Retired power battery residual value evaluation method based on full life cycle data
By combining full lifecycle data and machine learning models with multi-scenario optimization algorithms, the problems of inaccurate assessment and lack of optimization decision-making in the residual value evaluation of retired power batteries have been solved, realizing the scientific and accurate assessment of the residual value of retired power batteries and the output of the optimal recycling scheme.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-17
- Publication Date
- 2026-04-03
AI Technical Summary
Existing methods for assessing the residual value of retired power batteries lack full life-cycle data support, fail to accurately characterize the nonlinear aging behavior of batteries, resulting in large deviations in the prediction of health status and remaining lifespan, weak assessment foundation, and lack of optimization decision-making capabilities.
By collecting data throughout the entire life cycle, a battery state assessment model based on machine learning is constructed. Combined with multi-scenario optimization algorithms, a scientific and accurate assessment of the residual value of retired power batteries is achieved, including data preprocessing, model training and verification, calculation of potential recycling value, and output of the optimal solution.
It enables accurate assessment of the residual value of retired power batteries, improves the reliability of assessment and the scientific nature of decision-making, provides customized optimal recycling solutions, and solves the problems of single assessment dimensions and inability to dynamically optimize in existing technologies.
Smart Images
Figure CN121784547A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of residual value assessment technology for power batteries, and in particular to a method for assessing the residual value of retired power batteries based on full life cycle data. Background Technology
[0002] With the increasing popularity of electric vehicles, the number of retired power batteries has risen sharply. Scientific recycling and reuse technologies can effectively extract metals such as nickel, cobalt, manganese, and lithium from batteries, reducing dependence on primary resources and minimizing the negative environmental impact of waste.
[0003] However, existing methods for assessing the residual value of retired batteries have significant limitations: traditional assessments often rely on empirical formulas or linear models, which cannot accurately characterize the complex nonlinear aging behavior of batteries, resulting in large prediction biases for remaining lifespan and health status. Furthermore, most methods depend only on limited current test data (such as voltage and internal resistance) or simple usage history, lacking comprehensive data support across the entire battery lifecycle, from material design and manufacturing processes to long-term operating conditions. This leads to a lack of clarity regarding the battery's "health profile" and a weak foundation for assessment. Therefore, there is an urgent need in this field for a novel residual value assessment method that can integrate information across the entire battery lifecycle, synthesize multi-dimensional value, and possess optimization decision-making capabilities. Summary of the Invention
[0004] (a) Technical problems to be solved
[0005] To address the shortcomings of existing technologies, the method of this invention collects data covering the entire life cycle of production, use, and recycling, and constructs a battery state assessment model based on machine learning, thereby achieving a scientific, accurate, and value-maximizing assessment of the residual value of retired power batteries.
[0006] (II) Technical Solution
[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for assessing the residual value of retired power batteries based on full life-cycle data, comprising the following steps: S1: Collect full lifecycle data of retired power batteries, including data from the production stage, usage stage, and recycling stage; S2: Based on the full life cycle data, construct a battery state assessment model to predict the battery's remaining capacity, health status, and cycle life; S3: Based on the results of the battery state assessment model, calculate the potential recycling value of the battery, including the value of extractable metals and the reuse value; S4: Combining recycling cost data and policy subsidy data, the potential recycling value is discounted to obtain the residual value assessment result; S5: Based on a multi-scenario optimization algorithm, simulate different combinations of recycling methods and output the optimal residual value assessment scheme.
[0008] As a preferred embodiment, the production stage data in step S1 includes battery material composition, manufacturing process parameters, and initial performance indicators; the usage stage data includes cycle count, charge / discharge history, temperature environment, and capacity decay rate; and the recycling stage data includes recycling method type, metal extraction efficiency, and environmental impact indicators.
[0009] As a preferred embodiment, the construction of the battery state assessment model in step S2 specifically includes: S21: Preprocess the collected full lifecycle data, including data cleaning, normalization and feature extraction, to form a model training dataset; S22: Input the training dataset into the machine learning algorithm to train the model, adjust the model hyperparameters through cross-validation, and use the actual remaining capacity and health status of the battery as the training target. S23: Validate the trained model using an independent test dataset. When the model's prediction accuracy reaches a preset threshold, the model is completed and put into use.
[0010] As a preferred option, the battery state assessment model constructed in step S2 adopts a machine learning algorithm, including support vector machine or neural network. The model is trained based on historical full life cycle data and updated in real time to adapt to battery aging characteristics.
[0011] As a preferred option, when calculating the potential recycling value in step S3, the market price fluctuations of metals such as nickel, cobalt, manganese, and lithium are taken into account, and the amount of recyclable metals is estimated through a metal mass balance model.
[0012] As a preferred option, the discounting calculation in step S4 adopts the net present value method, the discount rate is adjusted based on the remaining battery life and risk factors, the recycling cost includes transportation, dismantling and processing costs, and the policy subsidy data includes government environmental subsidies and carbon trading revenue.
[0013] As a preferred option, the specific process for step S5, which outputs the optimal residual value evaluation scheme, is as follows: S51: Define the optimization objective and constraints, wherein the optimization objective is to maximize the residual value or minimize the negative environmental impact, and the constraints include technical feasibility, regulatory requirements, and resource availability; S52: Based on the defined optimization objectives and constraints, generate multiple different combinations of recycling methods using a multi-scenario optimization algorithm; S53: For each generated scheme, call the model and method from steps S2 to S4 to calculate its corresponding residual value assessment results and environmental benefit indicators; S54: By iteratively searching the optimization algorithm, the evaluation results of all schemes are compared, and the optimal scheme that satisfies the optimization objective is selected. S55: Output the optimal solution and its detailed residual value assessment report.
[0014] As a preferred embodiment, in step S3, the evaluation of the reuse value specifically includes the suitability judgment and value estimation for the battery cascade utilization scenarios; the suitability judgment is based on the battery health status and cycle life prediction results obtained in step S2, and is matched with the minimum technical standards of different cascade utilization scenarios; the value estimation is based on the successfully matched scenarios, estimating the economic benefits that can be generated within their remaining service life and discounting them.
[0015] As a preferred embodiment, the evaluation method is based on an evaluation system and includes: The data acquisition module is used to collect and store data throughout the entire lifecycle. The state assessment module is used to execute the battery state assessment model; The residual value calculation module is used to calculate the potential recovery value and the discounted residual value; The optimization module is used to run multi-scenario optimization algorithms and output the optimal solution.
[0016] (III) Beneficial Effects
[0017] Compared with existing technologies, this invention provides a method for evaluating the residual value of retired power batteries based on full life-cycle data, which has the following beneficial effects: I. This invention provides a solid data foundation for evaluation by introducing full lifecycle data, especially in-depth information from the production and usage stages. Combined with machine learning models, it can more accurately predict the future performance of batteries, fundamentally improving the reliability of residual value assessment. Through multi-scenario optimization algorithms, it upgrades the evaluation from a single numerical output to the comparison and selection of multiple solutions. Users can obtain customized optimal solutions according to their own needs, realizing the intelligent and scientific decision-making process. It solves the shortcomings of existing technologies, such as lack of a full lifecycle perspective, single evaluation dimensions, and inability to dynamically optimize decisions. Attached Figure Description
[0018] Figure 1 This is a flowchart of the evaluation method steps of the present invention; Figure 2 This is a schematic diagram of the technical process of the method of the present invention. Detailed Implementation
[0019] To better understand the purpose, structure, and function of this invention, the following will further explain the residual value assessment method for retired power batteries based on full life cycle data, in conjunction with the accompanying drawings and specific embodiments. Example 1
[0020] refer to Figure 1-2The present invention provides a method for assessing the residual value of decommissioned power batteries based on full life-cycle data, comprising the following steps: S1: Collect full lifecycle data of retired power batteries, including data from the production stage, usage stage, and recycling stage; S2: Based on the full life cycle data, construct a battery state assessment model to predict the battery's remaining capacity, health status, and cycle life; S3: Based on the results of the battery state assessment model, calculate the potential recycling value of the battery, including the value of extractable metals and the reuse value; S4: Combining recycling cost data and policy subsidy data, the potential recycling value is discounted to obtain the residual value assessment result; S5: Based on a multi-scenario optimization algorithm, simulate different combinations of recycling methods and output the optimal residual value assessment scheme.
[0021] Specifically, step S1 of this invention involves collecting full lifecycle data, and its specific implementation process is as follows: First, production stage data is obtained from battery manufacturers, including material composition such as positive and negative electrode material ratios, separator type, and electrolyte composition; manufacturing process parameters such as formation regime and compaction density; and initial performance indicators such as capacity, internal resistance, and open-circuit voltage at the time of delivery. Second, usage stage data is obtained from vehicle BMS logs and cloud platforms, including charge / discharge history such as total cycle count, average depth of discharge, and fast charge rate; temperature environment such as average operating temperature and temperature standard deviation; and capacity decay rate calculated based on capacity testing. Finally, recycling stage data comes from recycling company databases and industry reports, including recycling method types (such as pyrometallurgical, hydrometallurgical, and physical methods) and their corresponding metal extraction efficiencies; and environmental impact indicators such as energy consumption, carbon emissions, and pollutant emissions for each method.
[0022] Furthermore, battery aging is a complex and non-linear process. Next, in step S2 of this invention, a battery state assessment model is constructed based on the full lifecycle data from step S1, specifically including: S21: Data preprocessing, cleaning the raw data collected in S1 to remove obvious anomalies and erroneous records; normalization to eliminate the influence of different units; feature extraction, such as extracting capacity increment curve features from cyclic charge-discharge curves and extracting high-temperature cumulative time from temperature history, to form a high-quality feature matrix as a training dataset.
[0023] S22: Model Training. This invention uses a neural network as the machine learning algorithm. The preprocessed dataset is divided into training and validation sets proportionally. The network is trained using full-lifecycle features as input and the actual measured battery remaining capacity and health status as output targets. Hyperparameters such as the number of network layers and neurons are adjusted through cross-validation to prevent overfitting and improve the model's generalization ability.
[0024] S23: Model Validation and Deployment. Validate the trained model using the reserved test dataset. Calculate the mean absolute error (MAE) between the predicted remaining capacity and the actual capacity. When the MAE < 3% (preset threshold), the model accuracy is considered satisfactory, and it can be deployed to the evaluation system for rapid and non-destructive prediction of the state of newly decommissioned batteries.
[0025] Furthermore, step S3 of this invention focuses on calculating the potential recycling value of the battery. The value of a retired battery is the result of optimizing either secondary use or a combination thereof, through parallel calculation of these two values, ensuring that no potential value sources are overlooked. Step S4 of this invention involves a discounted calculation combining recycling costs and policy subsidies.
[0026] The first step is to extract the value of the metals. Based on the S2-predicted remaining battery composition and metal mass balance model, the mass of recyclable metals such as nickel, cobalt, manganese, and lithium is estimated. Then, a real-time updated metal market price database is connected to calculate the total material value.
[0027] Secondly, there's the reuse value. First, an applicability assessment is performed: the battery health status and cycle life predicted by S2 are matched with known secondary use scenarios (e.g., home energy storage requires SOH > 80%, low-speed electric vehicles require SOH > 70%). If the match is successful, a value estimation is performed: assessing the electrical energy the battery can generate or the peak grid power it can replace within its remaining lifespan in the target scenario, estimating its future cash flow accordingly, and then discounting it to its present value.
[0028] In existing technologies, a single evaluation path may not be optimal. This invention employs an optimization algorithm to efficiently search a vast solution space, automatically identifying the solution that optimizes the objective function under given constraints. The specific process for outputting the optimal residual evaluation solution is as follows: S51: Define the optimization objective and constraints, wherein the optimization objective is to maximize the residual value or minimize the negative environmental impact, and the constraints include technical feasibility, regulatory requirements, and resource availability; S52: Based on the defined optimization objectives and constraints, generate multiple different combinations of recycling methods using a multi-scenario optimization algorithm; S53: For each generated scheme, call the model and method from steps S2 to S4 to calculate its corresponding residual value assessment results and environmental benefit indicators; S54: By iteratively searching the optimization algorithm, the evaluation results of all schemes are compared, and the optimal scheme that satisfies the optimization objective is selected. S55: Output the optimal solution and its detailed residual value assessment report. Example
[0029] This invention relates to a method for assessing the residual value of decommissioned power batteries based on full life-cycle data. The assessment method is based on an assessment system and includes: The data acquisition module is used to collect and store data throughout the entire lifecycle. The state assessment module is used to execute the battery state assessment model; The residual value calculation module is used to calculate the potential recovery value and the discounted residual value; The optimization module is used to run multi-scenario optimization algorithms and output the optimal solution.
[0030] Specifically, in this embodiment, the data acquisition module automatically captures and aggregates data from dispersed data sources through API interfaces, IoT protocols, and other methods. Its core function is to integrate and standardize multi-source heterogeneous data, providing consistent and reliable data input for upstream models.
[0031] The condition assessment module incorporates a trained machine learning model (such as the model built by S2). When new retired battery data is input, this module calls the model to perform calculations and outputs intelligent diagnostic results for the battery's health status. Its function is to achieve non-destructive, rapid, and accurate insight into the battery's condition.
[0032] The residual value calculation module integrates a value calculation model, a cost model, a financial discount model, and an environmental benefit quantification model. It receives the results from the state assessment module and performs complex value accounting and discounting calculations based on built-in economic and environmental rules. Its function is to transform the physical state of the battery into precise economic and environmental value.
[0033] The optimization module encapsulates multi-scenario optimization algorithms. It receives various outputs from the residual calculation module and performs large-scale, automated simulations and deductions within the user-defined policy space to find the optimal decision. Its role is to find the "best" among the "possibilities," thereby improving the quality of decision-making.
[0034] The purpose of this invention is to overcome the shortcomings of the prior art and provide a scientific, accurate, and comprehensive method and system for evaluating the residual value of retired power batteries based on full life cycle data.
[0035] It is understood that the present invention has been described through some embodiments, and those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.
Claims
1. A method for assessing the residual value of decommissioned power batteries based on full life-cycle data, characterized in that, Includes the following steps: S1: Collect full lifecycle data of retired power batteries, including data from the production stage, usage stage, and recycling stage; S2: Based on the full life cycle data, construct a battery state assessment model to predict the battery's remaining capacity, health status, and cycle life; S3: Based on the results of the battery state assessment model, calculate the potential recycling value of the battery, including the value of extractable metals and the reuse value; S4: Combining recycling cost data and policy subsidy data, the potential recycling value is discounted to obtain the residual value assessment result; S5: Based on a multi-scenario optimization algorithm, simulate different combinations of recycling methods and output the optimal residual value assessment scheme.
2. The method for assessing the residual value of decommissioned power batteries based on full life-cycle data as described in claim 1, characterized in that, The production stage data in step S1 includes battery material composition, manufacturing process parameters, and initial performance indicators; the usage stage data includes cycle count, charge / discharge history, temperature environment, and capacity decay rate; and the recycling stage data includes recycling method type, metal extraction efficiency, and environmental impact indicators.
3. The method for assessing the residual value of decommissioned power batteries based on full life-cycle data as described in claim 1, characterized in that, The construction of the battery state assessment model in step S2 specifically includes: S21: Preprocess the collected full lifecycle data, including data cleaning, normalization and feature extraction, to form a model training dataset; S22: Input the training dataset into the machine learning algorithm to train the model, adjust the model hyperparameters through cross-validation, and use the actual remaining capacity and health status of the battery as the training target. S23: Validate the trained model using an independent test dataset. When the model's prediction accuracy reaches a preset threshold, the model is completed and put into use.
4. The method for assessing the residual value of decommissioned power batteries based on full life-cycle data as described in claim 3, characterized in that, In step S2, the battery state assessment model is constructed using machine learning algorithms, including support vector machines or neural networks. The model is trained based on historical full life cycle data and updated in real time to adapt to battery aging characteristics.
5. The method for assessing the residual value of decommissioned power batteries based on full life-cycle data as described in claim 1, characterized in that, In step S3, when calculating the potential recycling value, market price fluctuations of metals such as nickel, cobalt, manganese, and lithium are taken into account, and the amount of recyclable metals is estimated through a metal mass balance model.
6. The method for assessing the residual value of decommissioned power batteries based on full life-cycle data as described in claim 1, characterized in that, In step S4, the discount calculation uses the net present value method. The discount rate is adjusted based on the remaining battery life and risk factors. The recycling costs include transportation, dismantling and processing costs. Policy subsidy data includes government environmental subsidies and carbon trading revenue.
7. The method for assessing the residual value of decommissioned power batteries based on full life-cycle data as described in claim 1, characterized in that, The specific process for outputting the optimal residual value evaluation scheme in step S5 is as follows: S51: Define the optimization objective and constraints, wherein the optimization objective is to maximize the residual value or minimize the negative environmental impact, and the constraints include technical feasibility, regulatory requirements, and resource availability; S52: Based on the defined optimization objectives and constraints, use a multi-scenario optimization algorithm to generate multiple different combinations of recycling methods; S53: For each generated scheme, call the model and method from steps S2 to S4 to calculate its corresponding residual value assessment results and environmental benefit indicators; S54: By iteratively searching the optimization algorithm, the evaluation results of all schemes are compared, and the optimal scheme that satisfies the optimization objective is selected. S55: Output the optimal solution and its detailed residual value assessment report.
8. The method for assessing the residual value of decommissioned power batteries based on full life-cycle data as described in claim 1, characterized in that, In step S3, the evaluation of the reuse value specifically includes the suitability judgment and value estimation for the battery cascade utilization scenarios; the suitability judgment is based on the battery health status and cycle life prediction results obtained in step S2, and is matched with the minimum technical standards of different cascade utilization scenarios; the value estimation is based on the successfully matched scenarios, estimating the economic benefits that can be generated within their remaining service life and discounting them.
9. The method for assessing the residual value of decommissioned power batteries based on full life-cycle data as described in claim 1, characterized in that, The evaluation method is based on an evaluation system and includes: The data acquisition module is used to collect and store data throughout the entire lifecycle. The state assessment module is used to execute the battery state assessment model; The residual value calculation module is used to calculate the potential recovery value and the discounted residual value; The optimization module is used to run multi-scenario optimization algorithms and output the optimal solution.