A method and system for state of charge prediction of an electrochemical energy storage system

Abstract

The application provides a state of charge prediction method and system of an electrochemical energy storage system, the method obtains a physical port voltage and a charge / discharge current of a battery pack, and a current temperature, a current humidity and an insulation defect associated characteristic gas content, performs parasitic parameter stripping calculation on the physical port voltage based on pre-calibrated external circuit parasitic parameters, inputs an internal terminal voltage, the current temperature and the charge / discharge current into a battery equivalent circuit model to obtain a current impedance parameter of the battery pack, corrects the current impedance parameter according to the current humidity, the current temperature and the insulation defect associated characteristic gas content, updates parameters in the battery equivalent circuit model through the current corrected impedance parameter, and inputs the internal terminal voltage and the charge / discharge current into the updated battery equivalent circuit model to obtain a state of charge of the battery pack; and the application can improve the accuracy and robustness of state of charge prediction of the electrochemical energy storage system.

Description

Technical Field

[0001] This application relates to the field of power energy storage management technology, and in particular to a method and system for predicting the state of charge of an electrochemical energy storage system. Background Technology

[0002] Electrochemical energy storage systems are crucial facilities in power systems for enabling new energy grid integration, peak shaving and frequency regulation, and ensuring power supply reliability. Their core energy storage units are typically installed inside sealed gas-filled cabinets, relying on sealed gas chambers to isolate external moisture and impurities, ensuring stable operation. During operation, it is necessary to predict the state of charge (SOC) of the electrochemical energy storage system to determine its remaining capacity, providing a reliable basis for scheduling, operation, maintenance, and management.

[0003] Currently, methods for predicting the state of charge (SOC) of electrochemical energy storage systems include: ampere-hour integration, open-circuit voltage methods, model-based methods such as Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF), and machine learning algorithms. The ampere-hour integration method, combined with an equivalent circuit model, calibrates the initial SOC using open-circuit voltage and then uses the cumulative integration of charging and discharging currents for real-time estimation. However, this method relies on accurate identification of battery equivalent circuit parameters and is susceptible to current measurement errors, changes in charge / discharge rates, and parameter drift due to battery aging. Prediction accuracy gradually decreases over time. Model-based methods, such as Extended Kalman Filter and Unscented Kalman Filter, rely on accurate battery electrochemical model parameters. These parameters shift with battery aging and temperature changes, requiring periodic recalibration. Their high algorithm complexity makes them unsuitable for the real-time prediction requirements of large-capacity, multi-module energy storage systems. Existing machine learning prediction methods mostly rely on large amounts of labeled historical operating data. When the operating conditions of the energy storage system change or the batteries age and degrade, the model's generalization ability is insufficient, and the prediction accuracy will decrease significantly. At the same time, as the core closed protective structure of the electrochemical energy storage system, the gas chamber seal failure and gas leakage will allow external moisture to enter, accelerating the aging of the energy storage unit's insulation and electrode corrosion, further aggravating the prediction error of the state of charge.

[0004] Therefore, it is necessary to improve the state of charge prediction methods for electrochemical energy storage systems in related technologies. Summary of the Invention

[0005] In view of the shortcomings of the prior art described above, this application provides a method and system for predicting the state of charge of an electrochemical energy storage system to solve the above-mentioned technical problems.

[0006] According to one aspect of the embodiments of this application, a method for predicting the state of charge (SOC) of an electrochemical energy storage system is provided. The method includes: acquiring the physical port voltage and charging / discharging current of a battery pack in the electrochemical energy storage system, as well as the current temperature, current humidity, and content of characteristic gases associated with insulation defects in the environment where the battery pack is located; performing parasitic parameter stripping calculation on the physical port voltage based on pre-calibrated external circuit parasitic parameters to obtain the internal terminal voltage of the battery pack; inputting the internal terminal voltage, the current temperature, and the charging / discharging current into a battery equivalent circuit model to obtain the current impedance parameters of the battery pack; correcting the current impedance parameters according to the current humidity, the current temperature, and the content of characteristic gases associated with insulation defects to obtain current corrected impedance parameters; updating the parameters in the battery equivalent circuit model using the current corrected impedance parameters, and inputting the internal terminal voltage and the charging / discharging current into the updated battery equivalent circuit model to obtain the SOC of the battery pack, and using the SOC as the SOC prediction result.

[0007] According to another aspect of the embodiments of this application, a state of charge prediction system for an electrochemical energy storage system is also provided, comprising: a data acquisition module for acquiring the physical port voltage and charging / discharging current of a battery pack in the electrochemical energy storage system, as well as the current temperature, current humidity, and content of characteristic gases associated with insulation defects in the environment where the battery pack is located; a voltage calculation module for performing parasitic parameter stripping calculation on the physical port voltage based on pre-calibrated external circuit parasitic parameters to obtain the internal terminal voltage of the battery pack; a parameter calculation module for inputting the internal terminal voltage, the current temperature, and the charging / discharging current into a battery equivalent circuit model to obtain the current impedance parameter of the battery pack; correcting the current impedance parameter according to the current humidity, the current temperature, and the content of characteristic gases associated with insulation defects to obtain a current corrected impedance parameter; and a state of charge prediction module for updating the parameters in the battery equivalent circuit model using the current corrected impedance parameter, inputting the internal terminal voltage and the charging / discharging current into the updated battery equivalent circuit model to obtain the state of charge of the battery pack, and using the state of charge as the state of charge prediction result.

[0008] The beneficial effects of this application are as follows: This application obtains the physical port voltage and charging / discharging current of the battery pack in the electrochemical energy storage system, as well as the current temperature, current humidity, and content of characteristic gases associated with insulation defects in the environment where the battery pack is located. Based on pre-calibrated external circuit parasitic parameters, the parasitic parameter stripping calculation is performed on the physical port voltage to obtain the internal terminal voltage of the battery pack. The internal terminal voltage, current temperature, and charging / discharging current are input into the battery equivalent circuit model to obtain the current impedance parameters of the battery pack. The current impedance parameters are corrected according to the current humidity, current temperature, and content of characteristic gases associated with insulation defects to obtain the current corrected impedance parameters. The parameters in the battery equivalent circuit model are updated using the current corrected impedance parameters, and the internal terminal voltage and charging / discharging current are input into the updated battery equivalent circuit model to obtain the state of charge (SOC) of the battery pack. The SOC is then used as the SOC prediction result. The above process can significantly improve the accuracy and robustness of SOC prediction for electrochemical energy storage systems, and effectively solves the problems of decreased prediction accuracy over time caused by parasitic parameter influence, model parameter drift, and insufficient generalization ability in related technologies.

[0009] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0010] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort. In the drawings:

[0011] Figure 1 This is a flowchart illustrating a method for predicting the state of charge of an electrochemical energy storage system, as shown in an exemplary embodiment of this application.

[0012] Figure 2 This is a flowchart illustrating the process of determining the internal terminal voltage of a battery pack, as shown in another exemplary embodiment of this application.

[0013] Figure 3 This is a flowchart illustrating the process of correcting the current impedance parameter, as shown in another exemplary embodiment of this application.

[0014] Figure 4 This is an exemplary embodiment of the present application illustrating the operating interface of the state of charge prediction system for an electrochemical energy storage system;

[0015] Figure 5 This is a block diagram illustrating a state-of-charge prediction system for an electrochemical energy storage system, as shown in an exemplary embodiment of this application. Detailed Implementation

[0016] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0017] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of this application. The drawings only show the components related to this application and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0018] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the present application. However, it will be apparent to those skilled in the art that embodiments of the present application may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the present application.

[0019] The implementation details of the technical solutions in the embodiments of this application are described in detail below:

[0020] Figure 1 This is a flowchart illustrating a method for predicting the state of charge (SOC) of an electrochemical energy storage system, as shown in an exemplary embodiment of this application. (Refer to...) Figure 1 As shown, the method for predicting the state of charge of this electrochemical energy storage system includes at least steps S110 to S140, which are described in detail below:

[0021] In step S110, the physical port voltage and charging / discharging current of the battery pack in the electrochemical energy storage system, as well as the current temperature, current humidity, and content of characteristic gases associated with insulation defects in the environment where the battery pack is located, are acquired. In one embodiment of this application, the physical port voltage is directly measured by a voltage acquisition unit, the charging / discharging current is acquired by a current transformer connected in series in the main circuit, the environment where the parallel battery pack is located is a closed gas chamber of a gas-filling cabinet, the current temperature is acquired by a temperature sensor arranged inside the closed gas chamber, the current humidity is acquired by a humidity sensor arranged inside the closed gas chamber, and the content of characteristic gases associated with insulation defects is detected by a gas analyzer or gas sensor array installed in the closed gas chamber. The gas sensor array includes at least a hydrogen sensor, an ethylene sensor, and a dimethyl carbonate sensor.

[0022] In step S120, based on pre-calibrated external circuit parasitic parameters, parasitic parameter stripping calculations are performed on the physical port voltage to obtain the internal terminal voltage of the battery pack. In one embodiment of this application, the pre-calibrated external circuit parasitic parameters include the parasitic resistance and parasitic inductance of the external acquisition lines. These parasitic parameters can be directly measured and acquired by an impedance analyzer before the energy storage system is put into operation, or they can be automatically calculated through an initialization calibration process deployed within the energy storage system. The parasitic parameter stripping calculation process is as follows: calculate the voltage drop generated by the parasitic resistance, calculate the inductance voltage drop generated by the parasitic inductance, and subtract the voltage drop and inductance voltage drop from the physical port voltage (assuming the charging current is positive and the discharging current is negative) to obtain the internal terminal voltage after stripping the influence of parasitic parameters. This process can eliminate voltage deviations caused by external lines and physical connection ports, avoid the continuous accumulation of systematic errors caused by parasitic parameters, and prevent them from affecting the final prediction results, thereby ensuring the accuracy of the voltage data used in subsequent calculations.

[0023] In step S130, the internal terminal voltage, current temperature, and charging / discharging current are input into the battery equivalent circuit model to obtain the current impedance parameters of the battery pack; the current impedance parameters are corrected according to the current humidity, current temperature, and the content of characteristic gases associated with insulation defects to obtain the current corrected impedance parameters. In one embodiment of this application, the process of inputting the internal terminal voltage, current temperature, and charging / discharging current into the battery equivalent circuit model to obtain the current impedance parameters of the battery pack includes: determining the initial model parameter reference of the battery equivalent circuit model based on the current temperature, and using the initial model parameter reference as the prior initial value of the parameter identification algorithm; using the internal terminal voltage and charging / discharging current as the real-time observation data of the parameter identification algorithm; the parameter identification algorithm performing recursive loop calculations on the real-time observation data based on the prior initial value to obtain the predicted terminal voltage; with the goal of minimizing the residual between the predicted internal terminal voltage and the actual internal terminal voltage, dynamically updating the current model parameters of the battery equivalent circuit model (including: ohmic impedance, SEI film impedance, charge transfer impedance, polarization capacitance corresponding to charge transfer impedance, diffusion impedance, polarization capacitance corresponding to diffusion impedance, etc.), until the difference between the recursively calculated predicted internal terminal voltage and the actual internal terminal voltage is less than a preset difference threshold, terminating the recursive loop, and using the dynamically updated ohmic impedance, dynamically updated SEI film impedance, dynamically updated charge transfer impedance, and dynamically updated diffusion impedance as the current impedance parameters. Parameter identification algorithms include recursive least squares, extended Kalman filter, dual Kalman filter, or particle filter, which can be selected based on actual computational performance and accuracy requirements. The battery equivalent circuit model can be either the PNGV (Partnership for a New Generation of Vehicles) model or a second-order RC equivalent circuit model.

[0024] In one embodiment of this application, the process of inputting the internal terminal voltage, current temperature, and charging / discharging current into the battery equivalent circuit model to obtain the current impedance parameters of the battery pack is based on the dynamic electrical response characteristics of the battery pack. The equivalent circuit parameters are inferred by observing the mapping relationship between the charging / discharging current and the internal terminal voltage in real time. This method relies only on the dynamic excitation and response of voltage and current, without introducing the state of charge (SOC) of the battery pack. This decouples the impedance parameter identification algorithm from the SOC state estimation algorithm, effectively avoiding the "error coupling" problem of parameter identification divergence caused by SOC estimation errors in related algorithms.

[0025] In one embodiment of this application, after obtaining the current impedance parameter, the current impedance parameter is corrected according to the current humidity, current temperature and the content of characteristic gases associated with insulation defects. This can effectively offset the drift problem of battery equivalent circuit model parameters caused by temperature and humidity changes and battery pack insulation aging. It is beneficial to ensure that the battery equivalent circuit model parameters always match the current actual operating state of the battery pack, and obtain the current corrected impedance parameter that can accurately reflect the actual operating state of the battery pack.

[0026] In step S140, the parameters in the battery equivalent circuit model are updated using the current corrected impedance parameters, and the internal terminal voltage and charging / discharging current are input into the updated battery equivalent circuit model to obtain the state of charge of the battery pack, and the state of charge is used as the state of charge prediction result. In one embodiment of this application, the process of inputting the internal terminal voltage and charging / discharging current into the updated battery equivalent circuit model to obtain the state of charge (SOC) of the battery pack includes: calculating a preliminary predicted SOC value of the battery pack using the ampere-hour integral method based on the charging / discharging current; inputting the charging / discharging current and the preliminary predicted SOC value into the updated battery equivalent circuit model, which calculates an estimated internal terminal voltage using the current corrected impedance parameter, the charging / discharging current, and the preliminary predicted SOC value, and calculates the voltage residual between the estimated internal terminal voltage and the internal terminal voltage; calculating the Kalman gain based on the updated battery equivalent circuit model and the voltage residual using a Kalman filtering algorithm; wherein the Kalman gain is used to characterize the compensation weight for the voltage residual; multiplying the Kalman gain by the voltage residual to obtain the SOC correction compensation amount; and superimposing the SOC correction compensation amount with the preliminary predicted SOC value to obtain the current SOC of the battery pack.

[0027] In one embodiment of this application, the process of calculating the estimated internal terminal voltage using the current corrected impedance parameters, charging and discharging current, and preliminary SOC prediction value in the battery equivalent circuit model includes: obtaining the theoretical open-circuit voltage by looking up a table based on the preliminary SOC prediction value and a preset open-circuit voltage-state of charge (OCV-SOC) mapping relationship; multiplying the charging and discharging current by the current ohmic impedance in the current corrected impedance parameters to obtain the ohmic voltage drop; inputting the charging and discharging current into the RC parallel network of the battery equivalent circuit model, wherein the discretized mathematical expression of the physical characteristics of the RC parallel network is characterized by a preset dynamic difference equation (e.g., polarization voltage state transition equation), using the charging and discharging current as the input excitation variable of the preset dynamic difference equation, and substituting the charge transfer impedance, the polarization capacitance corresponding to the charge transfer impedance, the diffusion impedance, and the polarization capacitance corresponding to the diffusion impedance in the current corrected impedance parameters as coefficient parameters of the preset dynamic difference equation, solving the preset dynamic difference equation to obtain the polarization voltage drop; and subtracting the ohmic voltage drop and the polarization voltage drop from the theoretical open-circuit voltage based on a preset positive and negative current orientation to obtain the estimated internal terminal voltage.

[0028] In one embodiment of this application, a parasitic parameter stripping mechanism is introduced. By using pre-calibrated external circuit parasitic parameters, parasitic parameter stripping calculations are performed on the compensated voltage to restore the internal terminal voltage of the battery pack. This solves the problem of physical port voltage measurement distortion and avoids the systematic errors introduced by directly using the physical port voltage interfered with by parasitic parameters for state of charge estimation in related technologies. The nonlinear effects of environmental factors (e.g., current temperature, current humidity, and the content of characteristic gases associated with insulation defects) on the electrochemical characteristics of the battery pack are fully considered, and the impedance parameters in the battery equivalent circuit model are dynamically corrected. This allows the battery equivalent circuit model to adapt to the current operating conditions and aging state in real time, overcoming the defect of model parameters drifting with aging and environmental changes. The mechanism-driven strategy of "updating model parameters first and then predicting state" is adopted to update the impedance parameters after environmental feature correction to the battery equivalent circuit model in real time, ensuring that the battery equivalent circuit model is always in the "best matching" state. Compared to purely data-driven machine learning algorithms, the method in this application does not rely on massive amounts of labeled data, avoiding the problem of poor generalization ability of the battery equivalent circuit model after changes in operating conditions or battery pack aging. Compared to the approach of predicting the state of charge solely through the ampere-hour integration method combined with the equivalent circuit model, this application eliminates the impact of current measurement error accumulation and parameter drift by accurately acquiring the internal terminal voltage and correcting the parameters in real time. Compared to model methods represented by Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF) in related technologies, this application overcomes the shortcomings of model parameter drift, fixed model parameters, or the need for frequent manual calibration by accurately acquiring the internal terminal voltage and correcting the parameters in real time. Even in large-capacity multi-module energy storage systems, facing complex scenarios such as battery pack aging, drastic changes in charge and discharge rates, or switching of operating conditions, this application can still maintain high prediction accuracy, significantly improving the operational safety and maintenance management efficiency of electrochemical energy storage systems throughout their entire life cycle.

[0029] Furthermore, compared to model methods represented by Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF) in related technologies, this application first obtains the current impedance parameter, which is independent of the SOC estimation process, through parameter identification, and then obtains the current corrected impedance parameter through environmental factor correction. The current corrected impedance parameter is then updated into the battery equivalent circuit model, and the SOC estimation is completed based on the updated battery equivalent circuit model. This avoids the mutual coupling between SOC estimation error and parameter identification error. Even after a long and multi-cycle estimation process, it effectively alleviates the problem of continuous accumulation and amplification of SOC calculation error, maintains the prediction accuracy of battery pack state of charge, and effectively reduces the situation of algorithm divergence and inaccuracy.

[0030] In one embodiment of this application, if there are multiple battery packs, after obtaining the state of charge of each battery pack, the method for predicting the state of charge of the electrochemical energy storage system further includes:

[0031] The historical charge / discharge conditions, charge / discharge modes, and pack temperature of each battery pack are acquired. In one embodiment of this application, the historical corrected impedance parameters are obtained in the same way as the current corrected impedance parameters. The charge / discharge modes include charging mode and discharging mode. Historical charge / discharge conditions include: deep charge / discharge conditions (i.e., the charge / discharge state when the depth of discharge of the battery pack in a single cycle is greater than a preset depth threshold), high-rate charge / discharge conditions (i.e., the charge / discharge state when the real-time charge / discharge rate of the battery pack is greater than a preset rate threshold), and high-temperature charge / discharge conditions (i.e., the charge / discharge state when the internal cell temperature of the battery pack is greater than a preset temperature threshold). The pack temperature is acquired through a temperature sensor.

[0032] Based on the state of charge (SOC) and temperature of each battery pack, the basic polarization resistance and basic polarization capacitance of each battery pack under the current charge / discharge conditions are determined. In one embodiment of this application, given the SOC and temperature of each battery pack, the basic polarization resistance and basic polarization capacitance corresponding to the SOC and temperature of each battery pack are determined from a polarization parameter lookup table. The polarization parameter lookup table is pre-calibrated through offline experiments and is used to characterize the multidimensional mapping relationship between polarization resistance, polarization capacitance, SOC, and temperature.

[0033] The dynamic polarization voltage of each battery pack is obtained by inputting the charging / discharging current, the duration of the charging / discharging current, and the basic polarization resistance and basic polarization capacitance of each battery pack under the current charging / discharging conditions into the updated battery equivalent circuit model. The ratio of the dynamic polarization voltage of each battery pack to the charging / discharging current of the corresponding battery pack is used to obtain the polarization internal resistance increment of each battery pack under the current charging / discharging conditions. In one embodiment of this application, the process of obtaining the dynamic polarization voltage of each battery pack by inputting the charging / discharging current, the duration of the charging / discharging current, and the basic polarization resistance and basic polarization capacitance of each battery pack under the current charging / discharging conditions into the updated battery equivalent circuit model is implemented through a zero-order hold discretization recursive algorithm.

[0034] The internal resistance reference value of each battery pack is obtained by subtracting the polarization internal resistance increment of the corresponding battery pack from the total impedance of each battery pack. Then, based on a standard reference temperature, the internal resistance reference value of each battery pack is normalized to obtain the current internal resistance reference value of each battery pack. In one embodiment of this application, the total impedance is characterized by a current corrected impedance parameter, which is obtained based on the complex superposition of the impedance components within the battery pack. The formula for calculating the current internal resistance reference value is as follows:

[0035] Equation (1)

[0036] in, This indicates the current internal resistance reference value. Indicates the reference value of internal resistance. The relative temperature coefficient represents resistance; it is a positive value and its unit is one-tenth of a degree Celsius. A standard reference temperature that indicates relative temperature (e.g., 25 degrees Celsius). The current temperature is expressed in degrees Celsius. Formula (1) is used to characterize that when the current temperature is less than the standard reference temperature, the value of the internal resistance reference value increases due to the temperature effect, and the increased internal resistance reference value can be converted back to the level of the standard reference temperature; when the current temperature is greater than the standard reference temperature, the value of the internal resistance reference value decreases due to the temperature effect, and the smaller internal resistance reference value can be converted back to the level of the standard reference temperature.

[0037] Based on the current internal resistance baseline value and the preset internal resistance reference value of each battery pack, the degree of internal resistance increase for each battery pack is calculated. In one embodiment of this application, the formula for calculating the degree of internal resistance increase is as follows:

[0038] Equation (2)

[0039] in, Indicates the degree of internal resistance growth. This indicates the current internal resistance reference value. This indicates the preset internal resistance reference value (the internal resistance value obtained by testing the battery pack under standard conditions at the factory).

[0040] In one embodiment of this application, by eliminating the influence of real-time polarization effect on the internal resistance reference value, the calculation accuracy of the current internal resistance reference value is improved, thereby improving the calculation accuracy of the degree of internal resistance growth.

[0041] Feature data of each battery pack under historical charge-discharge conditions is extracted. The degree of internal resistance growth and feature data of each battery pack are input into the mapping curve between the degree of internal resistance growth and capacity decay to obtain the health status value of each battery pack. In one embodiment of this application, the feature data includes: cumulative charge-discharge cycle count, duration of charge-discharge exceeding a preset rate, and cumulative charge-discharge duration exceeding a preset temperature. The cumulative charge-discharge cycle count is the number of deep charge-discharge cycles, the duration of charge-discharge exceeding the preset rate is the duration of high-rate charge-discharge conditions, and the cumulative charge-discharge duration exceeding the preset temperature is the duration of high-temperature charge-discharge conditions. The curve relating internal resistance growth to capacity decay is obtained based on historical sample data through regression fitting or machine learning. The process involves: performing cyclic aging experiments on multiple battery pack samples at different battery pack temperatures, charge / discharge rates, and charge / discharge depths; collecting capacity retention data and corresponding electrochemical impedance spectroscopy (EIS) data within a preset cycle period as historical sample data; extracting features from the historical sample data: extracting features related to internal resistance growth, cumulative charge / discharge cycle counts, duration of charge / discharge exceeding a preset rate, and cumulative charge / discharge duration exceeding a preset temperature from the EIS data, as a feature dataset; and performing regression fitting on the feature dataset or training a machine learning model using the feature dataset to obtain the curve relating internal resistance growth to capacity decay.

[0042] The predicted state of charge (SOC) is determined based on the charging / discharging mode, the health status value of each battery pack, and the SOC. In one embodiment of this application, the accuracy of the health status value of each battery pack is improved by increasing the accuracy of the calculation of the internal resistance growth. Simultaneously, in determining the predicted SOC based on the charging / discharging mode, the health status value of each battery pack, and the SOC, in charging mode, the SOC of the battery pack with the smallest charging voltage drop margin and the larger SOC among all SOCs are selected as the calculation benchmark for the entire system. In discharging mode, the SOC of the battery pack with the smallest discharging voltage drop margin and the smaller SOC among all SOCs are selected as the calculation benchmark for the entire system. Through this method, the weakest battery pack most affected by internal resistance aging or inconsistent charge levels can be effectively identified, avoiding the risk of overcharging or over-discharging in the energy storage system and improving the overall operational safety of the multi-battery-pack energy storage system.

[0043] In one embodiment of this application, for a battery pack composed of multiple battery packs, the electrochemical energy storage system state of charge prediction method provided in this application identifies the differences in the health status and capacity inconsistencies of each battery pack. It uses the state of the bottleneck battery pack with limited charge and discharge boundaries as the calculation benchmark for the overall system SOC prediction. This method can better fit the actual operating characteristics of large-capacity multi-module energy storage systems. It can avoid the prediction errors caused by directly averaging the SOC of all battery packs in traditional methods, and can also identify system capacity bottlenecks in advance, providing data support for the balanced maintenance of energy storage systems. This further improves the reliability and practicality of SOC prediction for multi-battery pack systems.

[0044] In one embodiment of this application, the process of determining the state of charge (SOC) prediction result based on the charge / discharge mode, the health status value of each battery pack, and the SOC includes:

[0045] The system acquires the port voltage fluctuation value and load current change amplitude of each battery pack when a step change occurs in the load current under the current charging and discharging conditions. In one embodiment of this application, a step change in load current is determined when the load current change amplitude is greater than a preset current change amplitude. The load current change amplitude is determined based on the load current values ​​before and after the step change, which are acquired by a current sensor. Similarly, the port voltage fluctuation value is determined based on the port voltage values ​​before and after the step change, which are acquired by a voltage sensor.

[0046] The measured internal resistance of each battery pack is obtained by comparing the port voltage fluctuation value with the load current change amplitude. The measured internal resistance of each battery pack is then corrected using its health status value, state of charge (SBC), and pack temperature to obtain the dynamic internal resistance. In one embodiment of this application, the process of correcting the measured internal resistance of each battery pack using its health status value, SBC, and pack temperature to obtain the dynamic internal resistance comprehensively considers the difference between the measured internal resistance under the current charging and discharging conditions and the theoretical internal resistance. The dynamic internal resistance is determined based on a comparison between this difference and a preset difference threshold. This process further improves the accuracy of the dynamic internal resistance calculation, provides reliable support for the calculation of voltage drop margin, and avoids inaccurate predictions of the overall system SBC due to internal resistance calculation deviations.

[0047] If the charging / discharging mode is discharge mode, the port voltage drop value of each battery pack is determined based on the dynamic internal resistance and discharge current. The difference between the physical port voltage of each battery pack and the corresponding port voltage drop value is used as the predicted port voltage of each battery pack. The difference between the predicted port voltage of each battery pack and the discharge cutoff protection voltage is used as the discharge voltage drop margin of each battery pack. The state of charge (SOC) of the battery pack with the smallest SOC is compared with the smallest SOC among all SOCs, and the smaller SOC is used as the predicted SOC result. In one embodiment of this application, the product of the dynamic internal resistance of each battery pack and the discharge current of the corresponding battery pack is the port voltage drop value of the corresponding battery pack. Using the difference between the physical port voltage of each battery pack and the corresponding port voltage drop value as the predicted port voltage of each battery pack effectively eliminates the instantaneous interference of dynamic load current on the port voltage, improves the accuracy of port voltage prediction, and thus improves the accuracy of discharge voltage drop margin calculation. By comparing the state of charge (SOC) of a battery pack with the minimum discharge voltage drop margin with the minimum SOC among all SOCs, the smaller SOC is taken as the predicted SOC result. By simultaneously taking into account the voltage boundary (i.e., discharge voltage drop margin) and capacity boundary (i.e., the SOC of the battery pack) under dynamic operating conditions and taking the smaller value of the two, it is ensured that the system evaluation result will not exceed the physical safety boundary of the battery pack under extreme operating conditions.

[0048] If the charging / discharging mode is charging mode, the port boost value of each battery pack is determined based on the dynamic internal resistance and charging current. The sum of the physical port voltage of each battery pack and the corresponding port boost value is used as the predicted port voltage of each battery pack. The difference between the charging cut-off protection voltage and the predicted port voltage of each battery pack is used as the charging voltage drop margin of each battery pack. The state of charge (SOC) of the battery pack with the smallest SOC is compared with the largest SOC among all SOCs, and the larger SOC is used as the predicted SOC result. In one embodiment of this application, the product of the dynamic internal resistance of each battery pack and the charging current of the corresponding battery pack is the port boost value of the corresponding battery pack. Using the sum of the physical port voltage of each battery pack and the corresponding port boost value as the predicted port voltage compensates for and predicts the voltage rise trend caused by the dynamic load current, improves the accuracy of port voltage prediction, and thus improves the accuracy of charging voltage drop margin calculation. By comparing the state of charge (SOC) of a battery pack with the smallest charging voltage drop margin with the largest SOC among all SOCs, the larger of the two SOCs is used as the predicted SOC result. By simultaneously considering both the voltage boundary (i.e., charging voltage drop margin) and the capacity boundary (i.e., the SOC of the battery pack) under dynamic operating conditions and taking the larger of the two values, the risk of local overcharging caused by a surge in internal resistance or high current charging can be captured, effectively avoiding battery pack overcharging and ensuring the safe operation of the energy storage system.

[0049] In one embodiment of this application, the internal resistance measurement includes static internal resistance and dynamic polarization impedance, which are used to characterize the overall voltage response characteristics of the battery pack when the load current changes suddenly. The internal resistance reference value is static internal resistance, which is used to evaluate the health status, aging degree, etc. of the battery pack.

[0050] In one embodiment of this application, the process of correcting the measured internal resistance of the corresponding battery pack by using the health status value, state of charge, and pack temperature of each battery pack to obtain the dynamic internal resistance includes:

[0051] Based on the state of health (SHS), state of charge (SBC), and pack temperature of each battery pack, the theoretical internal resistance of each battery pack under the current charge-discharge conditions is determined. In one embodiment of this application, the process of determining the theoretical internal resistance of each battery pack under the current charge-discharge conditions includes: based on the SBC and pack temperature, looking up the reference internal resistance of the battery pack from a mapping table between the reference internal resistance of the battery pack and the SBC and pack temperature; based on the SHS, looking up the internal resistance aging correction coefficient from a mapping table between the SHS and the internal resistance aging correction coefficient; and multiplying the found reference internal resistance of the battery pack by the found internal resistance aging correction coefficient as the theoretical internal resistance.

[0052] If the difference between the measured internal resistance and the theoretical internal resistance is less than or equal to a preset difference threshold, the measured internal resistance and the theoretical internal resistance are weighted and fused to obtain the dynamic internal resistance. In one embodiment of this application, the preset difference threshold is determined according to the actual situation. By obtaining the dynamic internal resistance through weighted fusion of the measured internal resistance and the theoretical internal resistance, the measured internal resistance reflects the current internal resistance of the battery pack, while the theoretical internal resistance corrects for random noise interference during the measurement process, further improving the stability and accuracy of the dynamic internal resistance calculation.

[0053] If the difference between the measured internal resistance and the theoretical internal resistance is greater than a preset difference threshold, the theoretical internal resistance is used as the dynamic internal resistance. In one embodiment of this application, when the difference between the measured internal resistance and the theoretical internal resistance is greater than the preset difference threshold, it indicates that the measured internal resistance has been subjected to a strong instantaneous interference, and the reliability of the measured internal resistance is greatly reduced. Therefore, directly using the theoretical internal resistance after calibration for health status, temperature, and SOC as the dynamic internal resistance can avoid the interference of abnormal measurement results on subsequent calculations and ensure the stability of the entire prediction process.

[0054] In one embodiment of this application, if the current impedance parameters include: current ohmic impedance, current SEI film impedance, current charge transfer impedance, and current diffusion impedance, then the process of correcting the current impedance parameters based on the current humidity, current temperature, and the content of characteristic gases associated with insulation defects to obtain the current corrected impedance parameters includes:

[0055] Based on the current temperature, temperature normalization compensation is performed on the current impedance parameters to obtain the temperature-compensated current ohmic impedance, temperature-compensated current SEI film impedance, temperature-compensated current charge transfer impedance, and temperature-compensated current diffusion impedance. In one embodiment of this application, the formula for calculating the temperature-compensated current ohmic impedance is as follows:

[0056] Equation (3)

[0057] in, This indicates the current ohmic impedance after temperature compensation. Indicates the current ohmic impedance. The temperature coefficient representing ohmic impedance is a positive value, and its unit is one-tenth of a degree Celsius. A standard reference temperature that indicates relative temperature (e.g., 25 degrees Celsius). The current temperature is expressed in degrees Celsius. Formula (3) is used to characterize that when the current temperature is lower than the standard reference temperature, the value of the current ohmic impedance increases due to the temperature effect, and the increased current ohmic impedance can be converted back to the level of the standard reference temperature; when the current temperature is higher than the standard reference temperature, the value of the current ohmic impedance decreases due to the temperature effect, and the smaller current ohmic impedance can be converted back to the level of the standard reference temperature.

[0058] The formula for calculating the current SEI film impedance after temperature compensation is shown below:

[0059] Equation (4)

[0060] in, This indicates the current SEI film impedance after temperature compensation. Indicates the current SEI film impedance. The apparent activation energy, expressed in J / mol, represents the SEI film impedance. It is determined by the battery pack material properties and the SEI film composition and is obtained through offline multi-temperature point EIS testing and calibration. This represents the universal gas constant, with units of J / (mol·K). This indicates the standard reference temperature, which is an absolute temperature (e.g., 298.15K, which is 25 degrees Celsius). The current absolute temperature is obtained by converting the current temperature, and the unit is Kelvin. Formula (4) is used to characterize that when the current temperature is less than the standard reference temperature, the value of the current SEI film impedance increases due to the temperature effect, and the increased current SEI film impedance can be converted back to the level of the standard reference temperature; when the current temperature is greater than the standard reference temperature, the value of the current SEI film impedance decreases due to the temperature effect, and the decreased current SEI film impedance can be converted back to the level of the standard reference temperature.

[0061] The formula for calculating the current charge transfer impedance after temperature compensation is shown below:

[0062] Equation (5)

[0063] in, This represents the current charge transfer impedance after temperature compensation. Indicates the current charge transfer impedance. The apparent activation energy, expressed in J / mol, represents the charge transfer process. It is determined by the electrode material properties and the electrolyte system, and is calibrated through offline multi-temperature point EIS testing. This represents the universal gas constant, with units of J / (mol·K). This indicates the standard reference temperature, which is an absolute temperature (e.g., 298.15K, which is 25 degrees Celsius). The current absolute temperature is obtained by converting the current temperature, and the unit is Kelvin. Formula (5) is used to characterize that when the current temperature is less than the standard reference temperature, the value of the current charge transfer impedance increases due to the temperature effect, and the increased current charge transfer impedance can be converted back to the level of the standard reference temperature; when the current temperature is greater than the standard reference temperature, the value of the current charge transfer impedance decreases due to the temperature effect, and the decreased current charge transfer impedance can be converted back to the level of the standard reference temperature.

[0064] The formula for calculating the current diffusion impedance after temperature compensation is shown below:

[0065] Equation (6)

[0066] in, This represents the current diffusion impedance after temperature compensation. Indicates the current diffusion impedance. The apparent activation energy, expressed in J / mol, represents the lithium-ion solid-phase diffusion process. It is determined by the microstructure of the electrode material and the electrolyte system, and is calibrated through offline multi-temperature point EIS testing. This represents the universal gas constant, with units of J / (mol·K). This indicates the standard reference temperature, which is an absolute temperature (e.g., 298.15K, which is 25 degrees Celsius). The current absolute temperature is obtained by converting the current temperature, and the unit is Kelvin. Formula (6) is used to characterize that when the current temperature is less than the standard reference temperature, the value of the current diffusion impedance increases due to the temperature effect, and the increased current diffusion impedance can be converted back to the level of the standard reference temperature; when the current temperature is greater than the standard reference temperature, the value of the current diffusion impedance decreases due to the temperature effect, and the decreased current diffusion impedance can be converted back to the level of the standard reference temperature.

[0067] Based on the current humidity, humidity compensation is applied to the temperature-compensated current ohmic impedance, temperature-compensated current charge transfer impedance, and temperature-compensated current diffusion impedance to obtain the humidity-compensated current ohmic impedance, humidity-compensated current charge transfer impedance, and humidity-compensated current diffusion impedance. In one embodiment of this application, the calculation formula for the humidity-compensated current ohmic impedance can be a linear compensation model or an exponential compensation model. Taking the linear compensation model as an example, the calculation formula for the humidity-compensated current ohmic impedance is as follows:

[0068] Equation (7)

[0069] in, This indicates the current ohmic impedance after humidity compensation. This indicates the current ohmic impedance after temperature compensation. The slope of impedance as a function of humidity is represented by a negative value, and the unit is . For specific battery models and electrolyte formulations, the results were obtained through EIS (electrochemical impedance spectroscopy) measurements at multiple humidity points in a constant temperature chamber. This indicates the standard reference humidity, for example, 50%RH or 30%RH. The current humidity is represented by Equation (7). Equation (7) is used to correct the equivalent effect of the current humidity on the current ohmic impedance calculated based on the parameter identification algorithm. During the parameter identification process, fluctuations in ambient humidity will be coupled into the measurement system as interference variables, causing the identification result to deviate from the true value. When the current humidity is less than the standard reference humidity (dry environment): the dry environment may cause microscopic shrinkage of the contact interface or a decrease in the conductivity of a specific interface film. This physical change will cause the impedance data in the high-frequency region to rise, prompting the parameter identification algorithm to calculate an excessively high ohmic impedance value during the fitting process. Equation (7) can convert this excessively high identification result due to the dry environment back to the benchmark level of the standard reference humidity through a compensation mechanism. When the current humidity is greater than the standard reference humidity (humid environment): In power system energy storage scenarios where the air chamber seal fails or there is a surface effect (such as open battery system or modules significantly affected by the environment), condensation on the surface of the battery pack or moisture absorption of the insulating material will produce a parallel leakage conduction effect. This effect will change the impedance response characteristics in the low-frequency band, causing the parameter identification algorithm to misjudge this part of the leakage current path as a decrease in internal resistance, thereby calculating a lower current ohmic impedance value. Formula (7) can also correct the algorithm calculation deviation caused by humidity through a compensation mechanism, and convert it back to the benchmark level of the standard reference humidity.

[0070] The formula for calculating the current charge transfer impedance after humidity compensation can be either a linear compensation model or an exponential compensation model. Taking the linear compensation model as an example, the formula for calculating the current charge transfer impedance after humidity compensation is as follows:

[0071] Equation (8)

[0072] in, This represents the current charge transfer impedance after humidity compensation. This represents the current charge transfer impedance after temperature compensation. The humidity sensitivity coefficient, representing charge transfer impedance, is a negative value and its unit is . The properties of the electrode material and the electrolyte system are determined by the calibration obtained through offline multi-humidity point EIS testing. This indicates the standard reference humidity, for example, 50%RH or 30%RH. The current humidity is represented by Equation (8). Equation (8) is used to characterize the equivalent correction of the current humidity fluctuation to the current charge transfer impedance calculated based on the parameter identification algorithm. During the parameter identification process of the battery equivalent circuit model, the change in current humidity will be coupled into the measurement system as an external interference factor. That is, the fluctuation of current humidity may cause distortion or shift in the original impedance spectrum data collected by changing the insulation characteristics of the battery pack surface (such as the surface leakage conduction effect under high humidity) or affecting the interface state of the semi-open system. When the parameter identification algorithm fits and solves based on these disturbed data, this non-intrinsic impedance change caused by the environment will be incorrectly mapped and allocated to the estimated value of the current charge transfer impedance by the algorithm, causing the current charge transfer impedance to deviate from the true reference level under the standard reference humidity. When the current humidity is less than the standard reference humidity (dry environment): Increased surface insulation or electrolyte evaporation and concentration may lead to an artificially high impedance value. The current charge transfer impedance is corrected by the humidity sensitivity coefficient of the charge transfer impedance and converted back to the standard level. When the current humidity is greater than the standard reference humidity (humid environment): Surface condensation leakage or moisture absorption dilution effect may lead to an underestimation of the impedance value (or exhibit specific nonlinear drift). The current charge transfer impedance is corrected by the humidity sensitivity coefficient of the charge transfer impedance and converted back to the standard reference humidity level.

[0073] The current diffusion impedance after humidity compensation can be either a linear compensation model or an exponential compensation model. Taking the exponential compensation model as an example, the formula for calculating the current diffusion impedance after humidity compensation is as follows:

[0074] Equation (9)

[0075] in, This represents the current diffusion impedance after humidity compensation. This represents the current diffusion impedance after temperature compensation. The humidity sensitivity coefficient, representing diffusion impedance, is a negative value, and its unit is . , This indicates the standard reference humidity, for example, 50%RH or 30%RH. The current humidity is indicated. Formula (9) is used to eliminate the equivalent correction of the current diffusion impedance calculated based on the parameter identification algorithm by the current humidity. When the current humidity is less than the standard reference humidity (dry environment): the dry environment may cause the wettability of the electrode interface to decrease or the contact micropores to shrink. This physical change will cause the overall impedance spectrum to rise. When the parameter identification algorithm performs spectrum fitting, this part of the additional impedance caused by the deterioration of the interface state is easily coupled into the diffusion impedance term in the low frequency region, resulting in the calculated current diffusion impedance being artificially high. Formula (9) uses the exponential decay characteristic to convert this artificially high value back to the standard level. When the current humidity is greater than the standard reference humidity (humid environment): under high humidity or condensation conditions, parallel leakage conduction paths may be formed on the surface of the battery pack or insulating components. Since the diffusion impedance is mainly manifested as low frequency characteristics, and surface leakage also significantly changes the impedance response in the low frequency band, the parameter identification algorithm often misjudges the decrease in impedance magnitude caused by leakage as the acceleration of the diffusion process during the decoupling process, thus causing the identified current diffusion impedance to be abnormally low. Formula (9) eliminates the interference of surface leakage on the fitting results of the parameter identification algorithm through reverse compensation, and restores the true diffusion impedance.

[0076] Based on the current temperature and humidity, the content of characteristic gases associated with insulation defects is corrected for temperature and humidity. A gas correction factor is determined using the equivalent characteristic gas content to characterize the degree of damage inside the battery pack. In one embodiment of this application, there is a mapping relationship between the equivalent characteristic gas content and the gas correction factor characterizing the degree of damage inside the battery pack: the higher the equivalent characteristic gas content, the smaller the gas correction factor; conversely, the lower the equivalent characteristic gas content, the larger the gas correction factor. This mapping relationship is pre-determined by fitting experimental data. The process of correcting the content of characteristic gases associated with insulation defects for temperature and humidity to obtain the equivalent characteristic gas content aims to eliminate the interference of gas diffusion rate fluctuations caused by current temperature changes and gas molecule competitive adsorption effects caused by humidity changes on sensor readings. Through the temperature and humidity correction process, the equivalent characteristic gas content under different operating conditions can be normalized to the equivalent content under standard conditions, thereby accurately reflecting the true degree of damage to the insulation defects inside the battery pack and avoiding false alarms or missed alarms caused by drastic fluctuations in ambient temperature and humidity.

[0077] Based on the gas correction factor, damage decoupling compensation is performed on the current SEI film impedance after temperature compensation, the current charge transfer impedance after humidity compensation, and the current diffusion impedance after humidity compensation to obtain the current SEI film impedance, current charge transfer impedance, and current diffusion impedance after gas compensation. In one embodiment of this application, the calculation formula for the current SEI film impedance after gas compensation is as follows:

[0078] Equation (10)

[0079] in, This indicates the current SEI film impedance after gas compensation. This indicates the current SEI film impedance after temperature compensation. This represents the gas correction factor, with a value range of... When the gas correction factor is 1, it indicates that the current SEI film impedance after temperature compensation does not need correction. When the gas correction factor is greater than 0 and less than 1 (i.e., an increase in the content of characteristic gases is detected), it indicates that specific insulation damage or side reactions have occurred inside the battery pack. During parameter identification, such changes in interface state caused by gas generation (such as local micro-short-circuit effects or abnormal electrolyte wettability) often lead to an artificially high calculated current SEI film impedance. Therefore, the gas correction factor is used to recalculate the current SEI film impedance to eliminate the non-intrinsic impedance increment caused by insulation defects, thereby restoring the accurate impedance value that only reflects the aging state of the SEI film itself.

[0080] The formula for calculating the current charge transfer impedance after gas compensation is shown below:

[0081] Equation (11)

[0082] in, Indicates the current charge transfer impedance after gas compensation. This represents the current charge transfer impedance after humidity compensation. This represents the gas correction factor, with a value range of... When the gas correction factor is 1, it means that the current charge transfer impedance after humidity compensation does not need correction. When the gas correction factor is greater than 0 and less than 1, it indicates that the detected charge transfer impedance includes spurious impedance increments (or measurement errors) caused by gas expansion or interfacial side reactions, resulting in an inflated identification result. In this case, the gas correction factor is used to attenuate and correct the current charge transfer impedance to eliminate gas interference components and obtain a charge transfer impedance that reflects the true electrochemical reaction kinetics.

[0083] The formula for calculating the current diffusion impedance after gas compensation is shown below:

[0084] Equation (12)

[0085] in, This represents the current diffusion impedance after gas compensation. This represents the current diffusion impedance after humidity compensation. This represents the gas correction factor, with a value range of... When the gas correction factor is 1, it means that the current diffusion impedance after humidity compensation does not need to be corrected. When the gas correction factor is greater than 0 and less than 1, it means that the current diffusion impedance after humidity compensation includes the non-intrinsic diffusion resistance increment (or spurious impedance component) caused by gas generation due to insulation defects, which leads to an overestimation of the identification result. This increment needs to be removed by the gas correction factor to restore the true diffusion impedance level.

[0086] The current SEI film impedance, current charge transfer impedance, current diffusion impedance, and current ohmic impedance after gas compensation are used as the current corrected impedance parameters. In one embodiment of this application, by correcting the current impedance parameters for temperature, humidity, and gas, the impedance changes caused by environmental factors and battery pack aging and damage are effectively decoupled. This allows the current corrected impedance parameters to accurately reflect the intrinsic electrochemical characteristics of the battery pack at the current SOC, providing a reliable input basis for subsequent high-precision SOC prediction.

[0087] In one embodiment of this application, the process of performing temperature and humidity compensation correction on the content of characteristic gases associated with insulation defects based on the current temperature and humidity to obtain the equivalent characteristic gas content includes:

[0088] Based on a pre-defined gas fingerprint feature library, the gas components in the characteristic gases associated with insulation defects are determined; and the gas components generated by damage inside the battery pack are used as target characteristic gases. In one embodiment of this application, the gas components in the characteristic gases associated with insulation defects, including target characteristic gases, include: hydrogen, ethylene, and dimethyl carbonate. Although the gas components associated with insulation defects may also contain conventional side reaction products such as methane, carbon monoxide, and ethane (generated by conventional electrolyte decomposition or micro-short circuits, with low specificity), the three target gases selected in this embodiment have clear fault indications: hydrogen serves as a sensitive indicator of early thermal runaway, ethylene serves as a confirming sign of high-temperature decomposition or violent reaction of insulating materials, and dimethyl carbonate serves as a direct tracer of electrolyte leakage or damage to the sealing structure. By selecting these three types of gases as target characteristic gases, the interference of conventional side reaction products can be eliminated, improving the accuracy of the gas correction factor in characterizing the degree of insulation damage. The process of determining the gas components in characteristic gases associated with insulation defects, based on a pre-defined gas fingerprint feature library, is achieved using a gas analyzer (e.g., a multi-component infrared gas analyzer or Raman spectrometer). This gas analyzer integrates an optical sensor (or spectral acquisition module) and a signal processing unit. During the gas component identification process, the optical sensor acquires characteristic signals (such as characteristic spectra) of the mixed gas. The signal processing unit matches and identifies these characteristic signals based on the pre-defined gas fingerprint feature library, confirming the types of each component. Combining this with the intensity information of the characteristic signals, the actual concentration of each component in the current mixed gas is calculated.

[0089] The current pressure of the environment where the battery pack is located is obtained, and the target characteristic gas content is corrected by the current temperature, current humidity, and current pressure to obtain the compensated target characteristic gas content. In one embodiment of this application, the current pressure is acquired by a pressure sensor, and the calculation formula for the compensated target characteristic gas content includes:

[0090] Equation (13)

[0091] in, This indicates the content of the target characteristic gas after compensation. Indicates the content of the target characteristic gas. This indicates the current pressure (absolute value, unit: kPa). This represents the standard reference pressure, for example, a value of 101.3 kPa (in absolute terms). This indicates the standard reference temperature, measured in Kelvin. This represents the current absolute temperature obtained by converting the current temperature, in Kelvin. This represents an empirical coefficient, ranging from 0.008 to 0.015, with units of m³ / g. This indicates the current absolute humidity (obtained by converting the current humidity and current temperature), with units of g / m³. This indicates the standard reference absolute humidity, for example, a value of 11 g / m³. This represents the pressure correction index, with a value between 1.0 and 1.5.

[0092] Based on preset target characteristic gas weights, the compensated target characteristic gas content is weighted and fused to obtain the equivalent characteristic gas content. In one embodiment of this application, the calculation formula for the equivalent characteristic gas content is as follows:

[0093] Equation (14)

[0094] in, Indicates the equivalent characteristic gas content, Indicates the first Weights of target characteristic gases Indicates the compensation after the first The content of each target characteristic gas is such that the sum of the weights of all target characteristic gases equals 1. The weights of the target characteristic gases are set according to the actual situation. For example, when the target characteristic gases include hydrogen, ethylene and dimethyl carbonate, considering that insulation defects are often accompanied by minor damage to the sealing structure, dimethyl carbonate is given the first priority weight as an extremely early and highly sensitive characteristic of minor leakage of electrolyte; hydrogen is given the second priority weight as a universal marker of abnormal gas production in the early stage of thermal runaway; and ethylene is given the third priority weight as a confirmation marker of accelerated thermal runaway chain reaction and high-temperature decomposition of insulating materials. Formula (14) is the first priority gas weight obtained after compensation for environmental factors such as temperature, pressure and humidity. The content of a target characteristic gas is calculated to obtain the equivalent characteristic gas content, which eliminates the interference of environmental fluctuations on sensor readings and improves the accuracy of the equivalent characteristic gas content.

[0095] In one embodiment of this application, if the pre-calibrated external circuit parasitic parameters include parasitic resistance and parasitic inductance, then the process of deconstructing the physical port voltage based on the pre-calibrated external circuit parasitic parameters to obtain the internal terminal voltage of the battery pack includes:

[0096] The voltage drop compensation value is calculated based on the parasitic resistance and the charging / discharging current. In one embodiment of this application, the product of the parasitic resistance and the charging / discharging current is used as the voltage drop compensation value.

[0097] The inductor voltage drop compensation value is calculated based on the parasitic inductance and the charging / discharging current. In one embodiment of this application, the formula for calculating the inductor voltage drop compensation value is as follows:

[0098] Equation (15)

[0099] in, This indicates the inductor voltage drop compensation value. Indicates parasitic inductance. Indicates the charging and discharging current. The rate of change of the charging and discharging current with time is represented by the formula (15). Before the inductor voltage drop compensation value is calculated, the charging and discharging current needs to be filtered by a low-pass filter to suppress high-frequency noise.

[0100] The internal terminal voltage is determined based on the physical port voltage, resistor voltage drop compensation value, and inductor voltage drop compensation value. In one embodiment of this application, when the charging current is set to a positive value and the discharging current to a negative value, the formula for calculating the internal terminal voltage is as follows:

[0101] Equation (16)

[0102] in, Indicates the internal terminal voltage. Indicates the physical port voltage. This indicates the voltage drop compensation value. This represents the inductance voltage drop compensation value. Formula (16) can effectively restore the internal terminal voltage of the battery pack by deducting the voltage drop caused by the external line impedance and inductance effect, eliminate the systematic deviation introduced by the physical connection port and the acquisition circuit, prevent the error from accumulating in the subsequent SOC estimation, and thus significantly improve the accuracy of state of charge prediction.

[0103] In one embodiment of this application, before inputting the internal terminal voltage, current temperature, and charge / discharge current into the battery equivalent circuit model to obtain the current impedance parameters of the battery pack, the method for predicting the state of charge of the electrochemical energy storage system further includes:

[0104] This invention acquires the parallel insulation monitoring impedance and internal absolute humidity of a battery pack in an electrochemical energy storage system. In one embodiment of this application, the parallel insulation monitoring impedance characterizes the insulation performance of the high-voltage side of the battery pack to ground. The parallel insulation monitoring impedance includes the insulation impedance between the high-voltage positive electrode of the battery pack and the battery pack casing ground, and the insulation impedance between the high-voltage negative electrode and the casing ground. The process of acquiring the parallel insulation monitoring impedance includes: real-time acquisition of the voltage response signal of the high-voltage side of the battery pack to ground and the current response signal flowing through the detection circuit; using the ratio of the voltage response signal to the current response signal as the real-time equivalent insulation impedance of the high-voltage side of the battery pack to ground; converting the real-time equivalent insulation impedance into the parallel insulation monitoring impedance, or mapping the real-time equivalent insulation impedance into the parallel insulation monitoring impedance based on a preset insulation state mapping table. The process of obtaining internal absolute humidity includes: collecting relative humidity and internal temperature data of the battery pack using temperature and humidity sensors installed inside the battery pack; calculating the equivalent saturated vapor pressure based on the internal temperature data and Magnus's formula; and calculating the internal absolute humidity based on the equivalent saturated vapor pressure and the internal relative humidity. The formula for calculating the equivalent saturated vapor pressure is shown below:

[0105] Equation (17)

[0106] in, The equivalent saturated vapor pressure, expressed as hectopascals, represents the internal temperature data of the battery pack. This indicates the internal temperature data of the pool, in degrees Celsius.

[0107] The formula for calculating internal absolute humidity is as follows:

[0108] Equation (18)

[0109] in, Indicates the internal absolute humidity. This indicates the relative humidity inside the battery pack. The equivalent saturated vapor pressure represents the internal temperature data of the battery pack.

[0110] If the current humidity is greater than the relative humidity threshold, the internal absolute humidity is greater than the absolute humidity threshold, and the parallel insulation monitoring impedance is less than the preset impedance threshold, then the battery pack is determined to be in a leakage state. In one embodiment of this application, the relative humidity threshold is set according to the actual situation, the absolute humidity threshold is set according to the actual situation, and the preset impedance threshold is set according to the actual situation. For example, the relative humidity threshold is set within the range of 70%-80%, the absolute humidity threshold is set within the range of 10g / m³-25g / m³, and the preset impedance threshold is set within the range of 200Ω / V-400Ω / V.

[0111] If the battery pack is in a leakage state, leakage compensation is performed on the physical port voltage based on the parallel insulation monitoring impedance and the reference ohmic impedance value to obtain the compensated physical port voltage. In one embodiment of this application, the reference ohmic impedance value is an ohmic impedance value set based on historical operating data before the battery pack experiences a leakage state; the calculation formula for the compensated physical port voltage is as follows:

[0112] Equation (19)

[0113] in, This represents the compensated physical port voltage. Indicates the physical port voltage. Indicates the reference ohmic impedance value. The parallel insulation monitoring impedance is represented. Formula (19) is used to estimate the leakage current flowing through the casing ground using the parallel insulation monitoring impedance and the physical port voltage, and to calculate the additional voltage drop generated by the leakage current on the reference ohmic impedance value. This additional voltage drop is then compensated back into the physical port voltage to restore the true physical port voltage. In formula (19), the charging current is set to a positive value and the discharging current is set to a negative value.

[0114] Based on pre-calibrated external circuit parasitic parameters, parasitic parameter stripping calculations are performed on the compensated physical port voltage to obtain the internal terminal voltage of the battery pack. In one embodiment of this application, the process of performing parasitic parameter stripping calculations on the compensated physical port voltage based on pre-calibrated external circuit parasitic parameters to obtain the internal terminal voltage of the battery pack is the same as the process of performing parasitic parameter stripping calculations on the physical port voltage based on pre-calibrated external circuit parasitic parameters to obtain the internal terminal voltage of the battery pack.

[0115] In one embodiment of this application, the process of performing leakage current compensation on the physical port voltage based on the parallel insulation monitoring impedance and the reference ohmic impedance value to obtain the compensated physical port voltage includes:

[0116] The leakage current is calculated based on the parallel insulation monitoring impedance and the physical port voltage. In one embodiment of this application, the leakage current is the ratio of the physical port voltage to the parallel insulation monitoring impedance.

[0117] The additional voltage drop generated by the physical port voltage is determined based on the leakage current and the reference ohmic impedance value, and this additional voltage drop is used as the compensation voltage. In one embodiment of this application, the additional voltage drop is the product of the leakage current and the reference ohmic impedance value.

[0118] The sum of the compensation voltage and the physical port voltage is used as the compensated physical port voltage. In one embodiment of this application, the above settings enable correction of the physical port voltage when insulation degradation and leakage occur in the battery pack, preventing voltage measurement errors caused by leakage from being transmitted to subsequent state of charge calculations. Furthermore, by using pre-calibrated external circuit parasitic parameters, parasitic parameter stripping calculations are performed on the compensated physical port voltage, further improving the accuracy of internal terminal voltage calculations.

[0119] Figure 2 This is a flowchart illustrating the process of determining the internal terminal voltage of a battery pack, as shown in another exemplary embodiment of this application. Figure 2 In the process of determining the internal terminal voltage of the battery pack, the following steps are taken: (1) Obtain the physical port voltage, charging and discharging current, parallel insulation monitoring impedance and internal absolute humidity of the battery pack in the electrochemical energy storage system, as well as the current temperature, current humidity and content of characteristic gases associated with insulation defects in the environment where the battery pack is located; (2) If the current humidity is greater than the relative humidity threshold, the internal absolute humidity is greater than the absolute humidity threshold, and the parallel insulation monitoring impedance is less than the preset impedance threshold, then the battery pack is determined to be in a leakage state; (3) If the battery pack is in a leakage state, the physical port voltage is compensated for leakage based on the parallel insulation monitoring impedance and the reference ohmic impedance value to obtain the compensated physical port voltage; (4) Based on the pre-calibrated external circuit parasitic parameters, the parasitic parameter stripping calculation is performed on the compensated physical port voltage to obtain the internal terminal voltage of the battery pack.

[0120] Figure 3 This is a flowchart illustrating the process of correcting the current impedance parameter, as shown in another exemplary embodiment of this application. Figure 3The process of correcting the current impedance parameters includes: (1) performing temperature normalization compensation on the current impedance parameters based on the current temperature to obtain the current ohmic impedance, the current SEI film impedance, the current charge transfer impedance, and the current diffusion impedance after temperature compensation; (2) performing humidity compensation on the current ohmic impedance, the current charge transfer impedance, and the current diffusion impedance after temperature compensation based on the current humidity to obtain the current ohmic impedance, the current charge transfer impedance, and the current diffusion impedance after humidity compensation; (3) adjusting the content of characteristic gases associated with insulation defects based on the current temperature and humidity. Temperature and humidity compensation correction is performed to obtain the equivalent characteristic gas content; the gas correction factor used to characterize the degree of damage inside the battery pack is determined by the equivalent characteristic gas content; (4) damage decoupling compensation is performed on the current SEI film impedance after temperature compensation, the current charge transfer impedance after humidity compensation and the current diffusion impedance after humidity compensation according to the gas correction factor, to obtain the current SEI film impedance after gas compensation, the current charge transfer impedance after gas compensation and the current diffusion impedance after gas compensation; (5) the current SEI film impedance after gas compensation, the current charge transfer impedance after gas compensation, the current diffusion impedance after gas compensation and the current ohmic impedance after humidity compensation are used as the current correction impedance parameters.

[0121] Figure 4 This is an exemplary embodiment of the present application illustrating the operating interface of a state of charge prediction system for an electrochemical energy storage system, as shown in the diagram. Figure 4 As shown, the operation interface includes: a battery pack selection window, an input data display window, an internal terminal voltage display window, a corrected impedance parameter display window, and a SOC prediction result display window. The battery pack selection window is used to select different individual battery packs. The input data display window displays electrical parameters, environmental parameters, and battery pack information. The internal terminal voltage display window shows the internal terminal voltage of the battery pack. The corrected impedance parameter display window shows the current corrected impedance parameter. The SOC prediction result displays the predicted state of charge (SOC) of a single battery pack, and also displays the SOC prediction result of a battery pack composed of multiple battery packs. This operation interface can display the parameters and prediction results of a single battery pack individually, or it can visualize the overall prediction results of the battery pack. This allows maintenance personnel to intuitively grasp the current SOC level of the electrochemical energy storage system, promptly detect abnormal SOC of individual cells, and provide intuitive decision-making basis for the balanced maintenance and safety management of the energy storage system.

[0122] The following describes an embodiment of the apparatus described in this application, which can be used to execute the state-of-charge prediction system for the electrochemical energy storage system described in the above embodiments of this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the state-of-charge prediction method for the electrochemical energy storage system described above in this application.

[0123] Figure 5 This is a block diagram illustrating a state-of-charge prediction system for an electrochemical energy storage system, as shown in an exemplary embodiment of this application.

[0124] like Figure 5 As shown, the exemplary state-of-charge prediction system 500 for electrochemical energy storage system includes:

[0125] The data acquisition module 501 is used to acquire the physical port voltage and charging / discharging current of the battery pack in the electrochemical energy storage system, as well as the current temperature, current humidity, and content of characteristic gases associated with insulation defects in the environment where the battery pack is located.

[0126] The voltage calculation module 502 is used to perform parasitic parameter stripping calculation on the physical port voltage based on pre-calibrated external circuit parasitic parameters to obtain the internal terminal voltage of the battery pack.

[0127] The parameter calculation module 503 is used to input the internal terminal voltage, current temperature and charging / discharging current into the battery equivalent circuit model to obtain the current impedance parameters of the battery pack; and to correct the current impedance parameters based on the current humidity, current temperature and the content of characteristic gases associated with insulation defects to obtain the current corrected impedance parameters.

[0128] The state of charge prediction module 504 is used to update the parameters in the battery equivalent circuit model through the current corrected impedance parameters, and input the internal terminal voltage and charging / discharging current into the updated battery equivalent circuit model to obtain the state of charge of the battery pack, and use the state of charge as the state of charge prediction result.

[0129] In one embodiment of this application, the physical port voltage is directly measured by a voltage acquisition unit, the charging and discharging current is acquired by a current transformer connected in series in the main circuit, the environment where the parallel battery pack is located is the closed gas chamber of the gas-filled cabinet, the current temperature is acquired by a temperature sensor arranged inside the closed gas chamber, the current humidity is acquired by a humidity sensor arranged inside the closed gas chamber, and the content of characteristic gases associated with insulation defects is detected by a gas analyzer or gas sensor array installed in the closed gas chamber. The gas sensor array includes at least a hydrogen sensor, an ethylene sensor, and a dimethyl carbonate sensor.

[0130] In one embodiment of this application, the pre-calibrated parasitic parameters of the external circuit include the parasitic resistance and parasitic inductance of the external acquisition line. These parasitic parameters can be directly measured and acquired using an impedance analyzer before the energy storage system is put into operation, or they can be automatically calculated through an initialization calibration process deployed within the energy storage system. The parasitic parameter stripping calculation process is as follows: calculate the voltage drop caused by the parasitic resistance, calculate the inductance voltage drop caused by the parasitic inductance, and subtract the voltage drop and inductance voltage drop from the physical port voltage (assuming the charging current is positive and the discharging current is negative). This yields the internal terminal voltage after stripping the influence of parasitic parameters. This process eliminates voltage deviations caused by external lines and physical connection ports, preventing the continuous accumulation of systematic errors caused by parasitic parameters that could affect the final prediction results, thereby ensuring the accuracy of the voltage data used in subsequent calculations.

[0131] In one embodiment of this application, the process of inputting the internal terminal voltage, current temperature, and charging / discharging current into the battery equivalent circuit model to obtain the current impedance parameters of the battery pack includes: determining the initial model parameter reference of the battery equivalent circuit model based on the current temperature, and using the initial model parameter reference as the prior initial value of the parameter identification algorithm; using the internal terminal voltage and charging / discharging current as the real-time observation data of the parameter identification algorithm; the parameter identification algorithm performing recursive loop calculations on the real-time observation data based on the prior initial value to obtain the predicted terminal voltage; with the goal of minimizing the residual between the predicted internal terminal voltage and the actual internal terminal voltage, dynamically updating the current model parameters of the battery equivalent circuit model (including: ohmic impedance, SEI film impedance, charge transfer impedance, polarization capacitance corresponding to charge transfer impedance, diffusion impedance, polarization capacitance corresponding to diffusion impedance, etc.), until the difference between the recursively calculated predicted internal terminal voltage and the actual internal terminal voltage is less than a preset difference threshold, terminating the recursive loop, and using the dynamically updated ohmic impedance, dynamically updated SEI film impedance, dynamically updated charge transfer impedance, and dynamically updated diffusion impedance as the current impedance parameters. Parameter identification algorithms include recursive least squares, extended Kalman filter, dual Kalman filter, or particle filter, which can be selected based on actual computational performance and accuracy requirements. The battery equivalent circuit model can be either the PNGV (Partnership for a New Generation of Vehicles) model or a second-order RC equivalent circuit model.

[0132] In one embodiment of this application, the process of inputting the internal terminal voltage, current temperature, and charging / discharging current into the battery equivalent circuit model to obtain the current impedance parameters of the battery pack is based on the dynamic electrical response characteristics of the battery pack. The equivalent circuit parameters are inferred by observing the mapping relationship between the charging / discharging current and the internal terminal voltage in real time. This method relies only on the dynamic excitation and response of voltage and current, without introducing the state of charge (SOC) of the battery pack. This decouples the impedance parameter identification algorithm from the SOC state estimation algorithm, effectively avoiding the "error coupling" problem of parameter identification divergence caused by SOC estimation errors in related algorithms.

[0133] In one embodiment of this application, after obtaining the current impedance parameter, the current impedance parameter is corrected according to the current humidity, current temperature and the content of characteristic gases associated with insulation defects. This can effectively offset the drift problem of battery equivalent circuit model parameters caused by temperature and humidity changes and battery pack insulation aging. It is beneficial to ensure that the battery equivalent circuit model parameters always match the current actual operating state of the battery pack, and obtain the current corrected impedance parameter that can accurately reflect the actual operating state of the battery pack.

[0134] In one embodiment of this application, the process of inputting the internal terminal voltage and charging / discharging current into the updated battery equivalent circuit model to obtain the state of charge (SOC) of the battery pack includes: calculating a preliminary predicted SOC value of the battery pack using the ampere-hour integral method based on the charging / discharging current; inputting the charging / discharging current and the preliminary predicted SOC value into the updated battery equivalent circuit model, which calculates an estimated internal terminal voltage using the current corrected impedance parameter, the charging / discharging current, and the preliminary predicted SOC value, and calculates the voltage residual between the estimated internal terminal voltage and the internal terminal voltage; calculating the Kalman gain based on the updated battery equivalent circuit model and the voltage residual using a Kalman filtering algorithm; wherein the Kalman gain is used to characterize the compensation weight for the voltage residual; multiplying the Kalman gain by the voltage residual to obtain the SOC correction compensation amount; and superimposing the SOC correction compensation amount with the preliminary predicted SOC value to obtain the current SOC of the battery pack.

[0135] In one embodiment of this application, the process of calculating the estimated internal terminal voltage using the current corrected impedance parameters, charging / discharging current, and preliminary SOC prediction value in the battery equivalent circuit model includes: obtaining the theoretical open-circuit voltage by looking up a table based on the preliminary SOC prediction value and a preset open-circuit voltage-state of charge (OCV-SOC) mapping relationship; multiplying the charging / discharging current by the current ohmic impedance in the current corrected impedance parameters to obtain the ohmic voltage drop; inputting the charging / discharging current into the RC parallel network of the battery equivalent circuit model, wherein the discretized mathematical expression of the physical characteristics of the RC parallel network is characterized by a preset dynamic difference equation (e.g., polarization voltage state transition equation), using the charging / discharging current as the input excitation variable of the preset dynamic difference equation, and substituting the polarization internal resistance (including diffusion impedance and charge transfer impedance) and polarization capacitance in the current corrected impedance parameters as coefficient parameters of the preset dynamic difference equation, solving the preset dynamic difference equation to obtain the polarization voltage drop; and subtracting the ohmic voltage drop and polarization voltage drop from the theoretical open-circuit voltage to obtain the estimated internal terminal voltage.

[0136] In one embodiment of this application, a parasitic parameter stripping mechanism is introduced. By using pre-calibrated external circuit parasitic parameters, parasitic parameter stripping calculations are performed on the compensated voltage to restore the internal terminal voltage of the battery pack. This solves the problem of physical port voltage measurement distortion and avoids the systematic errors introduced by directly using the physical port voltage interfered with by parasitic parameters for state of charge estimation in related technologies. The nonlinear effects of environmental factors (e.g., current temperature, current humidity, and the content of characteristic gases associated with insulation defects) on the electrochemical characteristics of the battery pack are fully considered, and the impedance parameters in the battery equivalent circuit model are dynamically corrected. This allows the battery equivalent circuit model to adapt to the current operating conditions and aging state in real time, overcoming the defect of model parameters drifting with aging and environmental changes. The mechanism-driven strategy of "updating model parameters first and then predicting state" is adopted to update the impedance parameters after environmental feature correction to the battery equivalent circuit model in real time, ensuring that the battery equivalent circuit model is always in the "best matching" state. Compared to purely data-driven machine learning algorithms, the method in this application does not rely on massive amounts of labeled data, avoiding the problem of poor generalization ability of the battery equivalent circuit model after changes in operating conditions or battery pack aging. Compared to the approach of predicting the state of charge solely through the ampere-hour integration method combined with the equivalent circuit model, this application eliminates the impact of current measurement error accumulation and parameter drift by accurately acquiring the internal terminal voltage and correcting the parameters in real time. Compared to model methods represented by Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF) in related technologies, this application overcomes the shortcomings of model parameter drift, fixed model parameters, or the need for frequent manual calibration by accurately acquiring the internal terminal voltage and correcting the parameters in real time. Even in large-capacity multi-module energy storage systems, facing complex scenarios such as battery pack aging, drastic changes in charge and discharge rates, or switching of operating conditions, this application can still maintain high prediction accuracy, significantly improving the operational safety and maintenance management efficiency of electrochemical energy storage systems throughout their entire life cycle.

[0137] Furthermore, compared to model methods represented by Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF) in related technologies, this application first obtains the current impedance parameter, which is independent of the SOC estimation process, through parameter identification, and then obtains the current corrected impedance parameter through environmental factor correction. The current corrected impedance parameter is then updated into the battery equivalent circuit model, and the SOC estimation is completed based on the updated battery equivalent circuit model. This avoids the mutual coupling between SOC estimation error and parameter identification error. Even after a long and multi-cycle estimation process, it effectively alleviates the problem of continuous accumulation and amplification of SOC calculation error, maintains the prediction accuracy of battery pack state of charge, and effectively reduces the situation of algorithm divergence and inaccuracy.

[0138] It should be noted that the state-of-charge prediction system for an electrochemical energy storage system provided in the above embodiments and the state-of-charge prediction method for an electrochemical energy storage system provided in the above embodiments belong to the same concept. The specific operation methods of each module and unit have been described in detail in the method embodiments and will not be repeated here. In practical applications, the state-of-charge prediction system for an electrochemical energy storage system provided in the above embodiments can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. This is not a limitation here.

[0139] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in this application should still be covered by the claims of this application.

Claims

1. A method for predicting the state of charge of an electrochemical energy storage system, characterized in that, The method includes: The physical port voltage and charging / discharging current of the battery pack in the electrochemical energy storage system are obtained, as well as the current temperature, current humidity, and content of characteristic gases associated with insulation defects in the environment where the battery pack is located. Based on pre-calibrated external circuit parasitic parameters, the parasitic parameter stripping calculation is performed on the physical port voltage to obtain the internal terminal voltage of the battery pack; The internal terminal voltage, the current temperature, and the charging / discharging current are input into the battery equivalent circuit model to obtain the current impedance parameters of the battery pack; the current impedance parameters are corrected according to the current humidity, the current temperature, and the content of characteristic gases associated with insulation defects to obtain the current corrected impedance parameters. The parameters in the battery equivalent circuit model are updated by using the current corrected impedance parameters, and the internal terminal voltage and the charging and discharging current are input into the updated battery equivalent circuit model to obtain the state of charge of the battery pack, and the state of charge is used as the state of charge prediction result.

2. The method for predicting the state of charge of an electrochemical energy storage system according to claim 1, characterized in that, If there are multiple battery packs, after obtaining the state of charge of each battery pack, the method further includes: Acquire historical charge and discharge conditions, charge and discharge modes, and pack temperature for each battery pack; Based on the state of charge and temperature of each battery pack, determine the basic polarization resistance and basic polarization capacitance of each battery pack under the current charge and discharge conditions. The charging and discharging current, the duration of the charging and discharging current, and the basic polarization resistance and basic polarization capacitance of each battery pack under the current charging and discharging conditions are input into the updated battery equivalent circuit model to obtain the dynamic polarization voltage of each battery pack; the ratio of the dynamic polarization voltage of each battery pack to the charging and discharging current of the corresponding battery pack is used to obtain the polarization internal resistance increment of each battery pack under the current charging and discharging conditions. The total impedance of each battery pack is subtracted from the polarization internal resistance increment of the corresponding battery pack to obtain the reference internal resistance value of each battery pack after eliminating the real-time polarization effect; and the reference internal resistance value of each battery pack is normalized and compensated based on the standard reference temperature to obtain the current reference internal resistance value of each battery pack; the total impedance is characterized by the current corrected impedance parameter. Based on the current internal resistance baseline value and the preset internal resistance reference value of each battery pack, the degree of internal resistance increase of each battery pack is calculated. Extract the characteristic data of each battery pack under the historical charge and discharge conditions, input the internal resistance growth rate and characteristic data of each battery pack into the internal resistance growth rate and capacity decay mapping curve to obtain the health status value of each battery pack; the characteristic data includes: cumulative charge and discharge cycle count, charge and discharge duration exceeding the preset rate, and cumulative charge and discharge duration exceeding the preset temperature. The state of charge prediction result is determined based on the charging and discharging mode, the health status value of each battery pack, and the state of charge.

3. The method for predicting the state of charge of an electrochemical energy storage system according to claim 2, characterized in that, The process of determining the state of charge prediction result based on the charging and discharging mode, the health status value of each battery pack, and the state of charge includes: Obtain the port voltage fluctuation value and load current change amplitude of each battery pack when the load current undergoes a step change under the current charging and discharging conditions; The measured internal resistance of each battery pack is obtained by taking the ratio between the port voltage fluctuation value and the load current change amplitude. The measured internal resistance of the corresponding battery pack is corrected by taking the health status value, state of charge and pack temperature of each battery pack to obtain the dynamic internal resistance. If the charging / discharging mode is the discharge mode, then based on the dynamic internal resistance and discharge current, the port voltage drop value of each battery pack is determined, and the difference between the physical port voltage of each battery pack and the corresponding port voltage drop value of the battery pack is taken as the predicted port voltage of each battery pack; the difference between the predicted port voltage of each battery pack and the discharge cutoff protection voltage is taken as the discharge voltage drop margin of each battery pack; the state of charge of the battery pack with the smallest discharge voltage drop margin is compared with the smallest state of charge among all states of charge, and the smaller state of charge is taken as the predicted state of charge result. If the charging / discharging mode is a charging mode, then based on the dynamic internal resistance and charging current, the port boost value of each battery pack is determined, and the sum of the physical port voltage of each battery pack and the corresponding port boost value is taken as the predicted port voltage of each battery pack; the difference between the charging cut-off protection voltage and the predicted port voltage of each battery pack is taken as the charging voltage drop margin of each battery pack; the state of charge of the battery pack with the smallest charging voltage drop margin is compared with the largest state of charge among all states of charge, and the larger state of charge is taken as the predicted state of charge result.

4. The method for predicting the state of charge of an electrochemical energy storage system according to claim 3, characterized in that, The process of correcting the measured internal resistance of each battery pack by using its health status, state of charge, and pack temperature to obtain the dynamic internal resistance includes: Based on the health status value, state of charge and pack temperature of each battery pack, determine the theoretical internal resistance of each battery pack under the current charge and discharge conditions. If the difference between the measured internal resistance and the theoretical internal resistance is less than or equal to a preset difference threshold, then the measured internal resistance and the theoretical internal resistance are weighted and fused to obtain the dynamic internal resistance; If the difference between the measured internal resistance and the theoretical internal resistance is greater than the preset difference threshold, then the theoretical internal resistance is taken as the dynamic internal resistance.

5. The method for predicting the state of charge of an electrochemical energy storage system according to any one of claims 1-4, characterized in that, If the current impedance parameters include: current ohmic impedance, current SEI film impedance, current charge transfer impedance, and current diffusion impedance, then the process of correcting the current impedance parameters based on the current humidity, the current temperature, and the content of characteristic gases associated with insulation defects to obtain the current corrected impedance parameters includes: Based on the current temperature, the current impedance parameters are normalized and compensated for temperature to obtain the current ohmic impedance, the current SEI film impedance, the current charge transfer impedance, and the current diffusion impedance after temperature compensation. Based on the current humidity, humidity compensation is applied to the current ohmic impedance after temperature compensation, the current charge transfer impedance after temperature compensation, and the current diffusion impedance after temperature compensation to obtain the current ohmic impedance after humidity compensation, the current charge transfer impedance after humidity compensation, and the current diffusion impedance after humidity compensation. Based on the current temperature and humidity, the content of the characteristic gas associated with the insulation defect is corrected for temperature and humidity to obtain the equivalent characteristic gas content; the equivalent characteristic gas content is used to determine a gas correction factor for characterizing the degree of damage inside the battery pack. Based on the gas correction factor, damage decoupling compensation is performed on the current SEI film impedance after temperature compensation, the current charge transfer impedance after humidity compensation, and the current diffusion impedance after humidity compensation to obtain the current SEI film impedance after gas compensation, the current charge transfer impedance after gas compensation, and the current diffusion impedance after gas compensation. The current SEI film impedance after gas compensation, the current charge transfer impedance after gas compensation, the current diffusion impedance after gas compensation, and the current ohmic impedance after humidity compensation are used as the current corrected impedance parameters.

6. The method for predicting the state of charge of an electrochemical energy storage system according to claim 5, characterized in that, The process of performing temperature compensation correction on the content of characteristic gases associated with insulation defects based on the current temperature and humidity to obtain the equivalent characteristic gas content includes: Based on a pre-set gas fingerprint feature library, the gas components in the characteristic gases associated with insulation defects are determined; and the gas components generated by damage inside the battery pack are used as target characteristic gases; the target characteristic gases include: hydrogen, ethylene and dimethyl carbonate; The current pressure of the environment where the battery pack is located is obtained, and the target characteristic gas content is corrected by the current temperature, the current humidity and the current pressure to obtain the compensated target characteristic gas content; Based on the preset target characteristic gas weights, the compensated target characteristic gas content is weighted and fused to obtain the equivalent characteristic gas content.

7. The method for predicting the state of charge of an electrochemical energy storage system according to any one of claims 1-4, characterized in that, If the pre-calibrated external circuit parasitic parameters include parasitic resistance and parasitic inductance, then the process of deconstructing the physical port voltage based on the pre-calibrated external circuit parasitic parameters to obtain the internal terminal voltage of the battery pack includes: Calculate the resistance voltage drop compensation value based on the parasitic resistance and the charging / discharging current; Calculate the inductor voltage drop compensation value based on the parasitic inductance and the charging / discharging current; The internal terminal voltage is determined based on the physical port voltage, the resistor voltage drop compensation value, and the inductor voltage drop compensation value.

8. The method for predicting the state of charge of an electrochemical energy storage system according to any one of claims 1-4, characterized in that, Before inputting the internal terminal voltage, the current temperature, and the charging / discharging current into the battery equivalent circuit model to obtain the current impedance parameters of the battery pack, the method further includes: Obtain the parallel insulation monitoring impedance and internal absolute humidity of the battery pack in the electrochemical energy storage system; the parallel insulation monitoring impedance characterizes the insulation performance of the high-voltage side of the battery pack to ground. If the current humidity is greater than the relative humidity threshold, the internal absolute humidity is greater than the absolute humidity threshold, and the parallel insulation monitoring impedance is less than the preset impedance threshold, then the battery pack is determined to be in a leakage state. If the battery pack is in a leakage state, leakage compensation is performed on the physical port voltage based on the parallel insulation monitoring impedance and the reference ohmic impedance value to obtain the compensated physical port voltage; the reference ohmic impedance value is an ohmic impedance value set based on historical operating data of the battery pack before the leakage state occurs. Based on the pre-calibrated parasitic parameters of the external circuit, the parasitic parameter stripping calculation is performed on the compensated physical port voltage to obtain the internal terminal voltage of the battery pack.

9. The method for predicting the state of charge of an electrochemical energy storage system according to claim 8, characterized in that, The process of performing leakage current compensation on the physical port voltage based on the parallel insulation monitoring impedance and the reference ohmic impedance value to obtain the compensated physical port voltage includes: Calculate the leakage current based on the parallel insulation monitoring impedance and the physical port voltage; Based on the leakage current and the reference ohmic impedance value, the additional voltage drop generated by the physical port voltage is determined, and the additional voltage drop is used as the compensation voltage. The sum of the compensation voltage and the physical port voltage is taken as the compensated physical port voltage.

10. A state-of-charge prediction system for an electrochemical energy storage system, characterized in that, include: The data acquisition module is used to acquire the physical port voltage and charging / discharging current of the battery pack in the electrochemical energy storage system, as well as the current temperature, current humidity, and content of characteristic gases associated with insulation defects in the environment where the battery pack is located. The voltage calculation module is used to perform parasitic parameter stripping calculation on the physical port voltage based on pre-calibrated external circuit parasitic parameters to obtain the internal terminal voltage of the battery pack; The parameter calculation module is used to input the internal terminal voltage, the current temperature and the charging and discharging current into the battery equivalent circuit model to obtain the current impedance parameters of the battery pack. Based on the current humidity, the current temperature, and the content of characteristic gases associated with insulation defects, the current impedance parameter is corrected to obtain the current corrected impedance parameter. State of charge prediction is used to update the parameters in the battery equivalent circuit model using the current corrected impedance parameters, and input the internal terminal voltage and the charging and discharging current into the updated battery equivalent circuit model to obtain the state of charge of the battery pack, and use the state of charge as the state of charge prediction result.

Images