A data processing-based hybrid energy storage battery state monitoring method and system

Abstract

This invention relates to the field of battery state monitoring technology, and more particularly to a method and system for monitoring the state of hybrid energy storage batteries based on data processing. The method includes the following steps: acquiring impedance measurement data of the hybrid energy storage battery pack at a preset frequency sequence; performing temperature correction processing on the impedance measurement data to obtain corrected impedance data under standard temperature conditions; establishing a health state impedance benchmark value for each energy storage unit based on the corrected impedance data, thus obtaining an impedance benchmark database; calculating the impedance difference value sequence between adjacent frequency points for each energy storage unit using the corrected impedance data; dividing the impedance difference value sequence by the characteristic impedance value of the corresponding energy storage unit in the impedance benchmark database and performing normalization processing to obtain a differential impedance characteristic vector. This invention establishes an accurate temperature compensation mechanism, eliminating the influence of ambient temperature changes on impedance measurement results, and ensuring the consistency and comparability of monitoring results under different operating conditions.

Description

Technical Field

[0001] This invention relates to the field of battery status monitoring technology, and in particular to a method and system for monitoring the status of hybrid energy storage batteries based on data processing. Background Technology

[0002] Hybrid energy storage battery state monitoring technology is a key technology for real-time tracking and management of the state of different energy storage units (such as lithium batteries, supercapacitors, or lead-acid batteries) in a hybrid energy storage system. Its core purpose is to ensure the efficient and safe operation of the energy storage system, while extending battery life and optimizing overall performance. Hybrid energy storage systems combine the advantages of different energy storage units to achieve a balance between energy density and power density. Lithium batteries provide high energy density to meet long-term power supply needs, while high-power-density units such as supercapacitors are suitable for short-term power compensation. However, monitoring methods based on voltage and current lack the ability to deeply perceive changes in the internal state of the battery. Voltage and current only reflect the surface characteristics of the battery and cannot accurately identify key fault precursors such as changes in internal impedance, electrolyte aging, and electrode material degradation. Especially in hybrid energy storage systems, the voltage plateaus and internal resistance characteristics of different types of energy storage units vary greatly, causing fault judgment methods based on a uniform voltage threshold to frequently result in misjudgments or omissions. Traditional monitoring methods lack consideration for the mutual influence between energy storage units. In actual hybrid energy storage systems, a fault in one energy storage unit often affects the operating state of adjacent units through electrical connection paths, and may even trigger a cascading failure. Existing methods typically monitor each energy storage unit as an independent entity, making it impossible to predict and assess the risk of fault propagation. Summary of the Invention

[0003] Therefore, it is necessary for the present invention to provide a hybrid energy storage battery status monitoring method and system based on data processing to solve at least one of the above-mentioned technical problems.

[0004] To achieve the above objectives, a data processing-based method for monitoring the state of hybrid energy storage batteries includes the following steps:

[0005] Step S1: Collect impedance measurement data of the hybrid energy storage battery pack at a preset frequency sequence; perform temperature correction processing on the impedance measurement data to obtain corrected impedance data; establish an impedance reference database based on the corrected impedance data;

[0006] Step S2: Calculate the impedance difference value sequence of each energy storage unit between adjacent frequency points; normalize the impedance difference value sequence to obtain the differential impedance characteristic vector; calculate the state deviation metric based on the deviation between the differential impedance characteristic vector and the reference vector.

[0007] Step S3: Identify abnormal energy storage units whose state deviation metric exceeds a preset threshold; perform frequency range analysis on the differential impedance characteristic vector of the abnormal energy storage unit, determine the fault type identifier based on the frequency distribution pattern, and then calculate the fault severity score.

[0008] Step S4: Rank the risk levels according to the fault severity score; establish a fault propagation path prediction model based on the risk level and the electrical connection relationship between energy storage units; use the fault propagation path prediction model to predict the fault risk index;

[0009] Step S5: When the fault risk index exceeds the safety threshold, output warning information including the location of the abnormal energy storage unit, fault type identifier and fault severity score.

[0010] The present invention also provides a hybrid energy storage battery state monitoring system based on data processing, used to execute the hybrid energy storage battery state monitoring method based on data processing as described above, wherein the hybrid energy storage battery state monitoring system based on data processing includes:

[0011] The impedance reference construction module is used to collect impedance measurement data of hybrid energy storage battery packs at a preset frequency sequence; perform temperature correction processing on the impedance measurement data to obtain corrected impedance data; and establish an impedance reference database based on the corrected impedance data.

[0012] The feature vector calculation module is used to calculate the impedance difference value sequence of each energy storage unit between adjacent frequency points; normalize the impedance difference value sequence to obtain the differential impedance feature vector; and calculate the state deviation metric based on the deviation between the differential impedance feature vector and the reference vector.

[0013] The fault identification and scoring module is used to identify abnormal energy storage units whose state deviation metric exceeds a preset threshold; to perform frequency range analysis on the differential impedance characteristic vector of the abnormal energy storage unit to determine the fault type identifier; and to calculate the fault severity score.

[0014] The risk prediction and ranking module is used to rank risk levels according to fault severity scores; establish a fault propagation path prediction model based on risk levels and electrical connection relationships between energy storage units; and use the fault propagation path prediction model to predict the fault risk index.

[0015] The fault warning output module is used to output warning information including the location of the abnormal energy storage unit, fault type identifier, and fault severity score when the fault risk index exceeds the safety threshold.

[0016] The hybrid energy storage battery status monitoring method based on data processing provided by this invention systematically collects impedance measurement data of the energy storage battery pack at a preset frequency sequence, performs temperature correction, and constructs a benchmark database to achieve accurate assessment of the health status of each energy storage unit. By calculating and normalizing the impedance difference sequence between adjacent frequency points for each energy storage unit, a differential impedance feature vector is formed. This method can accurately reflect the subtle deviations in unit impedance with frequency changes, thereby quantifying the degree of state deviation of the energy storage unit. Based on this, the system can identify abnormal energy storage units and, combined with the deviation distribution patterns in low, medium, and high frequency regions, automatically determine the fault type identifier and calculate the fault severity score, making fault diagnosis more refined and quantifiable.

[0017] Furthermore, by ranking energy storage units according to their fault severity scores and constructing a fault propagation path prediction model based on the series and parallel connections between these units, the system can predict potential fault propagation paths and the overall risk index of the battery pack, providing maintenance personnel with intuitive risk assessment results. When the risk index exceeds a preset safety threshold, the system automatically generates standardized early warning data packets, including the location of the abnormal energy storage unit, fault type, and severity score, and sends them in real time to the monitoring terminal or mobile device via a communication interface, enabling rapid response and precise location. This method not only enables real-time monitoring and dynamic assessment of the hybrid energy storage battery pack's status but also allows for risk prediction before anomalies occur, assisting maintenance personnel in developing scientific maintenance and replacement strategies and improving the safety and reliability of the battery pack's operation. In addition, by constructing a temperature correction and impedance benchmark database, the influence of ambient temperature on measurement results can be effectively eliminated, improving the accuracy and consistency of monitoring data. The use of differential impedance eigenvectors and Euclidean distance to calculate state deviation metrics makes the identification of abnormal energy storage units more sensitive and accurate. Through risk level classification and propagation path modeling, quantitative management of fault risks is achieved. Overall, this method organically combines energy storage battery status monitoring, fault identification, risk assessment, and early warning output to form a complete closed-loop monitoring system from data acquisition and processing to early warning, which can effectively improve the stability, operating efficiency, and operation and maintenance management level of energy storage systems. Attached Figure Description

[0018] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0019] Figure 1 This is a flowchart illustrating the steps of a hybrid energy storage battery state monitoring method based on data processing according to the present invention.

[0020] Figure 2 This is a schematic diagram of a hybrid energy storage battery pack status monitoring system according to an embodiment of the present invention;

[0021] Figure 3 This is a comparison chart of impedance spectrum analysis of an energy storage unit according to an embodiment of the present invention. Detailed Implementation

[0022] The technical method of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0023] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0024] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0025] To achieve the above objectives, please refer to Figures 1 to 3 This invention provides a method for monitoring the state of hybrid energy storage batteries based on data processing, the method comprising the following steps:

[0026] Step S1: Collect impedance measurement data of the hybrid energy storage battery pack at a preset frequency sequence; perform temperature correction processing on the impedance measurement data to obtain corrected impedance data; establish an impedance reference database based on the corrected impedance data;

[0027] Step S2: Calculate the impedance difference value sequence of each energy storage unit between adjacent frequency points; normalize the impedance difference value sequence to obtain the differential impedance characteristic vector; calculate the state deviation metric based on the deviation between the differential impedance characteristic vector and the reference vector.

[0028] Step S3: Identify abnormal energy storage units whose state deviation metric exceeds a preset threshold; perform frequency range analysis on the differential impedance characteristic vector of the abnormal energy storage unit, determine the fault type identifier based on the frequency distribution pattern, and then calculate the fault severity score.

[0029] Step S4: Rank the risk levels according to the fault severity score; establish a fault propagation path prediction model based on the risk level and the electrical connection relationship between energy storage units; use the fault propagation path prediction model to predict the fault risk index;

[0030] Step S5: When the fault risk index exceeds the safety threshold, output warning information including the location of the abnormal energy storage unit, fault type identifier and fault severity score.

[0031] See Figure 2 This invention provides a schematic diagram of a hybrid energy storage battery pack status monitoring system.

[0032] In one embodiment, the system includes multiple energy storage units B01 to B24. The system monitors the status of each energy storage unit through impedance measurement and data processing. Each energy storage unit is classified into a normal unit, a warning unit, or an abnormal unit based on real-time measurement results, and is labeled with the corresponding fault type, such as abnormal temperature, abnormal impedance, or abnormal discharge. By analyzing the electrical connections between energy storage units, the system establishes fault propagation paths and predicts potential fault transmission between high-risk energy storage units and their adjacent units, thereby achieving an assessment of the overall battery pack's operational risk.

[0033] For example, when energy storage units B04, B09, and B10 are identified as abnormal units, the system can automatically calculate the fault propagation path based on the connection diagram, such as B04→B10→B15, and assign corresponding propagation weights to the path for subsequent risk index calculation.

[0034] Further, step S1 includes the following steps:

[0035] Step S11: Set the measurement interval and scanning order of the frequency sequence; assign a unique number to each energy storage unit in the hybrid energy storage battery pack, and record the physical location corresponding to each number;

[0036] In one embodiment, based on the type, capacity, and target monitoring accuracy of the hybrid energy storage battery pack, frequency sequence parameters for impedance measurement are pre-set, including the measurement interval between adjacent frequency points and the overall scanning sequence. The frequency sequence can cover a low-frequency to high-frequency range to ensure that impedance characteristics corresponding to different physical processes are collected. Simultaneously, each energy storage unit in the hybrid energy storage battery pack is assigned a unique identification number, and this number is correlated with the installation location of the energy storage unit within the battery pack, thus forming a one-to-one mapping between the number and the physical location.

[0037] For example, for an energy storage battery pack consisting of multiple series-parallel battery modules, the battery packs can be numbered sequentially according to their physical arrangement, and the rack position, level, or slot information corresponding to each number can be recorded.

[0038] Step S12: Apply AC test signals to each energy storage unit one by one according to the frequency sequence, record the impedance response values ​​at each frequency point, and record the ambient temperature during the measurement.

[0039] In one embodiment, AC test signals are applied to each energy storage unit in the hybrid energy storage battery pack sequentially according to a frequency sequence, and corresponding impedance response data is collected at each frequency point. Simultaneously with the impedance response acquisition, temperature information under the current measurement environment is recorded. This ambient temperature can be obtained through temperature sensors deployed inside or near the battery pack. By acquiring impedance and temperature data within the same time window, temporal consistency between the two is ensured.

[0040] For example, when performing an impedance scan on a certain energy storage unit, the frequency can be gradually switched from low frequency to high frequency according to a preset scanning sequence. After stabilizing at each frequency point, the impedance response value is recorded, and the ambient temperature value at the corresponding time is read and stored.

[0041] Step S13: Calculate the temperature correction factor based on the difference between the ambient temperature and the preset standard reference temperature.

[0042] In one embodiment, the collected ambient temperature data is compared with a pre-set standard reference temperature in the system to calculate a temperature correction factor for subsequent corrections. This temperature correction factor characterizes the degree of temperature deviation of the current measurement environment relative to the standard reference state, providing a basis for the comparability of impedance data under uniform temperature conditions.

[0043] For example, when the ambient temperature is higher than the standard reference temperature, a corresponding correction factor can be calculated according to the preset temperature correction rule to offset the effect of temperature rise on impedance measurement results.

[0044] It should be noted that the calculation method of the temperature correction coefficient can be set according to the actual application requirements. This case does not limit its specific mathematical form, but emphasizes that the temperature deviation is quantified and used for subsequent processing.

[0045] Furthermore, step S1 also includes the following steps:

[0046] Step S14: Correct the impedance measurement data using a temperature correction factor to obtain corrected impedance data under standard temperature conditions;

[0047] In one embodiment, a temperature correction factor is used to uniformly correct the impedance measurement data, thereby obtaining corrected impedance data under standard temperature conditions. This correction process allows impedance data acquired at different times and under different environmental conditions to be compared and analyzed at the same reference temperature.

[0048] For example, the raw impedance data of a certain energy storage unit collected at multiple frequency points can be processed one by one with the corresponding temperature correction coefficient to generate a set of corrected impedance data that reflects the impedance characteristics under standard temperature conditions.

[0049] It should be noted that this correction process does not change the relative relationship between impedance data and frequency; its purpose is to eliminate systematic biases introduced by temperature factors.

[0050] Step S15: Calculate the average impedance and variability of each energy storage unit at each frequency point based on the corrected impedance data; use the average impedance as the health status benchmark value, establish a correspondence between the energy storage unit number and the frequency point, and construct an impedance benchmark database.

[0051] In one embodiment, statistical analysis is performed on the corrected impedance data to calculate the average impedance and corresponding variability of each energy storage unit at each frequency point, which is used to characterize the central tendency and dispersion of the impedance characteristics. Subsequently, the average impedance is used as a reference impedance value to characterize the health status of the energy storage unit, and an impedance reference database is constructed by establishing a correspondence between the energy storage unit number and the frequency point.

[0052] For example, during the initial operation of the system or when the battery pack is confirmed to be in a healthy state, the average impedance value at each frequency point can be calculated based on multiple measurement results and stored as a reference benchmark for subsequent state assessment.

[0053] It should be noted that the impedance reference database is used to describe typical impedance characteristics under healthy conditions, and its contents can be updated as the system runs and data accumulates to improve the accuracy of condition monitoring.

[0054] Furthermore, the process of calculating the impedance difference value sequence in step S2 is as follows:

[0055] The corrected impedance data are grouped by energy storage unit number, and within each group, they are sorted by frequency from low to high.

[0056] Within each group, adjacent frequency points are paired up, and the frequency value and corresponding correction impedance value of each pair of adjacent frequency points are recorded.

[0057] For each pair of adjacent frequency points, the impedance difference in that frequency range is obtained by subtracting the impedance value of the low frequency point from the impedance value of the high frequency point.

[0058] The impedance differences of each energy storage unit in each frequency range are arranged in order to form the impedance difference sequence of the energy storage unit.

[0059] In one embodiment, the obtained corrected impedance data is first grouped according to the unique identifier of the energy storage unit, so that each group corresponds to a specific energy storage unit. Then, within each group, the corrected impedance data is sorted in ascending order of frequency value to ensure continuity of the impedance data in the frequency dimension. After sorting, adjacent frequency points of the same energy storage unit are paired sequentially, and the frequency range corresponding to each pair of adjacent frequency points and the corresponding corrected impedance value are recorded.

[0060] After pairing adjacent frequency points, an impedance differential operation is performed on each pair of frequency points. This involves subtracting the corrected impedance value corresponding to the low frequency point from the corrected impedance value corresponding to the high frequency point, thereby obtaining the impedance difference value reflecting the impedance variation amplitude within that frequency range. Subsequently, the impedance differences obtained by the same energy storage unit in each adjacent frequency range are arranged sequentially according to the frequency range order to form the impedance difference value sequence of the energy storage unit.

[0061] For example, when a certain energy storage unit obtains corrected impedance data at multiple discrete frequency points, these data can be arranged in ascending order of frequency, and the impedance change between adjacent frequency points can be calculated sequentially to obtain a complete impedance difference value sequence.

[0062] It should be noted that this differential processing emphasizes the local trend characteristics of impedance variation with frequency, rather than the absolute impedance magnitude at a single frequency point.

[0063] Furthermore, the normalization process in step S2 is as follows:

[0064] Based on the frequency range information in the impedance difference value sequence, the reference impedance value of each energy storage unit in the corresponding frequency range is found from the impedance reference database.

[0065] The normalized difference value is obtained by dividing the difference in each frequency range in the impedance difference value sequence by the reference impedance value corresponding to that frequency range.

[0066] Arrange the normalized differential values ​​of each energy storage unit into a vector according to the frequency range; this is the differential impedance characteristic vector.

[0067] In one embodiment, after obtaining the impedance difference value sequence of each energy storage unit, a reference impedance value for the corresponding energy storage unit in the corresponding frequency range is retrieved from a pre-constructed impedance reference database based on the correspondence between frequency ranges in the difference value sequence. The reference impedance value is used to characterize the typical impedance level of that frequency range under healthy conditions. Subsequently, the impedance difference value of each frequency range in the impedance difference value sequence is proportionalized to its corresponding reference impedance value to obtain a normalized difference value.

[0068] By performing uniform normalization on the impedance differences across frequency ranges, the differences in impedance amplitude among different energy storage units can be reduced, allowing the differential results to focus more on the impedance variation patterns themselves. After normalization, the normalized differential values ​​obtained for each energy storage unit in different frequency ranges are arranged in order of frequency range to form a differential impedance eigenvector for subsequent analysis.

[0069] For example, when there are significant differences in the impedance variation amplitude of a certain energy storage unit in the low-frequency and high-frequency ranges, normalization can make the difference results in different frequency ranges within a comparable order of magnitude.

[0070] It should be noted that the normalization process does not change the relative trend of the impedance difference value; its purpose is to enhance the comparability of the eigenvectors between different energy storage units.

[0071] Furthermore, the process of calculating the state deviation metric in step S2 is as follows:

[0072] Extract the reference differential impedance vector of each energy storage unit under healthy conditions from the impedance reference database;

[0073] The deviation vector is obtained by subtracting the corresponding elements of the differential impedance characteristic vector obtained from the current measurement from the baseline differential impedance vector of the healthy state one by one.

[0074] Squaring each element in the deviation vector and summing them, then taking the square root of the sum, yields the Euclidean distance value, which is the state deviation metric.

[0075] In one embodiment, a reference differential impedance vector corresponding to each energy storage unit in a healthy state is extracted from an impedance reference database. This reference differential impedance vector is used to describe the typical distribution of impedance differential characteristics in each frequency range under healthy conditions. Subsequently, the currently measured differential impedance characteristic vector is compared item by item with the corresponding healthy state reference differential impedance vector. The difference between the two at the same frequency range position is calculated to obtain a deviation vector reflecting the degree of deviation in each frequency range.

[0076] After obtaining the deviation vector, the degree of deviation of each element in the deviation vector is comprehensively processed. By squaring the deviation of each element, summing them, and taking the square root, a single value is obtained to represent the overall degree of deviation of the current energy storage unit impedance characteristics from the healthy state. This value is the state deviation metric.

[0077] For example, when the impedance differential characteristics of a certain energy storage unit deviate significantly from the healthy state reference characteristics in multiple frequency ranges, its calculated state deviation metric will increase accordingly, thus being identified as a potential abnormal unit in subsequent steps.

[0078] It should be noted that the state deviation metric is a comprehensive indicator, and its magnitude reflects the degree of overall characteristic deviation, rather than local anomalies in a single frequency range.

[0079] Further, step S3 includes the following steps:

[0080] Step S31: Set three threshold parameters: normal threshold, warning threshold, and abnormal threshold;

[0081] In one embodiment, to classify the operating status of an energy storage unit, three threshold parameters—normal threshold, warning threshold, and abnormal threshold—are pre-set to divide the deviation values ​​of the state into intervals. These threshold parameters can be set based on historical operating data, statistical results of health status samples, or engineering experience, forming multi-level threshold intervals in ascending order. The normal threshold characterizes the deviation range of the energy storage unit from a healthy operating state, the warning threshold indicates the state has begun to deviate but has not yet reached a serious abnormality, and the abnormal threshold indicates that the energy storage unit has a significant risk of abnormality.

[0082] For example, during the system initialization phase, the distribution of state deviation metrics of multiple healthy energy storage units can be statistically analyzed, and their upper limit can be used as the normal threshold. Then, higher warning thresholds and abnormal thresholds can be set based on this.

[0083] Step S32: Compare the state deviation metric of each energy storage unit with three threshold parameters, and select energy storage units that exceed the abnormal threshold as abnormal energy storage units;

[0084] In one embodiment, after obtaining the state deviation metric value corresponding to each energy storage unit, the state deviation metric value is compared with a normal threshold, a warning threshold, and an abnormal threshold to determine the current state level of each energy storage unit. When the state deviation metric value of a certain energy storage unit exceeds the abnormal threshold, the energy storage unit is identified as an abnormal energy storage unit, and its corresponding energy storage unit number is recorded to form an abnormal energy storage unit set.

[0085] This method allows energy storage units whose condition deviates significantly from the health benchmark to be screened out from the overall battery pack, providing a basis for subsequent fault type identification and severity assessment.

[0086] For example, when the state of an energy storage unit deviates significantly from the metric value above the anomaly threshold, the energy storage unit will be directly marked as an abnormal energy storage unit and enter the subsequent analysis process.

[0087] Step S33: Extract the differential impedance characteristic vector of the abnormal energy storage unit, analyze the deviation distribution of the vector in the low frequency range, medium frequency range and high frequency range, and determine the fault type identifier based on the frequency range where the deviation is mainly concentrated.

[0088] In one embodiment, for the identified abnormal energy storage unit, its corresponding differential impedance feature vector is extracted, and the deviation distribution of the feature vector in different frequency ranges is analyzed. Specifically, the differential impedance feature vector is divided into low-frequency, mid-frequency, and high-frequency ranges according to a preset frequency division rule, and the concentration of feature deviations in each frequency range is statistically analyzed.

[0089] By comparing the distribution of deviation amplitude or number of deviations in each frequency range, the frequency range in which abnormal deviations mainly occur is determined, and the abnormal mode corresponding to the frequency range is used as the fault type identifier of the abnormal energy storage unit.

[0090] For example, if the differential impedance characteristics of an abnormal energy storage unit show obvious deviation concentration in the low-frequency range, while the deviation is smaller in the mid-to-high frequency range, then the abnormality can be identified as a fault type dominated by low-frequency characteristics.

[0091] It should be noted that the frequency range is divided according to a preset rule, the purpose of which is to enhance the distinguishability of different types of anomalies at the feature level.

[0092] Step S34: Obtain the basic score weight of this type of fault from the preset scoring standard table according to the fault type identifier;

[0093] In one embodiment, after obtaining the fault type identifier corresponding to the abnormal energy storage unit, the corresponding basic score weight is extracted from a pre-established fault severity scoring standard table based on the fault type identifier. The scoring standard table presets different basic score weights for different fault types to reflect the degree of impact of different types of faults on the operational safety and reliability of the energy storage unit.

[0094] By introducing basic scoring weights, the risk levels of different fault types can be distinguished in subsequent scoring calculations, making the scoring results more in line with the needs of actual engineering applications.

[0095] For example, when a certain fault type is defined as a high-risk type in the scoring criteria table, its corresponding basic score weight will be higher than that of a general fault type.

[0096] Step S35: Multiply the state deviation metric by the base score weight and weight it according to the number of abnormal frequency bands to obtain the fault severity score.

[0097] In one embodiment, the severity of faults in abnormal energy storage units is quantitatively calculated based on state deviation metrics and base score weights. Specifically, the state deviation metric of the abnormal energy storage unit is multiplied by the base score weight of its corresponding fault type, and weighted by the number of frequency intervals involved in the abnormal deviation, thereby obtaining a fault severity score that comprehensively reflects the impact of the degree of abnormality and the fault type.

[0098] In this way, the severity score can simultaneously reflect multiple factors such as the magnitude of the abnormal deviation, the fault type attribute, and the scope of the abnormal impact.

[0099] For example, when a certain abnormal energy storage unit not only has a high state deviation metric value, but the abnormal deviation is distributed in multiple frequency ranges, the final calculated fault severity score will be increased accordingly.

[0100] Further, step S4 includes the following steps:

[0101] Step S41: Sort all energy storage units from highest to lowest according to their fault severity scores, and record the number, score and ranking of each energy storage unit;

[0102] In one embodiment, to clarify the risk priority of each energy storage unit, the fault severity score calculated in step S3 is used as the basis for sorting all energy storage units. Specifically, the number, fault severity score, and ranking of each energy storage unit are recorded in a data table and arranged from high to low scores, so that energy storage units with higher scores are ranked higher in the sorting list, thereby intuitively reflecting the risk level of each energy storage unit.

[0103] For example, in a battery pack containing 50 energy storage units, there may be an energy storage unit numbered 12 with a fault severity score of 0.85 and an energy storage unit numbered 7 with a score of 0.72. In this case, number 12 is ranked first and number 7 is ranked second.

[0104] It should be noted that the sorting operation is a data processing behavior, the purpose of which is to provide a basis for subsequent risk level classification and fault propagation path analysis. The actual state of the energy storage unit is not changed during the sorting process.

[0105] Step S42: Divide the energy storage units into three levels: high risk, medium risk, and low risk, based on the sorting results;

[0106] In one embodiment, energy storage units are divided into three levels—high-risk, medium-risk, and low-risk—based on the ranking results. The specific classification method can be based on a preset percentage rule, for example, classifying the top 20% of energy storage units as high-risk, the next 30% as medium-risk, and the remaining 50% as low-risk.

[0107] For example, if a battery pack has 50 energy storage units, the top 10 units with the highest scores are classified as high-risk, the middle 15 units as medium-risk, and the remaining 25 units as low-risk.

[0108] It should be noted that the risk level classification ratio and rules can be adjusted according to the battery pack size, system safety requirements, or engineering experience, and are not fixed values.

[0109] Step S43: Based on the series and parallel connection relationship of energy storage units in the battery pack, establish a connection diagram between adjacent energy storage units; combine the risk level to assign propagation weights to the connection paths between high-risk energy storage units and their adjacent units, and construct a fault propagation path prediction model.

[0110] In one embodiment, a connection graph is established based on the series and parallel connections of energy storage units within the battery pack to simulate the potential propagation of faults along electrical connection paths. Nodes in the connection graph represent energy storage units, and edges represent direct electrical connections between them. Simultaneously, propagation weights are assigned to the connection paths between high-risk energy storage units and their adjacent units; these weights reflect the likelihood of a fault propagating from one energy storage unit to a neighboring unit. Combining these weights and node risk levels, a complete fault propagation path prediction model is constructed.

[0111] For example, for three sets of energy storage units connected in series, the high-risk energy storage units are numbered 5, 6, and 7. In the connection diagram, the edge weights of 5 and 6, and 6 and 7 can be set to high values, indicating that the fault may propagate rapidly along this path.

[0112] Step S44: For each propagation path, calculate the propagation attenuation coefficient based on the risk level of the starting unit and the propagation distance; multiply the path weight by the propagation attenuation coefficient to obtain the propagation probability value of that path;

[0113] In one embodiment, for each established fault propagation path, a propagation attenuation coefficient is calculated based on the risk level of the starting unit and the path length. Specifically, the higher the risk level of the starting unit, the stronger the initial propagation capability; the longer the path, the more significant the attenuation during propagation. Multiplying the propagation weight of the path by the attenuation coefficient yields the final propagation probability value of that path, which is used to quantify the possibility of fault propagation between energy storage units.

[0114] For example, if energy storage unit number 5 is high-risk, its direct path weight with unit number 6 is 0.9, and its path attenuation coefficient is 0.95, then the propagation probability value of this path is 0.855.

[0115] Step S45: Calculate the number and distribution density of high-risk energy storage units, and combine them with the propagation probability values ​​of each propagation path to obtain the failure risk index of the entire battery pack.

[0116] In one embodiment, the number and spatial distribution density of high-risk energy storage units within the battery pack are statistically analyzed, and combined with the propagation probability values ​​of each path, a comprehensive fault risk index for the entire battery pack is calculated. This fault risk index reflects the degree of overall fault propagation risk present in the battery pack under its current state and can be used for system early warning or safety strategy adjustments.

[0117] For example, when a battery pack has 5 high-risk energy storage units and their connection path propagation probability values ​​are all high, the calculated failure risk index may be 0.78 (out of 1), indicating that the battery pack has a high overall risk.

[0118] It should be noted that the failure risk index is a comprehensive evaluation indicator that can be quantified according to different engineering needs. It is mainly used to assist decision-making and risk management, and does not directly reflect the absolute safety status of a single energy storage unit.

[0119] See Figure 3 This invention provides a comparative impedance spectrum analysis chart for energy storage units. In one embodiment, the system measures the impedance of each energy storage unit within a preset frequency range and compares it with a healthy reference value to identify abnormal frequency bands. In the chart, the green curve represents the healthy reference impedance value, the blue curve represents the measured value of a normal energy storage unit, and the red curve represents the measured value of an abnormal energy storage unit. At a frequency of 1 kHz, the impedance value of the abnormal energy storage unit is 48 mΩ, while the reference value is 35 mΩ, showing a significant deviation.

[0120] For example, the system can calculate the differential impedance eigenvector based on this deviation and determine the fault severity score through the state deviation metric, which can be used for subsequent risk ranking and early warning output.

[0121] It should be noted that impedance measurements should be performed under standard temperature conditions and the voltage range should be between 2.5 and 4.2 V to ensure the comparability and accuracy of the data.

[0122] Further, step S5 includes the following steps:

[0123] Step S51: When the fault risk index exceeds the preset safety threshold, record the time of the over-threshold event and collect the complete status data of the current battery pack;

[0124] In one embodiment, when the battery pack failure risk index exceeds a preset safety threshold, the system records the occurrence time of the over-threshold event and collects complete status data of the current battery pack, including impedance measurement values, differential impedance feature vectors, fault severity scores, and ambient temperature for each energy storage unit.

[0125] For example, if the safety threshold is set to 0.7 and the calculated risk index is 0.78, the system will automatically trigger event logging and save the current status information of all energy storage units for subsequent analysis and early warning processing.

[0126] It should be noted that the recording operation is mainly used for data tracking and fault analysis, and will not change the actual operating status of the energy storage unit.

[0127] Step S52: Based on the number of the abnormal energy storage unit, find the coordinate position of each abnormal unit in the battery pack from the correspondence between the number and the physical location;

[0128] In one embodiment, based on the selected abnormal energy storage unit numbers, the coordinate position of each abnormal energy storage unit in the battery pack is found from the correspondence between the numbers and physical positions established in step S11, so as to obtain the specific position of each abnormal unit in the actual battery pack structure.

[0129] For example, if energy storage units numbered 12 and 15 are identified as abnormal, by querying the correspondence between the number and the physical location, it can be found that number 12 is located in the 2nd row and 3rd column of the battery pack, and number 15 is located in the 3rd row and 1st column, thus providing accurate information for subsequent fault location and maintenance.

[0130] It should be noted that the coordinate positions are queried based on the initial number correspondence to ensure that the position of each abnormal unit in the battery pack is unique and traceable.

[0131] Step S53: Compile the abnormal energy storage unit's number, physical coordinates, fault type identifier, and fault severity score into a fault information table;

[0132] In one embodiment, the number, physical coordinates, fault type identifier, and fault severity score of each abnormal energy storage unit are compiled into a structured fault information table for unified management and rapid querying. The fault information table can be stored in tabular form, with each row corresponding to an abnormal energy storage unit, and columns including fields such as number, location, fault type, and severity score.

[0133] For example, for energy storage unit number 12, its fault type is identified as "low-frequency impedance anomaly" and its fault severity score is 0.85. Then, a complete record is formed in the fault information table: No. 12, 2nd row, 3rd column, low-frequency impedance anomaly, 0.85.

[0134] Step S54: Format the fault information table according to the preset data format, add event timestamps and system identification codes, and generate a standardized early warning data package;

[0135] In one embodiment, the fault information table is standardized according to a preset data format, such as by adding event timestamps, system identification codes, and data signatures, to generate standardized early warning data packets so that they can be reliably transmitted to monitoring terminals or mobile devices through a communication interface.

[0136] For example, the information of the abnormal energy storage unit numbered 12 is combined with the recording time 202X-XX-XX XX:XX:XX and the system ID "SYS001" to form an early warning data packet in JSON or XML format, which is convenient for the monitoring system to automatically parse and display.

[0137] Step S55: Select the sending method according to the severity of the fault risk index, and send the warning data packet to the monitoring terminal or mobile device through the communication interface.

[0138] In one embodiment, an appropriate method for sending early warning information is selected based on the severity of the fault risk index. For example, for events with a high-risk index (e.g., exceeding 0.8), the information can be sent directly to the monitoring terminal via wired network or cellular communication, while simultaneously triggering push notifications from mobile devices; for medium-risk events, the information can be stored in a local database and sent periodically.

[0139] For example, if the failure risk index of energy storage unit number 12 is 0.85, the system can immediately push the early warning data packet to the maintenance personnel's mobile app through the monitoring center server, and at the same time display a red early warning icon on the monitoring terminal.

[0140] It should be noted that the sending method and frequency can be adjusted according to specific application scenarios and security policies, with the aim of ensuring timely transmission of important early warning information to assist in operation and maintenance decisions.

[0141] The present invention also provides a hybrid energy storage battery state monitoring system based on data processing, used to execute the hybrid energy storage battery state monitoring method based on data processing as described above, wherein the hybrid energy storage battery state monitoring system based on data processing includes:

[0142] The impedance reference construction module is used to collect impedance measurement data of hybrid energy storage battery packs at a preset frequency sequence; perform temperature correction processing on the impedance measurement data to obtain corrected impedance data; and establish an impedance reference database based on the corrected impedance data.

[0143] The feature vector calculation module is used to calculate the impedance difference value sequence of each energy storage unit between adjacent frequency points; normalize the impedance difference value sequence to obtain the differential impedance feature vector; and calculate the state deviation metric based on the deviation between the differential impedance feature vector and the reference vector.

[0144] The fault identification and scoring module is used to identify abnormal energy storage units whose state deviation metric exceeds a preset threshold; to perform frequency range analysis on the differential impedance characteristic vector of the abnormal energy storage unit to determine the fault type identifier; and to calculate the fault severity score.

[0145] The risk prediction and ranking module is used to rank risk levels according to fault severity scores; establish a fault propagation path prediction model based on risk levels and electrical connection relationships between energy storage units; and use the fault propagation path prediction model to predict the fault risk index.

[0146] The fault warning output module is used to output warning information including the location of the abnormal energy storage unit, fault type identifier, and fault severity score when the fault risk index exceeds the safety threshold.

[0147] Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0148] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A method for monitoring the state of a hybrid energy storage battery based on data processing, characterized in that, Includes the following steps: Step S1: Collect impedance measurement data of the hybrid energy storage battery pack at a preset frequency sequence; perform temperature correction processing on the impedance measurement data to obtain corrected impedance data; An impedance reference database was established based on corrected impedance data. Step S2: Calculate the impedance difference sequence of each energy storage unit between adjacent frequency points; normalize the impedance difference sequence to obtain the differential impedance characteristic vector; calculate the state deviation metric based on the deviation between the differential impedance characteristic vector and the reference vector; the process of calculating the impedance difference sequence is as follows: The corrected impedance data is grouped by energy storage unit number, and sorted by frequency from low to high within each group; adjacent frequency points are paired within each group, and the frequency value and corresponding corrected impedance value of each pair of adjacent frequency points are recorded. For each pair of adjacent frequency points, the impedance difference in that frequency range is obtained by subtracting the impedance value of the low frequency point from the impedance value of the high frequency point. Arrange the impedance differences of each energy storage unit in each frequency range in order to form the impedance difference sequence of the energy storage unit. Step S3: Identify abnormal energy storage units whose state deviation metric exceeds a preset threshold; perform frequency range analysis on the differential impedance characteristic vector of the abnormal energy storage unit to determine the fault type identifier, and then calculate the fault severity score. Step S4: Sort the risk levels according to the severity score of the fault; A fault propagation path prediction model is established based on risk level and electrical connection relationship between energy storage units; the fault risk index is predicted using the fault propagation path prediction model. Step S5: When the fault risk index exceeds the safety threshold, output warning information including the location of the abnormal energy storage unit, fault type identifier and fault severity score.

2. The hybrid energy storage battery state monitoring method based on data processing according to claim 1, characterized in that, Step S1 includes the following steps: Step S11: Set the measurement interval and scanning order of the frequency sequence; assign a unique number to each energy storage unit in the hybrid energy storage battery pack, and record the physical location corresponding to each number; Step S12: Apply AC test signals to each energy storage unit one by one according to the frequency sequence, record the impedance response values ​​at each frequency point, and record the ambient temperature during the measurement. Step S13: Calculate the temperature correction factor based on the difference between the ambient temperature and the preset standard reference temperature.

3. The hybrid energy storage battery state monitoring method based on data processing according to claim 2, characterized in that, Step S1 also includes the following steps: Step S14: Correct the impedance measurement data using a temperature correction factor to obtain corrected impedance data under standard temperature conditions; Step S15: Calculate the average impedance and variability of each energy storage unit at each frequency point based on the corrected impedance data; use the average impedance as the health status benchmark value, establish a correspondence between the energy storage unit number and the frequency point, and construct an impedance benchmark database.

4. The hybrid energy storage battery state monitoring method based on data processing according to claim 3, characterized in that, The normalization process in step S2 is as follows: Based on the frequency range information in the impedance difference value sequence, the reference impedance value of each energy storage unit in the corresponding frequency range is found from the impedance reference database. The normalized difference value is obtained by dividing the difference in each frequency range in the impedance difference value sequence by the reference impedance value corresponding to that frequency range. Arrange the normalized differential values ​​of each energy storage unit into a vector according to the frequency range; this is the differential impedance characteristic vector.

5. The hybrid energy storage battery state monitoring method based on data processing according to claim 4, characterized in that, The process of calculating the state deviation metric in step S2 is as follows: Extract the reference differential impedance vector of each energy storage unit under healthy conditions from the impedance reference database; The deviation vector is obtained by subtracting the corresponding elements of the differential impedance characteristic vector obtained from the current measurement from the baseline differential impedance vector of the healthy state one by one. Squaring each element in the deviation vector and summing them, then taking the square root of the sum, yields the Euclidean distance value, which is the state deviation metric.

6. The hybrid energy storage battery state monitoring method based on data processing according to claim 5, characterized in that, Step S3 includes the following steps: Step S31: Set three threshold parameters: normal threshold, warning threshold, and abnormal threshold; Step S32: Compare the state deviation metric of each energy storage unit with three threshold parameters, and select energy storage units that exceed the abnormal threshold as abnormal energy storage units; Step S33: Extract the differential impedance characteristic vector of the abnormal energy storage unit, analyze the deviation distribution of the vector in the low frequency range, medium frequency range and high frequency range, and determine the fault type identifier based on the frequency range where the deviation is mainly concentrated. Step S34: Obtain the basic score weight of this type of fault from the preset scoring standard table according to the fault type identifier; Step S35: Multiply the state deviation metric by the base score weight and weight it according to the number of abnormal frequency bands to obtain the fault severity score.

7. The hybrid energy storage battery state monitoring method based on data processing according to claim 6, characterized in that, Step S4 includes the following steps: Step S41: Sort all energy storage units from highest to lowest according to their fault severity scores, and record the number, score and ranking of each energy storage unit; Step S42: Divide the energy storage units into three levels: high risk, medium risk, and low risk, based on the sorting results; Step S43: Based on the series and parallel connection relationship of energy storage units in the battery pack, establish a connection diagram between adjacent energy storage units; combine the risk level to assign propagation weights to the connection paths between high-risk energy storage units and their adjacent units, and construct a fault propagation path prediction model. Step S44: For each propagation path, calculate the propagation attenuation coefficient based on the risk level of the starting unit and the propagation distance; multiply the path weight by the propagation attenuation coefficient to obtain the propagation probability value of that path; Step S45: Calculate the number and distribution density of high-risk energy storage units, and combine them with the propagation probability values ​​of each propagation path to obtain the failure risk index of the entire battery pack.

8. The hybrid energy storage battery state monitoring method based on data processing according to claim 7, characterized in that, Step S5 includes the following steps: Step S51: When the fault risk index exceeds the preset safety threshold, record the time of the over-threshold event and collect the complete status data of the current battery pack; Step S52: Based on the number of the abnormal energy storage unit, find the coordinate position of each abnormal unit in the battery pack from the correspondence between the number and the physical location; Step S53: Compile the abnormal energy storage unit's number, physical coordinates, fault type identifier, and fault severity score into a fault information table; Step S54: Format the fault information table according to the preset data format, add event timestamps and system identification codes, and generate a standardized early warning data package; Step S55: Select the sending method according to the severity of the fault risk index, and send the warning data packet to the monitoring terminal or mobile device through the communication interface.

9. A hybrid energy storage battery status monitoring system based on data processing, characterized in that, For performing the data processing-based hybrid energy storage battery state monitoring method as described in claim 1, the data processing-based hybrid energy storage battery state monitoring system includes: The impedance reference construction module is used to collect impedance measurement data of hybrid energy storage battery packs at a preset frequency sequence; perform temperature correction processing on the impedance measurement data to obtain corrected impedance data; and establish an impedance reference database based on the corrected impedance data. The feature vector calculation module is used to calculate the impedance difference value sequence of each energy storage unit between adjacent frequency points; normalize the impedance difference value sequence to obtain the differential impedance feature vector; and calculate the state deviation metric based on the deviation between the differential impedance feature vector and the reference vector. The fault identification and scoring module is used to identify abnormal energy storage units whose state deviation metric exceeds a preset threshold; to perform frequency range analysis on the differential impedance characteristic vector of the abnormal energy storage unit to determine the fault type identifier; and to calculate the fault severity score. The risk prediction and ranking module is used to rank risk levels according to fault severity scores; establish a fault propagation path prediction model based on risk levels and electrical connection relationships between energy storage units; and use the fault propagation path prediction model to predict the fault risk index. The fault warning output module is used to output warning information including the location of the abnormal energy storage unit, fault type identifier, and fault severity score when the fault risk index exceeds the safety threshold.

Images