A cascade battery data monitoring method and system

Abstract

The application discloses a kind of cascade battery data monitoring method and system, it is related to electric digital data processing field, the method comprises the following steps: obtaining the initial information of battery of new access cascade battery package;Retrieve matching historical battery sample in historical database, extract to obtain same batch historical monitoring data;The time sequence variation rate of each monitoring parameter in same batch historical monitoring data is calculated, and key parameter set is identified, and difference collection frequency table is generated;According to difference collection frequency table, the operation data of cascade battery package is collected, and real-time monitoring data flow is obtained;Data distribution interval is determined based on same batch historical monitoring data, and based on data distribution interval, abnormal data in real-time monitoring data flow is marked, and data monitoring record is generated;Based on same batch historical monitoring data and the data monitoring record of cascade battery package in a plurality of complete charge-discharge cycles, the health state data of cascade battery package is calculated.The application can reduce the data monitoring cost of cascade battery.

Description

Technical Field

[0001] This application relates to the field of digital data processing, and in particular to a method and system for monitoring data of cascaded batteries. Background Technology

[0002] To further improve the economics of charging stations, using retired batteries from new energy vehicles to build energy storage systems is feasible. These batteries, due to their secondary use, are also called secondary batteries. Although secondary batteries no longer meet automotive standards, their 70%-80% remaining capacity makes them valuable in scenarios like charging station energy storage where energy density requirements are relatively low. Because secondary batteries come from diverse sources and have varying usage histories, their internal state is uncertain, making data monitoring crucial.

[0003] In related technologies, data monitoring of cascaded batteries employs a unified monitoring strategy. Specifically, all battery packs collect basic parameters such as voltage, current, and temperature at a specified frequency. When any collected data exceeds a pre-set fixed threshold in the BMS (e.g., voltage above 4.2V or temperature above 60°C), the BMS executes protection logic, such as disconnecting the main circuit relay, to interrupt the battery charging and discharging process, thereby preventing immediate risks such as overcharging, over-discharging, or overheating. Simultaneously, the data monitoring system in these technologies also calculates the battery state of health (SOH) based on this data to assess battery lifespan.

[0004] However, with the increasing number of cascaded batteries, the data monitoring used in related technologies generates a large amount of monitoring data, which consumes a lot of computing, storage resources and communication bandwidth, resulting in high data monitoring costs. Summary of the Invention

[0005] This application provides a method and system for monitoring data of cascaded batteries, which can reduce the cost of monitoring data of cascaded batteries.

[0006] Firstly, this application provides a method for monitoring cascaded battery data, applied to a battery monitoring system. The method includes: acquiring the batch identifier and initial state parameters of a newly connected cascaded battery pack to obtain initial battery information; retrieving matching historical battery samples from a historical database based on the initial battery information, extracting monitoring parameter sequences from the historical battery samples, and obtaining historical monitoring data for the same batch; calculating the time-series change rate of each monitoring parameter in the historical monitoring data for the same batch, identifying monitoring parameters with time-series change rates exceeding a preset fluctuation threshold as key monitoring parameters, and obtaining a set of key parameters; generating a collection frequency configuration for the monitoring parameters corresponding to the cascaded battery pack based on the set of key parameters, and obtaining a differentiated collection frequency table; collecting the operating data of the cascaded battery pack according to the differentiated collection frequency table and uploading it to the cloud to obtain a real-time monitoring data stream; determining the data distribution range based on the historical monitoring data for the same batch, and marking abnormal data in the real-time monitoring data stream based on the data distribution range, generating data monitoring records with abnormal markers; and calculating the health status data of the cascaded battery pack based on the historical monitoring data for the same batch and the data monitoring records collected by the cascaded battery pack over multiple complete charge-discharge cycles.

[0007] In the above embodiments, the battery monitoring system identifies key parameters with high time-series change rates by analyzing historical data from the same batch and generates a differentiated acquisition frequency table. This allows for high-frequency acquisition of key parameters such as voltage and temperature that change drastically during high-rate discharge at the charging station to support fast charging, while low-frequency acquisition of non-critical parameters that change slowly during off-peak charging at night. This ensures the safe operation of the charging station's energy storage system and prevents the loss of critical status information, while reducing the total amount of data collected, lowering data transmission and storage resource consumption, and reducing the cost of tiered battery monitoring.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, after calculating the health status data of the secondary battery pack based on historical monitoring data of the same batch and data monitoring records collected by the secondary battery pack over multiple complete charge-discharge cycles, the method further includes: generating an authentication data package containing a unique battery identifier, batch identifier, monitoring data summary, and current remaining capacity based on the health status data, and uploading the authentication data package to the blockchain ledger; when the secondary battery pack ends its current application and is planned for retirement, reading the authentication data package corresponding to the secondary battery pack from the blockchain ledger and generating batch authentication information; determining the applicable scope of the secondary battery in the next application based on the authentication data package, and generating a retirement assessment report containing the applicable scope and batch authentication information.

[0009] In the above embodiments, the battery monitoring system utilizes the immutability and decentralization of blockchain technology to create a trusted digital lifecycle profile for each cascaded battery pack used for energy storage in charging stations. It uploads certified data packets containing health status to the blockchain, solving the problems of information opacity and lack of trust when cascaded battery assets are transferred between charging station operators, battery suppliers, and recyclers, thereby improving the liquidity and value of cascaded battery assets.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, the step of generating a sampling frequency configuration for monitoring parameters corresponding to the secondary battery pack based on a set of key parameters to obtain a differentiated sampling frequency table specifically includes: obtaining the charge / discharge rate value of the secondary battery pack in the current application scenario to obtain the current rate parameter; extracting the charge / discharge rate value of the historical application scenario from historical monitoring data of the same batch to obtain the historical rate parameter; calculating the difference between the current rate parameter and the historical rate parameter to obtain the rate difference; when the rate difference is greater than a preset rate threshold, adding the instantaneous voltage drop parameter and the single-cell instantaneous temperature difference parameter to the set of key parameters to obtain an extended parameter set; and determining multiple monitoring parameters and corresponding sampling frequency configurations corresponding to the secondary battery pack based on the extended parameter set to obtain a differentiated sampling frequency table.

[0011] In the above embodiments, the battery monitoring system dynamically adjusts the set of key parameters by comparing the charge and discharge rates of the current charging station (such as a high-power fast charging station) with those of historical application scenarios (such as low-speed vehicle energy storage). When the battery switches from a low-load application to a high-load fast charging power support application, the system incorporates parameters that reflect changes in the internal state of the battery under high-rate impact, such as instantaneous voltage drop and instantaneous temperature difference of individual cells, into high-frequency monitoring, thereby improving the targeting and safety of the monitoring.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, before the step of collecting the operating data of the cascaded battery packs according to the differentiated acquisition frequency table and uploading it to the cloud to obtain the real-time monitoring data stream, the method further includes: reading the batch identifiers of all battery packs in the battery system where the cascaded battery packs are located to obtain a system batch identifier set; when there are different batch identifiers in the system batch identifier set, extracting the internal resistance parameters of each battery pack in the battery system to obtain a system internal resistance data set; calculating the difference between the internal resistance value of the cascaded battery pack and the internal resistance value of other battery packs in the system internal resistance data set to obtain an internal resistance deviation value; and calculating the current distribution correction coefficient of the cascaded battery pack based on the current shunting principle of parallel circuits, according to the internal resistance deviation value and the internal resistance value of the cascaded battery pack.

[0013] In the above embodiments, the battery monitoring system solves the problem of uneven current distribution caused by inconsistent internal resistance when different batches of tiered battery packs are connected in parallel in a large energy storage system of a charging station. By calculating the current distribution correction coefficient, the measured current can be calibrated to obtain a current value that is closer to the actual load borne by the battery pack, avoiding misjudgment caused by current monitoring deviation and improving the accuracy of the entire monitoring system.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, the step of determining the statistical range where the real-time value deviates from the historical value and generating an anomaly marker based on the real-time monitoring data stream and the historical monitoring data of the same batch, and obtaining a data monitoring record with an anomaly marker, specifically includes: multiplying the measured current value in the real-time monitoring data stream by a current distribution correction coefficient to obtain a corrected current value; extracting the upper and lower limits of the historical current from the historical monitoring data of the same batch to obtain a historical current range; generating an anomaly marker when the corrected current value exceeds the historical current range, and attaching the anomaly marker to the corresponding monitoring data record to obtain a data monitoring record with an anomaly marker.

[0015] In the above embodiments, the battery monitoring system applies the calculated current distribution correction coefficient to the real-time monitoring data stream, which can effectively identify the real current anomalies caused by uneven parallel current distribution. For example, when supporting fast charging, the actual output of a certain battery pack far exceeds expectations, which improves the reliability of abnormal data marking and makes the data monitoring records more realistically reflect the battery's operating status.

[0016] In conjunction with some embodiments of the first aspect, in some embodiments, the step of collecting operational data of the cascaded battery pack according to a differentiated acquisition frequency table and uploading it to the cloud to obtain a real-time monitoring data stream specifically includes: calculating the resting time based on the last decommissioning timestamp and the current access timestamp of the cascaded battery pack; when the resting time exceeds a preset resting threshold, setting the working mode of the cascaded battery pack to activation mode; in activation mode, collecting operational data of the cascaded battery pack during the first charge and discharge process at an activation acquisition frequency higher than the frequency of each parameter in the differentiated acquisition frequency table to obtain activation period monitoring data; calculating the standard deviation of the voltage change rate and temperature change rate in the activation period monitoring data to obtain fluctuation statistics; when the fluctuation statistics are all less than a preset steady-state threshold within a preset time range, switching the working mode of the cascaded battery pack back to normal monitoring mode.

[0017] In the above embodiments, the battery monitoring system introduces an activation mode for the cascaded battery packs that have been taken out of the warehouse and placed for a long time and newly installed at the charging station. In the activation mode, high-frequency monitoring can capture the subtle changes in the internal electrochemical state of the battery from passivation to activation. By judging whether the fluctuation statistics of voltage and temperature change rate have reached a steady state, it can be confirmed that the battery has safely and stably recovered to the normal working state, thus ensuring electrical safety.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, after collecting the operating data of the cascaded battery pack during the first charge and discharge process at an activation collection frequency higher than the frequency of each parameter in the differentiated collection frequency table to obtain activation period monitoring data, the method further includes: extracting the historical voltage curve and historical temperature curve of the first activation period from the historical monitoring data of the same batch to obtain the activation period reference curve; calculating the safe threshold range of voltage and temperature during the activation period based on the activation period reference curve; and generating an abnormal warning signal when it is determined that the real-time collected voltage value or temperature value exceeds the safe threshold range during the activation collection frequency collection process.

[0019] In the above embodiments, the battery monitoring system introduces a dynamic safety threshold based on historical data in the activation mode. By comparing the activation curves of the same batch of batteries in history, a safety boundary that is more in line with the actual characteristics of the current battery activation process can be constructed. Abnormal behavior in the activation process can be identified earlier and more accurately, and early warning signals can be generated, thereby improving the safety of the charging station energy storage system.

[0020] In a second aspect, embodiments of this application provide a battery monitoring system, the battery monitoring system comprising: one or more processors and a memory; the memory is coupled to the one or more processors, the memory being used to store computer program code, the computer program code including computer instructions, the one or more processors calling the computer instructions to cause the battery monitoring system to perform the method described in the first aspect and any possible implementation thereof.

[0021] Thirdly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a battery monitoring system, cause the battery monitoring system to perform the method described in the first aspect and any possible implementation thereof.

[0022] Fourthly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a battery monitoring system, cause the battery monitoring system to perform the method described in the first aspect and any possible implementation thereof.

[0023] Understandably, the battery monitoring system provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the methods provided in the embodiments of this application. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods, and will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0025] 1. By adopting a technical solution that dynamically generates differentiated acquisition frequency tables based on historical data, the initial information of the cascaded battery packs of newly connected charging station energy storage systems is obtained. Based on this, historical monitoring data of the same batch is retrieved from the historical database to identify key parameters that change drastically. In order to customize the acquisition frequency configuration, this method can accurately allocate monitoring resources to the parameters that need the most attention. For example, high-frequency acquisition is carried out for parameters that have large voltage fluctuations when supporting fast charging, and low-frequency acquisition is carried out for parameters that change slowly when charging during off-peak hours at night. This enables charging station operators to effectively control the monitoring cost of cascaded battery energy storage systems.

[0026] 2. By employing a technical solution that combines battery health status data with blockchain technology, after calculating the health status data of the tiered battery pack, an authentication data package containing core information such as the battery's unique identifier, batch identifier, monitoring data summary, and current remaining capacity is generated and uploaded to a blockchain ledger with immutable and traceable characteristics. Therefore, when the battery ends its tiered application at the current charging station, this credible digital certificate can be obtained from the blockchain. Thus, this method establishes a transparent and reliable value and status transfer chain for tiered batteries used for energy storage at charging stations, providing a scientific basis for the next tiered application of the battery (such as transferring from a fast charging station to a slow charging station), and realizing standardized management and value maximization of tiered battery assets.

[0027] 3. Because it adopts a technical solution that dynamically expands the set of key parameters according to changes in application scenarios, that is, firstly, it obtains and compares the charging and discharging rates of the current charging station's operating mode with those of historical application scenarios. When the rate difference exceeds a preset threshold, it dynamically adds the instantaneous voltage drop and the instantaneous temperature difference of individual units, two parameters that are sensitive to high-rate operating conditions, to the set of key parameters. Therefore, this method enables the monitoring strategy to adapt to the adjustment of the charging station's operating strategy in real time, avoiding the problem of missing instantaneous risks caused by sudden changes in operating conditions due to the changing operating conditions of the charging station. Attached Figure Description

[0028] Figure 1 This is a flowchart illustrating a method for monitoring cascade battery data in an embodiment of this application;

[0029] Figure 2 This is another flowchart illustrating the method for monitoring cascade battery data in this application embodiment;

[0030] Figure 3 This is a schematic diagram of the physical device structure of a battery monitoring system in an embodiment of this application. Detailed Implementation

[0031] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification of this application, the singular expressions “a,” “an,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0032] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0033] In the specific implementation scenario of this application, namely when cascaded batteries are applied to charging station energy storage systems, the technical terms involved have specific meanings. Among them, a cascaded battery pack refers to an energy storage unit that has been retired from a new energy vehicle and integrated into a photovoltaic-energy storage-charging integrated charging station or a high-power fast charging station. Batch identification is a code that identifies battery production information and is crucial for ensuring the initial performance consistency of numerous battery packs connected in parallel within the charging station energy storage system. Initial state parameters describe the basic state of the battery pack when it is installed in the charging station energy storage cabinet. The historical database is a cloud platform that records the historical operating data of cascaded batteries in all the company's charging stations, serving as the basis for big data analysis. Application scenarios specifically refer to the operating strategies of the charging station energy storage system, such as peak shaving and valley filling (charging during off-peak hours at night and supplying power to charging piles during peak hours to reduce electricity costs), high-power fast charging support (providing instantaneous power boost when multiple vehicles are fast charging simultaneously to avoid grid tripping), or participating in grid demand-side response to obtain additional revenue. The differentiated acquisition frequency table is the core output of this application, specifying different acquisition frequencies for different parameters under different operating strategies. Status of Health (SOH) data is directly related to whether the energy storage system can continuously meet the operational requirements of the charging station, and is a key basis for charging station operators to assess asset value and formulate replacement plans.

[0034] The following describes the process of the method provided in this implementation. Please refer to [link / reference]. Figure 1 This is a flowchart illustrating a method for monitoring cascade battery data in an embodiment of this application.

[0035] S101. Obtain the batch identifier and initial status parameters of the newly connected tiered battery pack to obtain the initial battery information.

[0036] Among them, batch identifier refers to a unique code used to identify the battery production batch; initial state parameters refer to the basic performance indicators of the battery pack when it is installed in the energy storage cabinet of the charging station and first connected to the system; battery initial information is a combination of batch identifier and initial state parameters.

[0037] Specifically, this step is triggered when a new batch of cascaded battery packs arrives at the charging station, is installed into the energy storage cabinet by maintenance personnel, and is connected to the battery monitoring system for the first time. The battery monitoring system establishes communication with the battery management system (BMS) of the battery pack through its data acquisition unit, reading the batch identifier and initial status parameters. This information forms the identity and status basis for personalized monitoring of the battery pack and is the starting point for all subsequent analyses.

[0038] In some embodiments, this step can be implemented in several ways: Optionally, the initial battery information can be obtained by scanning a QR code or barcode on the battery pack casing. The QR code encodes the batch identifier and initial state parameters at the time of manufacture, and maintenance personnel can complete the information entry by scanning the code with a handheld terminal. Optionally, contactless data reading can be performed using near field communication (NFC) or radio frequency identification (RFID) tags. An electronic tag containing the initial battery information is affixed to the battery pack, and the monitoring terminal can automatically obtain it when it approaches. It is understood that other methods can also be used to obtain the initial battery information, such as manual entry by maintenance personnel through a human-machine interface, which is not limited here.

[0039] In some embodiments, newly connected battery packs may have incomplete or corrupted initial information stored in their BMS. In response, the battery monitoring system initiates an information completion process. This process first marks the battery pack as "to be identified," then guides it through a standard, low-rate full charge-discharge cycle at the charging station (e.g., utilizing off-peak electricity at night). During this process, the system measures its key parameters and, combined with readable partial batch information, performs a fuzzy match in the historical database to generate a set of the closest initial state parameters.

[0040] S102. Based on the initial battery information, retrieve matching historical battery samples from the historical database, extract the monitoring parameter sequences of the historical battery samples, and obtain historical monitoring data for the same batch.

[0041] Among them, the historical database refers to the central cloud platform database that stores all monitoring data of a large number of battery packs of the same or different types in all the charging stations under the company during the past operating cycle; the historical battery sample refers to the set of battery packs in the historical database that have the same batch identifier as the newly connected battery packs; the historical monitoring data of the same batch is the set of monitoring parameter sequences of all the extracted matching samples.

[0042] Specifically, after obtaining the initial battery information in step S101, the battery monitoring system uses the batch identifier in this information as the primary key to perform a query operation in the cloud-based historical database. The query target is all historical battery samples that completely match the batch identifier. Upon retrieving a matching sample, the system extracts all monitoring parameter sequences recorded when these samples were previously deployed at other charging stations. This data reflects the historical performance of other batteries from the same "origin" as the new battery pack under various charging station operating conditions.

[0043] In some embodiments, this step can be implemented in several ways: Optionally, a distributed database architecture, such as Hadoop HDFS, can be used, with the batch identifier as the partition key to accelerate the retrieval process. The system extracts the monitoring parameter sequences distributed across different nodes in parallel through MapReduce tasks. Optionally, a time-series database (such as InfluxDB) can be used. This database is optimized for time-series data, and by indexing the batch identifier, it can efficiently query and extract all monitoring data for a specified time range and a specified batch of samples. It is understood that other methods can also be used to implement data retrieval, such as traditional SQL database queries, which are not limited here.

[0044] In some embodiments, introducing a completely new supplier or model of cascaded battery to a charging station may result in no perfectly matching samples in the historical database. To address this, the battery monitoring system employs a similar batch matching strategy. This strategy searches the database for the most similar batch sample based on secondary parameters such as manufacturer, battery chemistry, and rated capacity. The system uses the data from this similar batch as an initial reference while simultaneously creating a separate, entirely new historical data archive for this new batch, which is continuously populated during subsequent operations.

[0045] S103. Calculate the time-series change rate of each monitoring parameter in the same batch of historical monitoring data, identify the monitoring parameters whose time-series change rate exceeds the preset fluctuation threshold as key monitoring parameters, and obtain the key parameter set.

[0046] Among them, the time-series change rate refers to the rate at which the value of the monitored parameter changes over time, such as the instantaneous voltage drop rate (dV / dt) when supporting fast charging; the preset fluctuation threshold is a pre-set threshold value used to judge the activity level of the parameter; the key monitoring parameter refers to those parameters whose time-series change rate often exceeds the threshold, indicating that they are highly dynamic, especially under the complex operating conditions of the charging station; the key parameter set is a list of all identified key monitoring parameters.

[0047] Specifically, after acquiring historical monitoring data from the same batch, the battery monitoring system calculates the first difference or derivative of the time series of each monitoring parameter (such as total voltage, highest single-cell voltage, average temperature, etc.) to obtain the time-series rate of change sequence. Subsequently, the system counts the percentage of values ​​in the rate of change sequence for each parameter that exceed a preset fluctuation threshold. If this percentage exceeds a certain set percentage, the parameter is identified as a critical monitoring parameter and added to the critical parameter set.

[0048] In some embodiments, this step can be implemented in several ways: Optionally, a sliding window method can be used to calculate the standard deviation of each parameter within a fixed-length time window. If the average or maximum value of the standard deviation exceeds a threshold, it is identified as a key parameter. Optionally, a Fourier transform can be performed on the time series of the parameters to analyze their spectral characteristics, and parameters with a large proportion of high-frequency components can be identified as key parameters, because high-frequency components represent drastic changes in the rapid power response of the charging station. It is understood that other methods can also be used to identify key parameters, such as analyzing their instantaneous change characteristics based on wavelet transform, which is not limited here.

[0049] It's important to note that the preset fluctuation threshold is not a fixed empirical value, but rather dynamically generated based on in-depth statistical analysis of historical monitoring data from the same batch. Specifically, after acquiring historical monitoring data from the same batch, the system first calculates the time-series rate of change sequence for each monitoring parameter (such as cell voltage, temperature, etc.) under all historical operating conditions. Subsequently, the system performs statistical distribution analysis on this rate of change sequence, such as calculating its probability density function or cumulative distribution function. The preset fluctuation threshold can be set at a high percentile of this distribution, such as the 95th or 99th percentile. This method of setting the threshold is adaptive, reflecting the normal fluctuation range of the batch of batteries over the vast majority of historical applications; only statistically significant and drastic changes exceeding this range are identified. For example, if the time-series rate of change (dV / dt) of the historical single-cell voltage data of a certain batch of batteries is calculated, and the 95th percentile value of the rate of change data is 0.05V / s, then the system can set 0.05V / s as the preset fluctuation threshold of the single-cell voltage parameter of that batch of batteries. Any parameter corresponding to a time point where the rate of change exceeds this value is considered as a candidate for key monitoring.

[0050] S104. Based on the set of key parameters, generate the acquisition frequency configuration of the monitoring parameters corresponding to the tiered battery packs to obtain a differentiated acquisition frequency table.

[0051] Among them, the data acquisition frequency configuration refers to the data acquisition time interval set for each monitoring parameter; the differentiated data acquisition frequency table is a configuration list containing all monitoring parameters and their corresponding acquisition frequencies, which is the core of realizing refined monitoring of the energy storage system of the charging station.

[0052] Specifically, the battery monitoring system formulates a data acquisition strategy based on the set of key parameters obtained in step S103. For parameters identified as key monitoring parameters (e.g., cell voltage, total current), the system assigns a higher acquisition frequency (e.g., 1Hz or higher) to ensure that its rapid dynamic changes during fast charging can be captured. For non-key parameters (e.g., parameters with slow changes such as ambient humidity inside the energy storage cabinet), the system assigns a lower acquisition frequency (e.g., 0.1Hz or lower).

[0053] In some embodiments, this step can be implemented in several ways: Optionally, a hierarchical frequency configuration can be adopted, with the system preset three frequency levels: high, medium, and low. Key parameters are assigned to high frequencies, secondary parameters to medium frequencies, and auxiliary parameters to low frequencies. Optionally, a dynamic frequency configuration related to the rate of change can be adopted, using the historical average time-series rate of change of the parameters as input and mapping it to specific acquisition frequency values ​​through a function to achieve more refined frequency allocation. It is understood that other methods can also be used to generate the frequency table, such as assigning higher acquisition frequencies to parameters with higher information entropy based on information entropy theory; this is not limited here.

[0054] S105. Collect the operating data of the tiered battery packs according to the differentiated acquisition frequency table and upload it to the cloud to obtain real-time monitoring data stream.

[0055] Among them, operational data refers to the real-time values ​​of various parameters generated when the battery pack actually performs tasks such as peak shaving and valley filling and power support at the charging station; uploading to the cloud refers to transmitting the collected data to a remote cloud monitoring platform through a 4G / 5G network; and real-time monitoring data stream is a data stream composed of continuously uploaded data points.

[0056] Specifically, the battery monitoring system sends the differentiated acquisition frequency table generated in step S104 to the local data acquisition unit (DTU) or BMS within the energy storage cabinet of the charging station. The acquisition unit reads data according to the frequency set in the table. For example, it reads the cell voltage once per second, but only the ambient temperature once every 10 seconds. The acquired data is packaged, timestamped, and sent to the battery monitoring platform in the cloud in real time.

[0057] In some embodiments, this step can be implemented in several ways: Optionally, a data packaging and uploading mechanism can be used, where the local acquisition unit first caches data collected at different frequencies over a period of time (e.g., 1 minute), and then packages this data into a data frame and uploads it to the cloud at once to save communication traffic costs for the charging station; Optionally, an edge computing strategy can be used to preprocess the data on the local acquisition unit, for example, uploading only when the data change exceeds a certain threshold (change reporting), further reducing the amount of data uploaded. It is understood that other methods can also be used to achieve data collection and uploading, such as using lightweight IoT protocols such as MQTT for data transmission, which is not limited here.

[0058] S106. Determine the data distribution range based on historical monitoring data of the same batch, and mark abnormal data in the real-time monitoring data stream based on the data distribution range, and generate data monitoring records with abnormal markers.

[0059] Among them, the data distribution range refers to the normal fluctuation range of each parameter under specific charging station operating conditions (such as 2C rate discharge supporting fast charging) as statistically derived from the historical monitoring data of the same batch; abnormal data refers to data points in the real-time monitoring data stream that fall outside the normal distribution range; data monitoring records with abnormal tags are records with "normal" or "abnormal" tags added on the basis of the original monitoring data.

[0060] Specifically, the battery monitoring system first categorizes historical monitoring data from the same batch under different operating conditions (e.g., based on current magnitude, into high-power discharge, low-power charging, and idle periods). Then, under each operating condition, it statistically analyzes the data distribution of various monitoring parameters to determine a dynamic, condition-dependent normal data distribution range. For example, when multiple electric vehicles simultaneously initiate high-power charging, the energy storage system is called upon to provide power support. At this time, the cloud platform selects the corresponding historical data distribution range based on the real-time operating condition (high-current discharge). If the system detects that the real-time voltage drop of a certain battery pack significantly exceeds this range, it marks it as "abnormal," which may indicate that the battery pack's internal resistance is too high and cannot effectively support fast-charging loads.

[0061] In some embodiments, normal battery aging can cause an overall drift in its operating parameters (such as a drop in voltage plateau due to increased internal resistance), resulting in situations where even normal data may be marked as abnormal by older data distribution ranges. To address this, the battery monitoring system implements a dynamic distribution range update mechanism. The system periodically (e.g., monthly) uses recently confirmed normal monitoring data to update or fine-tune the distribution range for the corresponding operating conditions in the historical database. This allows the range to slowly adjust as the battery's health deteriorates, thereby improving the long-term accuracy of anomaly marking.

[0062] It's important to note that determining the data distribution range is a dynamic process coupled with operating conditions. The system first needs to cluster the massive amounts of historical monitoring data from the same batch under different operating conditions. Based on key dimensions such as current, state of charge (SOC), and ambient temperature, the data is divided into different operating condition clusters, such as "high-rate discharge supporting fast charging (SOC 50%-80%)" and "low-rate nighttime charging (SOC 20%-50%)". Within each specific operating condition cluster, the system statistically models the data samples of various monitoring parameters (such as cell voltage, voltage drop, and temperature rise rate). A common method is to assume that the parameter values ​​follow a Gaussian distribution under that operating condition. By calculating the mean (μ) and standard deviation (σ) of all historical data points under that condition, a dynamic normal data distribution range is determined, for example, set as [μ-3σ, μ+3σ]. This range theoretically covers 99.7% of normal data points. When real-time monitoring data streams enter the system, the system first matches the current operating conditions (real-time current, SOC, etc.) to the corresponding historical operating condition cluster. Then, it retrieves the distribution range of specific parameters under that operating condition cluster. If the real-time data value falls outside this range, it is marked as abnormal. For example, under the 2C discharge condition supporting fast charging, historical data shows that the average voltage drop of a certain batch of batteries is 0.1V, and the standard deviation is 0.01V. The normal voltage drop range can be set as [0.07V, 0.13V]. If a newly connected battery pack from the same batch has a real-time voltage drop of 0.15V under this operating condition, then this data point will be marked as abnormal.

[0063] S107. Based on historical monitoring data of the same batch and data monitoring records collected by the secondary battery pack in multiple complete charge and discharge cycles, calculate the health status data of the secondary battery pack.

[0064] Among them, multiple complete charge-discharge cycles refer to the process in which the battery has undergone several cycles of being fully charged from off-peak hours at night to being fully discharged during peak hours in the charging station; State of Health (SOH) data is the core indicator for assessing the degree of battery aging, which directly affects whether it can continue to meet the operational needs of the charging station.

[0065] Specifically, the battery monitoring system accumulates data records of the battery pack during its operation at the charging station. The system uses this data, combined with a degradation model based on historical data from the same batch, examining how the health status changes with cycle count and the number of abnormal events, to calculate the current state of health (SOH) of the battery pack. For example, the system can calculate the actual discharged capacity for each complete charge-discharge cycle using the ampere-hour integral method and combine this with capacity degradation curves from historical data to estimate the current state of equilibrium (SOH). Voltage anomaly markers that frequently occur during fast charging are also used as correction factors to accelerate SOH degradation estimation.

[0066] In some embodiments, this step can be implemented in several ways: Optionally, based on incremental capacity analysis (ICA), the peak position and area changes of the voltage-capacity curve during charging and discharging can accurately reflect the internal aging mechanism of the battery by analyzing the differential (dQ / dV). Alternatively, a data-driven approach can be used to train a neural network model, with the input being the statistical characteristics of recent monitoring data (such as average voltage, internal resistance increment, and anomaly counts), and the output being an estimated value of SOH. This model is trained using historical data from the same batch of all charging stations under the company. It is understood that other methods can also be used to calculate SOH, such as a Kalman filter algorithm based on an electrochemical model; this is not limited here.

[0067] In some embodiments, the energy storage systems of some charging stations are primarily used for emergency backup power and rarely undergo complete charge-discharge cycles, making it difficult to accurately calculate the State of Health (SOH) using traditional methods. To address this, the battery monitoring system employs an SOH estimation method based on local data segments. This method searches a historical database for local charge-discharge curve segments within the same SOC range as the current battery pack. By comparing the characteristic differences between real-time and historical data segments and combining these characteristics with a correlation model of SOH in historical data, the current SOH is estimated.

[0068] In practical applications, the "data-driven model" used to estimate the state of health (SOH) of cascaded batteries is typically a time-series prediction model based on deep learning, such as a recurrent neural network (RNN) employing a Long Short-Term Memory (LSTM) network or a gated recurrent unit (GRU). This type of model can effectively capture the nonlinear degradation behavior of batteries over long time scales.

[0069] The model's training relies on a large, high-quality historical database containing extensive monitoring data from the same or similar batches of batteries throughout their entire lifecycle, from initial use to retirement. Preparing the training data is a crucial step; each training sample consists of two parts: an input feature sequence and an output label. The input feature sequence is one or more representative time-series segments extracted from historical monitoring data, such as voltage, current, temperature, and SOC sequences for one or more complete charge-discharge cycles, as well as derived features calculated from this raw data, such as incremental capacity curves (dQ / dV), internal resistance changes, and anomaly data marker counts. The output label is the "true" SOH value of the battery at the corresponding time point in the feature sequence. This true SOH value typically cannot be directly measured online; it comes from periodic, high-precision offline laboratory calibration tests performed on historical battery samples (e.g., capacity obtained through standard 1C charge-discharge testing and internal resistance obtained through hybrid pulse power characteristic testing (HPPC)). The training criterion is to minimize the error between the model-predicted SOH value and the laboratory-calibrated true SOH value. Commonly used loss functions are Mean Squared Error (MSE) or Mean Absolute Error (MAE). The training process employs the backpropagation algorithm and a gradient descent optimizer (such as the Adam optimizer). By iteratively inputting a large number of training samples, the weight parameters inside the network are continuously adjusted until the prediction error of the model on an independent validation dataset converges to an acceptable range.

[0070] At the heart of this SOH estimation model are one or more stacked LSTM layers. The special gating structure of the LSTM units (input gate, forget gate, output gate) enables them to learn and remember long-term dependencies in time-series data, which is crucial for simulating the slow and complex degradation process of batteries. The overall architecture of the model is typically as follows: an input layer that receives a preprocessed sequence of multidimensional features (e.g., a tensor of shape [time step, number of features]); followed by LSTM layers that extract dynamic evolution features of the battery state over time; the output of the LSTM layers may then pass through one or more fully connected (dense) layers to map and integrate the high-dimensional temporal features; finally, an output layer that outputs a single scalar value, the predicted SOH (e.g., a floating-point number between 0 and 100 representing the percentage of capacity retention). The entire model is trained as an end-to-end mapping function that can directly decode the current health state of the battery from the running data.

[0071] After training, the model is deployed to a cloud-based battery monitoring system. In actual use, the system collects real-time operational data streams from newly connected battery packs. To estimate State of Health (SOH), the system extracts a data window from the real-time data stream that matches the feature sequence format used during training, such as data from the most recent complete charge-discharge cycle. This data undergoes the same preprocessing and feature engineering steps as during training before being input into the trained LSTM model. The model performs a forward propagation calculation, and its output is the current SOH estimate for the battery pack. This estimated health status data is then used in subsequent steps of this application, such as generating an authentication data packet uploaded to the blockchain (S212) and generating a retirement assessment report when the battery is scheduled for retirement (S214). By periodically calling the model (e.g., daily or after each complete cycle), the system can achieve continuous tracking and accurate prediction of the SOH of the battery pack.

[0072] The following provides a more detailed description of the process of the method provided in this implementation. Please refer to [link / reference]. Figure 2 This is another flowchart illustrating the cascade battery data monitoring method in this application.

[0073] S201. Obtain the batch identifier and initial status parameters of the newly connected tiered battery pack to obtain the initial battery information.

[0074] Refer to step S101, which will not be repeated here.

[0075] S202. Based on the initial battery information, retrieve matching historical battery samples from the historical database, extract the monitoring parameter sequences of the historical battery samples, and obtain historical monitoring data for the same batch.

[0076] Refer to step S102, which will not be repeated here.

[0077] S203. Calculate the time-series change rate of each monitoring parameter in the same batch of historical monitoring data, identify the monitoring parameters whose time-series change rate exceeds the preset fluctuation threshold as key monitoring parameters, and obtain the key parameter set.

[0078] Refer to step S103, which will not be repeated here.

[0079] S204. Obtain the charge / discharge rate value of the current application scenario of the secondary battery pack to obtain the current rate parameter.

[0080] The charge / discharge rate (C-rate) refers to the multiple of the charge / discharge current relative to the battery's rated capacity; the current rate parameter refers to the typical or maximum operating rate corresponding to the main operating mode set for the battery pack at the current charging station.

[0081] Specifically, when configuring monitoring strategies for the energy storage system of a charging station, the battery monitoring system needs to know the upcoming workload it will face. This information can be obtained from the charging station's energy management system (EMS). For example, a charging station deployed in a highway service area, primarily used to support high-power fast charging, might have an energy storage system with a discharge rate of 2C; while a charging station deployed in a residential area, primarily used for peak shaving and valley filling at night, might only have an operating rate of 0.2C. The battery monitoring system reads this configuration parameter as the current rate parameter.

[0082] In some embodiments, this step can be implemented in several ways: Optionally, when deploying the battery pack, maintenance personnel can manually input the application scenario (such as "fast charging power support" or "peak shaving and valley filling") through a configuration interface. The system internally maintains a mapping table from scenario to rate, and uses this table to retrieve the current rate parameter. Optionally, during the initial operation phase, the battery monitoring system performs short-term monitoring and statistics on current data, automatically identifies the peak current and average current, and calculates the equivalent rate parameter for actual operation. It is understood that other methods can also be used to obtain the rate parameter, and this is not limited here.

[0083] In some embodiments, the operating mode of a charging station may be dynamically changing. For example, an integrated "photovoltaic-storage-charging" station may store excess photovoltaic power at a low rate during the day when photovoltaic output is sufficient, and switch to high-rate discharge to support fast charging during the evening peak charging period. To this end, the battery monitoring system maintains continuous communication with the energy management system (EMS) to obtain future scheduling plans or the current operating mode in real time. Upon detecting a change in operating mode, the system immediately re-acquires the new rate parameters and triggers subsequent parameter set adjustment procedures.

[0084] S205. Extract the charge / discharge rate values ​​of historical application scenarios from the historical monitoring data of the same batch to obtain the historical rate parameters.

[0085] Among them, the charge / discharge rate values ​​of historical application scenarios refer to the typical operating rates of the same batch of battery samples recorded in the historical database when they were deployed in other charging stations in the past; the historical rate parameters are the collection or statistical values ​​of these historical rate values.

[0086] Specifically, after the battery monitoring system obtains historical monitoring data for the same batch in step S202, it extracts the operating condition information associated with this data, including the charge and discharge rates in historical application scenarios. The system then performs statistical analysis on these historical rate data, such as calculating their average value or 95th percentile, to represent the typical load intensity of this batch of batteries in history. This statistical result is the historical rate parameter.

[0087] In some embodiments, this step can be implemented in several ways: Optionally, if application scenario tags (such as "energy storage at a bus terminal") are directly recorded in the historical data, the system can directly extract the standard rate corresponding to these tags; alternatively, the system can iterate through historical current data, combine it with the battery's rated capacity, calculate the instantaneous rate at each historical time point, and then perform statistical analysis on all historical rate data to obtain a comprehensive historical rate parameter. It is understood that other methods can also be used to extract the historical rate parameter, and this is not limited here.

[0088] In some embodiments, historical battery samples from the same batch may have been used in multiple completely different charging station scenarios, resulting in a very wide distribution of historical rate parameters. To address this, the battery monitoring system employs cluster analysis. The system processes the historical rate data using clustering algorithms such as K-means, dividing it into several typical rate clusters (e.g., a "low-rate peak-shaving and valley-filling cluster" and a "high-rate fast-charging support cluster"). In subsequent comparisons, the system compares the current rate parameter with the center of all historical rate clusters.

[0089] S206. Calculate the difference between the current magnification parameter and the historical magnification parameter to obtain the magnification difference.

[0090] Among them, the rate difference is the numerical difference between the charging and discharging rate of the current charging station operation mode and the historical typical rate, which is used to quantify the degree of change in application conditions.

[0091] Specifically, the battery monitoring system subtracts the current rate parameter obtained in step S204 from the historical rate parameter obtained in step S205, and takes the absolute value to obtain the rate difference. For example, if a batch of batteries was historically used primarily for 0.5C peak shaving and valley filling, but is now deployed to a fast charging station requiring 2C rate support, the rate difference is 1.5C. This difference directly reflects the magnitude of the impact of the new application scenario on the battery compared to its historical experience.

[0092] In some embodiments, this step can be implemented in several ways: Optionally, the relative difference can be calculated, i.e., (current multiplier - historical multiplier) / historical multiplier, which better reflects the relative magnitude of the change; alternatively, if the historical multiplier parameter is an interval or set, the shortest distance from the current multiplier to that interval or set can be calculated as the multiplier difference. It is understood that other methods can also be used to calculate the difference, which are not limited here.

[0093] S207. When the ratio difference is greater than the preset ratio threshold, add the instantaneous pressure drop parameter and the single-unit instantaneous temperature difference parameter to the key parameter set to obtain the extended parameter set.

[0094] Among them, the preset rate threshold is a threshold value used to determine whether the rate change is positive; the instantaneous voltage drop parameter refers to the instantaneous drop in battery terminal voltage when supporting the start of high-power fast charging; the instantaneous temperature difference of a single cell refers to the temperature difference between the cell with the highest temperature and the cell with the lowest temperature inside the battery pack at the same moment; the extended parameter set is a set of new parameters added to the original key parameter set.

[0095] Specifically, the battery monitoring system compares the rate difference calculated in step S206 with a preset rate threshold (e.g., 0.5C). If the difference is greater than the threshold, it indicates that the battery will be operating under a much more demanding load condition than its historical experience. In this case, the battery's polarization resistance and inconsistent heat generation become prominent. Therefore, the system automatically adds the two parameters that reflect these problems, "instantaneous voltage drop" and "instantaneous temperature difference of individual cells," to the key parameter set generated in step S203, forming an extended parameter set.

[0096] In some embodiments, this step can be implemented in several ways: Optionally, in addition to adding instantaneous voltage drop and temperature difference, the target sampling frequency of these two newly added parameters can be dynamically adjusted according to the magnitude of the rate difference; the larger the difference, the higher the allocated sampling frequency. Optionally, the list of added parameters can be more comprehensive, for example, including "current change rate (dI / dt)" itself, because the power support conditions of fast charging stations are often accompanied by rapid current fluctuations. It is understood that other more suitable monitoring parameters can also be added according to the specific battery chemistry system and application risk points, which are not limited here.

[0097] In some embodiments, even if the rate difference is small, the battery's State of Harmony (SOH) may already be low. In such cases, even medium-rate fast charging may trigger drastic internal changes. To address this, the battery monitoring system employs a dynamic, SOH-related preset rate threshold. This threshold decreases as the battery's SOH decreases. This means that for a severely aged battery, even if its operating rate changes little, the system may trigger an expansion of the parameter set, thereby achieving more cautious and sensitive monitoring of aging batteries.

[0098] S208. Based on the extended parameter set, determine the multiple monitoring parameters and corresponding acquisition frequency configurations for the tiered battery packs to obtain a differentiated acquisition frequency table.

[0099] The extended parameter set is a set generated in step S207 that may contain newly added parameters; the differentiated collection frequency table here is a collection strategy that is finally generated based on the extended parameter set and adapted to the current charging station operation mode.

[0100] Specifically, this step is similar to step S104, but its input is a dynamically adjusted set of extended parameters. The battery monitoring system allocates a higher sampling frequency to all parameters in the extended parameter set (including existing key parameters and newly added parameters such as instantaneous voltage drop and single-cell instantaneous temperature difference). For other non-key parameters, a lower sampling frequency is still allocated. Finally, the system outputs an updated, differentiated sampling frequency table that is better adapted to the current high-rate fast charging support conditions.

[0101] In some embodiments, frequent adjustments to the acquisition frequency can burden the local acquisition unit. To address this, the battery monitoring system employs a "smooth transition" strategy when issuing new differentiated acquisition frequency tables. Instead of instantly switching the frequencies of all parameters, the system sets a short transition period during which the frequency is gradually increased or decreased to the target value. This ensures a smooth transition of the monitoring system's state and avoids potential hardware and software instability issues caused by sudden configuration changes.

[0102] S209. Collect the operating data of the tiered battery packs according to the differentiated acquisition frequency table and upload them to the cloud to obtain real-time monitoring data streams.

[0103] Refer to step S105, which will not be repeated here.

[0104] In some embodiments, before collecting operating data in step S209, the battery monitoring system, in order to address the problem of uneven current distribution that may be caused by the mixed parallel connection of different batches of battery packs in a large-scale charging station energy storage system, reads the batch identifiers of all battery packs in the battery system containing the secondary battery packs to obtain a system batch identifier set; when different batch identifiers exist in the system batch identifier set, the internal resistance parameters of each battery pack in the battery system are extracted to obtain a system internal resistance data set; the difference between the internal resistance value of the secondary battery pack and the internal resistance value of other battery packs in the system internal resistance data set is calculated to obtain an internal resistance deviation value; based on the current shunting principle of parallel circuits, the current distribution correction coefficient of the secondary battery pack is calculated according to the internal resistance deviation value and the internal resistance value of the secondary battery pack.

[0105] Among them, the system batch identifier set refers to the list of batch identifiers of all battery packs connected in parallel under the same energy storage converter (PCS) in the charging station; the system internal resistance data set refers to the set of internal resistance values ​​of each of these parallel battery packs; and the current distribution correction factor is a multiplier factor used to correct the measured current.

[0106] Specifically, this process is executed when the charging station's energy storage system is expanded or some battery packs are replaced. The system first polls all battery packs on the parallel branch to obtain their batch identifiers. If different batches are found, the system determines it to be a mixed-batch parallel scenario. Next, the system instructs the BMS of each battery pack to report its current DC internal resistance. After obtaining all internal resistance values, for the monitored tiered battery pack, the system calculates the theoretical current distribution ratio of that battery pack based on the current shunting formula for parallel circuits. The current distribution correction factor is the ratio of the theoretical current distribution ratio to the average distribution ratio of the battery pack.

[0107] In some embodiments, this step can be implemented in several ways: Optionally, the internal resistance parameter is actively measured by the central monitoring system. The system controls the PCS to apply a brief, high-current pulse and simultaneously acquires the voltage and current changes of each branch battery pack at high frequency. The dynamic internal resistance of each pack is obtained by calculating ΔV / ΔI, which provides better consistency. Optionally, the calculation of the current distribution correction coefficient can be simplified. If only one target battery pack is monitored, it can be assumed that the internal resistance of all other battery packs is approximately equal and their average value is taken. Then, the correction coefficient is calculated based on the relationship between the internal resistance of the target battery pack and this average value. It is understood that more complex circuit models can also be used to calculate the correction coefficient, such as considering the effects of contact resistance and cable resistance, which is not limited here.

[0108] In some embodiments, the internal resistance of the battery pack may dynamically change with temperature and SOC, causing the static correction coefficient to become invalid after prolonged operation of the charging station. To address this, the battery monitoring system implements a "dynamic refresh mechanism for correction coefficients." The system continuously monitors the temperature and SOC of each battery pack at low frequency and has a built-in lookup table of internal resistance-temperature-SOC. When the temperature or SOC change of any battery pack exceeds a preset threshold, the system updates its estimated internal resistance value according to the lookup table and recalculates the current distribution correction coefficients for all battery packs.

[0109] It should be noted that the calculation logic of the current distribution correction coefficient strictly follows the physical laws of parallel circuits, that is, the current distribution is inversely proportional to the resistance of each branch. Specifically, in an energy storage unit composed of n battery packs connected in parallel, the target battery pack being monitored is the i-th pack, with an internal resistance of Ri. The internal resistances of the other battery packs are R1, R2, ..., Rn (excluding Ri). According to Kirchhoff's current law, the relationship between the theoretical current Ii flowing through the i-th battery pack and the total current Itotal is: Ii = Itotal * (1 / Ri) / (Σ(1 / Rj)), where Σ(1 / Rj) is the sum of the conductivities of all n battery packs. If we assume that the current is evenly distributed, then the average current Iavg of each battery pack is Itotal / n. Therefore, the current distribution correction coefficient (Coefficient_i) of the i-th battery pack is defined as the ratio of its theoretically distributed current to its average distributed current, that is:

[0110] The coefficient `Coefficient_i = Ii / Iavg = [Itotal * (1 / Ri) / (Σ(1 / Rj))] / [Itotal / n] = n * (1 / Ri) / (Σ(1 / Rj))`. This coefficient reflects how much more or less current the battery pack will carry compared to the average level. For example, in a dual-pack parallel system, pack A has an internal resistance of 10mΩ and pack B has an internal resistance of 15mΩ. The correction coefficient for pack A is 2 * (1 / 10) / (1 / 10 + 1 / 15) = 1.2, meaning it will carry 60% of the current, 20% higher than the average of 50%. In subsequent data processing, multiplying the measured current by this coefficient yields a "normalized" equivalent current, which is used for fair comparison with historical data distribution ranges based on standard battery performance.

[0111] S210. Determine the data distribution range based on historical monitoring data of the same batch, and mark abnormal data in the real-time monitoring data stream based on the data distribution range, and generate data monitoring records with abnormality marks.

[0112] Refer to step S106, which will not be repeated here.

[0113] Following step S209 above, in some embodiments, when the battery monitoring system marks abnormal data in step S210, in order to improve the accuracy of abnormal detection in the energy storage system of a charging station with mixed batches connected in parallel, the battery monitoring system will multiply the measured current value in the real-time monitoring data stream by the current distribution correction coefficient to obtain the corrected current value; extract the upper and lower limits of the historical current from the historical monitoring data of the same batch to obtain the historical current range; generate an abnormal mark when the corrected current value exceeds the historical current range, and attach the abnormal mark to the corresponding monitoring data record to obtain the data monitoring record with the abnormal mark.

[0114] Among them, the measured current value is the value directly read by the local acquisition unit from the current sensor; the corrected current value is the estimated value after correction of the measured current value, which represents the "equivalent" current under the influence of no parallel inconsistency; the historical current range is the normal fluctuation range of current parameters under the specific charging station operating conditions, which is based on the statistical analysis of historical data from the same batch.

[0115] Specifically, in the real-time data processing pipeline, for each piece of monitoring data, the battery monitoring system first performs current correction. The system multiplies the measured current value by the latest current allocation correction coefficient to obtain the corrected current value. Subsequently, the system executes the original anomaly detection logic, that is, based on the current operating conditions (such as supporting fast charging), it queries the corresponding historical current range and compares the corrected current value with this range. Only when the corrected current value exceeds the normal range will the system generate an anomaly flag.

[0116] In some embodiments, this step can be implemented in several ways: Optionally, instead of directly correcting the current value, the historical current range is corrected in reverse. That is, based on the correction coefficient, the general upper and lower limits extracted from the historical database are scaled to generate a personalized anomaly judgment threshold range for the current specific battery pack. Optionally, the anomaly marking can be graded. For example, when the measured current value exceeds the historical range but the corrected current value is still within the range, it is marked as "attention" level, indicating that there may be uneven parallel operation; only when the corrected current value also exceeds the historical range is it marked as "serious" level anomaly. It is understood that the specific comparison and marking logic can be designed according to the risk management strategy, and is not limited here.

[0117] S211. Based on historical monitoring data of the same batch and data monitoring records collected by the secondary battery pack in multiple complete charge and discharge cycles, calculate the health status data of the secondary battery pack.

[0118] Refer to step S107, which will not be repeated here.

[0119] S212. Based on the health status data, generate an authentication data package containing the battery's unique identifier, batch identifier, monitoring data summary, and current remaining capacity, and upload the authentication data package to the blockchain ledger.

[0120] Among them, the battery unique identifier is a globally unique ID; the monitoring data summary is a statistical summary of key monitoring data of the battery during the operation of the charging station, such as the number of times it supports fast charging and the maximum discharge rate; the authentication data package is a data structure that encapsulates the above information and is digitally signed; the blockchain ledger is a decentralized, tamper-proof distributed database that provides a trust basis for the transfer of charging station assets.

[0121] Specifically, after calculating the latest health status data in step S211, the battery monitoring system will trigger this step periodically (e.g., every quarter or when the SOH drops below a set threshold). The system combines the battery's unique identifier, batch identifier, latest SOH, current remaining capacity, and a summary of recent charging station operation data into an authentication data packet. Then, the system uses its private key to digitally sign the data packet and, by calling a smart contract, writes this signed data packet as a transaction record into the designated blockchain network.

[0122] In some embodiments, this step can be implemented in several ways: Optionally, to save on blockchain storage costs, the monitoring data summary can be in the form of a hash value. That is, detailed operational reports are stored on the company's cloud storage, and only the hash value of the report, along with core data such as SOH, is uploaded to the blockchain, achieving on-chain data notarization and off-chain storage. Optionally, smart contracts can be used to automate SOH updates. When a smart contract is invoked, it not only records data but also automatically sends a warning notification to the charging station operator based on preset rules (such as SOH falling below a certain value, indicating it is no longer suitable for fast charging). It is understood that the specific technical implementation for interacting with the blockchain can be diverse and is not limited here.

[0123] In some embodiments, frequent blockchain updates can lead to excessive costs. To address this, the battery monitoring system employs a milestone-based upload strategy. Instead of uploading data after each State of Health (SOH) calculation, the system sets trigger conditions, such as "cumulative discharge supported by fast charging reaches 1000 kWh," "SOH decreases by more than 2%," or "a serious anomaly occurs." Only when these milestone events occur will the system generate an authentication data packet and upload it to the blockchain.

[0124] S213. When the cascaded battery pack ends its current application and is scheduled for retirement, read the authentication data packet corresponding to the cascaded battery pack from the blockchain ledger and generate batch authentication information.

[0125] Among them, ending the current tier application means that the battery pack can no longer meet the current charging station's operational requirements due to capacity decay (such as being unable to effectively support fast charging); planned retirement means preparing to sell it to a recycler or transfer it to a slow charging station with lower performance requirements for continued use; batch certification information is a collection of all historical certification data packets read from the blockchain.

[0126] Specifically, when a charging station operator decides to decommission a battery pack in an energy storage cabinet, they can initiate a query request to the battery monitoring system or directly to the blockchain network through an authorized client application. The request contains the unique identifier of the battery pack. The system or application will traverse the blockchain ledger, retrieving all historical authentication data packets associated with that identifier. After organizing and unpacking these data packets in chronological order, an immutable batch authentication information is formed, covering the period from when the battery pack was put into use at the charging station to the present moment.

[0127] In some embodiments, this step can be implemented in several ways: Optionally, a public query portal can be provided, allowing any potential downstream buyer (such as other charging station operators or battery recyclers) to query the batch certification information of the battery after obtaining its unique identifier, increasing the transparency of asset transactions; alternatively, the generated batch certification information can be automatically formatted into a visual report, using charts to display the changes in SOH, remaining capacity over time or cycle count, and records of significant anomalies during charging station operation. It is understood that the data reading and display methods can be customized according to business needs and are not limited here.

[0128] In some embodiments, there may be situations where it is necessary to verify the authenticity and validity of a particular authentication message. For this purpose, the batch authentication message may contain the digital signature and transaction hash of the original data packet. The verifier can independently retrieve the original transaction data on the blockchain based on the transaction hash and use the public key of the battery monitoring system to verify the correctness of the digital signature. Only when the signature verification passes can the authenticity and tamper-proof nature of the authentication message be confirmed.

[0129] S214. Determine the applicable scope of the cascaded battery in the next stage of application based on the certification data package, and generate a decommissioning assessment report containing the applicable scope and batch certification information.

[0130] The scope of application in the next tier refers to assessing which charging station scenarios with lower requirements are suitable for the battery based on its current health status. For example, after being retired from a high-power fast charging station, is it suitable for use in a regular slow charging station in a commercial building, or can it only be used as backup power for low-power charging piles in rural areas? The retirement assessment report is a comprehensive document that provides a basis for decision-making regarding the disposal of assets for charging station operators.

[0131] Specifically, after acquiring complete batch certification information, the battery monitoring system activates an evaluation decision engine. This engine has a built-in rule base that associates different health status data ranges with different charging station application scenarios. For example, the rule base might define batteries with "SOH > 60%" as suitable for ordinary 7kW AC slow charging stations. The system matches the latest health status data from the certification information with the rule base to determine the next applicable tier of scenarios. Finally, the system integrates this applicable scope, complete batch certification information, and residual value estimation into a formatted retirement assessment report.

[0132] In some embodiments, this step can be implemented in several ways: Optionally, the retirement assessment report may include an economic analysis, i.e., calculating the return on investment in various applicable scenarios based on the potential revenue of different next-generation charging station applications and the predicted remaining battery life, providing economic reference for operator decision-making; Optionally, the report may be automatically pushed to a secondary battery trading platform, listing the battery asset and its assessment report for downstream buyers to browse and bid on. It is understood that the report content and subsequent processes can be deeply integrated with multiple systems such as asset management and trading, and this is not limited here.

[0133] It should be noted that the rule base in the aforementioned evaluation decision engine is an expert knowledge system. Its rule construction integrates battery engineering, industry standards, historical data mining, and economic analysis. The specific construction process is as follows: First, core battery performance requirements for different application scenarios (such as high-power fast charging stations, slow charging in commercial buildings, residential energy storage, and backup power for communication base stations) are collected, including but not limited to minimum capacity thresholds, maximum permissible internal resistance, charge / discharge rate capability, and required cycle life, forming a scenario requirement profile. Second, by analyzing a large amount of battery lifecycle data from new to retired in a historical database, the correlation between battery health status data (such as capacity SOH and internal resistance SOH) and their actual performance under different operating conditions is mined. Based on this, a series of "IF-THEN" form decision rules can be established.

[0134] For example, a rule could be:

[0135] "IF(Capacity SOH>65%ANDInternal Resistance SOH<180%ANDNumber of Historical High Rate Discharge Anomaly Markings<5 times)THEN(Recommended Application Scenarios='Commercial Building Peak Shaving and Valley Filling', 'Slow Charging Station Energy Storage')."

[0136] These rules not only consider the single SOH value but also integrate the battery's historical performance stability. Furthermore, the rule base incorporates economic models to calculate the residual value of the battery based on its expected remaining lifespan and cost per kilowatt-hour under different recommended scenarios. This ensures that the retirement assessment report includes not only a technical suitability assessment but also an economic value evaluation, providing comprehensive decision support for asset disposal.

[0137] In some embodiments, before collecting data, the battery monitoring system ensures that newly installed secondary battery packs, which have been removed from the warehouse and left to stand for a long time, can be safely put back into use. Specifically, the system calculates the standby time based on the previous decommissioning timestamp and the current access timestamp of the secondary battery pack. If the standby time exceeds a preset standby threshold, the system sets the secondary battery pack's operating mode to activation mode. In activation mode, the system collects operational data of the secondary battery pack during its first charge and discharge process at an activation collection frequency higher than the frequencies of each parameter in the differentiated collection frequency table, obtaining activation period monitoring data. The system calculates the standard deviation of the voltage change rate and temperature change rate in the activation period monitoring data to obtain fluctuation statistics. If the fluctuation statistics are all less than a preset steady-state threshold within a preset time range, the system switches the secondary battery pack's operating mode back to normal monitoring mode.

[0138] Among them, the last retirement timestamp is the time recorded when the battery was last stopped being used; the preset resting threshold is a threshold for judging whether the resting time is too long (such as 90 days); the activation mode is a temporary, high-specification monitoring state in which the battery will perform a complete, controlled charge and discharge cycle; the activation acquisition frequency is a much higher acquisition frequency than the normal operation mode (such as 5Hz); and the fluctuation statistics are indicators that quantify the smoothness of the voltage and temperature curves.

[0139] Specifically, when a new battery pack is connected to the charging station's energy storage system, the system calculates its resting time. If the resting time exceeds a threshold, the system determines that the battery needs to be activated. The system sets its operating mode to "activation mode" and instructs the local acquisition unit to collect all key parameters using a preset high frequency during the first complete charge-discharge cycle. The cloud system then analyzes the activation period monitoring data in real time, calculating the standard deviation of voltage and temperature change rates within a sliding time window. If this standard deviation remains below a small steady-state threshold for a period of time, it indicates that the electrochemical reaction inside the battery has stabilized, and the system automatically switches the operating mode back to "normal monitoring mode" defined by the differentiated acquisition frequency table, allowing it to officially participate in the operation of the charging station.

[0140] In some embodiments, this step can be implemented in several ways: Optionally, the activation process can be carried out in stages, for example, by first pre-charging with a small current to observe the stability of its voltage response, and then proceeding to standard current charge-discharge activation after the voltage stabilizes. Optionally, the calculation of fluctuation statistics can be more complex, for example, using the roughness or entropy value of the voltage curve as an indicator of volatility. It is understood that the activation strategy and stability judgment method can be customized according to the battery type and the length of resting time, and are not limited here.

[0141] In some embodiments, the battery may exhibit unstable characteristics during activation, but the fluctuation statistics may oscillate around the steady-state threshold, causing the system to frequently switch between "activation mode" and "normal monitoring mode." To address this, the battery monitoring system introduces a "state switching hysteresis" mechanism. That is, when the system determines that the conditions for switching to normal mode are met, it does not switch immediately but instead starts a confirmation timer. Only if the fluctuation statistics remain below the steady-state threshold throughout this confirmation period will the system perform the switching operation.

[0142] In some embodiments, during the activation mode acquisition process, in order to further enhance the early warning capability for potential risks during activation, the battery monitoring system extracts the historical voltage curve and historical temperature curve of the first activation period from the historical monitoring data of the same batch to obtain the activation period reference curve; based on the activation period reference curve, the safe threshold range of voltage and temperature during the activation period is calculated; during the activation acquisition frequency acquisition process, when it is determined that the real-time acquired voltage value or temperature value exceeds the safe threshold range, an abnormal early warning signal is generated.

[0143] Among them, the activation period reference curve refers to the set of curves recording the changes in voltage / temperature over time or SOC of batteries in the same batch during the historical activation process; the safety threshold range is a dynamic boundary generated based on the statistics of these historical curves; the abnormal warning signal is a high-priority alarm that will be immediately pushed to the operation and maintenance personnel of the charging station.

[0144] Specifically, when a battery pack enters activation mode, the battery monitoring system not only initiates high-frequency data acquisition but also simultaneously retrieves historical activation data for all battery packs in the same batch from the historical database. The system aligns these historical curves and calculates the maximum and minimum values ​​at each alignment point, thus forming a dynamically changing safety threshold range that changes with the activation process. During real-time activation, the system compares the real-time voltage and temperature values ​​acquired at high frequency with the safety threshold range corresponding to the current activation progress. If the real-time values ​​deviate from the safety range, the system can generate an abnormal warning signal and may automatically isolate the battery pack from the charging station's energy storage system.

[0145] In some embodiments, this step can be implemented in several ways: Optionally, the calculation of the safety threshold range can employ statistical methods, such as calculating the mean and standard deviation of historical data at each alignment point, using the mean ± 3 * standard deviation as the dynamic upper and lower boundaries. Optionally, in addition to comparing the numerical values ​​themselves, the shape of the curve can also be compared. The system can utilize algorithms such as Dynamic Time Warping (DTW) to calculate the similarity distance between the real-time activated curve segment and the historical reference curve set. When the distance exceeds the threshold, it can also be judged as a morphological anomaly and an early warning can be generated. It is understood that the early warning algorithm can combine multiple features for comprehensive judgment, which is not limited here.

[0146] In this embodiment, a comprehensive technical solution is adopted, which uses historical data analysis to dynamically generate differentiated collection frequency tables and combines this with the specific operating mode of the charging station (such as peak shaving and valley filling, fast charging support) and the inconsistencies in parallel operation within the energy storage system to adjust strategies and correct data. Therefore, this method can achieve refined and adaptive data monitoring based on the uniqueness of each tiered battery pack used for energy storage in a charging station and its specific operating conditions. This method effectively solves the problems of data redundancy, high cost, inaccurate monitoring, and insufficient security caused by the fixed monitoring strategy in existing technologies. Furthermore, it significantly reduces data monitoring costs while improving the security, transparency, and economic benefits of the entire lifecycle management of energy storage assets in charging stations through multi-dimensional and in-depth data analysis and verification (such as activation modes and blockchain evidence storage).

[0147] The battery monitoring system in the embodiments of this invention is described below from the perspective of hardware processing. Please refer to [link / reference]. Figure 3 This is a schematic diagram of the physical device structure of a battery monitoring system in an embodiment of this application.

[0148] It should be noted that, Figure 3 The structure of the battery monitoring system shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0149] like Figure 3 As shown, the battery monitoring system includes a CPU 301, which can perform various appropriate actions and processes according to a program stored in ROM 302 or a program loaded from storage section 308 into RAM 303, such as executing the methods described in the above embodiments. RAM 303 also stores various programs and data required for system operation. The CPU 301, ROM 302, and RAM 303 are interconnected via bus 304. I / O interface 305 is also connected to bus 304.

[0150] The following components are connected to I / O interface 305: input section 306 including audio input devices, push-button switches, etc.; output section 307 including liquid crystal display (LCD) and audio output devices, indicator lights, etc.; storage section 308 including hard disks, etc.; and communication section 309 including network interface cards such as LAN (Local Area Network) cards, modems, etc. Communication section 309 performs communication processing via a network such as the Internet. Drive 310 is also connected to I / O interface 305 as needed. Removable media 311, such as disks, optical disks, magneto-optical disks, semiconductor memories, etc., are installed on drive 310 as needed so that computer programs read from them can be installed into storage section 308 as needed.

[0151] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by CPU 301, it performs the various functions defined in the present invention.

[0152] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0153] Specifically, the battery monitoring system of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, it implements the cascade battery data monitoring method provided in the above embodiment.

[0154] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the battery monitoring system described in the above embodiments; or it may exist independently and not assembled into the battery monitoring system. The storage medium carries one or more computer programs that, when executed by a processor of the battery monitoring system, cause the battery monitoring system to implement the cascade battery data monitoring method provided in the above embodiments.

[0155] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0156] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

Claims

1. A method for monitoring data of cascaded batteries, characterized in that, The method, applied to a battery monitoring system, includes: Obtain the batch identifier and initial status parameters of the newly connected tiered battery packs to obtain the initial battery information; Based on the initial battery information, a matching historical battery sample is retrieved from the historical database, and the monitoring parameter sequence of the historical battery sample is extracted to obtain the historical monitoring data of the same batch. Calculate the time-series change rate of each monitoring parameter in the same batch of historical monitoring data, identify the monitoring parameters whose time-series change rate exceeds a preset fluctuation threshold as key monitoring parameters, and obtain a set of key parameters; Based on the set of key parameters, a sampling frequency configuration for the monitoring parameters corresponding to the secondary battery pack is generated to obtain a differentiated sampling frequency table. Specifically, this includes: obtaining the charge / discharge rate value of the secondary battery pack in the current application scenario to obtain the current rate parameter; extracting the charge / discharge rate value of the historical application scenario from the historical monitoring data of the same batch to obtain the historical rate parameter; calculating the difference between the current rate parameter and the historical rate parameter to obtain the rate difference; when the rate difference is greater than a preset rate threshold, adding the instantaneous voltage drop parameter and the single-cell instantaneous temperature difference parameter to the set of key parameters to obtain an extended parameter set; and determining multiple monitoring parameters and corresponding sampling frequency configurations corresponding to the secondary battery pack based on the extended parameter set to obtain a differentiated sampling frequency table. The operating data of the tiered battery packs are collected according to the differentiated collection frequency table and uploaded to the cloud to obtain a real-time monitoring data stream; Based on the historical monitoring data of the same batch, the data distribution range is determined, and based on the data distribution range, abnormal data in the real-time monitoring data stream is marked to generate a data monitoring record with abnormal markings. Based on the historical monitoring data of the same batch and the data monitoring records collected by the cascade battery pack in multiple complete charge and discharge cycles, the health status data of the cascade battery pack is calculated; Based on the health status data, an authentication data package containing a unique battery identifier, batch identifier, monitoring data summary, and current remaining capacity is generated, and the authentication data package is uploaded to the blockchain ledger. When the current application of the battery pack ends and it is planned to be decommissioned, the authentication data packet corresponding to the battery pack is read from the blockchain ledger to generate batch authentication information; Based on the certification data package, the applicable scope of the cascaded battery in the next stage of application is determined, and a decommissioning assessment report containing the applicable scope and the batch certification information is generated.

2. The method according to claim 1, characterized in that, Before the step of collecting the operating data of the tiered battery pack according to the differentiated acquisition frequency table and uploading it to the cloud to obtain a real-time monitoring data stream, the method further includes: Read the batch identifiers of all battery packs in the battery system containing the tiered battery pack to obtain the system batch identifier set; When different batch identifiers exist in the system batch identifier set, the internal resistance parameters of each battery pack in the battery system are extracted to obtain the system internal resistance data set; The difference between the internal resistance value of the cascaded battery pack and the internal resistance value of other battery packs in the system internal resistance data group is calculated to obtain the internal resistance deviation value. Based on the principle of current shunting in parallel circuits, the current distribution correction coefficient of the secondary battery pack is calculated according to the internal resistance deviation value and the internal resistance value of the secondary battery pack.

3. The method according to claim 2, characterized in that, The step of determining the statistical range of real-time values ​​deviating from historical values ​​and generating anomaly markers based on the real-time monitoring data stream and the historical monitoring data of the same batch, to obtain data monitoring records with anomaly markers, specifically includes: Multiply the measured current value in the real-time monitoring data stream by the current distribution correction coefficient to obtain the corrected current value; The upper and lower limits of the historical current are extracted from the historical monitoring data of the same batch to obtain the historical current range; When the corrected current value exceeds the historical current range, an anomaly marker is generated and attached to the corresponding monitoring data record to obtain a data monitoring record with an anomaly marker.

4. The method according to claim 1, characterized in that, The step of collecting the operating data of the tiered battery packs according to the differentiated acquisition frequency table and uploading it to the cloud to obtain a real-time monitoring data stream specifically includes: The resting time is calculated based on the timestamp of the last decommissioning and the timestamp of the current access of the tiered battery pack; When the resting time exceeds a preset resting threshold, the working mode of the cascade battery pack is set to the activation mode; In activation mode, the operating data of the cascaded battery pack during the first charge and discharge process is collected at an activation collection frequency higher than the frequency of each parameter in the differentiated collection frequency table to obtain activation period monitoring data. Calculate the standard deviation of the voltage change rate and temperature change rate in the monitoring data during the activation period to obtain the fluctuation statistics; When the fluctuation statistics are all less than the preset steady-state threshold within a preset time range, the working mode of the cascade battery pack is switched back to normal monitoring mode.

5. The method according to claim 4, characterized in that, After the step of collecting the operational data of the cascaded battery pack during its first charge and discharge process at an activation acquisition frequency higher than the frequency of each parameter in the differentiated acquisition frequency table to obtain activation period monitoring data, the method further includes: The historical voltage curve and historical temperature curve of the first activation period are extracted from the historical monitoring data of the same batch to obtain the activation period reference curve; Based on the activation period reference curve, the safe threshold range of voltage and temperature during the activation period is calculated. During the activation acquisition frequency acquisition process, if the real-time acquired voltage or temperature value exceeds the safety threshold range, an abnormal warning signal is generated.

6. A battery monitoring system, characterized in that, The battery monitoring system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code including computer instructions, and the one or more processors call the computer instructions to cause the battery monitoring system to perform the method as described in any one of claims 1-5.

7. A computer-readable storage medium comprising instructions, characterized in that, When the instructions are executed on the battery monitoring system, the battery monitoring system performs the method as described in any one of claims 1-5.

8. A computer program product, characterized in that, When the computer program product is run on the battery monitoring system, it causes the battery monitoring system to perform the method as described in any one of claims 1-5.

Images