A new energy battery residual value electrochemical-risk coupling dynamic valuation method

Abstract

The application relates to the field of battery state evaluation, and discloses an electrochemical-risk coupling dynamic evaluation method for the residual value of a new energy battery. The health state prediction mean value, the health state prediction standard deviation, the inventory state, the sorting grade and the rated capacity of an electric core to be evaluated are determined based on electrochemical detection data. A state correction coefficient and a quality premium coefficient are determined, and a mapping relationship between the health state prediction uncertainty and the value discount degree is established. An uncertainty discount factor is determined based on the mapping relationship, and a risk adjusted value is formed. The benchmark unit price is dynamically updated by using historical transaction data of the same type of electric core, and market fluctuation parameters are determined. The residual value evaluation and the evaluation interval are determined according to the risk adjusted value, the updated benchmark unit price and the market fluctuation parameters. The application can couple the electrochemical state of the electric core, predict the risk and market fluctuation, and improve the rationality and dynamic adaptability of the residual value evaluation of the new energy battery.

Description

Technical Field

[0001] This application relates to the field of battery state assessment, specifically to an electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries. Background Technology

[0002] With the large-scale application of new energy vehicles, energy storage systems, and mobile energy devices, a large-scale demand for battery cell assets is gradually forming during the production, storage, use, and post-retirement circulation of new energy batteries. For new battery cells, long-term inventory battery cells, and retired battery cells, their remaining usability, subsequent application value, and transaction value are closely related to the health status of the battery cells themselves.

[0003] Current battery valuation methods typically use a single indicator such as State of Health (SOH), rated capacity, or number of cycles as the basis for value calculation. For example, the remaining value of a battery is directly calculated based on the measured SOH and a preset unit price, or a fixed depreciation rate is set based on the battery's service life. However, the predicted value of the same battery cell may have different uncertainties; at the same time, even if new cells, long-term inventory cells, and retired cells that have been recycled have similar health conditions, they may have different aging mechanisms and remaining usage risks. Valuation based solely on the average health condition or a uniform depreciation rate is insufficient to fully reflect the actual differences in the condition of different cells.

[0004] In addition, existing residual value assessments of new energy batteries typically use a pre-set benchmark unit price for value conversion, making it difficult to adjust the valuation benchmark in a timely manner according to changes in the actual transaction price of similar cells; at the same time, actual market transaction prices fluctuate to some extent, and outputting only a single valuation result is insufficient to reflect a reasonable range of value fluctuations. Summary of the Invention

[0005] In view of the above problems, this application provides an electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries, which solves the problems of difficulty in differentiating the valuation of cells with different risks and difficulty in dynamically adjusting the valuation results with market changes in the existing new energy battery residual value assessment process.

[0006] To achieve the above objectives, the inventors provide an electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries, comprising:

[0007] Obtain the mean and standard deviation of the health status of the battery cell to be valued, determined based on electrochemical detection data, and obtain the stock status, sorting grade, and rated capacity of the battery cell to be valued.

[0008] The status correction coefficient is determined based on the stock status, the quality premium coefficient is determined based on the sorting grade, and a mapping relationship between the uncertainty of health status prediction and the degree of value discount is established based on historical transaction data.

[0009] The uncertainty discount factor is determined based on the standard deviation of the health status prediction and the mapping relationship, and the risk-adjusted value is determined by combining the mean of the health status prediction, the rated capacity, the status correction coefficient, and the quality premium coefficient.

[0010] The benchmark unit price is dynamically updated based on historical transaction data of similar battery cells, and market fluctuation parameters are determined.

[0011] Based on the risk-adjusted value, the updated benchmark unit price, and the market volatility parameters, the residual value and valuation range of the battery cell to be valued are determined.

[0012] Unlike existing technologies, the above-mentioned technical solution incorporates the cell's health status and its predictive uncertainty into residual value assessment by acquiring the mean and standard deviation of the health status prediction based on electrochemical testing data. Furthermore, it determines a status correction coefficient and a quality premium coefficient based on the cell's stock status and sorting grade, and establishes a mapping relationship between the predictive uncertainty of health status and the degree of value discount based on historical transaction data. This allows cells with different storage conditions, usage levels, sorting quality, and predictive reliability to generate differentiated risk-adjusted values. Simultaneously, it dynamically updates the benchmark unit price and determines market fluctuation parameters using historical transaction data of similar cells, enabling residual value valuation to adjust according to actual market transaction conditions and price fluctuations. Therefore, it couples the cell's internal electrochemical health status, predictive uncertainty, and external market risks into a unified valuation process, reducing valuation bias caused by relying solely on a single SOH or fixed benchmark price, and improving the differentiation, rationality, and dynamic adaptability of new energy battery residual value pricing.

[0013] The above description of the invention is merely an overview of the technical solution of this application. In order to enable those skilled in the art to better understand the technical solution of this application and to implement it based on the description and drawings, and to make the above-mentioned objectives and other objectives, features and advantages of this application easier to understand, the following description is provided in conjunction with the specific embodiments and drawings of this application. Attached Figure Description

[0014] The accompanying drawings are only used to illustrate the principles, implementation methods, applications, features, and effects of specific embodiments of the present invention and other related contents, and should not be considered as limitations on this application.

[0015] In the accompanying drawings of the instruction manual:

[0016] Figure 1 This is a flowchart illustrating an electrochemical-risk coupled dynamic valuation method for the residual value of a new energy battery in an implementation scheme. Detailed Implementation

[0017] To illustrate the possible application scenarios, technical principles, implementable specific solutions, and achievable objectives and effects of this application in detail, the following description, in conjunction with the listed specific embodiments and accompanying drawings, provides a detailed explanation. The embodiments described herein are merely illustrative of the technical solutions of this application and are therefore intended to limit the scope of protection of this application.

[0018] Please refer to Figure 1 In one embodiment, the present invention provides an electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries, including steps S100 to S500.

[0019] S100: Obtain the mean and standard deviation of the estimated health status of the battery cell to be valued based on electrochemical detection data, and obtain the stock status, sorting grade and rated capacity of the battery cell to be valued.

[0020] S200: Determine the status correction coefficient based on the stock status, determine the quality premium coefficient based on the sorting grade, and establish a mapping relationship between the uncertainty of health status prediction and the degree of value discount based on historical transaction data.

[0021] S300: Determine the uncertainty discount factor based on the predicted standard deviation of the health status and the mapping relationship, and determine the risk-adjusted value by combining the predicted mean of the health status, the rated capacity, the status correction coefficient, and the quality premium coefficient.

[0022] S400: Dynamically updates the benchmark unit price and determines market fluctuation parameters based on historical transaction data of similar battery cells.

[0023] S500: Determine the residual value and valuation range of the battery cell to be valued based on the risk-adjusted value, the updated benchmark unit price, and the market volatility parameters.

[0024] By introducing the above settings into a single valuation chain, electrochemical health status, predictive uncertainty, stock status, sorting quality, and market transaction changes are introduced sequentially. This allows risk-adjusted value to work in conjunction with dynamic market benchmark prices. Furthermore, by outputting the valuation range through market fluctuation parameters, the problem of being unable to distinguish between different state risks and market changes when relying solely on SOH or fixed unit prices for valuation can be avoided.

[0025] In step S100, the battery cell to be valued can be a new cell, a stock cell, or a retired cell within the batch. The cell to be valued has undergone pre-electrochemical testing, and the electrochemical testing data can include impedance data obtained from electrochemical impedance spectroscopy. The electrochemical testing data can be input into a Gaussian process regression model, which outputs a posterior distribution of the health status. Based on the posterior distribution, the predicted mean and predicted standard deviation of the health status are obtained.

[0026] For the first batch The average estimated health status of the individual battery cells to be valued is expressed as follows:

[0027]

[0028] The standard deviation of health status prediction is expressed as:

[0029]

[0030] in, For the first The estimated average health status of each battery cell to be valued; For the first The standard deviation of the estimated health status of each battery cell to be valued; This refers to the serial number of the battery cell to be valued.

[0031] The sorting level can be represented as:

[0032]

[0033] The stock status can be represented as:

[0034]

[0035] in, For the first The sorting grade of each battery cell to be valued; For the first The inventory status of each battery cell to be valued; Indicates the status of the new battery cell; Indicates the status of the inventory cells; This indicates the status of a retired battery cell.

[0036] The rated capacity is expressed as follows:

[0037]

[0038] in, For the first The rated capacity of each battery cell to be valued can be expressed in Ah.

[0039] In one specific implementation, a validity check is performed on the estimated mean of health status, when:

[0040]

[0041] or:

[0042]

[0043] When the data is abnormal, the corresponding cell to be valued will be marked as abnormal, and the electrochemical detection system will be triggered to re-detect it in order to prevent abnormal prediction data from entering the subsequent valuation process.

[0044] By setting up the above parameters, the uncertainty of the health status prediction results can be retained while obtaining the central value of the health status, and obviously abnormal prediction results can be filtered out before valuation. This can provide effective input for subsequent uncertainty risk discounting, and at the same time reduce the probability of residual value valuation distortion caused by abnormal detection results.

[0045] In one embodiment of step S200, determining the state correction coefficient based on the stock status includes: when the stock status of the battery cell to be valued is a new battery cell, determining a first state correction coefficient based on the storage duration of the battery cell to be valued; when the stock status of the battery cell to be valued is a stock battery cell, determining a second state correction coefficient based on the storage duration of the battery cell to be valued; and when the stock status of the battery cell to be valued is a retired battery cell, determining a third state correction coefficient based on the degree of recycling of the battery cell to be valued.

[0046] Through the above settings, the stock status directly determines the corresponding status correction path. The new cell status and the stock cell status correspond to the correction method based on the storage time, respectively, while the retired cell status corresponds to the correction method based on the degree of recycling. This ensures that the status correction coefficient corresponds to the actual life cycle status of the cell to be valued.

[0047] In step S200, when the storage state of the battery cell to be valued is that of a new battery cell, a first state correction coefficient is determined based on the storage time of the battery cell to be valued. In one specific embodiment, a new battery cell is defined as the factory resting time. For battery cells that have been used for one month and have no usage record, the first-state correction factor meets the following requirements:

[0048]

[0049] in, The correction coefficient for the first state; Storage duration; The monthly attenuation rate.

[0050] The monthly static attenuation rate can be determined by long-term static test data of the same type of battery cell, and the typical value satisfies:

[0051]

[0052] When the battery cell to be valued is in the state of being a stock cell, a second state correction coefficient is determined based on the storage time of the battery cell to be valued. In one specific embodiment, a stock cell is defined as the factory resting time. For battery cells that have been used for one month and have no usage record, the second state correction factor meets the following requirements:

[0053]

[0054] in, This is the correction coefficient for the second state; Initial depreciation factor for inventory; Storage duration; The monthly attenuation rate.

[0055] The typical values ​​of the initial depreciation factor for the inventory satisfy the following:

[0056]

[0057] When the battery cell to be valued is in a retired battery cell state, a third state correction coefficient is determined based on the recycling rate of the battery cell. In one specific embodiment, a retired battery cell is defined as a cell with a cumulative number of cycles. For used cells with a secondary or state of harmless osmosis (SOH) below 90%, the third-state correction factor meets the following requirements:

[0058]

[0059] in, This is the correction coefficient for the third state; This is the cumulative number of loops; To design cycle life; It is a nonlinear accelerating decay exponent; It is the lower limit of the residual value, and satisfies:

[0060]

[0061] as well as:

[0062]

[0063] When the lower limit of the residual value is taken as the above value, the third state correction coefficient can also be equivalently expressed as:

[0064]

[0065] Using the three state correction formulas mentioned above, continuous state correction results can be generated based on the most representative time or cyclic variable under the corresponding state, given that the existing state is already determined. The third state correction coefficient is obtained through... The nonlinear relationship accelerates the correction in the later stages of the cycle, thus better reflecting the engineering characteristics of accelerated aging in the later stages of retired battery cells.

[0066] For example, when:

[0067]

[0068] and:

[0069]

[0070] hour:

[0071]

[0072] when:

[0073]

[0074] hour:

[0075]

[0076] It can be seen that as the cumulative number of cycles increases from the middle to the later part of the designed cycle life, the decrease in the third state correction coefficient increases, which can reflect the state change of more and more severe decay as the cycle progresses.

[0077] In one embodiment of step S200, determining the quality premium coefficient based on the sorting grade includes: acquiring historical transaction data corresponding to each sorting grade; determining the corresponding normalized transaction value based on the historical transaction data corresponding to each sorting grade; determining the quality premium coefficient corresponding to each sorting grade based on the normalized transaction value corresponding to different sorting grades, and ensuring that the quality premium coefficients corresponding to different sorting grades meet preset grade constraints.

[0078] Through the above settings, the sorting grade and the quality premium coefficient are directly correlated through historical transaction data, and the normalized transaction value is used to reduce the impact of differences in capacity and health status on the comparison of market value of different grades. At the same time, the reasonable order of the quality premium coefficients corresponding to different sorting grades is maintained by preset grade constraints, so that the quality premium coefficient can be updated with market transaction data and is not easily reversed due to local sample fluctuations.

[0079] When there are insufficient historical transaction samples or the quality premium coefficient has not yet been dynamically calibrated, an initial grade premium coefficient can be used:

[0080]

[0081]

[0082]

[0083]

[0084]

[0085] in, For sorting levels; For sorting levels The corresponding quality premium coefficient.

[0086] In one embodiment of step S200, the normalized transaction value and the quality premium coefficient satisfy:

[0087]

[0088] in, This represents the actual transaction value of battery cells traded historically. This corresponds to the rated capacity of the battery cells traded in the historical records. This is the estimated average health status of the corresponding historically traded battery cells; This corresponds to the sorting grade of the battery cells sold in the past. For sorting levels The corresponding quality premium coefficient; For level indicator functions; This represents the regression error; This refers to the sample serial number of the battery cells traded in the past. For sorting levels.

[0089] Apply a monotonically decreasing constraint to the quality premium coefficients corresponding to different sorting grades:

[0090]

[0091]

[0092]

[0093] The regression sample size satisfies:

[0094]

[0095] in, For dynamic calibration Sorting grade quality premium coefficient; For dynamic calibration Sorting grade quality premium coefficient; For dynamic calibration Sorting grade quality premium coefficient; For dynamic calibration Sorting grade quality premium coefficient; This represents the historical number of battery cells traded.

[0096] By setting up the above, the actual transaction value of historically traded battery cells is normalized according to the estimated average of rated capacity and health status. The normalized transaction value is then associated with the corresponding sorting grade using a grade indicator function. Under the condition of satisfying the grade monotonicity constraint, the quality premium coefficient is obtained, which can improve the identifiability of value differences between different sorting grades. At the same time, the use of sufficient historical samples reduces the impact of individual abnormal transactions on the grade premium calibration results.

[0097] In one embodiment of step S200, establishing a mapping relationship between the uncertainty of health status prediction and the degree of value discount based on historical transaction data includes: determining the deterministic valuation and actual transaction value corresponding to each historically traded battery cell based on the historical transaction data; determining the observed discount degree based on the actual transaction value and corresponding deterministic valuation of each historically traded battery cell; obtaining the standard deviation of the health status prediction corresponding to each historically traded battery cell, and establishing a correspondence between the standard deviation of the health status prediction and the observed discount degree; and determining the mapping relationship between the uncertainty of health status prediction and the degree of value discount based on the correspondence.

[0098] Through the above settings, the deviation of the actual transaction value from the certain valuation is converted into the degree of observation discount, and further paired with the standard deviation of the health status prediction corresponding to the same historically traded battery cells. This allows us to learn from historical real transaction data the impact of uncertainty on market value discount, providing a basis for determining the uncertainty discount factor for the current battery cells to be valued.

[0099] In one specific implementation, to avoid a circular dependency between the risk aversion coefficient calibration and subsequent dynamic learning of the benchmark unit price, a step-by-step decoupling calibration strategy is adopted. First, the average market price over the past three months is used... As the benchmark price for agents .

[0100] Certainty valuation satisfies:

[0101]

[0102] in, This is a certainty valuation for battery cells corresponding to historical transactions; This is the benchmark price for agents; This corresponds to the rated capacity of the battery cells traded in the historical records. This is the estimated average health status of the corresponding historically traded battery cells; This represents the quality premium coefficient corresponding to the sorting grade; This refers to the stock status of battery cells corresponding to historical transactions; This is the state correction coefficient corresponding to the existing stock status; This refers to the sample serial number of the battery cells that were traded in the past.

[0103] In one embodiment of step S200, the degree of observation discount satisfies:

[0104]

[0105] The risk aversion coefficient is determined based on the observed discount level and the estimated standard deviation of the health status corresponding to each historically traded battery cell.

[0106]

[0107] The mapping relationship satisfies:

[0108]

[0109] in, This represents the actual transaction value of battery cells traded historically. This corresponds to a certainty valuation; The corresponding observation discount level; Risk aversion coefficient; This refers to the historical number of battery cells traded. The standard deviation of the estimated health status of the corresponding historically traded battery cells; To observe the noise variance of the discount level; It is the probability density function of a normal distribution; This refers to the sample serial number of the battery cells traded in the past. Uncertainty discount factor corresponding to the uncertainty in predicting health status; The standard deviation is estimated for the health status of the battery cell to be valued.

[0110] With the above settings, the risk aversion coefficient The value was determined by combining the observed discount rate of historically traded battery cells with the standard deviation of their health status estimates. This leads directly to the exponential mapping relationship between the uncertainty of health status prediction and the degree of value discount, thus enabling the risk aversion coefficient to have a clear data source and subsequent use, avoiding its isolated use as an independent empirical parameter.

[0111] In another specific implementation, the valuation user can directly set the risk aversion coefficient based on their risk preference. A conservative risk preference corresponds to:

[0112]

[0113] Aggressive risk appetite corresponds to:

[0114]

[0115] In step S300, an uncertainty discount factor is determined based on the predicted standard deviation of the health status and the mapping relationship, and a risk-adjusted value is determined by combining the predicted mean of the health status, the rated capacity, the status correction coefficient, and the quality premium coefficient.

[0116] The uncertainty discount factor satisfies:

[0117]

[0118] in, Uncertainty discount factor corresponding to the uncertainty in predicting health status; The standard deviation of the health status estimate for the battery cell to be valued; Risk aversion coefficient; For the natural constant An exponential function with base 0.

[0119] With the above settings, when the standard deviation of the health status prediction increases, the uncertainty discount factor decreases accordingly, thereby enabling a greater risk discount to be applied to cells with higher prediction uncertainty. Furthermore, this discount result, together with the mean of the health status prediction, rated capacity, status correction coefficient, and quality premium coefficient, forms a risk-adjusted value, allowing the electrochemical health status and its prediction reliability to simultaneously affect the cell value evaluation.

[0120] When further combining the market benchmark unit price to obtain the basic risk-adjusted valuation of a single battery cell, the following conditions are met:

[0121]

[0122] in, For the first The basic risk adjustment valuation of each individual battery cell to be valued; The base unit price; For the first The rated capacity of the individual battery cells to be valued; For the first The estimated average health status of each battery cell to be valued; This represents the quality premium coefficient corresponding to the sorting grade; This is the state correction coefficient corresponding to the existing stock status; This is the corresponding uncertainty discount factor; This refers to the serial number of the battery cell to be valued.

[0123] In one embodiment of step S400, the dynamic update of the benchmark unit price based on historical transaction data of the same type of battery cells includes: determining an initial probability distribution of the benchmark unit price based on historical transaction unit prices of the same type of battery cells within a first preset time range; obtaining newly added transaction unit prices of the same type of battery cells, and updating the initial probability distribution based on the newly added transaction unit prices of the same type of battery cells to obtain an updated probability distribution; and determining the updated benchmark unit price based on the updated probability distribution.

[0124] By setting up the above, historical transaction prices are used to establish a probabilistic prior for the market benchmark price. Then, new transaction prices are used to continuously correct this probability distribution, so that the updated benchmark price can be dynamically adjusted according to changes in market transaction status. At the same time, the probability distribution can also retain the degree of uncertainty caused by market price fluctuations.

[0125] In one embodiment of step S400, the benchmark unit price is initialized using conjugate priors, satisfying:

[0126]

[0127] in:

[0128]

[0129]

[0130] in, The base unit price; The initial average of the benchmark unit price; The initial variance of the benchmark unit price; This refers to the average transaction price per unit of the same type of battery cell within the corresponding historical time period. This refers to the sequence number of the historical time unit; The number of historical time units within the first preset time range; This represents a normal distribution.

[0131] In one specific implementation:

[0132]

[0133] The initial probability distribution of the benchmark unit price is formed by using the average transaction price of similar battery cells over the past three months.

[0134] With the above settings, the initial probability distribution of the benchmark unit price is determined by the average transaction unit price and its dispersion of multiple historical time units. This ensures that the initial benchmark unit price not only includes the market price center level but also historical market price fluctuation information, providing a stable prior basis for subsequent recursive updates using newly added transaction data.

[0135] In step S400, the newly added transaction price of the same type of battery cell is obtained, and the initial probability distribution is updated based on the newly added transaction price of the same type of battery cell. The newly added actual transaction price can be used as new observation data, and a recursive update is performed based on the benchmark price probability distribution obtained in the previous valuation period. The updated probability distribution is then used as the benchmark price probability distribution for the next valuation period. During the update process, the transaction price observation noise can also be updated based on the degree of deviation of the newly added transaction price from the current probability distribution, thereby reducing the excessive impact of a single abnormal transaction on the benchmark price.

[0136] In step S400, the updated benchmark unit price is determined based on the updated probability distribution, and market fluctuation parameters are determined based on historical transaction data of the same type of battery cells. The market fluctuation parameters characterize the degree of change in market transaction prices of the same type of battery cells within a first preset time range. These parameters can be determined based on the statistical fluctuation results of historical transaction price sequences and are updated synchronously with new transaction data.

[0137] In one embodiment of step S500, determining the residual value and valuation range of the battery cell to be valued based on the risk-adjusted value, the updated benchmark unit price, and the market volatility parameters includes: determining the residual value based on the risk-adjusted value and the updated benchmark unit price; determining the fluctuation range corresponding to the residual value based on the market volatility parameters; and determining the valuation range based on the residual value and the fluctuation range.

[0138] Through the above settings, the risk-adjusted value is first combined with the dynamically updated benchmark unit price to obtain the residual value center value under the current market conditions. Then, the residual value center value is expanded into a valuation range through market fluctuation parameters. Therefore, it can simultaneously represent the state value of the battery cell to be valued and the uncertainty of market price, avoiding the problem that the residual value of the battery cell cannot reflect the reasonable range of transaction fluctuations when expressed by a single fixed price.

[0139] In a further embodiment, the valuation confidence level of the battery cell to be valued can also be output. The valuation confidence level satisfies:

[0140]

[0141] in, For the first The confidence level of the valuation of each battery cell to be valued; For the first Uncertainty discount factor corresponding to each unvalued battery cell; For the first The standard deviation of the estimated health status of each battery cell to be valued; To estimate the standard deviation threshold for acceptable health status, statistical calibration can be performed using test-retest experiments with the same type of battery cells. This indicates taking the smaller value.

[0142] when:

[0143]

[0144] hour:

[0145]

[0146] when:

[0147]

[0148] hour:

[0149]

[0150] Therefore, when the health status prediction is nearly certain, the confidence level of the valuation is at its highest; when the standard deviation of the health status prediction reaches the allowable threshold, the confidence level of the valuation drops to zero, which can trigger the cell to be valued to re-perform electrochemical testing and health status prediction.

[0151] In another implementation, the confidence level of the estimate can also be determined using the Gaussian decay method:

[0152]

[0153] in, For the first The confidence level of the valuation of each battery cell to be valued; For the first Uncertainty discount factor corresponding to each unvalued battery cell; For the first The standard deviation of the estimated health status of each battery cell to be valued; Estimate the standard deviation threshold for acceptable health conditions; For the natural constant An exponential function with base 0.

[0154] By continuing to use the uncertainty of health status prediction in the valuation confidence output, the valuation results can be further indicated by the degree of reliability in addition to the residual value and valuation range, and a retest can be triggered when the prediction uncertainty is too high, thereby forming a feedback relationship between electrochemical detection and residual value assessment.

[0155] In a further implementation, the actual transaction value can be obtained after the battery cell to be valued completes the actual transaction. The actual transaction value is then compared with the residual value valuation output at the corresponding time, and the subsequent valuation model parameters are updated based on the valuation deviation between the two. For sorting grades that show valuation deviations in the same direction over a long period, the quality premium coefficient or sorting threshold of the corresponding sorting grade can be corrected based on actual transaction feedback, so that battery cell testing, sorting, residual value assessment, and actual transaction results form a closed-loop feedback.

[0156] An electrochemical-risk coupled dynamic valuation terminal for the residual value of new energy batteries includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps in the electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries.

[0157] The processor described in the embodiments of this application can be implemented by hardware, firmware, software, or a combination thereof. It can be a circuit, one or more of an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field-programmable gate array (FPGA), a central processing unit (CPU), a controller, a microcontroller, or a microprocessor. It also includes other physical, biological, or chemical structures that can implement the same or equivalent functions as the processors listed above, such as biological neurons, quantum computing units, DNA computing units, etc., so that the processor can execute some or all of the steps in the computer program or method involved in the various embodiments of this application, or any combination of the steps mentioned therein.

[0158] The principle of this invention lies in using the electrochemical health status of the battery cell to be valued as the valuation basis. The mean of the predicted health status is used to characterize the current remaining health level of the cell, and the standard deviation of the predicted health status is used to characterize the uncertainty of the predicted health status. Based on this, corresponding condition correction coefficients are determined according to the stock status of the battery cell to be valued (new cells, inventory cells, or retired cells) to reflect different value decays caused by static aging, inventory depreciation, and recycling. A quality premium coefficient is determined based on the cell sorting grade and corresponding historical transaction data to characterize the value differences between cells of different quality grades. Simultaneously, a mapping relationship is established based on the correspondence between the uncertainty of the predicted health status and the actual value discount in historical transaction data, and an uncertainty discount factor is determined by the standard deviation of the predicted health status of the battery cell to be valued, thereby converting the uncertainty of the electrochemical prediction result into a risk discount that can participate in the value calculation. Thus, the mean of the predicted health status, rated capacity, condition correction coefficient, quality premium coefficient, and uncertainty discount factor are coupled to form a risk-adjusted value.

[0159] Furthermore, this invention utilizes historical transaction data of similar battery cells to establish an initial probability distribution of the benchmark unit price, and continuously updates the benchmark unit price as new transaction data emerges, enabling the battery cell value benchmark to dynamically change with the actual market transaction status. Simultaneously, it determines market volatility parameters based on the degree of fluctuation in historical transaction prices, and uses the risk-adjusted value, the dynamically updated benchmark unit price, and the market volatility parameters together to determine the residual value valuation and valuation range of the battery cell to be valued. Therefore, this invention allows the battery cell's own technical condition and its predictive reliability to work in conjunction with actual market value changes to ultimately determine the residual value pricing.

[0160] In summary, this invention differs from traditional cell residual value assessment methods that rely solely on fixed-ratio depreciation based on State of Health (SOH), rated capacity, or cycle count. Instead, it utilizes the mean and standard deviation of the predicted health status to characterize the cell's health level and predictive reliability. Furthermore, it employs appropriate condition correction methods for different stock conditions and incorporates sorting grade adjustments to generate quality premiums. Therefore, even cells with similar predicted health status means can exhibit differentiated risk-adjusted values ​​based on their resting state, cycle life, sorting quality, and the uncertainty of their predicted health status. This reduces value convergence caused by relying solely on a single health indicator and enhances the discriminative power and rationality of residual value assessments for cells in different conditions.

[0161] Meanwhile, this invention utilizes historical transaction data to establish a mapping relationship between the uncertainty of health status prediction and the degree of value discount, further converting the uncertainty in electrochemical status prediction into a risk discount in the residual value pricing process, allowing the reliability of the prediction to directly participate in value calculation; and dynamically updating the benchmark unit price through historical transaction data of similar cells, combined with market fluctuation parameters to generate residual value valuations and valuation ranges, enabling the valuation results to be continuously adjusted according to the actual market transaction status and price fluctuations. Therefore, this invention integrates the internal electrochemical health status of the cell and external market transaction risks into the same dynamic valuation system, improving the accuracy and dynamic adaptability of new energy battery residual value valuation, while also providing a more objective quantitative basis for the sorting, tiered utilization, and asset transfer of new cells, inventory cells, and retired cells.

[0162] Finally, it should be noted that although the above embodiments have been described in the text and drawings of this application, this should not limit the scope of patent protection of this application. Any technical solutions that are based on the essential concept of this application and utilize the content described in the text and drawings of this application, resulting in equivalent structural or procedural substitutions or modifications, as well as the direct or indirect application of the technical solutions of the above embodiments to other related technical fields, are all included within the scope of patent protection of this application.

Claims

1. An electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries, characterized in that, include: Obtain the mean and standard deviation of the estimated health status of the battery cell to be valued, determined based on electrochemical detection data, and obtain the stock status, sorting grade, and rated capacity of the battery cell to be valued. The status correction coefficient is determined based on the stock status, the quality premium coefficient is determined based on the sorting grade, and a mapping relationship between the uncertainty of health status prediction and the degree of value discount is established based on historical transaction data. The uncertainty discount factor is determined based on the standard deviation of the health status prediction and the mapping relationship, and the risk-adjusted value is determined by combining the mean of the health status prediction, the rated capacity, the status correction coefficient, and the quality premium coefficient. The benchmark unit price is dynamically updated based on historical transaction data of similar battery cells, and market fluctuation parameters are determined. Based on the risk-adjusted value, the updated benchmark unit price, and the market volatility parameters, the residual value and valuation range of the battery cell to be valued are determined.

2. The electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries according to claim 1, characterized in that, The step of determining the state correction coefficient based on the stock status includes: When the existing state of the battery cell to be valued is that of a new battery cell, the first state correction coefficient is determined based on the storage time of the battery cell to be valued. When the stock status of the battery cell to be valued is that of a stock cell, a second status correction coefficient is determined based on the storage duration of the battery cell to be valued. When the existing state of the battery cell to be valued is that of a retired battery cell, the third state correction coefficient is determined based on the degree of recycling of the battery cell to be valued.

3. The electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries according to claim 2, characterized in that, The first state correction coefficient satisfies: The second state correction coefficient satisfies: The third state correction coefficient satisfies: in, Storage duration; This is the static attenuation coefficient; Initial depreciation factor for inventory; This represents the cumulative number of loops. To design cycle life; It is a nonlinear accelerating decay exponent, and its value is greater than 1; This is the lower limit of the residual value.

4. The electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries according to claim 1, characterized in that, The step of determining the quality premium coefficient based on the sorting grade includes: Obtain historical transaction data corresponding to each of the sorting levels; The normalized transaction value is determined based on the historical transaction data corresponding to each of the sorting levels. Based on the normalized transaction value corresponding to different sorting levels, determine the quality premium coefficient corresponding to each sorting level, and ensure that the quality premium coefficient corresponding to different sorting levels meets the preset level constraints.

5. The electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries according to claim 4, characterized in that, The normalized transaction value and the quality premium coefficient satisfy the following: And the quality premium coefficient corresponding to different sorting grades Apply preset level constraints; in, This represents the actual transaction value of battery cells traded historically. This corresponds to the rated capacity of the battery cells traded in the historical records. This is the estimated average health status of the corresponding historically traded battery cells; This corresponds to the sorting grade of the battery cells sold in the past. For sorting levels The corresponding quality premium coefficient; For level indicator functions; This represents the regression error; This refers to the sample serial number of the battery cells traded in the past. For sorting levels.

6. The electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries according to claim 1, characterized in that, The establishment of a mapping relationship between the uncertainty of health status prediction and the degree of value discount based on historical transaction data includes: Based on the historical transaction data, determine the deterministic valuation and actual transaction value corresponding to each historically transacted battery cell; The degree of observed discount is determined based on the actual transaction value of each historically traded battery cell and its corresponding deterministic valuation. Obtain the estimated standard deviation of the health status for each of the historically traded battery cells, and establish the correspondence between the estimated standard deviation of the health status and the observed discount level; The mapping relationship between the uncertainty of the health status prediction and the degree of value discount is determined based on the correspondence.

7. The electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries according to claim 6, characterized in that, The observed discount rate satisfies: The risk aversion coefficient is determined based on the observed discount level and the estimated standard deviation of the health status corresponding to each historically traded battery cell. The mapping relationship satisfies: in, This represents the actual transaction value of battery cells traded historically. This corresponds to a certainty valuation; The corresponding observation discount level; Risk aversion coefficient; This refers to the historical number of battery cells traded. The standard deviation of the estimated health status of the corresponding historically traded battery cells; To observe the noise variance of the discount level; It is the probability density function of a normal distribution; This refers to the sample serial number of the battery cells traded in the past. Uncertainty discount factor corresponding to the uncertainty in predicting health status; The standard deviation is estimated for the health status of the battery cell to be valued.

8. The electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries according to claim 1, characterized in that, The dynamic updating of the benchmark unit price based on historical transaction data of the same type of battery cells includes: The initial probability distribution of the benchmark unit price is determined based on the historical transaction unit price of the same type of battery cell within a first preset time range; Obtain the newly added transaction price of the same type of battery cell, and update the initial probability distribution based on the newly added transaction price of the same type of battery cell to obtain the updated probability distribution; The updated benchmark unit price is determined based on the updated probability distribution.

9. The electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries according to claim 8, characterized in that, The initial probability distribution of the benchmark unit price satisfies: in, The base unit price; The initial average of the benchmark unit price; The initial variance of the benchmark unit price; The unit price of the same type of battery cell corresponding to the historical time unit; This refers to the sequence number of the historical time unit; The number of historical time units within the first preset time range; This represents a normal distribution.

10. The electrochemical-risk coupled dynamic valuation method for the residual value of new energy batteries according to claim 1, characterized in that, The step of determining the residual value and valuation range of the battery cell to be valued based on the risk-adjusted value, the updated benchmark unit price, and the market volatility parameters includes: The residual value is determined based on the risk-adjusted value and the updated benchmark unit price; The fluctuation range corresponding to the residual value valuation is determined based on the market fluctuation parameters. The valuation range is determined based on the residual value and the fluctuation range.

Images