Method, system, electronic device and storage medium for dynamic prediction of shell corrosion

Abstract

The application relates to the technical field of corrosion life prediction, and discloses a dynamic prediction method and system for shell corrosion, electronic equipment and a storage medium, the method specifically comprises the following steps: determining a corrosion acceleration correction term of a shell to be predicted; obtaining a corrosion acceleration coefficient based on the corrosion acceleration correction term; substituting the corrosion acceleration coefficient into a corrosion kinetics model to obtain corrosion degree data; obtaining a predicted resistance according to the corrosion degree data and the initial thickness of the shell to be predicted; updating the corrosion acceleration correction parameter by using a recursive least square method based on the error between the measured resistance and the predicted resistance, so that the corrosion acceleration coefficient is adjusted; and predicting the future value of the corrosion degree data by using the corrosion kinetics model based on the adjusted corrosion acceleration coefficient, and determining the predicted failure time according to the future value. The method provided by the application can continuously correct the model parameters to realize dynamic prediction of the failure time, and is suitable for adapting to the influence of environmental changes and working condition drifts of the shell during actual storage.

Description

Technical Field

[0001] This invention relates to the field of corrosion lifetime prediction technology, and in particular to dynamic prediction methods, systems, electronic devices, and storage media for casing corrosion. Background Technology

[0002] Cylindrical lithium-ion batteries often require storage for several years in scenarios such as energy storage power stations, backup power supplies, and inventory management. During this period, the battery casing is exposed to the environment for extended periods, and surface corrosion of the casing is a significant factor affecting the structural integrity and long-term reliability of the battery. When casing corrosion develops to a certain extent, it can lead to casing perforation, causing electrolyte leakage and safety risks. Therefore, predicting the casing corrosion process under storage conditions is a practical requirement in battery reliability management. However, current technologies for predicting casing corrosion are typically based on laboratory conditions, relying on testing or analyzing the chemical or physical properties of materials. These calculations are only rough based on the characteristics of the coating and casing materials, without considering the accelerating factors of the environment in which the casing is located. The failure prediction results provided are often highly inaccurate, leading to premature battery damage beyond the expectations of maintenance personnel, resulting in additional losses and safety risks. Summary of the Invention

[0003] This invention aims to address at least one of the technical problems existing in the prior art. To this end, this invention proposes a dynamic prediction method, system, electronic device, and storage medium for casing corrosion, capable of dynamically adjusting the predicted failure time by taking into account accelerating factors in the casing environment.

[0004] A dynamic prediction method for shell corrosion according to a first aspect embodiment of the present invention includes: Based on the initial values ​​of the environmental parameters, contact surface state parameters, and corrosion acceleration correction parameters of the microenvironment in which the shell to be predicted is located, the time-varying corrosion acceleration correction term is determined. The corrosion acceleration correction term is used to correct the baseline corrosion rate of the shell to be predicted, and the corrosion acceleration coefficient is obtained. The corrosion acceleration coefficient is substituted into a preset corrosion kinetic model to obtain the corrosion degree data of the shell to be predicted; wherein, the corrosion kinetic model can solve the corrosion degree data based on the corrosion acceleration coefficient and a pre-calibrated environmental correction constant; Based on the corrosion degree data and the initial thickness of the shell to be predicted, the predicted resistance of the shell to be predicted is obtained; The measured resistance of the shell to be predicted is obtained. Based on the error between the measured resistance and the predicted resistance, the corrosion acceleration correction parameter is identified and updated online by recursive least squares method to adjust the corrosion acceleration correction term and the corrosion acceleration coefficient. Based on the adjusted corrosion acceleration coefficient, the future value of the corrosion degree data is recursively predicted through the corrosion kinetic model, and the predicted failure time of the shell to be predicted is determined according to the future value and the preset failure criterion.

[0005] The embodiments of this invention have at least the following beneficial effects: This invention incorporates actual environmental parameters under storage conditions into a corrosion kinetic model, and uses the error between the measured resistance and the predicted resistance as a feedback signal. Through online identification and updating, the model maintains its predictive ability throughout the entire actual storage process. Compared to existing technologies that can only rely on structural protection design to reduce corrosion risks at the front end and cannot provide quantitative predictions of the corrosion process of manufactured batteries, this invention can output corrosion degree data and predict failure time in real time for manufactured and stored batteries based on their actual environmental parameters. Furthermore, it can continuously correct model parameters through online identification and updating, adapting to environmental changes and operating condition drift of the casing during actual storage.

[0006] According to some embodiments of the present invention, the corrosion acceleration correction term includes an environmental corrosion intensity index; when the corrosion acceleration correction term includes the environmental corrosion intensity index, the step of determining the time-varying corrosion acceleration correction term based on the environmental parameters of the microenvironment in which the shell to be predicted is located, the contact surface state parameters, and the initial values ​​of the corrosion acceleration correction parameter includes: Set an initial value for the first weight associated with the environmental corrosion intensity index, wherein the first weight is a corrosion acceleration correction parameter corresponding to the environmental corrosion intensity index; The environmental parameters of the microenvironment in which the shell to be predicted is located and the nominal resistance of the shell to be predicted are obtained; wherein, the environmental parameters include temperature, relative humidity, bottom liquid thickness and bottom local oxygen partial pressure; For each of the environmental parameters, a microenvironmental correction term is determined relative to its respective reference value; wherein the microenvironmental correction term is determined based on exponential weighting or exponential relationships based on the Arrhenius formula. Based on all the microenvironment correction terms, the microenvironment correction function is obtained; The instantaneous resistance change rate of the shell to be predicted is normalized based on the nominal resistance, and the normalization result is corrected by the microenvironment correction function to obtain the environmental corrosion intensity index.

[0007] According to some embodiments of the present invention, the corrosion acceleration correction term includes a historical cumulative effect index; when the corrosion acceleration correction term includes the historical cumulative effect index, the step of determining the time-varying corrosion acceleration correction term based on the environmental parameters, contact surface state parameters, and initial values ​​of the corrosion acceleration correction parameter of the storage microenvironment where the shell to be predicted is located further includes: The initial values ​​of the second weight, short-term memory time constant, and long-term memory time constant associated with the historical cumulative effect index are set; wherein, the second weight, the short-term memory time constant, and the long-term memory time constant are the corrosion acceleration correction parameters corresponding to the historical cumulative effect index; Based on the short-term memory time constant and its corresponding first cumulative weight, and the long-term memory time constant and its corresponding second cumulative weight, a double exponential decay kernel is constructed. Obtain the environmental corrosion intensity index corresponding to each moment within a preset historical time period. Based on the dual-exponential decay kernel, perform weighted accumulation processing on the environmental corrosion intensity index within the preset historical time period to obtain the historical cumulative effect index at the current moment.

[0008] According to some embodiments of the present invention, the corrosion acceleration correction term includes the contact surface galvanic corrosion index; when the corrosion acceleration correction term includes the contact surface galvanic corrosion index, the step of determining the time-varying corrosion acceleration correction term based on the environmental parameters of the microenvironment in which the housing to be predicted is located, the contact surface state parameters, and the initial values ​​of the corrosion acceleration correction parameter further includes: The initial value of the third weight associated with the contact surface galvanic corrosion index is set; wherein, the third weight is the corrosion acceleration correction parameter corresponding to the contact surface galvanic corrosion index; The contact surface state parameters of the housing to be predicted are obtained, including the initial contact surface insulation resistance after the housing to be predicted is first installed and the online contact surface insulation resistance of the housing to be predicted is obtained in real time. The galvanic corrosion index of the contact surface is obtained based on the initial contact surface insulation resistance, the online contact surface insulation resistance, the contact corrosion reference coefficient, and the crevice corrosion development function; wherein, the crevice corrosion development function characterizes the change law of the crevice corrosion damage degree of the shell to be predicted over time.

[0009] According to some embodiments of the present invention, the corrosion degree data includes coating porosity, substrate corrosion depth, and pitting area ratio; the step of substituting the corrosion acceleration coefficient into a preset corrosion kinetic model to obtain the corrosion degree data of the shell to be predicted includes: The corrosion rate constant of the coating, the corrosion rate constant of the substrate, and the pitting corrosion propagation rate constant are obtained; wherein, the corrosion rate constant of the coating, the corrosion rate constant of the substrate, and the pitting corrosion propagation rate constant are all obtained based on accelerated corrosion experiments; Based on the corrosion acceleration coefficient, the coating corrosion rate constant, and the preset first reaction order, the rate of change of coating porosity is obtained, and the rate of change of coating porosity is numerically integrated to obtain the coating porosity. The rate of change of the substrate corrosion depth is obtained based on the corrosion acceleration coefficient, the coating porosity, the substrate corrosion rate constant, and the preset second reaction order. The rate of change of the substrate corrosion depth is then numerically integrated to obtain the substrate corrosion depth. Based on the corrosion acceleration coefficient, the pitting corrosion propagation rate constant, and the preset third reaction order, the propagation rate of the pitting corrosion area ratio is obtained, and the propagation rate of the pitting corrosion area ratio is numerically integrated to obtain the pitting corrosion area ratio.

[0010] According to some embodiments of the present invention, the step of obtaining the predicted resistance of the shell to be predicted based on the corrosion degree data and the initial thickness of the shell to be predicted includes: A first resistive corrosion mapping coefficient and a second resistive corrosion mapping coefficient are determined for the shell to be predicted; wherein the first resistive corrosion mapping coefficient is related to the resistance of the corrosion products of the shell to be predicted, and the second resistive corrosion mapping coefficient is related to the resistance of the pitting corrosion site of the shell to be predicted. Based on the first resistance corrosion mapping coefficient and the substrate corrosion depth, a first resistance correction term is obtained; Based on the second resistance corrosion mapping coefficient and the pitting area ratio, a second resistance correction term is obtained; The nominal resistance is corrected based on the first resistance correction term and the second resistance correction term to obtain the predicted resistance.

[0011] According to some embodiments of the present invention, after the steps of obtaining the measured resistance of the shell to be predicted, and based on the error between the measured resistance and the predicted resistance, identifying and updating the corrosion acceleration correction parameter online using the recursive least squares method to adjust the corrosion acceleration correction term and the corrosion acceleration coefficient, the method further includes: Physical constraint projection is performed on the updated corrosion acceleration correction parameters to ensure that the updated corrosion acceleration correction parameters are within the corresponding physically reasonable range. Determine whether the updated corrosion acceleration correction parameter meets the convergence condition; wherein, the convergence condition includes: the error between the predicted resistance and the measured resistance is less than a preset error threshold within a first number of consecutive cycles, and the fluctuation amplitude of the corrosion acceleration correction parameter is less than a preset fluctuation threshold within a second number of consecutive cycles. When the convergence condition is met, a confidence index is determined based on the statistics of the error; the confidence index characterizes the reliability of the prediction result of the predicted failure time.

[0012] A dynamic prediction system for shell corrosion according to a second aspect embodiment of the present invention includes: The corrosion correction module is used to determine a time-varying corrosion acceleration correction term based on the environmental parameters, contact surface state parameters, and initial values ​​of corrosion acceleration correction parameters of the microenvironment in which the shell to be predicted is located; the corrosion correction module is also used to correct the reference corrosion rate of the shell to be predicted based on the corrosion acceleration correction term to obtain a corrosion acceleration coefficient. The corrosion state acquisition module is used to substitute the corrosion acceleration coefficient into a preset corrosion kinetic model to obtain corrosion degree data of the shell to be predicted; wherein, the corrosion kinetic model can solve the corrosion degree data based on the corrosion acceleration coefficient and a pre-calibrated environmental correction constant; the corrosion state acquisition module is also used to obtain the predicted resistance of the shell to be predicted based on the corrosion degree data and the initial thickness of the shell to be predicted. The recursive update module is used to obtain the measured resistance of the shell to be predicted, and based on the error between the measured resistance and the predicted resistance, to identify and update the corrosion acceleration correction parameters online by recursive least squares method, so as to adjust the corrosion acceleration correction term and the corrosion acceleration coefficient. The corrosion prediction module is used to predict the future value of the corrosion degree data based on the adjusted corrosion acceleration coefficient through the corrosion kinetic model, and to determine the predicted failure time of the shell to be predicted based on the future value and the preset failure criterion.

[0013] An electronic device according to a third aspect of the present invention includes a memory and a processor, the memory being used to store at least one program, and the processor being used to load the at least one program to execute the dynamic prediction method for casing corrosion described in the above-described aspects.

[0014] According to a fourth aspect of the present invention, a computer-readable storage medium stores computer-executable instructions for causing a computer to perform the dynamic prediction method for shell corrosion as described in the above-described aspects.

[0015] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0016] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a flowchart of the dynamic prediction method for shell corrosion according to an embodiment of the present invention; Figure 2 This is a schematic block diagram of the dynamic prediction system for shell corrosion according to an embodiment of the present invention; Figure 3This is a schematic diagram illustrating the process of obtaining the environmental corrosion intensity index according to an embodiment of the present invention; Figure 4 This is the convergence curve of the first weight in this embodiment of the invention; Figure 5 This is the convergence curve of the second weight in this embodiment of the invention; Figure 6 This is the convergence curve of the third weight in this embodiment of the invention; Figure 7 The convergence curves of the short-term memory time constant and the long-term memory time constant in an embodiment of the present invention; Figure 8 This is a curve showing the corrosion acceleration coefficient in a variable environmental intensity experiment according to an embodiment of the present invention; Figure 9 This is a comparison curve of the measured resistance and the predicted resistance in an embodiment of the present invention; Figure 10 This is a structural diagram of an electronic device provided in another embodiment of the present invention. Detailed Implementation

[0017] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0018] In the description of this invention, "several" means one or more, "multiple" means two or more, "greater than," "less than," "exceeding," etc. are understood to exclude the stated number, and "above," "below," "within," etc. are understood to include the stated number. If "first," "second," etc. are used in the description, they are only for the purpose of distinguishing technical features and should not be construed as indicating or implying relative importance or implicitly indicating the number of indicated technical features or the order of the indicated technical features.

[0019] Cylindrical lithium-ion batteries often require storage for several years in scenarios such as energy storage power stations, backup power supplies, and inventory management. During this period, the battery casing is exposed to the environment for extended periods, and surface corrosion of the casing is a significant factor affecting the structural integrity and long-term reliability of the battery. When casing corrosion develops to a certain extent, it can lead to casing perforation, causing electrolyte leakage and safety risks. Therefore, predicting the casing corrosion process under storage conditions is a practical requirement in battery reliability management. However, current technologies for predicting casing corrosion are typically based on laboratory conditions, relying on testing or analyzing the chemical or physical properties of materials. These calculations are only rough based on the characteristics of the coating and casing materials, without considering the accelerating factors of the environment in which the casing is located. The failure prediction results provided are often highly inaccurate, leading to premature battery damage beyond the expectations of maintenance personnel, resulting in additional losses and safety risks.

[0020] The concepts involved in this invention are explained below: Salt spray test: An environmental test that uses artificially simulated salt spray environment conditions created by salt spray test equipment to assess the corrosion resistance of products or metal materials; RLS: Recursive Least Square, is a recursive implementation of the least squares algorithm. It is a fast algorithm of the least squares algorithm. The recursive least squares adaptive filter is the best filter for a set of known data. BMS: Battery Management System, is a core electronic unit that monitors the battery status (such as SOC, voltage, and temperature) in real time and controls the charging and discharging process to ensure the safe and efficient operation of the battery. Pitting corrosion, also known as pitting corrosion, is a form of corrosion that is concentrated on a very small area of ​​the metal surface and penetrates deep into the metal. It is generally characterized by a small diameter and a deep depth. The ratio of the maximum depth of the pit to the average corrosion depth of the metal is called the pitting coefficient. The larger the pitting coefficient, the more severe the pitting corrosion.

[0021] Please refer to Figures 1-10 , Figure 1 This is a flowchart of a dynamic prediction method for shell corrosion according to an embodiment of the present invention. This embodiment discloses a dynamic prediction method for shell corrosion, which includes, but is not limited to, the following steps: Step S100: Determine the time-varying corrosion acceleration correction term based on the initial values ​​of the environmental parameters, contact surface state parameters, and corrosion acceleration correction parameters of the microenvironment in which the shell to be predicted is located. It should be noted that in step S100, environmental parameters include, but are not limited to, parameters in the storage environment that can affect the corrosion process of the shell, such as temperature, humidity, oxygen partial pressure, and chloride ion concentration; contact surface state parameters include, but are not limited to, the contact area between the shell and the support, and the gap width between the shell and the support; the corrosion acceleration correction parameter can be the influence weight of each corrosion acceleration correction term on the corrosion acceleration coefficient. The actual operating conditions of the shell to be predicted during the storage period are transformed into a corrosion acceleration correction term that varies with time, so as to adjust the baseline corrosion rate of the shell to be predicted. It can be understood that the corrosion acceleration correction term is not a fixed value that remains unchanged after it is determined, but a time-varying variable jointly determined by the environmental parameters, contact surface state parameters, and the initial value of the corrosion acceleration correction parameter, thereby realizing the dynamic influence on the corrosion acceleration prediction; in practical applications, the corrosion acceleration correction parameter can be determined by experience or the results of specific experiments during the first run of the system.

[0022] Step S200: Correct the baseline corrosion rate of the shell to be predicted based on the corrosion acceleration correction term to obtain the corrosion acceleration coefficient; It should be noted that step S200 applies the time-varying corrosion acceleration correction term determined in step S100 to the baseline corrosion rate to obtain a corrosion acceleration coefficient with actual operating condition response capability. The baseline corrosion rate is the intrinsic corrosion rate of the material obtained experimentally under laboratory standard conditions, reflecting the basic corrosion tendency of the shell material in the reference environment; while the corrosion acceleration correction term output by S100 carries the acceleration information of the corrosion process by the actual storage environment. After the two are combined, the baseline corrosion rate completes the conversion from laboratory reference conditions to actual storage conditions.

[0023] Specifically, in this embodiment of the invention, the corrosion acceleration coefficient is shown in the following formula ①: ···①, In formula ①, As a baseline corrosion rate, under standard laboratory conditions, =1, however, in the actual storage environment of the shell to be predicted, there must be multiple types of influencing factors, making >1, therefore step S200 quantifies the various types of influencing factors. In this embodiment of the invention, the corrosion acceleration correction term includes the environmental corrosion intensity index. Historical cumulative effect index Surface galvanic corrosion index , among them , , The corresponding corrosion acceleration correction parameters are represented as weights in this embodiment of the invention. It is understood that, in a specific environment, the corrosion acceleration correction items are not limited to the types mentioned in the above embodiments, and can be additionally added and calibrated according to environmental requirements.

[0024] Step S300: Substitute the corrosion acceleration coefficient into the preset corrosion kinetic model to obtain the corrosion degree data of the shell to be predicted; wherein, the corrosion kinetic model can solve the corrosion degree data based on the corrosion acceleration coefficient and the pre-calibrated environmental correction constant. It should be noted that in step S300, the corrosion acceleration coefficient is substituted into the preset corrosion kinetic model to obtain the corrosion degree data of the shell to be predicted. In addition to the corrosion acceleration coefficient, this corrosion kinetic model introduces a pre-calibrated environmental correction constant. The environmental correction constant is an intrinsic property parameter of the material-environment system, predetermined through accelerated corrosion tests, and remains unchanged during the actual operation of the method in this embodiment of the invention. Based on this, the model solving process is driven by both the time-varying corrosion acceleration coefficient and the fixed calibration constant. The results obtained reflect both real-time operating condition changes and are anchored to the physical calibration basis of material corrosion behavior. Therefore, the current corrosion degree data of the shell to be predicted obtained by the model is dynamically corrected by the environment, rather than being a coarse data solved solely based on the shell material.

[0025] Step S400: Based on the corrosion degree data and the initial thickness of the shell to be predicted, the predicted resistance of the shell to be predicted is obtained; It should be noted that, in step S300 of the above embodiment, although the current corrosion level data of the shell to be predicted is solved, because... , , The corrosion acceleration correction parameters, in the current step S400, are either initial values ​​that have not been identified and updated in subsequent steps, or values ​​from the previous round of identification and updates. Taking the initial value as an example... , , The corrosion acceleration correction parameters are set based on empirical values ​​that have not been tested in the current environment. These empirical values ​​may be inaccurate in the current environment, leading to deviations in the predicted current corrosion level data. Therefore, this embodiment of the invention selects to identify and update the corrosion acceleration correction coefficient using the RLS algorithm, so that the corrosion acceleration correction coefficient can more objectively reflect the influence of the current environment on the corrosion acceleration coefficient (for example, in a specific storage environment, the influence weight of environmental parameters should have the greatest impact on the corrosion acceleration coefficient, but the pre-set empirical values ​​actually minimize the influence weight of environmental parameters on the corrosion acceleration coefficient). However, the corrosion level parameters obtained from the model solution in step S300, such as coating porosity, substrate corrosion depth, and pitting area ratio, cannot be accurately measured in any storage scenario unless measured using specialized instruments and after the shell to be predicted is destroyed. Therefore, after obtaining the corrosion level data, the corrosion kinetic model of this embodiment of the invention will also fit and predict the current resistance value of the shell to be predicted based on the corrosion level data and the initial thickness determined before storage, thereby utilizing the ease of obtaining the measured resistance value of the shell to be predicted to obtain the update driver for the RLS algorithm.

[0026] Step S500: Obtain the measured resistance of the shell to be predicted. Based on the error between the measured resistance and the predicted resistance, identify and update the corrosion acceleration correction parameters online using the recursive least squares method to adjust the corrosion acceleration correction term and corrosion acceleration coefficient. It should be noted that step S500 compares the predicted resistance output in step S400 with the measured resistance acquired online, using the error between the two as a feedback signal. The corrosion acceleration correction parameters are then updated online using the recursive least squares method, thereby adjusting the corrosion acceleration correction term and corrosion acceleration coefficient. The measured resistance is the external observation value that the shell to be predicted can continuously acquire from the BMS during the storage period, while the predicted resistance is the model's calculation result for the same physical quantity under the current parameter state. When there is a deviation between the two, this deviation directly reflects the degree of mismatch between the current corrosion acceleration correction parameters and the actual corrosion behavior of the shell. It can be understood that step S500 enables the prediction method of this embodiment to have closed-loop self-correction capability. The corrosion acceleration correction parameters are updated based on newly arrived measured data in each identification step. The updated parameters then enter the next round of calculation of the corrosion acceleration correction term and corrosion acceleration coefficient, thereby gradually approximating the corrosion degree data and predicted resistance output by the model to the actual corrosion state of the shell. This recursive mechanism of "prediction-comparison-correction-re-prediction" ensures that even if the corrosion characteristics of the shell change due to the migration of operating conditions during the actual storage process, the model can continue to track it in subsequent operations without the need for offline recalibration. In addition, step S500 is computationally efficient, avoiding the repeated calculation of all historical data, with a complexity of O(n) → O(1), which can be applied to devices with limited computing power such as BMS / edge computing devices.

[0027] Specifically, the error between the predicted resistance and the measured resistance is: ...Form ②, In formula ②, This represents the prediction error (error) at the current moment. For actual measured resistance, To predict the resistance, the error is based on the difference between the measured resistance and the predicted resistance. The process of performing the least squares method is as follows: ·····Form ③, ...Form 4, ...Form ⑤, In equations ②, ③, ④, and ⑤, t-1 represents the current time (or iteration number), and t-1 represents the previous time. The parameter vector representing the current moment, in the embodiments of the present invention ; Here, the gain matrix represents the learning step size / gain. The larger the value, the larger the adjustment step size. The mathematical essence of Kalman gain is an equivalent form of Kalman gain, which determines what proportion of the "measured error" is used to correct the parameters. P(t) is the covariance matrix, representing the uncertainty of the current parameter estimate. As the amount of data increases, P(t) gradually decreases, indicating that the parameter estimate becomes more accurate. Characterization of regression vector yes The transpose, in embodiments of the present invention includes , , In the embodiments of the present invention ,when The closer the value is to 1 (e.g., 0.995), the slower the forgetting, the stronger the system's noise resistance, and the more stable the estimation, but the slower the tracking speed of parameter changes. The smaller the value (e.g., 0.92), the faster the forgetting occurs, the more sensitive the system is to new data, and the faster the tracking speed, but it is easily affected by noise and fluctuates.

[0028] Step S600: Based on the adjusted corrosion acceleration coefficient, the future value of the corrosion degree data is recursively predicted through the corrosion kinetic model, and the predicted failure time of the shell to be predicted is determined according to the future value and the preset failure criterion.

[0029] It should be noted that step S600 identifies the corrosion acceleration coefficient corresponding to the updated corrosion acceleration correction parameter in step S500, recursively solves for the corrosion degree data on the future time axis using the corrosion kinetic model to obtain the future value, and compares the future value with the preset failure criterion to determine the predicted time point corresponding to the shell reaching the failure condition. It can be understood that the future value represents the recursive prediction value of the corrosion kinetic model along the time axis, that is, the corrosion degree data corresponding to the future time predicted based on the corrosion kinetic model. This step is executed after the online identification step; therefore, the acceleration coefficient used for extrapolation already includes the correction information of all measured data up to the current time, rather than relying solely on the initial setting value. During the recursive process, the corrosion kinetic model calculates the corrosion degree data hourly based on the adjusted acceleration coefficient. When the future value first meets the preset failure criterion, the corresponding time coordinate is the predicted failure time. Because the aforementioned steps establish an error feedback channel between the predicted resistance and the measured resistance, and the basic parameters used by S600 have been calibrated, the extrapolation process is based on a close approximation of the actual corrosion behavior of the casing. The output results can serve as a basis for battery maintenance and replacement scheduling, and are continuously updated as measured data is continuously corrected. In some embodiments of the present invention, the failure time is predicted. The criteria for judgment are as follows: or ;in, For the thickness of the shell to be predicted, =0.1, meaning failure is determined when the corrosion depth reaches 90% of the shell thickness, and the corresponding time is... This refers to predicting the failure time; Take 0.05~0.1, that is, the pitting area accounts for 5%~10%. Since the depth of pitting corrosion is usually 3~5 times that of conventional substrate corrosion, even if the proportion is small, it may cause the substrate to be penetrated.

[0030] It should be noted that the method provided in steps S100 to S600 forms a closed-loop prediction chain of "actual operating condition driven - physical model solution - measurable resistance verification - error feedback correction - future state extrapolation". For batteries that have been manufactured and put into storage, even if the corrosion resistance of their casing has been determined by the manufacturing process and cannot be changed by structural design, this method can still continuously output the prediction results of their corrosion process and failure time based on the actual environmental parameters of the battery and online resistance measurement data. At the same time, the corrosion degree data obtained is constrained by the corrosion kinetic model and has physical interpretability. It will not produce predictions that contradict the corrosion mechanism due to abnormal changes in individual environmental parameters. In addition, due to the existence of the online identification mechanism, the model parameters can be adaptively adjusted as the actual operating conditions drift. This makes the prediction results effective over a storage period of several years and can directly provide quantifiable decision-making basis for battery maintenance and replacement arrangements in scenarios such as energy storage power stations and backup power supplies.

[0031] Reference Figure 3 Furthermore, step S100 in the above-described embodiments also includes, but is not limited to, the following steps: Step S111: Set the initial value of the first weight associated with the environmental corrosion intensity index, where the first weight is the corrosion acceleration correction parameter corresponding to the environmental corrosion intensity index. Step S112: Obtain the environmental parameters of the microenvironment in which the shell to be predicted is located and the nominal resistance of the shell to be predicted; wherein, the environmental parameters include temperature, relative humidity, bottom liquid thickness and bottom local oxygen partial pressure; Step S113: For each environmental parameter, determine the micro-environment correction term relative to its respective reference value; wherein the micro-environment correction term is determined based on exponential weighting or exponential relationship based on the Arrhenius formula. Step S114: Based on all microenvironment correction terms, obtain the microenvironment correction function; Step S115: Normalize the instantaneous resistance change rate of the shell to be predicted based on the nominal resistance, and correct the normalization result through the microenvironment correction function to obtain the environmental corrosion intensity index.

[0032] It should be noted that the process for obtaining the environmental corrosion intensity index is as follows: Figure 3As shown, the initial value of the first weight associated with the environmental corrosion intensity index is first determined, i.e., the value in the above-described embodiment. The initial value is not the final value used, but rather provides an iterative base state for subsequent online identification and updates, enabling the model to be driven during the first calculation and avoiding computational interruptions due to empty parameters. Simultaneously, the first weight ensures that the environmental corrosion intensity index does not act as a fixed proportion when entering the corrosion acceleration correction term, but rather participates in the construction of the acceleration coefficient with an online adjustable contribution weight, providing a physical channel for subsequent inverse adjustment of this weight by measured resistance error. Subsequently, a microenvironment correction term and a microenvironment correction function are constructed based on temperature, relative humidity, bottom liquid thickness, and bottom local oxygen partial pressure, as follows: ·····Form 6, In formula ⑥, The measured relative humidity at the bottom of the shell to be predicted; For reference humidity; Humidity sensitivity index; The corrosion activation energy can be determined through designed experiments or literature values; The measured temperature at the bottom of the shell to be predicted; For reference temperature; The thickness of the liquid accumulation at the bottom; For reference fluid thickness; The fluid accumulation thickness sensitivity index; This refers to the local oxygen partial pressure at the bottom. Reference oxygen partial pressure; This is the oxygen sensitivity index. In equation ⑥... Each product term is a microenvironment correction term. After obtaining the above data, the system corresponding to this invention processes the data in the following way: Based on temperature, the corrosion activation energy of the matrix material, and a preset reference temperature, a bottom temperature correction term is obtained using the Arrhenius formula. A bottom relative humidity correction term is obtained by exponentially weighting the ratio of relative humidity to a preset reference humidity, based on a preset humidity sensitivity index. A bottom liquid accumulation thickness correction term is obtained by exponentially weighting the ratio of bottom liquid accumulation thickness to a preset reference liquid accumulation thickness, based on a preset liquid accumulation thickness sensitivity index. Finally, an oxygen partial pressure correction term is obtained by exponentially weighting the ratio of bottom local oxygen partial pressure to a preset reference oxygen partial pressure, based on a preset oxygen sensitivity index.

[0033] For each environmental parameter, an environmental correction term relative to its respective reference value is determined. This term transforms the absolute observed values ​​of the original environmental quantities into deviations from baseline conditions. Each parameter employs exponential weighting or an Arrhenius-type exponential relationship, ensuring that the accelerating effect of each environmental parameter on corrosion is no longer directly calculated based on its original dimensions, but rather converted into a comparable correction factor under the unified reference of its respective reference value. Finally, the microenvironment correction function is obtained by multiplying the bottom temperature correction term, bottom relative humidity correction term, liquid accumulation thickness correction term, and oxygen partial pressure correction term. The microenvironment correction function converts the actual abandoned microenvironment to the equivalent corrosion intensity under reference conditions. The corrosion process is essentially the result of the combined effects of multiple environmental factors; corrections for a single factor cannot characterize the combined amplification effect between factors in the actual bottom microenvironment. The microenvironment correction function synthesizes the correction terms through a product, making the synergistic accelerating relationships between environmental parameters mutually influential, consistent with the physical laws of corrosion reactions.

[0034] In some embodiments of the present invention, the above-described... A humidity sensor can be integrated at the bottom of the housing. The temperature can be obtained through a temperature sensor integrated at the bottom of the housing; Liquid that accumulates at the bottom due to factors such as humid weather can be detected by a liquid level sensor integrated at the bottom of the housing. Oxygen can be acquired through an electrochemical oxygen sensor or an optical oxygen sensor integrated at the bottom of the casing. The acquired data can be processed by the BMS system of the battery corresponding to the casing to be predicted, or remotely by a host computer. Furthermore, in the above data... This data is based on the industry standard GB / T 36276-2023 "Lithium-ion Batteries for Power Storage", which stipulates that the humidity of the storage environment for energy storage batteries should be controlled below 60%. The humidity sensitivity index ranges from [1.0 to 2.5] and is calibrated based on accelerated corrosion testing at multiple humidity points. The range of the liquid accumulation sensitivity index is determined based on corrosion comparisons of different liquid accumulation thicknesses; The range of the oxygen sensitivity index is determined based on corrosion comparisons at different oxygen concentrations; The corrosion activation energy ranges from [30, 50] kJ / mol, and is calibrated based on Arrhenius fitting at multiple temperature points.

[0035] After determining the microenvironment correction function, it is necessary to normalize the instantaneous resistance change rate of the shell to be predicted based on the nominal resistance, and then correct the normalization result using the microenvironment correction function to obtain the environmental corrosion intensity index, which specifically includes: ...Form ⑦, In formula ⑦, The nominal resistance characterizing the casing to be predicted, i.e., the reference value of the casing's body resistance measured under standard reference conditions at the time of manufacture and before significant corrosion occurs, is used as the normalization denominator in this embodiment. The instantaneous rate of change of shell resistance is used to reflect the corrosion progress of the shell to be predicted. It is understood that, since the degree of shell corrosion is related to the shell resistance, the ratio of the instantaneous rate of change of shell resistance to the nominal resistance is used to convert the invisible changes in the shell to be predicted into visible parameters, which are then corrected by a microenvironment correction function. The instantaneous rate of change of shell resistance can be achieved through the battery's BMS system and a preset four-wire test circuit. The measurement point is fixed on the shell (e.g., near the bottom of the shell, which is the location with the highest corrosion risk and the natural location for the four-wire method). The BMS completes resistance sampling at certain intervals to obtain the time series of shell resistance, and then the time series is differentially analyzed to obtain the final value.

[0036] Furthermore, step S100 in the above-described embodiments also includes, but is not limited to, the following steps: Step S121: Set the initial values ​​of the second weight, short-term memory time constant, and long-term memory time constant associated with the historical cumulative effect index; wherein, the second weight, short-term memory time constant, and long-term memory time constant are corrosion acceleration correction parameters corresponding to the historical cumulative effect index; Step S122: Construct a double exponential decay kernel based on the short-term memory time constant and its corresponding first cumulative weight, and the long-term memory time constant and its corresponding second cumulative weight; Step S123: Obtain the environmental corrosion intensity index corresponding to each moment within the preset historical time period. Based on the double exponential decay kernel, perform weighted accumulation processing on the environmental corrosion intensity index within the preset historical time period to obtain the historical cumulative effect index at the current moment.

[0037] It should be noted that in the actual storage environment of the battery casing, the humidity is likely not constant and will change with battery transfer and weather changes. High humidity environments can have a significant impact on the casing, which will be reflected in the predicted resistance of the casing through the environmental corrosion intensity index in the above-described embodiments. However, even after the battery casing is transferred from a high-humidity environment to a dry environment, the impact of the previous high-humidity environment on the casing will not disappear in a short time, but will have a lag effect over a period of time. Based on this, steps S121 to S123 first determine the second weight μ associated with the historical cumulative effect index, as well as the initial values ​​of the short-term memory time constant, long-term memory time constant, first cumulative weight, and second cumulative weight. After determination, the corrosion memory kernel function shown below is provided as the historical cumulative effect index. To describe the delayed effects of shelved history on subsequent corrosion: ...Form ⑧, Furthermore, the double exponential decay kernel (memory kernel function) is: ...Form 9, In equations ⑧ and ⑨, The environmental influence at historical moment τ on the corrosion of the predicted shell is characterized. Representation memory kernel function, This is the coating degradation memory time constant (typical value: 30-90 days). The time constant for matrix corrosion memory (typical value: 180-365 days). As the first cumulative weight, For the second cumulative weight and Understandably, A smaller value indicates that the effect decays quickly (e.g., coating peeling, initial surface moisture absorption), and the longer the time, the smaller the early coating stress effect. The larger value indicates that the hysteresis effect of a humid environment on the substrate takes longer than the hysteresis effect on the coating. , By allocating weights, the total contribution of short-term and long-term mechanisms is ensured to remain at 100%, with mechanisms having a larger weight having a stronger dominant effect on the current state. , The initial value can be pre-calibrated based on the material of the shell to be predicted; the corrosion acceleration correction parameter related to the historical cumulative effect index includes, in addition to the second weight, the short-term memory time constant. Long-term memory time constant Short-term memory time constant and long-term memory time constant The memory decay rate of the historical environmental corrosion intensity index is controlled separately in the coating degradation stage and the substrate corrosion stage. The two together determine the dynamic response of the historical cumulative effect index through the dual exponential decay kernel. Therefore, it is also used as a parameter that can be identified and corrected online, and together with the second weight, it participates in the time-varying construction of the corrosion acceleration correction term.

[0038] In general, steps S121-S123 use a double-exponential decay kernel as the time weight to weight and accumulate the environmental corrosion intensity index within a preset historical time period. This allows the decay of historical environmental influences to be carried out according to both short-term and long-term memory time constants, corresponding to the response characteristics of coating degradation and substrate corrosion at two different time scales. Thus, even if the shell is transferred to a milder environment, the corrosion inertia formed in the previous harsh environment continues at both time scales, and the resulting historical cumulative effect index can characterize this nonlinear memory behavior.

[0039] Furthermore, step S100 in the above-described embodiments also includes, but is not limited to, the following steps: Step S131: Set the initial value of the third weight associated with the galvanic corrosion index of the contact surface; wherein, the third weight is the corrosion acceleration correction parameter corresponding to the galvanic corrosion index of the contact surface. Step S132: Obtain the contact surface state parameters of the housing to be predicted. The contact surface state parameters include the initial contact surface insulation resistance after the housing to be predicted is first installed and the online contact surface insulation resistance of the housing to be predicted obtained in real time. Step S133: Based on the initial contact surface insulation resistance, online contact surface insulation resistance, contact corrosion reference coefficient, and crevice corrosion development function, the contact surface galvanic corrosion index is obtained; wherein, the crevice corrosion development function characterizes the change law of the crevice corrosion damage degree of the shell to be predicted over time.

[0040] It should be noted that, in addition to the environmental corrosion intensity index and the historical cumulative effect index accelerating corrosion, the battery casing will inevitably come into contact with the bracket or other supporting structures during storage. Under the influence of the galvanic corrosion effect between different materials and the effect of liquid accumulation in the contact gaps, this embodiment of the invention also introduces the contact surface galvanic corrosion index as a correction term for the corrosion acceleration coefficient. In steps S131 to S133, the initial value of the third weight associated with the contact surface galvanic corrosion index is first determined, i.e., the value in the above-mentioned embodiments. The galvanic corrosion index of the contact surface is then calculated using the following formula. : ...Form 10, In formula ⑩, As a reference coefficient for contact corrosion, it is calibrated by salt spray test (ASTM B117, 3.5% NaCl, 35°C, 500h) of shell-tray paired samples, which comprehensively reflects the material's galvanic properties, initial area and surface condition; The initial contact surface insulation resistance after the battery is installed in place; The contact surface insulation resistance for online monitoring can be obtained by injecting a microampere-level test current through a preset BMS multiplexing voltage sampling circuit; Here is the crevice corrosion development function, which is based on the following formula: ······ , Mode middle, Characterizing the development level of crevice corrosion, Characterizing a saturated growth function, The cumulative amount of liquid exposure between the shell and the supporting structure to be predicted is represented by the integral, where the integral represents the cumulative effect (corrosion is not instantaneous, but rather the cumulative result of the liquid residence time). The ratio describes the degree to which the crevice is filled with liquid; Characterizing the crevice corrosion rate constant. Furthermore, The greater the thickness of the accumulated liquid, the greater the amount of solution in the crevice, the more abundant the corrosive medium, and the more severe the galvanic corrosion effect. The denominator represents the width of the gap between the shell and the supporting structure to be predicted. A narrower gap indicates a thicker liquid film and a more significant oxygen concentration difference for the same amount of liquid. It is understood that the crevice corrosion development function in this embodiment of the invention is used to characterize corrosion as not instantaneous, but rather as the cumulative result of the liquid's residence time.

[0041] In summary, steps S131-S133, by acquiring the initial contact surface insulation resistance and the online contact surface insulation resistance, incorporate the interface state of the initial installation state and the current operating state of the casing into the same comparison framework, enabling continuous online reflection of the contact surface deterioration. Based on this, and combined with the contact corrosion baseline coefficient, the change in interface insulation resistance is transformed into a contact surface galvanic corrosion index. Furthermore, the crevice corrosion development function is used to describe the temporal evolution of crevice corrosion damage, allowing this index to not only reflect the current interface state but also the dynamic characteristics of the cumulative effect of local damage over time. Thus, the contact surface galvanic corrosion index, as a local dimension input to the corrosion acceleration correction term, works in conjunction with the overall environmental corrosion intensity index and the historical cumulative effect index, enabling the corrosion acceleration coefficient to characterize the spatial non-uniformity of corrosion under bottom-mounted conditions, compensating for the shortcomings of being driven solely by overall environmental parameters.

[0042] Furthermore, step S300 in the above-described embodiments also includes, but is not limited to, the following steps: Step S310: Obtain the coating corrosion rate constant, the substrate corrosion rate constant, and the pitting corrosion propagation rate constant; wherein, the coating corrosion rate constant, the substrate corrosion rate constant, and the pitting corrosion propagation rate constant are all obtained based on accelerated corrosion experiments; Step S320: Based on the corrosion acceleration coefficient, the coating corrosion rate constant and the preset first reaction order, the rate of change of coating porosity is obtained, and the rate of change of coating porosity is numerically integrated to obtain the coating porosity. Step S330: Based on the corrosion acceleration coefficient, coating porosity, substrate corrosion rate constant and preset second reaction order, the rate of change of substrate corrosion depth is obtained, and the rate of change of substrate corrosion depth is numerically integrated to obtain the substrate corrosion depth. Step S340: Based on the corrosion acceleration coefficient, the pitting corrosion propagation rate constant, and the preset third reaction order, the propagation rate of the pitting corrosion area is obtained, and the propagation rate of the pitting corrosion area is numerically integrated to obtain the pitting corrosion area percentage.

[0043] It should be noted that steps S310 to S340, through pre-calibrated coating corrosion rate constants, substrate corrosion rate constants, and pitting corrosion rate constants, transform the corrosion acceleration coefficient into the evolution process of three corrosion degree data: coating porosity, substrate corrosion depth, and pitting area ratio. In solving for each state variable, firstly, based on the corrosion acceleration coefficient and the corresponding rate constant, and combined with the preset reaction order, the rate of change of the corresponding state variable is determined; then, each rate of change is numerically integrated to obtain the value of the state variable over time. The rate of change of substrate corrosion depth also incorporates coating porosity, allowing the accelerating effect of coating degradation on substrate corrosion to be dynamically transmitted, reflecting the coupled and progressive relationship between coating degradation, substrate corrosion, and pitting corrosion propagation. Therefore, the corrosion degree data output by the corrosion kinetic model are driven by the time-varying corrosion acceleration coefficient while maintaining the physical correlation between the internal state evolution and the corrosion acceleration coefficient.

[0044] The corrosion kinetic model of this invention includes the following set of differential equations: ······Mode , ······Mode , ······Mode , In the above formula, the expression Characterizing the degradation of the coating porosity of the shell to be predicted, among which The corrosion acceleration coefficient, The corrosion constant of the nickel plating is determined based on preset experimental conditions. The corrosion constant of the iron matrix is ​​determined based on preset experimental conditions. The pitting corrosion propagation rate constant is determined based on preset experimental conditions. This represents the reaction order that causes the coating porosity to degrade. The reaction order is the depth of corrosion of the substrate. This represents the reaction order for the expansion of the pitting corrosion area. (Equation) The physical significance of determining the plating porosity lies in the fact that the nickel plating layer is the first line of defense for the battery casing. Increased porosity means decreased protection and increased exposed area of ​​the iron substrate. The data processing flow is to obtain real-time temperature and relative humidity → look up tables / calculate. → multiplied by Obtain the actual degradation rate → Integrate to obtain the coating porosity This formula can be used to determine when the coating fails; The physical significance of determining the depth of matrix corrosion lies in the fact that matrix corrosion is the main cause of shell thinning. Approximately the thickness of the shell When structural strength is lost, the data processing flow involves inputting and acquiring real-time temperature, real-time relative humidity, and the current coating porosity. calculate → multiplied by →Integration yields the depth of substrate corrosion. (i.e., the output of the first formula) (This becomes the input to the second formula), which can be used to determine when the substrate is corroded to a minimum; Formula The physical significance of determining the percentage of pitting corrosion area lies in the parallel development of uniform corrosion (Equation 2) and localized corrosion (Equation 3). Pitting corrosion is the direct cause of "perforation failure"—even if the average corrosion depth is not large, localized penetration can lead to electrolyte leakage. The data processing flow involves obtaining real-time temperature, real-time relative humidity, and real-time chloride ion concentration → calculation. → multiplied by →Integration yields the percentage of pitting area. This formula can be used to calculate the risk of localized perforation. Furthermore, the reaction order... , , These are used to describe the rate of change of the corrosion degree data for different coating porosities, substrate corrosion depths, and pitting area ratios. When each level = 1, the corrosion rate is proportional to the current corrosion degree data; when each level < 1, it indicates that the reaction rate is limited by the diffusion-controlled process, resulting in a slow rate increase; when each level > 1, it indicates that the reaction rate conforms to the autocatalytic process, and the rate increases rapidly.

[0045] Mode ,Mode ,Mode The three models simulated the entire process of battery casing corrosion. The specific formula is as follows: ······ , Mode middle, Characterize the nickel corrosion rate determined by RH experiments at 85°C / 85% conditions; The calibration temperature is 85°C. The calibrated humidity is 85%. The activation energy for nickel oxidation (literature value); ; represents the nickel ground humidity sensitivity index (an index determined by fitting multiple sets of preset experimental conditions); R is a constant, taken as 8.314 in this embodiment; T is the measured temperature; RH is the measured relative humidity. Based on formula According to embodiments of the present invention, the corrosion constant of the nickel plating can be obtained by determining the temperature T and relative humidity RH of the current environment. The temperature T and relative humidity RH can be obtained in the same way as the environmental corrosion intensity index obtained in the above-mentioned embodiments.

[0046] Furthermore, Obtained based on the following formula: ······ , Mode middle, The corrosion rate of nickel-iron determined by RH experiments under 85°C / 85% conditions was characterized. The calibration temperature is 85°C. The calibrated humidity is 85%. The activation energy for iron oxidation (literature value); The humidity sensitivity index for iron corrosion (more sensitive than nickel) is given by formula [formula missing]. In Specifically, it includes: ······ , Mode middle, This is the coating failure acceleration factor; It is a porosity sensitivity index; This is a coating coupling term, and its physical meaning is that the higher the porosity, the more substrate is exposed. Specifically, it can be calibrated by comparing the substrate corrosion rates of samples with different coating integrity (intact / scratched / partially coated / bare steel) under the same corrosion conditions. This allows for the preparation of shell samples with different coating integrity. = 0, 0.2, 0.5, 0.8, 1.0; , It is also an index determined by fitting based on multiple sets of preset experimental conditions.

[0047] Furthermore, the pitting rate constant Obtained based on the following formula: ·Mode , Mode middle, The pitting corrosion propagation rate determined by RH experiments at 85°C / 85% was characterized. The calibration temperature is 85°C. The calibrated humidity is 85%. The activation energy for pitting corrosion (literature value); The humidity sensitivity index (the specific value is between that of nickel and iron). The chloride ion concentration can be obtained in real time by using a chloride ion concentration sensor installed on the shell to be predicted; To calibrate chloride ion concentration; This is the chloride ion sensitivity index. Understandably, chloride ions can be considered the "switch" for pitting corrosion—without them, pitting corrosion almost never occurs, but their presence drastically accelerates the corrosion process. Its mechanism begins with the destruction of the passivation film on the metal surface: chloride ions preferentially adsorb onto the metal surface, displacing oxygen atoms from the passivation film, thereby disintegrating the protective layer mainly composed of nickel and chromium oxides; subsequently, in crevices or pits, chloride ions continuously accumulate and undergo hydrolysis, forming an acidic microenvironment. This localized acidity further accelerates metal dissolution, releasing large amounts of Fe²⁺. + ions, while Fe² + The increase in chloride ions attracts more chloride ions to maintain charge balance, thus forming a vicious autocatalytic cycle of "acidification accelerating corrosion - more metal ions dissolving - more chloride ions migrating inward - further exacerbating corrosion". Therefore, in addition to temperature and relative humidity, chloride ion concentration data is also needed to more accurately describe the pitting corrosion propagation of the shell to be predicted.

[0048] It is understandable that the formula , , The process of obtaining the coating corrosion rate constant, the substrate corrosion rate constant, and the pitting corrosion propagation rate constant is illustrated. These three constants are obtained through accelerated corrosion experiments, and the type of accelerated corrosion experiment can be adjusted according to the type of corrosion severity data. In this embodiment of the invention, the coating corrosion rate constant, the substrate corrosion rate constant, and the pitting corrosion propagation rate constant are all calibrated based on accelerated experiments under specific implementation conditions. During the actual operation of the model, as long as real-time temperature, real-time relative humidity, and real-time chloride ion concentration data are obtained, the definite values ​​of the above three equations can be obtained for subsequent calculation of specific corrosion severity data.

[0049] Furthermore, step S400 in the above-described embodiments also includes, but is not limited to, the following steps: Step S410: Determine the first resistive corrosion mapping coefficient and the second resistive corrosion mapping coefficient of the shell to be predicted; wherein, the first resistive corrosion mapping coefficient is related to the resistance of the corrosion products of the shell to be predicted, and the second resistive corrosion mapping coefficient is related to the resistance of the pitting corrosion site of the shell to be predicted. Step S420: Based on the first resistance corrosion mapping coefficient and the substrate corrosion depth, the first resistance correction term is obtained; Step S430: Based on the second resistance corrosion mapping coefficient and the pitting area ratio, the second resistance correction term is obtained; Step S440: Correct the nominal resistance based on the first resistance correction term and the second resistance correction term to obtain the predicted resistance.

[0050] It should be noted that, since the corrosion degree data calculated based on the corrosion kinetic model cannot be obtained in real time under normal resting conditions, the predicted resistance is obtained from the substrate corrosion data in steps S410 to S440, serving as the error benchmark for subsequent least squares recursive prediction. Using the first and second resistive corrosion mapping coefficients, the two internal state variables—substrate corrosion depth and pitting area ratio—are converted into correction contributions to the shell resistance, and then the predicted resistance is obtained by correcting the nominal resistance. Since the first resistive corrosion mapping coefficient relates to the resistivity characteristics of the corrosion products, and the second resistive corrosion mapping coefficient relates to the local resistivity characteristics of the pitting location, the above mapping simultaneously characterizes both the overall resistance increase caused by uniform corrosion and the local cross-sectional shrinkage caused by pitting, making the predicted resistance a valuable data point for observing the corrosion degree. It is understood that the method of obtaining the predicted resistance is also considered part of the corrosion kinetic model in this embodiment.

[0051] Specifically, the predicted resistance is determined based on the substrate corrosion depth and the proportion of pitting area as shown in the following formula: ······Mode , Mode middle, The initial thickness of the shell to be predicted; The first resistance corrosion mapping coefficient; This is the second resistance corrosion mapping coefficient; The depth of corrosion of the substrate; Percentage of pitting area; The initial value characterizing the resistance of the casing to be predicted, i.e., the reference resistance at t=0, can be obtained through a four-wire test circuit installed on the casing, via a BMS system. Among these, Acting on uniform corrosion depth Furthermore, because the resistivity of the matrix corrosion products (iron oxide / iron hydroxide in this embodiment of the invention) is much higher than that of metallic iron, the overall resistance increases. Usually >1; Calibration was performed by four-wire resistance measurement of a uniformly corroded sample. Acting on pitting area Furthermore, the geometric shrinkage effect at the pit and the contact resistance of the corrosion products within the pit are typically much greater than... ; Calibration was performed by resistance measurement of a sample containing artificial pitting defects.

[0052] Furthermore, step S500 in the above-described embodiments also includes, but is not limited to, the following steps: Step S510: Perform physical constraint projection on the updated corrosion acceleration correction parameters so that the updated corrosion acceleration correction parameters are within the corresponding physical reasonable range. Step S520: Determine whether the updated corrosion acceleration correction parameters meet the convergence conditions; wherein, the convergence conditions include: the error between the predicted resistance and the measured resistance is less than a preset error threshold within a first number of consecutive cycles, and the fluctuation amplitude of the corrosion acceleration correction parameters is less than a preset fluctuation threshold within a second number of consecutive cycles. Step S530: When the convergence condition is met, the confidence index is determined based on the error statistics; the confidence index characterizes the reliability of the prediction result of the predicted failure time.

[0053] It should be noted that steps S510 to S530 address the issues of parameters obtained from online identification and updates deviating from physically reasonable ranges due to data noise or algorithm iterations, and the lack of reliability metrics for prediction results. After each identification and update, this step applies a physical constraint projection to the updated corrosion acceleration correction parameters, forcibly constraining the parameters within their corresponding physically reasonable ranges to prevent recursive least squares from generating parameter values ​​contradicting the corrosion mechanism during data fluctuations. Then, based on the magnitude of the error between the predicted resistance and the measured resistance over multiple consecutive periods, as well as the fluctuation magnitude of the corrosion acceleration correction parameters themselves, it is determined whether the parameters have entered a convergence state. When the convergence condition is met, a confidence index is determined based on the error statistics. This confidence index provides a reliability metric for the final output predicted failure time, adding credibility information to the prediction results beyond the time coordinate, providing a reference for subsequent maintenance decisions.

[0054] Specifically, embodiments of physically constrained projection include: (The microenvironment in which the battery casing is placed only accelerates, not inhibits, corrosion.) sky, sky; ; , Furthermore, when continued After one period, the parameter convergence is determined. The specific confidence indicators are as follows: ······ , Mode middle, As a standard for prediction error, This represents the average error.

[0055] Reference Figure 2 According to an embodiment of a second aspect of the present invention, the present invention also provides a dynamic prediction system for shell corrosion, comprising: The corrosion correction module is used to determine the time-varying corrosion acceleration correction term based on the environmental parameters, contact surface state parameters, and initial values ​​of the corrosion acceleration correction parameter of the microenvironment in which the shell to be predicted is located. The corrosion correction module is also used to correct the reference corrosion rate of the shell to be predicted based on the corrosion acceleration correction term to obtain the corrosion acceleration coefficient. The corrosion state acquisition module is used to substitute the corrosion acceleration coefficient into a preset corrosion kinetic model to obtain the corrosion degree data of the shell to be predicted. The corrosion kinetic model can solve the corrosion degree data based on the corrosion acceleration coefficient and the pre-calibrated environmental correction constant. The corrosion state acquisition module is also used to obtain the predicted resistance of the shell to be predicted based on the corrosion degree data and the initial thickness of the shell to be predicted. The recursive update module is used to obtain the measured resistance of the shell to be predicted. Based on the error between the measured resistance and the predicted resistance, the corrosion acceleration correction parameters are identified and updated online by recursive least squares method to adjust the corrosion acceleration correction term and corrosion acceleration coefficient. The corrosion prediction module is used to predict the future value of corrosion degree data by recursively using a corrosion kinetic model based on the adjusted corrosion acceleration coefficient, and to determine the predicted failure time of the shell to be predicted based on the future value and the preset failure criteria.

[0056] It should be noted that, Figure 2 The diagram below is a schematic block diagram of the system. The data flow between the modules within the system provided in this embodiment of the invention is consistent with the processing timing of the method. The connection relationship can be summarized as a closed-loop link of "environmental input - coefficient construction - state solution - resistance mapping - error feedback - parameter update - failure prediction". Specifically, the corrosion correction module is located at the front end of the link. It continuously receives environmental parameters, contact surface state parameters, and initial settings of corrosion acceleration correction parameters from the bottom microenvironment sensor. It outputs time-varying corrosion acceleration correction terms and, based on this, completes the conversion from the baseline corrosion rate to the corrosion acceleration coefficient. The corrosion state acquisition module receives the corrosion acceleration coefficient, calls the corrosion kinetic model to solve for the corrosion degree data inside the shell, and couples these indirect physical quantities with the shell thickness into a predicted resistance, completing the mapping output from the internal state to the measurable electrical quantity. The recursive update module uses the measured resistance as an external reference and the error between the predicted resistance and the measured resistance as a feedback signal. It performs online identification and update of the corrosion acceleration correction parameters through the recursive least squares method. The update result is sent back to the corrosion correction module, so that the construction of the next round of correction terms and acceleration coefficients is based on parameter states that approximate the actual corrosion behavior. The corrosion prediction module, based on the updated corrosion acceleration coefficients output by the preceding module, recursively predicts the corrosion degree data along the time axis and determines and outputs the predicted failure time using a preset failure criterion as a boundary condition. Through the modular architecture described above, the system can operate continuously without manual intervention after deployment: environmental data arrives daily, the corrosion acceleration coefficient is updated accordingly, corrosion degree data and predicted resistance are updated synchronously with the acquisition of environmental parameters, error feedback drives parameter approximation round by round, and extrapolation prediction gradually approaches the actual failure time as parameters converge. Thus, the system can continuously track the shell corrosion process over a period of several years, and continuously output the prediction results as time-coordinated data that can directly support maintenance decisions.

[0057] The following are overall embodiments of the present invention: After the battery is installed, first initialize the parameters: Calibration based on 85°C / 85%RH accelerated corrosion test. , , , Calibration based on salt spray test (ASTM B117) , Based on contact corrosion tests, calibration was performed. ;initialization , , , , The specific calibration experiment is shown below: Experiment 1: Accelerated Corrosion Test at 85°C / 85%RH Principle: High temperature and high humidity are used to accelerate the corrosion process, and data equivalent to several years of corrosion at normal temperature and humidity can be obtained in a short period of time (90 days).

[0058] Experiment 2: ASTM B117 Salt Spray Test Principle: Simulates the high salt spray environment of the ocean / coastal area to induce pitting corrosion.

[0059] Experiment 3: Contact Corrosion Reference Coefficient Calibration Prepare 10 sets of standard shell-tray paired samples, controlling the gap width. =0.1mm; 500h salt spray test according to ASTM B117; insulation resistance of contact surface measured every 24h. The actual corrosion acceleration coefficient was determined by the weight loss method, and the result was obtained by fitting. .

[0060] Experiment 4: Crevice Corrosion Rate Constant Calibration Four different gap widths were set (0.05, 0.10, 0.20, 0.50 mm); a 1000-hour salt spray test was conducted to track the results. Attenuation curve; via ln(1- vs. linear regression, slope inverse calculation .

[0061] After parameter initialization is complete, online monitoring and verification are performed: A simulated BMS platform was built, and a microcurrent of 1~10μA was injected. The difference was compared with that of a standard insulation resistance meter (500VDC). The deviation was verified to be <10%, and the optimal injection current was determined to be 5μA with a pulse width of 100ms.

[0062] After the above experiments, calibrations, and verifications are completed, the specific implementation examples applied to actual batteries are as follows: Battery specifications: 21700 cylindrical cells, nickel-plated steel casing, casing thickness 0.25mm.

[0063] Step 1: Baseline parameter calibration (85°C / 85%RH accelerated test) Purely stationary corrosion test: RH=85%, T=85°C, for 90 days; Sample size: ≥5 battery cells; The fitting yielded: , ; The fitting yielded: , ; Step 2: Contact Corrosion Benchmark Test Housing-aluminum alloy tray contact, 3.5% NaCl salt spray, 35°C, for 500 hours. The fitting yielded: , Step 3: Coupled testing under idle conditions and initial parameter values Design of variable environmental strength experiments: Phase A (1-60 days): RH=80%, T=45°C, liquid accumulation at the bottom. Simulated scenario: High-temperature and high-humidity energy storage; Phase B (61-120 days): RH=40%, T=25°C, dry. Simulated scenario: Transfer to an air-conditioned warehouse; Phase C (121-180 days): RH=70%, T=35°C, condensation on the contact surface. Simulated scenario: Rainy season dampness; The initial empirical values ​​for the corrosion acceleration correction parameters are: , , , , .

[0064] Step 4: Online Identification Results After 12 months of operation: (a) such as Figure 4 As shown, the microenvironment intensity sensitivity coefficient (First weight): Converged to 0.58; (b) such as Figure 5 As shown, the cumulative effect coefficient of historical data is... (Second weight): Converges to 0.22; (c) such as Figure 6 As shown, the galvanic corrosion coefficient of the contact surface (Third weight): Converged to 0.31; (d) such as Figure 7 As shown, the coating degradation memory time constant (Short-term memory time constant, Figure 7 a): Converged to 46; Matrix corrosion memory time constant (Long-term memory time constant, Figure 7 (b): converges to 318.

[0065] The above embodiments, experimental design and results, three stages Changes such as Figure 8 As shown, the measured resistance is compared with the predicted resistance. Figure 9 As shown.

[0066] Step 5: Prediction accuracy verification Compared with the traditional linear corrosion model (error 23%), the method of this invention reduces the prediction error of 10-year storage corrosion life to 6.7%.

[0067] like Figure 10 As shown, Figure 10 This is a structural diagram of an electronic device provided in one embodiment of the present invention. The present invention also provides an electronic device, comprising: The processor 801 can be implemented using a general-purpose central processing unit (CPU), microprocessor, application specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 802 can be implemented as a read-only memory (ROM), static storage device, dynamic storage device, or random access memory (RAM). The memory 802 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 802 and called by the processor 801 to execute the dynamic prediction method for shell corrosion of the embodiments of this application. The 803 input / output interface is used to implement information input and output. The communication interface 804 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 805 transmits information between various components of the device (e.g., processor 801, memory 802, input / output interface 803, and communication interface 804); The processor 801, memory 802, input / output interface 803, and communication interface 804 are connected to each other within the device via bus 805.

[0068] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0069] It will be understood by those skilled in the art that all or some of the steps and systems in the methods disclosed above can be implemented as software, firmware, hardware, and suitable combinations thereof. Some or all of the physical components can be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which can include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, as is known to those skilled in the art, communication media typically include computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0070] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those skilled in the art, various changes can be made without departing from the spirit of the present invention.

Claims

1. A dynamic prediction method for shell corrosion, characterized in that, include: Based on the initial values ​​of the environmental parameters, contact surface state parameters, and corrosion acceleration correction parameters of the microenvironment in which the shell to be predicted is located, the time-varying corrosion acceleration correction term is determined. The corrosion acceleration correction term is used to correct the baseline corrosion rate of the shell to be predicted, and the corrosion acceleration coefficient is obtained. The corrosion acceleration coefficient is substituted into a preset corrosion kinetic model to obtain the corrosion degree data of the shell to be predicted; wherein, the corrosion kinetic model can solve the corrosion degree data based on the corrosion acceleration coefficient and a pre-calibrated environmental correction constant; Based on the corrosion degree data and the initial thickness of the shell to be predicted, the predicted resistance of the shell to be predicted is obtained; The measured resistance of the shell to be predicted is obtained. Based on the error between the measured resistance and the predicted resistance, the corrosion acceleration correction parameter is identified and updated online by recursive least squares method to adjust the corrosion acceleration correction term and the corrosion acceleration coefficient. Based on the adjusted corrosion acceleration coefficient, the future value of the corrosion degree data is recursively predicted through the corrosion kinetic model, and the predicted failure time of the shell to be predicted is determined according to the future value and the preset failure criterion.

2. The dynamic prediction method for shell corrosion according to claim 1, characterized in that, The corrosion acceleration correction term includes an environmental corrosion intensity index; when the corrosion acceleration correction term includes the environmental corrosion intensity index, the step of determining the time-varying corrosion acceleration correction term based on the environmental parameters, contact surface state parameters, and initial values ​​of the corrosion acceleration correction parameter of the microenvironment in which the shell to be predicted is located includes: Set an initial value for the first weight associated with the environmental corrosion intensity index, wherein the first weight is a corrosion acceleration correction parameter corresponding to the environmental corrosion intensity index; The environmental parameters of the microenvironment in which the shell to be predicted is located and the nominal resistance of the shell to be predicted are obtained; wherein, the environmental parameters include temperature, relative humidity, bottom liquid thickness and bottom local oxygen partial pressure; For each of the environmental parameters, a microenvironmental correction term is determined relative to its respective reference value; wherein the microenvironmental correction term is determined based on exponential weighting or exponential relationships based on the Arrhenius formula. Based on all the microenvironment correction terms, the microenvironment correction function is obtained; The instantaneous resistance change rate of the shell to be predicted is normalized based on the nominal resistance, and the normalization result is corrected by the microenvironment correction function to obtain the environmental corrosion intensity index.

3. The dynamic prediction method for shell corrosion according to claim 2, characterized in that, The corrosion acceleration correction term includes a historical cumulative effect index; when the corrosion acceleration correction term includes the historical cumulative effect index, the step of determining the time-varying corrosion acceleration correction term based on the environmental parameters, contact surface state parameters, and initial values ​​of the corrosion acceleration correction parameter of the microenvironment in which the shell to be predicted is located further includes: The initial values ​​of the second weight, short-term memory time constant, and long-term memory time constant associated with the historical cumulative effect index are set; wherein, the second weight, the short-term memory time constant, and the long-term memory time constant are the corrosion acceleration correction parameters corresponding to the historical cumulative effect index; Based on the short-term memory time constant and its corresponding first cumulative weight, and the long-term memory time constant and its corresponding second cumulative weight, a double exponential decay kernel is constructed. Obtain the environmental corrosion intensity index corresponding to each moment within a preset historical time period. Based on the dual-exponential decay kernel, perform weighted accumulation processing on the environmental corrosion intensity index within the preset historical time period to obtain the historical cumulative effect index at the current moment.

4. The dynamic prediction method for shell corrosion according to claim 2 or 3, characterized in that, The corrosion acceleration correction term includes the contact surface galvanic corrosion index; when the corrosion acceleration correction term includes the contact surface galvanic corrosion index, the step of determining the time-varying corrosion acceleration correction term based on the environmental parameters of the microenvironment in which the housing to be predicted is located, the contact surface state parameters, and the initial values ​​of the corrosion acceleration correction parameter further includes: The initial value of the third weight associated with the contact surface galvanic corrosion index is set; wherein, the third weight is the corrosion acceleration correction parameter corresponding to the contact surface galvanic corrosion index; The contact surface state parameters of the housing to be predicted are obtained, including the initial contact surface insulation resistance after the housing to be predicted is first installed and the online contact surface insulation resistance of the housing to be predicted is obtained in real time. The galvanic corrosion index of the contact surface is obtained based on the initial contact surface insulation resistance, the online contact surface insulation resistance, the contact corrosion reference coefficient, and the crevice corrosion development function; wherein, the crevice corrosion development function characterizes the change law of the crevice corrosion damage degree of the shell to be predicted over time.

5. The dynamic prediction method for shell corrosion according to claim 4, characterized in that, The corrosion degree data includes coating porosity, substrate corrosion depth, and pitting area ratio; the step of substituting the corrosion acceleration coefficient into a preset corrosion kinetic model to obtain the corrosion degree data of the shell to be predicted includes: The corrosion rate constant of the coating, the corrosion rate constant of the substrate, and the pitting corrosion propagation rate constant are obtained; wherein, the corrosion rate constant of the coating, the corrosion rate constant of the substrate, and the pitting corrosion propagation rate constant are all obtained based on accelerated corrosion experiments; Based on the corrosion acceleration coefficient, the coating corrosion rate constant, and the preset first reaction order, the rate of change of coating porosity is obtained, and the rate of change of coating porosity is numerically integrated to obtain the coating porosity. The rate of change of the substrate corrosion depth is obtained based on the corrosion acceleration coefficient, the coating porosity, the substrate corrosion rate constant, and the preset second reaction order. The rate of change of the substrate corrosion depth is then numerically integrated to obtain the substrate corrosion depth. Based on the corrosion acceleration coefficient, the pitting corrosion propagation rate constant, and the preset third reaction order, the propagation rate of the pitting corrosion area ratio is obtained, and the propagation rate of the pitting corrosion area ratio is numerically integrated to obtain the pitting corrosion area ratio.

6. The dynamic prediction method for shell corrosion according to claim 5, characterized in that, The step of obtaining the predicted resistance of the shell to be predicted based on the corrosion degree data and the initial thickness of the shell to be predicted includes: A first resistive corrosion mapping coefficient and a second resistive corrosion mapping coefficient are determined for the shell to be predicted; wherein the first resistive corrosion mapping coefficient is related to the resistance of the corrosion products of the shell to be predicted, and the second resistive corrosion mapping coefficient is related to the resistance of the pitting corrosion site of the shell to be predicted. Based on the first resistance corrosion mapping coefficient and the substrate corrosion depth, a first resistance correction term is obtained; Based on the second resistance corrosion mapping coefficient and the pitting area ratio, a second resistance correction term is obtained; The nominal resistance is corrected based on the first resistance correction term and the second resistance correction term to obtain the predicted resistance.

7. The dynamic prediction method for shell corrosion according to claim 6, characterized in that, After the steps of obtaining the measured resistance of the shell to be predicted, and based on the error between the measured resistance and the predicted resistance, identifying and updating the corrosion acceleration correction parameter online using the recursive least squares method to adjust the corrosion acceleration correction term and the corrosion acceleration coefficient, the method further includes: Physical constraint projection is performed on the updated corrosion acceleration correction parameters to ensure that the updated corrosion acceleration correction parameters are within the corresponding physically reasonable range. Determine whether the updated corrosion acceleration correction parameter meets the convergence condition; wherein, the convergence condition includes: the error between the predicted resistance and the measured resistance is less than a preset error threshold within a first number of consecutive cycles, and the fluctuation amplitude of the corrosion acceleration correction parameter is less than a preset fluctuation threshold within a second number of consecutive cycles. When the convergence condition is met, a confidence index is determined based on the statistics of the error; the confidence index characterizes the reliability of the prediction result of the predicted failure time.

8. A dynamic prediction system for shell corrosion, characterized in that, include: The corrosion correction module is used to determine a time-varying corrosion acceleration correction term based on the environmental parameters, contact surface state parameters, and initial values ​​of corrosion acceleration correction parameters of the microenvironment in which the shell to be predicted is located; the corrosion correction module is also used to correct the reference corrosion rate of the shell to be predicted based on the corrosion acceleration correction term to obtain a corrosion acceleration coefficient. The corrosion state acquisition module is used to substitute the corrosion acceleration coefficient into a preset corrosion kinetic model to obtain corrosion degree data of the shell to be predicted; wherein, the corrosion kinetic model can solve the corrosion degree data based on the corrosion acceleration coefficient and a pre-calibrated environmental correction constant; the corrosion state acquisition module is also used to obtain the predicted resistance of the shell to be predicted based on the corrosion degree data and the initial thickness of the shell to be predicted. The recursive update module is used to obtain the measured resistance of the shell to be predicted, and based on the error between the measured resistance and the predicted resistance, to identify and update the corrosion acceleration correction parameters online by recursive least squares method, so as to adjust the corrosion acceleration correction term and the corrosion acceleration coefficient. The corrosion prediction module is used to predict the future value of the corrosion degree data based on the adjusted corrosion acceleration coefficient through the corrosion kinetic model, and to determine the predicted failure time of the shell to be predicted based on the future value and the preset failure criterion.

9. An electronic device, characterized in that, It includes a memory and a processor, the memory being used to store at least one program, and the processor being used to load the at least one program to execute the dynamic prediction method for shell corrosion as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions for causing a computer to perform the dynamic prediction method for shell corrosion as described in any one of claims 1 to 7.

Images