SOC differentiation state evaluation method of energy storage module

Abstract

The invention relates to the technical field of battery state evaluation, and discloses an SOC differentiated state evaluation method for an energy storage module, which comprises the following steps: applying symmetric electrical excitation pulses with equal electric quantity and opposite polarities to a battery module, and respectively obtaining a first voltage relaxation curve and a second voltage relaxation curve after the symmetric electrical excitation pulses; furthermore, a differentiation index for directly quantifying the internal unbalance degree is generated by calculating and integrating the asymmetric response between the two curves, and an instant symmetric electrical disturbance reference system is constructed in the battery module, so that the internal unbalance degree can be directly quantified under the condition of not depending on any historical baseline data or external compensation model. According to the method, the asymmetric state information of the battery is directly extracted from the internal contrast relation generated by one-time detection action, the evaluation result avoids the influence of common drift introduced by factors such as environment temperature change and uniform aging, and a basis is provided for distinguishing unbalanced physical sources.

Description

Technical Field

[0001] This invention relates to a method for assessing the state of charge (SOC) of an energy storage module, belonging to the field of battery state assessment technology. Background Technology

[0002] Currently, in the state assessment of vanadium redox flow battery energy storage systems, the common technical approach is to use the stable function correspondence between the battery's open-circuit voltage and its state of charge, i.e., the OCV-SOC curve, as the basis for all state estimation algorithms. The internal energy storage state of the battery is inferred by measuring the open-circuit voltage value that can be directly obtained from the outside.

[0003] When this scheme is applied to the long-term cyclic operation of a large-scale vanadium redox flow battery energy storage power station, a situation determined by the battery's physical structure is that, during continuous operation, due to the small pressure difference and concentration gradient on both sides of the ion exchange membrane, ions and water will slowly migrate between the positive and negative electrodes. The direct consequence of this process is that the actual state of charge of the positive and negative electrolytes will gradually deviate from the initial equilibrium state, resulting in differences. At this time, the only measurable open-circuit voltage value outside the system is actually an apparent value under the combined effect of two different internal potentials. This value can no longer accurately correspond to the true overall energy storage state of the system, nor can it reflect its internal asymmetric state information.

[0004] To improve estimation accuracy, those skilled in the art have employed methods such as optimized filtering algorithms or state observers. However, the calibration benchmark for these methods remains the aforementioned apparent open-circuit voltage. Therefore, their role is to optimize the tracking and estimation of this apparent value, but they cannot obtain the internal state differences that have already occurred behind this apparent value. Specifically, the existing technical approach has the following limitations: 1. The OCV-SOC model on which state assessment is based is a static symmetric model, and the basis of this model has changed in the actual dynamic operation of the battery; 2. All algorithmic processing based on apparent voltage cannot separate the feature information that directly characterizes the degree of asymmetry within the system. This limitation is not only reflected at the state estimation algorithm level, but also in the more macroscopic overall design of the battery management system (BMS). Existing technologies generally lack the ability to actively perceive and deeply diagnose internal state imbalances within the system. For example, the authorization announcement number is C Chinese invention patent N118099476A discloses a vanadium redox flow battery management system and a vanadium redox flow battery energy storage system. By monitoring a series of macroscopic operating parameters such as pipeline pressure, temperature, DC bus voltage, and current, as well as switch status signals such as liquid level and valves, it executes corresponding control commands such as frequency converter speed regulation and switch on / off to ensure that the battery operates under safe conditions. However, the control strategy of this type of management system is entirely based on passive response to externally measurable parameters. It does not have an inherent mechanism to actively detect and quantify the degree of state of charge (SOC) imbalance between the positive and negative electrolytes. The voltage, current, and other data on which all its decisions depend are precisely the apparent values ​​that can no longer accurately reflect the true internal state. Although such systems can achieve routine management and safety protection of the battery operation process, they cannot fundamentally diagnose and solve the capacity decay and performance degradation problems caused by internal state imbalance. Therefore, the technical problem to be solved by this invention is how to establish an evaluation method that does not rely on the static OCV-SOC correspondence and can bypass the limitation of apparent open circuit voltage, thereby obtaining an evaluation method that can directly characterize the degree of asymmetry of the positive and negative electrode charge states inside the battery system. Summary of the Invention

[0005] This invention provides a method for assessing the state of charge (SOC) of an energy storage module. Its main purpose is to solve the problem that existing technologies rely on apparent open-circuit voltage and static models, which cannot obtain asymmetric information about the positive and negative states of charge inside the battery system.

[0006] To achieve the above objectives, the present invention provides a method for assessing the state of charge (SOC) differentiation of energy storage modules, comprising:

[0007] Step a: Apply a first electrical excitation pulse with a first polarity to the battery module, and after the first electrical excitation pulse ends, obtain a first voltage relaxation curve;

[0008] Step b: Apply a second electrical excitation pulse with a second polarity opposite to the first polarity to the battery module, wherein the magnitude of the second electrical excitation pulse is equal to the magnitude of the first electrical excitation pulse, and obtain a second voltage relaxation curve after the second electrical excitation pulse ends;

[0009] Step c, through the following steps, a real-time differentiation index is calculated based on the difference between the first voltage relaxation curve and the second voltage relaxation curve: Step c1, calculate an asymmetric response function curve between the two voltage relaxation curves; Step c2, perform an integral operation on the asymmetric response function curve to generate the real-time differentiation index.

[0010] Step d: Based on the correspondence between a calibration differentiation index established through offline calibration and the SOC differentiation state of the battery module, and using the real-time differentiation index calculated in step c as the input value of the calibration differentiation index, the SOC differentiation state of the battery module is determined.

[0011] Preferably, after step a of obtaining the first voltage relaxation curve and before step b of applying the second electrical excitation pulse, the method further includes: measuring and obtaining a voltage drift rate characterizing the potential change of the battery module within a quiet period; and in step c, firstly generating a linear compensation vector based on the voltage drift rate, and subtracting the linear compensation vector from the second voltage relaxation curve to generate a compensated second voltage relaxation curve, and then using the first voltage relaxation curve and the compensated second voltage relaxation curve for subsequent calculations.

[0012] Preferably, the method further includes: performing steps a to d in a first flow state where the electrolyte circulation stops to obtain a first differentiation index; and performing steps a to d again in a second flow state where the electrolyte circulation is normal to obtain a second differentiation index; wherein the SOC differentiation state is further determined based on a comparison of the first differentiation index and the second differentiation index.

[0013] Preferably, the comparison based on the first differentiation index and the second differentiation index includes: if the value of the first differentiation index obtained in the first flow state is less than the value of the second differentiation index obtained in the second flow state, then the imbalance is determined to originate from the electrolyte in the storage tank; if the value of the first differentiation index is close to the value of the second differentiation index, then the imbalance is determined to originate from inside the fuel cell stack.

[0014] Preferably, the specific calculation rules for steps c1 and c2 are as follows: by subtracting the stable voltage at the end of relaxation from the first voltage relaxation curve and the second voltage relaxation curve respectively, two change curves are obtained. and Asymmetric response function curve It is generated by calculating the difference in absolute values ​​of two change curves at the same moment, point by point; the differentiation index. By analyzing the asymmetric response function curve It is generated by integration over the entire recording duration T, and its calculation rules are as follows: in, .

[0015] Preferably, before step b of applying the second electrical excitation pulse, the method further includes: retrieving a first excitation thermal effect index from a preset lookup table based on the initial voltage drop value of the first voltage relaxation curve to quantify the instantaneous thermal effect caused by the first electrical excitation pulse; and dynamically calculating and adjusting the magnitude of the second electrical excitation pulse according to the first excitation thermal effect index.

[0016] Preferably, the method further includes: calculating the quadratic residual signal between the two asymmetric response function curves obtained under the first flow state and the second flow state; and extracting the peak occurrence time and the amplitude after stabilization from the quadratic residual signal as fluid state indicators for evaluating the health status of the electrolyte circulation system.

[0017] Preferably, the method further includes: continuously acquiring and recording a series of SOC-differentiated states and a series of health states at different time points to form two time series; calculating the cross-correlation function and the ratio of change rate of the two time series; and matching the cross-correlation function and the ratio of change rate with a preset battery aging mode library to determine the health status evolution trend of the battery module.

[0018] Preferably, before steps a and b of acquiring the first voltage relaxation curve and the second voltage relaxation curve, the method further includes: acquiring the current operating mode of the energy storage converter; determining the corresponding noise spectrum characteristics from a preset noise spectrum library based on the current operating mode; and configuring an adaptive notch filter for the acquisition process of the voltage relaxation curve based on the noise spectrum characteristics.

[0019] Preferably, the method is automatically triggered when the battery module is in an idle state.

[0020] Compared with the prior art, the beneficial effects of the present invention are:

[0021] 1. By applying a first electrical excitation pulse with a first polarity to the battery module and then obtaining a first voltage relaxation curve; subsequently, applying a second electrical excitation pulse with the same magnitude but opposite polarity as the first electrical excitation pulse and then obtaining a second voltage relaxation curve, this series of operations with specific timing and charge relationships constructs an instantaneous and symmetrical electrical disturbance reference system within the battery module itself. As a result, when determining the state based on the difference between the first and second voltage relaxation curves, any common drift introduced by factors such as ambient temperature changes, battery uniformity aging, etc., will be canceled out in the differential processing of the two curves. The establishment of this evaluation method makes it no longer dependent on any historical baseline data or external compensation model, but directly extracts the asymmetric state information of the battery from the internal comparison relationship generated in a complete detection action.

[0022] 2. During the above assessment process, this method further places the sample under two preset flow conditions: one where the electrolyte circulation is stopped and the other where it is in normal circulation. This operation method, which combines a unified electrical detection procedure with different material transport states, results in two state indicators. One mainly reflects the local electrochemical state inside the stack due to the relative restriction of the electrolyte, while the other reflects the overall electrochemical state of the system after sufficient material exchange between the stack and the storage tank. Therefore, by comparing the state indicators obtained under these two different physical conditions, a direct basis is provided for determining whether the physical source of the difference in the state of charge originates from the electrolyte of the entire system or from a local problem in the stack, thus providing clear direction for subsequent maintenance work.

[0023] 3. By continuously acquiring and recording a series of differentiated states of charge and a series of health states determined by the above method at different time points, two parallel state time series are formed. By analyzing the mutual influence patterns between these two time series, such as whether there is a fixed sequential relationship between the accelerated change of values ​​in one series and the inflection point of values ​​in the other series, the intrinsic evolution path of the battery system's health state throughout its entire life cycle is revealed. This transforms the judgment of battery state from the assessment of the current isolated state point to the identification of system aging characteristic patterns, establishing an objective data correlation basis for predictively identifying potential degradation trends and formulating proactive maintenance strategies. Attached Figure Description

[0024] Figure 1 This is a flowchart of the SOC differential state assessment and imbalance source diagnosis process of the present invention;

[0025] Figure 2 This is a diagram showing the co-evolution relationship between the differentiation index and battery health status in this invention.

[0026] Figure 3 This is a diagram of the vanadium redox flow battery system architecture used for SOC differentiation evaluation according to the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described below. Obviously, the described embodiments are only some embodiments of this invention, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0028] This invention provides a method for assessing the state of charge (SOC) of an energy storage module, applicable to scenarios such as vanadium redox flow battery energy storage power stations. In such power stations, the SOC of the positive and negative electrolytes deviates during long-term operation, rendering traditional assessment methods based on open-circuit voltage (OCV) ineffective. This assessment method provides a dynamic detection procedure, mainly consisting of a symmetrical electrical excitation and response acquisition step, an asymmetrical signal extraction and quantization step, and a differential state determination step based on offline calibration relationships. All steps are controlled and executed by the battery management system (BMS). In one embodiment, the assessment method is configured to be executed automatically by the BMS when the battery module is idle, for example, during periods of low grid load each day. Before detection, to reduce interference from the high-frequency switching operation of the power storage converter (PCS) on voltage signal acquisition, the BMS obtains the current operating temperature of the PCS. The system operates in a pattern and retrieves the corresponding noise spectrum characteristics (i.e., the dominant noise frequency and harmonic frequencies) from a pre-established noise spectrum library. Based on this, the BMS configures one or more adaptive notch filters in the digital signal processor of its voltage signal acquisition channel. The center frequency of the notch filter corresponds to the noise frequency, used to filter out specific PCS noise during acquisition, thereby improving the quality of the original signal used for subsequent processing. After completing the configuration of the signal acquisition front-end, the evaluation method begins to execute the symmetrical electrical excitation and response acquisition steps. The BMS controls the PCS to apply a first electrical excitation pulse with a first polarity to the battery module; in this embodiment, the first polarity is the charging polarity. The parameters of this excitation pulse are set to charge at a current of 0.05C for 120 seconds. After the first electrical excitation pulse ends, the BMS records the battery terminal voltage for the following 300 seconds at a sampling frequency of 10Hz, obtaining the first voltage relaxation curve, denoted as... Subsequently, the BMS controls the PCS to apply a second electrical excitation pulse. This pulse has a second polarity opposite to the first, i.e., a discharge polarity, and its magnitude is equal to that of the first electrical excitation pulse, i.e., discharging at a current of 0.05C for 120 seconds. After the second electrical excitation pulse ends, the BMS acquires the second voltage relaxation curve using the same acquisition parameters, denoted as... .

[0029] After acquiring two original voltage relaxation curves, the evaluation method performs asymmetric signal extraction and quantization steps to generate a differentiation index. The calculation rule is as follows: the processor in the BMS subtracts the stable voltage at the end of the acquisition from the first and second voltage relaxation curves respectively to obtain two change curves. and Then, by calculating the difference in absolute values ​​of the two change curves at the same moment point by point, an asymmetric response function curve is generated. The calculation method is as follows Finally, the asymmetric response function curve was analyzed throughout the entire recording time. Integrating within the system generates a real-time differentiation index. The calculation rules are as follows: ,in, It is a differentiation index, and its unit is volt-second; The curve represents an asymmetric response function. The recording duration for voltage relaxation is 300 seconds in this example; the difference index is then obtained. Subsequently, the evaluation method performs the differential status determination step; due to As an indirect quantitative indicator, it needs to be converted into a physically meaningful SOC difference state through an offline calibration to establish a correspondence. This calibration procedure is performed on a test module with the ability to independently adjust the SOC of the positive and negative electrolytes. The steps include: First, setting a known SOC difference percentage. For example, 2%; the second step is to set the ambient temperature, for example, 25°C. After performing the aforementioned detection and calculation process, the result is obtained that... corresponding Value; the third step is to systematically change The values ​​of the ambient temperature and the total ambient temperature are measured repeatedly to obtain a value containing... A database of triplet data is then established. Finally, by fitting the database with data, a correspondence is established between the BMS callable differentiation index and the battery module SOC differentiation state.

[0030] In one specific implementation, to improve detection accuracy, a calibration step is added after acquiring the first voltage relaxation curve and before applying the second electrical excitation pulse; during a 60-second silent period after the first voltage relaxation curve recording ends, the BMS measures and calculates a voltage drift rate characterizing the potential change of the battery module itself. The unit is microvolts per second; in the calculation step of asymmetric signal extraction, the BMS generates a linear compensation vector based on this voltage drift rate. The linear compensation vector is subtracted from the original second voltage relaxation curve to generate a compensated second voltage relaxation curve. This compensated second voltage relaxation curve is then used for subsequent differentiation index calculation. This procedure corrects the calculation bias introduced by the background drift of the battery module itself. To distinguish the physical sources of SOC differentiation, the evaluation method can also perform a flow-modulation-based diagnostic. This diagnostic first performs the aforementioned steps to obtain the first differentiation index under the first flow state where the electrolyte circulation stops. This index primarily reflects the asymmetry within the fuel cell stack; subsequently, under the second flow state of normal electrolyte circulation, the aforementioned steps are repeated to obtain the second differentiation index. This index reflects the asymmetry of the entire battery module; by comparing the two indices, the source of the imbalance can be determined. A specific judgment rule is: if... The value is significantly smaller than The value, for example Then the imbalance is determined to originate from the electrolyte in the storage tank; if The value and The values ​​are close, for example This confirms that the imbalance originates from within the fuel cell stack.

[0031] Example 1: In a commercially operational 10MW / 40MWh vanadium redox flow battery energy storage power station, which performs grid peak shaving tasks, operation and maintenance data shows that the actual dispatchable power of the system consistently deviates from the theoretical value estimated by the battery management system (BMS) based on the traditional OCV-SOC curve by more than 8%. This situation causes the power station to be unable to reliably deliver its declared power when participating in electricity market transactions. To diagnose this problem, a SOC differential state assessment method was executed during a nighttime idle period of the energy storage power station. This method was executed automatically by the BMS. The BMS controlled the PCS to apply a 0.05C charging pulse to the battery module for 120 seconds. After the pulse ended, the first voltage relaxation curve was recorded for the following 300 seconds. Subsequently, a discharge pulse of equal magnitude but opposite polarity was applied, and the second voltage relaxation curve was recorded. Based on these two relaxation curves, the BMS processor calculated an asymmetric response function and performed integration to obtain a differential index. ; this After substituting the values ​​into the correspondence established through offline calibration in the BMS, the SOC difference status output by the system is 6.5%. This result directly correlates the observed power deviation with a quantified internal state imbalance index.

[0032] After confirming the existence of SOC differences in the system, the BMS then executed a decoupling diagnostic procedure based on flow modulation to determine its physical source. First, under the first flow condition where the electrolyte circulation pump was instructed to stop running, the aforementioned symmetrical electrical excitation and response acquisition steps were repeated once to obtain the first difference index. Subsequently, under the second flow condition where the command circulation pump resumes normal operation, this step is performed again to obtain the second differentiation index. The measurement results show that The value is close to zero, while The value corresponds to a 6.5% difference from the previous measurement. The values ​​are at the same level, according to the judgment rules described in the specific implementation method, that is, when The value is significantly smaller than At that time, the imbalance originated from the electrolyte in the storage tank. Based on this, maintenance personnel focused their maintenance efforts on rebalancing the systemic electrolyte, without needing to inspect the fuel cell stack. After completing the maintenance operations targeting the storage tank electrolyte, the assessment method was repeated, and the resulting difference index was calculated. The levels have fallen back to near zero; in subsequent charge-discharge cycles, the deviation between the dispatchable power reported by the BMS of the energy storage power station and the power delivered during the actual discharge process was controlled within 1%, and the power station regained its reliable dispatch capability in the electricity market.

[0033] Example 2: This example aims to quantitatively verify the effectiveness and accuracy of the SOC (State of Charge) differential assessment method. The test platform uses a 5kW rated power, 20kWh all-vanadium redox flow battery test module. The positive and negative electrolyte storage tanks of this module have independently operable sampling and titration ports. The test module is placed in a temperature control environment with an accuracy of ±0.5℃. The test was conducted in an environmental test chamber, controlled by a battery testing device with a current control accuracy of ±0.1% of full scale and a voltage measurement accuracy of ±0.05% of full scale. The data acquisition system had a sampling frequency of 10Hz. A control group and a test group using the method of this invention were set up to compare the evaluation results under different SOC differential states. The initial conditions of the test were as follows: the positive and negative electrolytes of the test module were adjusted to a 50% SOC equilibrium state through chemical titration and charge-discharge operations. Subsequently, a certain amount of reducing agent was added to the negative electrolyte tank, and an equivalent amount of oxidant was added to the positive electrolyte tank. While maintaining a constant total system charge, a series of known, gradient-changing SOC differential states were artificially set. In each preset At the state point, the battery modules were left to rest for 2 hours, and then tests were performed on the control group and the test group respectively. For the control group, after the resting period, the apparent open-circuit voltage of the battery module was measured, and the overall system SOC was calculated using the OCV-SOC curve applicable to the SOC equilibrium state of the module. The estimated deviation between this curve and 50% of the true average SOC value was recorded. For the test group, the evaluation method implemented in the specific implementation method was executed, namely, applying a symmetrical electrical excitation pulse with a charge size of 1 / 600 of the rated capacity and opposite polarity, acquiring two subsequent voltage relaxation curves, and calculating the difference index. All trials were conducted at 25°C. The data were collected at the ambient temperature and are recorded in Table 1.

[0034] Table 1: Comparison of test data under different SOC (State of Charge) conditions

[0035] The test results are shown in Table 1. The data shows that when the preset SOC differentiation state is... As the SOC estimation deviation increased from 0.0% to 10.0%, the OCV method used in the control group consistently showed a deviation within ±0.6%, failing to effectively respond to changes in SOC differentiation within the system. In contrast, the test group using the method of this invention measured a significantly higher differentiation index. Its value varies The increase in V·s shows a monotonically increasing relationship, growing from 0.02 V·s to 3.58 V·s. The experimental results show that the evaluation method can convert the SOC differentiation state inside the battery module into an external measurable index that has a clear correspondence with the degree of differentiation.

[0036] To further verify the necessity and non-obviousness of the evaluation method of the present invention from a reverse perspective, the following comparative examples are provided to illustrate the technical problems that could not be solved by those skilled in the art using conventional technical paths in the background art before the disclosure of the present invention.

[0037] Comparative Example 1: This comparative example aims to illustrate that, on the same test platform as Example 2 and using the same preset SOC differential state for testing, if the symmetrical electrical excitation and asymmetric response analysis method claimed in this invention is not used, but instead the conventional technical path of state assessment based on apparent open-circuit voltage (OCV) is used, the assessment results cannot effectively characterize the SOC differential state inside the battery module. The test platform, test module, environmental conditions, and the setting steps for the preset SOC differential state are all completely consistent with the conditions and procedures described in Example 2. The only difference is that, in this comparative example, the state assessment step is replaced by: in each preset At the state point, after the battery module has been left to stand for 2 hours to reach an electrochemically stable state, the apparent open-circuit voltage across the battery module is measured and recorded using the same high-precision battery testing equipment as in Example 2. Subsequently, based on the test module in the initial SOC equilibrium state (i.e., A standard OCV-SOC correlation curve was calibrated below, and the measured apparent open-circuit voltage value was used to reverse-look up the values ​​to calculate the current overall SOC estimate of the system. Finally, this SOC estimate was compared with the actual average SOC value of the system (which was maintained at 50% throughout the experiment) to obtain the SOC estimation deviation of the OCV method. All experiments were conducted at an ambient temperature of 25℃, and the data are recorded in Table 2.

[0038] Table 2: Test data using the conventional OCV method under different SOC differentiation states

[0039] The test results are shown in Table 2. This data indicates that, with the preset SOC differentiation state... As the value was gradually increased from 0.0% to 10.0%, although the asymmetry state inside the battery module changed significantly, the actual change in the externally measurable apparent open-circuit voltage was extremely small. Therefore, the absolute value of the SOC estimation deviation obtained by the conventional OCV method never exceeded 0.6%. This result objectively shows that the conventional technical approach, which relies entirely on measuring the external apparent voltage, cannot extract characteristic information that can directly characterize the degree of asymmetry inside the system. Therefore, it cannot effectively respond to the actual SOC differences inside the battery module, and it cannot provide a reliable basis for subsequent maintenance work.

[0040] Example 3: This example combines Figures 1 to 3 The method for assessing the state of charge (SOC) differentiation of energy storage modules is explained, such as... Figure 1As shown, the process starts with an idle battery module. First, an adaptive notch filter is configured according to the acquired PCS operating mode to improve signal quality. Then, symmetrical electrical excitation is applied to the module and two voltage relaxation curves are acquired. The difference between the two curves is calculated and integrated to extract and quantify the asymmetric response, generating a differentiation index. This index is ultimately determined as the SOC differentiation output to quantify the degree of imbalance of the battery's internal state of charge based on offline calibration. The core evaluation process can be performed under two preset flow states: evaluation is performed under the first flow state where the electrolyte circulation stops to obtain the first differentiation index, which mainly reflects the internal state of the battery stack; evaluation is performed under the second flow state where the electrolyte circulation is normal to obtain the second differentiation index, which reflects the overall state of the system. Finally, by comparing the numerical relationship between the two indices, if the first index is significantly smaller than the second index, it is determined that the imbalance originates from the electrolyte in the storage tank; if the two index values ​​are close, it is determined that the imbalance originates from inside the battery stack.

[0041] like Figure 2 As shown, the horizontal axis represents the number of days the system is running, and the vertical axis on the left represents the unit (in meters). Differentiation Index The vertical axis on the right represents health status as a percentage. The curve formed by the circular data points connected by solid lines in the figure represents the differentiation index. Its value increases monotonically with the number of operating days, and the curve formed by the triangular data points connected by dashed lines represents a healthy state. The value of the curve decreases monotonically with the increase of the number of operating days. The co-evolution relationship between the two curves reveals the intrinsic connection between the internal imbalance state and the overall health state of the battery system.

[0042] like Figure 3 As shown, the system uses a fuel cell stack as its core. The stack contains a negative electrode chamber and a positive electrode chamber separated by an ion exchange membrane, and these chambers are connected to a negative electrode electrolyte storage tank (containing...) via circulation pipelines. (electrode pair) and positive electrode electrolyte storage tank (containing) (Electrical pairs) are connected, and the operation of the entire system is controlled by one The control system provides unified management, and is responsible for executing electrical excitation and response acquisition, performing SOC differential assessment and fluid health monitoring. At the same time, it achieves the switching between the electrolyte in two states: stopped circulation and normal circulation by controlling the flow state of the circulating pump P.

[0043] Example 4: This example discloses a method for establishing a differentiation index. Difference from actual SOC status A calibration method for determining the quantitative correspondence between them, which is used to solve... To address the issue of numerical values ​​being affected by operating temperature, a calibration model covering a predetermined operating range is established to provide a data foundation for the application of the evaluation method. This calibration method is executed on a test platform with controllable initial conditions. The platform is used with a 5kW / 20kWh vanadium redox flow battery test module, which has ports for independent sampling and chemical titration of the positive and negative electrolytes. The enabling environment is one capable of maintaining an ambient temperature of -10°C. Up to 40 The test chamber is a temperature-controlled test chamber that can be set and maintained within a certain range, and a battery testing device and data recording system that is integrated with the BMS to be calibrated and has a voltage measurement accuracy of 0.05%.

[0044] The calibration process uses a nested loop test sequence, with the outer loop being a temperature scan. First, the temperature of the temperature-controlled test chamber is set to -10°C. And wait until the temperature difference between the battery module and the set temperature of the test chamber is less than 0.5°C. Within a certain range and remaining stable for 30 minutes; at each temperature point, perform an internal cycle scan of the SOC differential state, the steps of which are as follows: first adjust the battery module to... In the equilibrium state, the symmetrical electrical excitation and response acquisition process described in this specific implementation is executed once, and the reference temperature is recorded. Value; then, through chemical titration, a known value is established. For example, a value of 1.0%. After the system stabilizes, repeat the same process and record the corresponding value. Value; this step uses a step size of 1.0%, in Repeat the process from 1.0% to 10.0%; once all values ​​at a given temperature point have been achieved... After the test, the external circulation will set the temperature to the next test point, such as 0. The entire internal loop is repeated until all preset temperature points have been tested; after completing the above test sequence, a result containing ( , , To transform the original database of triples into a mapping that can be directly accessed by the BMS, this embodiment uses polynomial surface fitting. The least squares method is used to fit these data points into a bivariate function model. , where P is about and The bivariate polynomial; ultimately, the coefficients of this function are stored in the non-volatile memory of the BMS, which, during actual operation, adjusts the value based on real-time measurements. By calculating the battery module temperature and other parameters, the compensated SOC differential state can be obtained by calling this function.

[0045] Example 5: To establish the data benchmark required for the evaluation method, after the initial deployment or replacement of the energy storage converter (PCS) in the energy storage system, the system executes a set of self-learning and calibration procedures during the commissioning phase. The BMS is placed in a noise self-learning mode. In this mode, the BMS controls the PCS to traverse its typical operating modes according to a preset sequence, including charging and discharging at 25%, 50%, 75%, and 100% of rated power, as well as standby mode. In each operating mode, the BMS performs a fast Fourier transform on the acquired voltage signal to identify the main frequency and harmonic frequencies of the noise signal introduced by the PCS switching operation, and stores the data containing the operating mode and the corresponding noise spectrum characteristics in a lookup table inside the BMS to establish a noise spectrum library.

[0046] After the aforementioned calibration is completed, the procedure calibrates the health baseline of the electrolyte circulation system. When the BMS confirms that the electrolyte pump system is in a healthy state, it establishes a primary fluid response baseline. This process controls the pump system to start from a stopped state and reach normal circulation flow rate. During this period, the system performs parallel symmetrical electrical excitation and response acquisition steps under the first and second flow states, and calculates the quadratic residual signal between the two asymmetric response function curves. The BMS extracts the peak occurrence time from this quadratic residual signal. The amplitude after reaching steady state This set of characteristic parameters is stored in the BMS as a fluid state index benchmark for comparison of the pump system state during subsequent operation.

[0047] Example 6: This example discloses a method for determining the health status evolution trend of a battery module. This method is applied in a scenario of full lifecycle management of energy storage assets. In this scenario, a core technical challenge is to distinguish between normal aging of the battery module and accelerated degradation caused by specific physicochemical mechanisms, in order to facilitate predictive maintenance. To address this challenge, after each execution of the aforementioned SOC differentiated state assessment, the BMS obtains the data points, i.e., {current timestamp}. Differentiation Index Health status indicators Continuous recording is carried out, including... These metrics are obtained through standard capacity testing or internal resistance measurement, thereby forming two parallel state-time series in the cloud or local database. and .

[0048] To perform correlation analysis on these two time series, the system has a built-in library of preset battery aging patterns. This library stores various typical battery aging paradigms, each defined by a set of quantified co-evolutionary features. For example, mode one, i.e., balanced aging, is characterized by... The rate of descent is approximately a small constant, while The value of [something] has remained stable at a low level for a long period; Mode 2, namely, imbalance accelerates aging, is characterized by [something]. The value reached an inflection point and continued to increase, and rate of change and The cross-correlation function of the rate of change over a specific time delay A significant negative correlation peak appears; the analysis algorithm is configured to periodically, such as every 30 days, analyze the data within the most recent time window. and The data is processed to calculate their respective rates of change and the cross-correlation function between them. The calculated co-evolutionary features are then matched with features in the aging pattern library. When the matching degree between the evolutionary features of a battery module and the second pattern exceeds a preset threshold, the system determines that the battery module has entered an unbalanced accelerated aging evolutionary trend and generates a corresponding maintenance warning.

[0049] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0050] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for SOC differentiated state assessment of an energy storage module, characterized by, The method comprises: Step a, applying a first electrical excitation pulse with a first polarity to the battery module, and obtaining a first voltage relaxation curve after the first electrical excitation pulse ends; Step b, applying a second electrical excitation pulse with a second polarity opposite to the first polarity to the battery module, wherein the second electrical excitation pulse has the same amount of electricity as the first electrical excitation pulse, and obtaining a second voltage relaxation curve after the second electrical excitation pulse ends; Step c, calculating a real-time differentiation index according to the difference between the first voltage relaxation curve and the second voltage relaxation curve by the following steps: step c1, calculating an asymmetric response function curve between the two voltage relaxation curves; and step c2, integrating the asymmetric response function curve to generate the real-time differentiation index; Step d, determining the SOC differentiation state of the battery module based on a corresponding relationship between a calibration differentiation index established through offline calibration and the SOC differentiation state of the battery module, and using the real-time differentiation index calculated in step c as an input value of the calibration differentiation index.

2. The method of claim 1, wherein, After step a of obtaining the first voltage relaxation curve and before step b of applying the second electrical excitation pulse, the method further comprises: measuring and obtaining a voltage drift rate representing the change of the battery module potential in a silent period; and in step c, first generating a linear compensation vector according to the voltage drift rate, and subtracting the linear compensation vector from the second voltage relaxation curve to generate a compensated second voltage relaxation curve, and then using the first voltage relaxation curve and the compensated second voltage relaxation curve for subsequent calculation.

3. The method of claim 1, wherein, The method further comprises: In a first flow state where the electrolyte stops circulating, steps a to d are performed to obtain a first differentiation index; And in a second flow state where the electrolyte normally circulates, steps a to d are performed again to obtain a second differentiation index; wherein the SOC differentiation state is further determined based on the comparison between the first differentiation index and the second differentiation index.

4. The method of claim 3, wherein the SOC of the energy storage module is determined by the following equation: ###0001### wherein, V is the voltage of the energy storage module, C is the capacitance of the energy storage module, I is the current of the energy storage module, and T is the time. The comparison between the first differentiation index and the second differentiation index comprises: if the value of the first differentiation index obtained in the first flow state is smaller than the value of the second differentiation index obtained in the second flow state, it is determined that the imbalance is caused by the tank electrolyte; and if the value of the first differentiation index is close to the value of the second differentiation index, it is determined that the imbalance is caused by the internal stack.

5. The method of claim 1, wherein, The specific calculation rules of step c1 and step c2 are: by subtracting the stable voltage at the end of relaxation of each from the first voltage relaxation curve and the second voltage relaxation curve respectively, two change amount curves are obtained With ; the asymmetric response function curve Generated by point-by-point calculation of the absolute value difference of the two change amount curves at the same time; the differentiation index Generated by the asymmetric response function curve Generated by integral operation in the entire recording duration T, and the calculation rule is: , wherein .

6. The method of claim 1, wherein, Before step b of applying the second electrical excitation pulse, the method further comprises: based on the initial voltage drop value of the first voltage relaxation curve, inversely deriving a first excitation thermal effect index quantifying the instantaneous thermal effect caused by the first electrical excitation pulse from a pre-set lookup table; and dynamically calculating and adjusting the amount of electricity of the second electrical excitation pulse according to the first excitation thermal effect index.

7. The method of claim 3, wherein the SOC of the energy storage module is determined by the following equation: ###0001### wherein, V is the voltage of the energy storage module, I is the current of the energy storage module, C is the capacitance of the energy storage module, and T is the time. The method further comprises: Calculating a quadratic residual signal between the two asymmetric response function curves obtained in the first flow state and the second flow state; And extracting the peak occurrence time and the stable amplitude from the quadratic residual signal as a fluid state index for evaluating the health state of the electrolyte circulation system.

8. The method of claim 1, wherein, The method further comprises: At different time points, a series of SOC differentiated states and a series of health states are continuously acquired and recorded to form two time series; a correlation function and a change rate of the two time series are calculated; and a preset battery aging mode library is matched based on the proportion relationship between the correlation function and the change rate.

9. The method for assessing the state of charge (SOC) differentiation of an energy storage module according to claim 1, characterized in that, Before the steps a and b of acquiring the first voltage relaxation curve and the second voltage relaxation curve, the method further comprises: acquiring a current working mode of the energy storage converter; and determining a corresponding noise spectrum feature from a preset noise spectrum library according to the current working mode.

Images