention can be achieved through the following technical solution: a method for sorting retired batteries based on multi-dimensional parameters, comprising: Construct a battery fault analysis model and train the constructed battery fault analysis model; Read the multidimensional operating parameters of the target retired battery, input the obtained multidimensional operating parameters into the trained battery fault analysis model, and output the fault analysis results of the target retired battery; Based on the fault analysis results, the performance of the target retired batteries is rated, and the application tier of the target retired batteries is determined based on the performance rating results.

[0006] Furthermore, the process of constructing a battery fault analysis model and training the constructed battery fault analysis model includes: Construct a battery fault analysis model and initialize the constructed battery fault analysis model; Collect retired batteries with various faults and obtain multi-dimensional operating parameters of each retired battery as sample data; The collected sample data is divided into training set and test set; The training set is input into the battery fault analysis model to train the battery fault analysis model and obtain the corresponding training results. The training results of the battery fault analysis model are tested using a test set. Based on the test results, it is determined whether the training of the battery fault analysis model has achieved the expected results. If it has not achieved the expected results, the training set is used to iteratively train the battery fault analysis model. This process is repeated until the test results of the battery fault analysis model achieve the expected results or the number of iterations reaches the preset number, thereby completing the training of the battery fault analysis model.

[0007] Furthermore, the process of reading the multi-dimensional operating parameters of the target retired battery, inputting the obtained multi-dimensional operating parameters into the trained battery fault analysis model, and outputting the fault analysis results of the target retired battery includes: Place the target decommissioned battery under specified test conditions; The test parameters of the target retired battery during the test process are obtained, including test input current, test input voltage, test output current, test output voltage, and test remaining capacity. Construct a time axis and generate corresponding parameter change curves based on the obtained multidimensional test parameters; Map the generated parameter change curves onto the time axis; Set a step window within the time axis, and extract the parameter time series characteristics of each parameter change curve within the time axis by sliding the step window; The obtained time-series features of the parameters are input into the trained fault analysis model, and the fault analysis results of the target retired battery are output.

[0008] Furthermore, the time-series characteristics of the parameters include the time period corresponding to the step window, the peak value of the parameter change curve within the step window, the trough value of the parameter change curve within the step window, the fluctuation coefficient of the parameter change curve within the step window, and the percentage coefficient of the parameter change curve within the step window that deviates from the healthy threshold range.

[0009] Furthermore, the process of performance rating of the target decommissioned battery based on the fault analysis results includes: The parameter time series characteristics corresponding to the fault analysis results are summarized to obtain the parameter time series characteristic set; Match the fault analysis results with each fault type node in the performance rating graph, and activate the performance tree associated with the corresponding fault type node in the performance rating graph that matches the fault analysis results. The average time-series features of each parameter are matched with the constraints corresponding to each child node in the performance tree to obtain constraints that match the average time-series features of each parameter, thereby obtaining the performance score of the target retired battery, and obtaining the corresponding performance rating based on the performance score.

[0010] Furthermore, the performance rating map is as follows: Create corresponding fault type nodes based on different fault types, and construct corresponding tree nodes based on the time period corresponding to the step window in the parameter feature set. Generate child nodes corresponding to tree nodes based on each parameter dimension, and connect the generated child nodes to the tree nodes; Set constraints for each child node, where the constraints are the parameter ranges of the parameter dimension corresponding to that child node; Set a corresponding base score for each constraint; By summarizing all tree nodes, child nodes, constraints, and corresponding basic scores connected to nodes of the same fault type, a performance tree corresponding to that fault type node is obtained. By summarizing all fault type nodes and their corresponding performance trees, the performance rating graph is constructed.

[0011] Furthermore, the process of obtaining the corresponding performance rating based on the performance score is as follows: Several scoring ranges are set, each scoring range corresponding to a scoring level. The performance score of the target retired battery is matched with each of the set scoring ranges to determine the scoring range to which the performance score belongs, and the scoring level corresponding to the scoring range is used as the performance rating of the target retired battery.

[0012] Furthermore, the process of determining the application tiers of target retired batteries based on performance rating results includes: Based on the fault type of the target retired battery, determine the applicable fields corresponding to the fault type of the target retired battery; Set corresponding reference ratings for each application area; Match the performance rating of the target retired battery with the reference rating to determine the corresponding application area; The identified application areas will be compiled to form the application tiers for retired batteries targeting this purpose.

[0013] Furthermore, a retired battery sorting system based on a multi-dimensional parameter-based method for sorting retired batteries includes: The data acquisition module is used to read the multi-dimensional operating parameters of the target decommissioned battery; The data analysis module is used to input the obtained multi-dimensional operating parameters into the trained battery fault analysis model and output the fault analysis results of the target retired battery. The tiered sorting module is used to perform performance rating on target retired batteries based on fault analysis results, and determine the application tier of target retired batteries based on the performance rating results.

[0014] Compared with the prior art, the beneficial effects of the present invention are: By constructing and training a battery fault analysis model, the intrinsic relationship between multi-dimensional operating parameters and battery faults can be fully explored. This allows for a comprehensive analysis of retired batteries from multiple dimensions, improving the accuracy and reliability of fault analysis results and providing a solid foundation for subsequent performance rating and tier determination. By reading the multi-dimensional operating parameters of the target retired battery and inputting them into the trained model, fault analysis results can be output quickly and accurately, improving sorting efficiency and reducing sorting costs. Based on the fault analysis results, performance ratings and application tier determination can be performed, allocating batteries with different performance characteristics to appropriate application scenarios and improving the overall battery tier utilization efficiency. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0016] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a system block diagram of the present invention. Detailed Implementation

[0017] like Figure 1 As shown, a method for sorting retired batteries based on multi-dimensional parameters includes: Construct a battery fault analysis model and train the constructed battery fault analysis model; Read the multidimensional operating parameters of the target retired battery, input the obtained multidimensional operating parameters into the trained battery fault analysis model, and output the fault analysis results of the target retired battery; Based on the fault analysis results, the performance of the target retired batteries is rated, and the application tier of the target retired batteries is determined based on the performance rating results.

[0018] It should be further explained that, in the specific implementation process, the process of constructing a battery fault analysis model and training the constructed battery fault analysis model includes: A battery fault analysis model is constructed and initialized; in an embodiment of the present invention, the battery fault analysis model adopts a clustering analysis model. Collect retired batteries with various faults and obtain multi-dimensional operating parameters for each retired battery as sample data; the multi-dimensional operating parameters include fault type, and the corresponding input current, output current, input voltage, output voltage and remaining capacity for each fault type; The collected sample data is divided into training set and test set; The training set is input into the battery fault analysis model to train the battery fault analysis model and obtain the corresponding training results. The training results of the battery fault analysis model are tested using a test set. Based on the test results, it is determined whether the training of the battery fault analysis model has achieved the expected results. If it has not achieved the prediction, the battery fault analysis model is iteratively trained using the training set. This process is repeated until the test results of the battery fault analysis model meet the expectations or the number of iterations reaches the preset number, thus completing the training of the battery fault analysis model. It should be further noted that the process of training the model using a training set and verifying the model training results using a test set is a common technique used by those skilled in the art and will not be elaborated here.

[0019] It should be further explained that, in the specific implementation process, the process of reading the multi-dimensional operating parameters of the target retired battery, inputting the obtained multi-dimensional operating parameters into the trained battery fault analysis model, and outputting the fault analysis results of the target retired battery includes: The target decommissioned battery is placed under specified test conditions; it should be noted that the specified test conditions are specifically a constant current environment that meets the requirements of the target decommissioned battery. The test parameters of the target retired battery during the test process are obtained, including test input current, test input voltage, test output current, test output voltage, and test remaining capacity. Construct a time axis and generate corresponding parameter change curves based on the obtained multidimensional test parameters; Map the generated parameter change curves onto the time axis; A step-size window is set within the time axis, and the time-series characteristics of the parameter variation curves of each parameter within the time axis are extracted by sliding the step-size window. Each slide of the step window is labeled as i, where i = 1, 2, ..., n; For each parameter change curve, a healthy threshold range is set. Based on the parameter change curve within the step window, the parameter characteristics corresponding to that step window are obtained. The parameter characteristics include the time period corresponding to the step window. Peak value of the parameter variation curve within the step window Valley of the parameter variation curve within the step window Fluctuation coefficient of parameter variation curve within step window and the percentage of parameter change curves deviating from the healthy threshold range within the step window. Where 'c' represents the parameter type, namely, test input current, test input voltage, test output current, test output voltage, and test remaining power. It should be further explained that the volatility coefficient The specific acquisition process is as follows: The start and end times of the time period corresponding to the step window are denoted as t1 and t2, respectively. Let the parameter change curve be denoted as ,but: ; Percentage coefficient The specific acquisition process is as follows: The portion of the parameter change curve that deviates from the healthy threshold range is marked, and the area of ​​the marked portion is obtained and denoted as S1. Let the total area of ​​the parameter variation curve be S2, that is... ; but ; The obtained parameter features are summarized to obtain the parameter time-series features corresponding to the step-size window labeled i. , , , , ]; The obtained time-series features of the parameters are input into the trained fault analysis model, and the fault analysis results of the target retired battery are output.

[0020] It should be further explained that, in the specific implementation process, the process of performance rating of the target decommissioned battery based on the fault analysis results includes: The parameter time series characteristics corresponding to the fault analysis results are summarized to obtain the parameter time series characteristic set; Based on the obtained parameter time series feature set, the normalized value of each parameter time series feature is obtained, that is, the mean of each parameter time series feature; Match the fault analysis results with each fault type node in the performance rating graph, and activate the performance tree associated with the corresponding fault type node in the performance rating graph that matches the fault analysis results. The average time-series features of each parameter are matched with the constraints corresponding to each child node in the performance tree to obtain constraints that match the average time-series features of each parameter, thereby obtaining the performance score of the target retired battery, and obtaining the corresponding performance rating based on the performance score.

[0021] It should be further explained that the performance rating chart is specifically as follows: Create corresponding fault type nodes based on different fault types, and construct corresponding tree nodes based on the time period corresponding to the step window in the parameter feature set. Generate child nodes corresponding to tree nodes based on each parameter dimension, and connect the generated child nodes to the tree nodes; Set constraints for each child node, where the constraints are the parameter ranges of the parameter dimension corresponding to that child node; Set a corresponding base score for each constraint; By summarizing all tree nodes, child nodes, constraints, and corresponding basic scores connected to nodes of the same fault type, a performance tree corresponding to that fault type node is obtained. By summarizing all fault type nodes and their corresponding performance trees, the performance rating graph is constructed.

[0022] In another embodiment of the present invention, it can be known that each fault type node in the performance rating graph corresponds to a tree node. The tree node has four child nodes connected to it, which correspond to... , , , ; If the number of constraints set for each child node is m, then each constraint is labeled and denoted as j, where j = 1, 2, ..., m; Let K denote the constraint labeled j. j ; Then, with constraint K j The corresponding baseline score is denoted as P. j ; Obtain the constraints that the mean temporal features of each parameter satisfy, and label the base scores corresponding to each constraint as follows: , , , , ; The performance score of the target retired battery is then denoted as Xp, where: .

[0023] It should be further explained that the process of obtaining the corresponding performance rating based on the performance score is as follows: Several scoring ranges are set, each scoring range corresponding to a scoring level. The performance score of the target retired battery is matched with each of the set scoring ranges to determine the scoring range to which the performance score belongs, and the scoring level corresponding to the scoring range is used as the performance rating of the target retired battery.

[0024] It should be further explained that, in the specific implementation process, the process of determining the application tiers of target retired batteries based on performance rating results includes: Based on the fault type of the target retired battery, determine the applicable fields corresponding to the fault type of the target retired battery; Set corresponding reference ratings for each application area; Match the performance rating of the target retired battery with the reference rating to determine the corresponding application area; The identified application areas will be compiled to form the application tiers for retired batteries targeting this purpose.

[0025] like Figure 2 As shown, in another embodiment of the present invention, a multi-dimensional parameter-based decommissioned battery sorting system is also disclosed, comprising: The data acquisition module is used to read the multi-dimensional operating parameters of the target decommissioned battery; The data analysis module is used to input the obtained multi-dimensional operating parameters into the trained battery fault analysis model and output the fault analysis results of the target retired battery. The tiered sorting module is used to perform performance rating on target retired batteries based on fault analysis results, and determine the application tier of target retired batteries based on the performance rating results.

[0026] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications or equivalent substitutions made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for sorting retired batteries based on multi-dimensional parameters, characterized in that, include: Construct a battery fault analysis model and train the constructed battery fault analysis model; Read the multidimensional operating parameters of the target retired battery, input the obtained multidimensional operating parameters into the trained battery fault analysis model, and output the fault analysis results of the target retired battery; Based on the fault analysis results, the performance of the target retired batteries is rated, and the application tier of the target retired batteries is determined based on the performance rating results.

2. The method for sorting retired batteries based on multi-dimensional parameters according to claim 1, characterized in that, The process of constructing a battery fault analysis model and training the constructed battery fault analysis model includes: Construct a battery fault analysis model and initialize the constructed battery fault analysis model; Collect retired batteries with various faults and obtain multi-dimensional operating parameters of each retired battery as sample data; The collected sample data is divided into training set and test set; The training set is input into the battery fault analysis model to train the battery fault analysis model and obtain the corresponding training results. The training results of the battery fault analysis model are tested using a test set. Based on the test results, it is determined whether the training of the battery fault analysis model has achieved the expected results. If it has not achieved the expected results, the training set is used to iteratively train the battery fault analysis model. This process is repeated until the test results of the battery fault analysis model achieve the expected results or the number of iterations reaches the preset number, thereby completing the training of the battery fault analysis model.

3. The method for sorting retired batteries based on multi-dimensional parameters according to claim 2, characterized in that, The process of reading the multidimensional operating parameters of the target decommissioned battery, inputting the obtained multidimensional operating parameters into the trained battery fault analysis model, and outputting the fault analysis results of the target decommissioned battery includes: Place the target decommissioned battery under specified test conditions; The test parameters of the target retired battery during the test process are obtained, including test input current, test input voltage, test output current, test output voltage, and test remaining capacity. Construct a time axis and generate corresponding parameter change curves based on the obtained multidimensional test parameters; Map the generated parameter change curves onto the time axis; Set a step window within the time axis, and extract the parameter time series characteristics of each parameter change curve within the time axis by sliding the step window; The obtained time-series features of the parameters are input into the trained fault analysis model, and the fault analysis results of the target retired battery are output.

4. The method for sorting retired batteries based on multi-dimensional parameters according to claim 3, characterized in that, The time-series characteristics of the parameters include the time period corresponding to the step window, the peak value of the parameter change curve within the step window, the trough value of the parameter change curve within the step window, the fluctuation coefficient of the parameter change curve within the step window, and the percentage coefficient of the parameter change curve within the step window that deviates from the healthy threshold range.

5. The method for sorting retired batteries based on multi-dimensional parameters according to claim 4, characterized in that, The process of rating the performance of target decommissioned batteries based on fault analysis results includes: The parameter time series characteristics corresponding to the fault analysis results are summarized to obtain the parameter time series characteristic set; Match the fault analysis results with each fault type node in the performance rating graph, and activate the performance tree associated with the corresponding fault type node in the performance rating graph that matches the fault analysis results. The average time-series features of each parameter are matched with the constraints corresponding to each child node in the performance tree to obtain constraints that match the average time-series features of each parameter, thereby obtaining the performance score of the target retired battery, and obtaining the corresponding performance rating based on the performance score.

6. The method for sorting retired batteries based on multi-dimensional parameters according to claim 5, characterized in that, The performance rating chart is as follows: Create corresponding fault type nodes based on different fault types, and construct corresponding tree nodes based on the time period corresponding to the step window in the parameter feature set. Generate child nodes corresponding to tree nodes based on each parameter dimension, and connect the generated child nodes to the tree nodes; Set constraints for each child node, where the constraints are the parameter ranges of the parameter dimension corresponding to that child node; Set a corresponding base score for each constraint; By summarizing all tree nodes, child nodes, constraints, and corresponding basic scores connected to nodes of the same fault type, a performance tree corresponding to that fault type node is obtained. By summarizing all fault type nodes and their corresponding performance trees, the performance rating graph is constructed.

7. The method for sorting retired batteries based on multi-dimensional parameters according to claim 6, characterized in that, The process of obtaining the corresponding performance rating based on the performance score is as follows: Several scoring ranges are set, each scoring range corresponding to a scoring level. The performance score of the target retired battery is matched with each of the set scoring ranges to determine the scoring range to which the performance score belongs, and the scoring level corresponding to the scoring range is used as the performance rating of the target retired battery.

8. The method for sorting retired batteries based on multi-dimensional parameters according to claim 7, characterized in that, The process of determining the application tiers of target retired batteries based on performance rating results includes: Based on the fault type of the target retired battery, determine the applicable fields corresponding to the fault type of the target retired battery; Set corresponding reference ratings for each application area; Match the performance rating of the target retired battery with the reference rating to determine the corresponding application area; The identified application areas will be compiled to form the application tiers for retired batteries targeting this purpose.

9. A retired battery sorting system applied to the retired battery sorting method based on multi-dimensional parameters as described in any one of claims 1 to 8, characterized in that, include: The data acquisition module is used to read the multi-dimensional operating parameters of the target decommissioned battery; The data analysis module is used to input the obtained multi-dimensional operating parameters into the trained battery fault analysis model and output the fault analysis results of the target retired battery. The tiered sorting module is used to perform performance rating on target retired batteries based on fault analysis results, and determine the application tier of target retired batteries based on the performance rating results.

Images