A thermal runaway early warning and health protection system and method for energy storage batteries

Abstract

The application provides a thermal runaway early warning and health protection system and method for an energy storage battery. A cycle distribution model of a battery health state is constructed by using voltage, temperature and internal resistance data of the energy storage battery in a continuous charge and discharge cycle. A health deviation index of the current battery is obtained by performing feature matching on response parameters of the battery in the current charge and discharge cycle and the cycle distribution model. When the health deviation index is greater than a safety threshold of the energy storage battery, the sampling frequency gradient of voltage, temperature and internal resistance is increased to a specified test frequency. A thermal runaway early warning threshold is determined according to the voltage reduction rate, temperature rise rate and internal resistance change rate of each battery cell after each increase in the sampling frequency. When the health deviation index is greater than the thermal runaway early warning threshold, an abnormal battery cell is located by using the deviation of the voltage of each battery cell and the average voltage in the energy storage battery, and then a graded alarm is given to each abnormal battery cell. The above scheme can realize dynamic frequency determination of the thermal runaway early warning threshold in the energy storage battery.

Description

Technical Field

[0001] This application relates to the field of battery safety technology, and more specifically, to a system and method for early warning and health protection of thermal runaway in energy storage batteries. Background Technology

[0002] Energy storage batteries are energy storage devices that convert electrical energy into chemical energy through electrochemical reactions. They consist of multiple cells connected in series or parallel to form battery modules and battery clusters. They are characterized by high energy density, long cycle life, and fast response speed, and are widely used in scenarios such as grid frequency regulation, peak-valley arbitrage, and renewable energy grid connection.

[0003] In the field of thermal runaway early warning for energy storage batteries, rapid dynamic characteristics such as voltage drops, temperature surges, or internal resistance mutations within a millisecond to second timescale precede thermal runaway. However, existing technologies, in order to reduce the burden of data storage and transmission, employ a uniform low sampling frequency for voltage, temperature, and internal resistance parameters. This fixed and low sampling frequency has fundamental limitations in terms of time resolution; the time interval between adjacent sampling points is much longer than the duration of the precursory thermal runaway characteristics. This results in the incomplete capture of the starting point of the voltage drop, the inflection point of the temperature surge, and the instantaneous value of the internal resistance mutation. Consequently, key dynamic processes are severely smoothed or directly lost in the discrete sampling sequence, making it impossible to obtain the dense data points required to calculate the voltage drop rate, temperature rise rate, and internal resistance change rate. Consequently, thermal runaway early warning systems experience detection blind spots when facing rapidly developing precursory thermal runaway characteristics, making it difficult to issue accurate alarms within the effective warning window. Therefore, how to achieve dynamic frequency determination of the thermal runaway early warning threshold in energy storage batteries, thereby improving the real-time capture of precursory thermal runaway characteristics, has become a challenge for the industry. Summary of the Invention

[0004] This application provides a thermal runaway early warning and health protection system and method for energy storage batteries, which can realize the dynamic frequency determination of the thermal runaway early warning threshold in energy storage batteries, thereby improving the real-time capture of thermal runaway precursor characteristics of energy storage batteries.

[0005] In a first aspect, this application provides a method for early warning and health protection of thermal runaway in energy storage batteries, including: Collect voltage, temperature and internal resistance data of energy storage batteries during continuous charge and discharge cycles, extract steady-state operating characteristics under each charge and discharge cycle, and construct a periodic distribution model of battery health status. Monitor the battery's response parameters during the current charge / discharge cycle, and perform feature matching between the response parameters and the cycle distribution model to obtain the current battery's health deviation index; When the health deviation index is greater than the safety threshold of the energy storage battery, the sampling frequency gradient of voltage, temperature and internal resistance is increased to the specified test frequency. The thermal runaway warning threshold of the battery health status is determined based on the voltage drop rate, temperature rise rate and internal resistance change rate of each cell after each increase in sampling frequency. When the health deviation index is greater than the thermal runaway warning threshold, the abnormal cell is located by the deviation of the voltage and average voltage of each cell in the energy storage battery, and then the abnormal cell is given a graded alarm.

[0006] In some embodiments, extracting steady-state operating characteristics under each charge-discharge cycle and constructing a cycle distribution model of battery health status specifically includes: Median filtering is applied to the voltage, temperature, and internal resistance data within each charge-discharge cycle to remove transient fluctuation segments at the beginning and end of the charge-discharge phases, thus obtaining a steady-state data window for the middle segment of each charge-discharge cycle. Within each steady-state data window, the mean voltage, median temperature, and mode of internal resistance are calculated respectively, serving as the steady-state characteristic vector for the corresponding charge-discharge cycle; The steady-state eigenvectors of multiple consecutive charge-discharge cycles are arranged in chronological order, and a periodic distribution model of the battery health state is generated by using a moving average method.

[0007] In some embodiments, performing feature matching between the response parameters and the periodic distribution model to obtain the current battery health deviation index specifically includes: Extract the response parameters of the current charge / discharge cycle in the steady state phase. The response parameters include the average voltage, median temperature, and mode of internal resistance of the current charge / discharge cycle. The response parameters of the current charge / discharge cycle are compared with the expected interval of the same cycle number in the cycle distribution model, and the voltage deviation, temperature deviation and internal resistance deviation are calculated respectively to obtain the preliminary deviation value of the current charge / discharge cycle.

[0008] In some embodiments, the specified test frequency includes a first test frequency and a second test frequency, wherein the reference sampling frequency is denoted as the reference frequency, the first test frequency is set to twice the reference frequency, and the second test frequency is set to three times the reference frequency.

[0009] In some embodiments, determining the thermal runaway early warning threshold for battery health status based on the voltage drop rate, temperature rise rate, and internal resistance change rate of each cell after each increase in sampling frequency specifically includes: The sampling frequencies for voltage, temperature, and internal resistance are sequentially increased from the reference frequency to the first test frequency and the second test frequency by fixed multiples. Data is continuously collected at each test frequency within a preset time window. The voltage drop rate, temperature rise rate, and internal resistance change rate of each cell within the time window are calculated, and the average value of the corresponding rates of all cells is obtained. The average voltage drop rate, average temperature rise rate, and average internal resistance change rate at each test frequency are compared with the corresponding rate tolerance upper limit. When any two of the three exceed the tolerance upper limit, the current health deviation index is calibrated as the thermal runaway warning threshold of the battery health status.

[0010] In some embodiments, locating abnormal cells by the deviation between the voltage of each cell and the average voltage within the energy storage battery specifically includes: Calculate the average voltage of all cells in the energy storage battery at the current sampling time, and calculate the absolute deviation and relative deviation percentage of each cell voltage from the average voltage; Cells with a relative deviation percentage greater than the abnormal threshold are marked as suspected abnormal cells, and the spatial distribution clusters formed by all suspected abnormal cells are counted. Abnormal cells are located by means of the spatial distribution clusters and individual absolute deviations.

[0011] In some embodiments, a battery monitoring slave is used to monitor the battery's response parameters during the current charge / discharge cycle.

[0012] Secondly, this application provides a thermal runaway early warning and health protection system for energy storage batteries, including a graded alarm unit, wherein the graded alarm unit includes: The data acquisition module is used to collect voltage, temperature and internal resistance data of the energy storage battery during continuous charge and discharge cycles, extract steady-state operating characteristics under each charge and discharge cycle, and construct a periodic distribution model of the battery health status. The processing module is used to monitor the response parameters of the battery during the current charge-discharge cycle, perform feature matching between the response parameters and the cycle distribution model, and obtain the current battery health deviation index. The processing module is also used to increase the sampling frequency gradient of voltage, temperature and internal resistance to a specified test frequency when the health deviation index is greater than the safety threshold of the energy storage battery, and determine the thermal runaway warning threshold of the battery health state based on the voltage drop rate, temperature rise rate and internal resistance change rate of each cell after each increase in sampling frequency. The execution module is used to locate abnormal cells by measuring the deviation between the voltage and average voltage of each cell in the energy storage battery when the health deviation index is greater than the thermal runaway warning threshold, and then to issue graded alarms for each abnormal cell.

[0013] Thirdly, this application provides a computer device, the computer device including a memory and a processor, the memory for storing computer programs, and the processor for calling and running the computer programs from the memory, so that the computer device executes the above-described method for thermal runaway warning and health protection of energy storage batteries.

[0014] Fourthly, this application provides a computer-readable storage medium storing instructions or code that, when executed on a computer, cause the computer to implement the aforementioned method for early warning and health protection of thermal runaway in energy storage batteries.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: This application provides a thermal runaway early warning and health protection system and method for energy storage batteries. The system collects voltage, temperature, and internal resistance data of the energy storage battery during continuous charge-discharge cycles, extracts steady-state operating characteristics under each charge-discharge cycle, and constructs a periodic distribution model of the battery's health state. It monitors the battery's response parameters during the current charge-discharge cycle, performs feature matching between the response parameters and the periodic distribution model, and obtains the current battery health deviation index. When the health deviation index exceeds the energy storage battery's safety threshold, the sampling frequency gradient of voltage, temperature, and internal resistance is increased to a specified test frequency. Based on the voltage drop rate, temperature rise rate, and internal resistance change rate of each cell after each increase in sampling frequency, a thermal runaway early warning threshold for the battery's health state is determined. When the health deviation index exceeds the thermal runaway early warning threshold, abnormal cells are located by the deviation between the voltage and average voltage of each cell within the energy storage battery, and then graded alarms are issued for each abnormal cell.

[0016] Therefore, in this application, when the health deviation index is greater than the thermal runaway warning threshold, the abnormal cell is located by the deviation of the voltage and average voltage of each cell in the energy storage battery, and then the abnormal cells are given graded alarms. First, by determining the health deviation index, the steady-state deviation of the current health state of the energy storage battery relative to its normal aging trajectory can be obtained, thus providing a trigger condition for the dynamic frequency determination of the thermal runaway warning threshold. The health deviation index can effectively filter out the slow drift generated during the normal aging process of the battery. Only when the actual response parameters deviate significantly from the expected range will a large index value be output. The thermal runaway warning threshold does not need to run continuously in a fixed high-frequency mode. Instead, the subsequent sampling frequency gradient increase process is only started after the steady-state abnormality of the health deviation index exceeding the safety threshold is confirmed. This avoids unnecessary high-frequency sampling when the battery is in the normal aging trajectory. At the same time, it ensures that once an abnormal signal exceeding the normal fluctuation range occurs, it can immediately switch to the dynamic rate detection mode. Without increasing the data processing burden during the normal operation phase, it significantly shortens the response time from the occurrence of the abnormality to the start of high-frequency sampling. The delay improves the real-time capture of dynamic signals in the precursor features of thermal runaway. Then, by determining the thermal runaway warning threshold, the critical discrimination boundary of the energy storage battery under dynamic deterioration conditions can be obtained. The thermal runaway warning threshold itself carries the dynamic response information of the cell under the current abnormal state; that is, this threshold reflects whether the battery has further exhibited rapid deterioration characteristics at the rate level after experiencing steady-state deviation. When the health deviation index exceeds this dynamically calibrated threshold, it can be confirmed that the abnormality has evolved from cumulative aging deviation to a rapid change stage with precursor features of thermal runaway. This avoids misjudgment or missed judgment caused by aging or environmental changes due to fixed thresholds, ensuring that an alarm is triggered only when any two of the true values ​​of the voltage drop rate, temperature rise rate, and internal resistance change rate exceed the tolerance limit. This allows the entire thermal runaway warning threshold determination process to focus on the dynamic deterioration window period with the highest time resolution, improving the real-time capture of transient change signals in the precursor features of thermal runaway. In summary, based on the above scheme, the dynamic frequency determination of the thermal runaway warning threshold in energy storage batteries can be realized, thereby improving the real-time capture of precursor features of thermal runaway in energy storage batteries. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is an exemplary flowchart of a method for early warning and health protection of thermal runaway in energy storage batteries according to some embodiments of this application; Figure 2 This is a flowchart illustrating the process of determining the thermal runaway early warning threshold according to some embodiments of this application; Figure 3 This is a schematic diagram of the structure of a hierarchical alarm unit according to some embodiments of this application; Figure 4 This is a schematic diagram of the structure of a computer device for implementing a method for early warning and health protection of thermal runaway of energy storage batteries, according to some embodiments of this application. Detailed Implementation

[0019] To better understand the technical solution of this application, the technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0020] refer to Figure 1 The figure is an exemplary flowchart of a method for early warning and health protection of thermal runaway in energy storage batteries according to some embodiments of this application. The method mainly includes the following steps: In step 101, voltage, temperature and internal resistance data of the energy storage battery are collected during continuous charge and discharge cycles, steady-state operating characteristics under each charge and discharge cycle are extracted, and a periodic distribution model of battery health status is constructed.

[0021] It should be noted that in this application, voltage data is a physical quantity that measures the potential difference between the positive and negative electrodes inside the energy storage battery; temperature data is a physical quantity that measures the internal thermal state of the energy storage battery; and internal resistance data is a physical quantity that measures the resistance to current inside the energy storage battery.

[0022] In practice, firstly, before the energy storage battery system starts operating, a reference sampling frequency is pre-set, with a default setting of 10 data collections per second. A single charge-discharge cycle is defined as the process from full charge to discharge and back to full charge. After the energy storage battery begins operation, the number of consecutive charge-discharge cycles experienced by the battery is continuously recorded, starting from the first cycle and sequentially numbered as cycle number one, cycle number two, cycle number three, and so on. Within each charge-discharge cycle, three categories of data are collected simultaneously according to the preset reference sampling frequency: the first category is the data between the positive and negative terminals of each cell at each sampling moment. The first category is voltage values; the second category is temperature values ​​collected by temperature sensors placed on the surface of each cell or at the terminals; the third category is to obtain the internal resistance value of each cell at the current sampling moment by injecting a short-term AC excitation current into the battery and measuring the response voltage, or by calculating the ratio of the voltage change to the current change at the moment of charge-discharge switching. Thus, the voltage, temperature, and internal resistance data at each sampling moment can be obtained. According to the chronological order of sampling time, these data are sequentially associated with the current cycle number and cell number for storage, thereby obtaining the voltage, temperature, and internal resistance data for multiple consecutive charge-discharge cycles.

[0023] In some embodiments, extracting the steady-state operating characteristics under each charge-discharge cycle and constructing a cycle distribution model of the battery health state can be achieved by the following steps: Median filtering is applied to the voltage, temperature, and internal resistance data within each charge-discharge cycle to remove transient fluctuation segments at the beginning and end of the charge-discharge phases, thus obtaining a steady-state data window for the middle segment of each charge-discharge cycle. Within each steady-state data window, the mean voltage, median temperature, and mode of internal resistance are calculated respectively, serving as the steady-state characteristic vector for the corresponding charge-discharge cycle; The steady-state eigenvectors of multiple consecutive charge-discharge cycles are arranged in chronological order, and a periodic distribution model of the battery health state is generated by using a moving average method.

[0024] It should be noted that in this application, the steady-state data window is a continuous data interval that can represent the stable working state of the battery within a charge-discharge cycle; the steady-state feature vector is a combination of central tendency statistics that characterize the health state of the battery under a single charge-discharge cycle; and the periodic distribution model is a statistical reference benchmark used to describe the value range and normal fluctuation boundary of the three steady-state features of voltage, temperature and internal resistance of the energy storage battery in multiple consecutive charge-discharge cycles.

[0025] In practice, firstly, for each charge / discharge cycle in which raw data has been collected, all voltage, temperature, and internal resistance values ​​recorded in chronological order of sampling time within that cycle are extracted. Taking a specific charge / discharge cycle as an example, assuming the cycle starts charging at 8:00 AM and ends discharging at 10:00 AM, with a total duration of two hours, the start time of the cycle is identified as 8:00 AM and the end time as 10:00 AM. A fixed removal time is set, and by default, the first ten minutes of the start phase and the last ten minutes of the end phase are marked as transient fluctuation segments. From the total data of the cycle, all data points collected within the last ten minutes after the start time are removed, and all data points collected within the last ten minutes before the end time are also removed. For the remaining intermediate data, median filtering is performed on the voltage, temperature, and internal resistance data sequences respectively. The specific method of median filtering is as follows: for the voltage data sequence, three consecutive sampling points are taken in sequence, and the voltage values ​​of the three points are arranged in ascending order. The value at the middle position is taken as the new voltage value at the center position of the three points, and the process is repeated. The above operation is repeated for each sampling point, and the temperature and internal resistance data are processed in the same way. After median filtering, the remaining continuous data interval with abnormal fluctuations filtered out for each charge-discharge cycle is taken as the steady-state data window for that cycle. Then, for each steady-state data window of the charge-discharge cycle, statistical calculations are performed on all voltage, temperature, and internal resistance values ​​within the window. The voltage mean is calculated by summing the voltage values ​​at all sampling times within the steady-state data window to obtain the voltage sum, and then dividing the voltage sum by... The quotient obtained by taking the total number of voltage sampling points within the window is the average voltage for that period. To calculate the temperature median: arrange all temperature values ​​within the steady-state data window in ascending order, and find the temperature value in the middle. If the total number of temperature sampling points within the window is odd, directly take the middle value; if the total number is even, take the average of the two middle values. To calculate the mode of internal resistance: group all internal resistance values ​​within the steady-state data window, for example, grouping the internal resistance values ​​according to the range from 0.5 milliohms to 2 milliohms, starting with 0.The system is divided into several groups with an interval of 0.01 milliohms. The number of internal resistance values ​​appearing in each interval is counted. The internal resistance value corresponding to the interval with the most occurrences, or the median value of that interval, is the mode of internal resistance. The average voltage, median temperature, and mode of internal resistance are arranged in a fixed order to form the steady-state feature vector for that charge-discharge cycle. Finally, the steady-state feature vectors corresponding to each cycle from the first, second, third, up to the Nth charge-discharge cycle are continuously collected. Arranging the steady-state feature vectors in ascending order of cycle number yields a time series. A moving average method is used to process this time series: a moving window length is set, for example, five cycles. Starting from the first cycle number, the steady-state feature vectors for each cycle are obtained. The arithmetic mean of the average voltage for each of the five cycles is calculated to obtain the first moving average voltage value; the arithmetic mean of the median temperature for each of the five cycles is calculated to obtain the second moving average voltage value. A moving average temperature value is obtained. The arithmetic mean of the mode of internal resistance over five cycles is calculated to obtain the first moving average internal resistance value. Simultaneously, the standard deviation of the mean voltage, median temperature, and mode of internal resistance over the five cycles is calculated to obtain the voltage fluctuation range, temperature fluctuation range, and internal resistance fluctuation range. The sliding window is then moved forward one cycle, and the steady-state characteristic vectors of cycles two, three, four, five, and six are taken. The above operations of calculating the average and standard deviation are repeated to obtain the second moving average voltage value, moving average temperature value, moving average internal resistance value, and corresponding fluctuation range. Each sliding movement generates a set of average values ​​and fluctuation range data. After the sliding window has traversed all consecutive cycles, the average value sequence and fluctuation range sequence corresponding to all sliding windows are combined to obtain a reference trajectory that varies with the cycle number. Each cycle number corresponds to a desired voltage range, desired temperature range, and desired internal resistance range. This reference trajectory serves as a periodic distribution model of the battery's health state.

[0026] In step 102, the response parameters of the battery during the current charge-discharge cycle are monitored, and the response parameters are matched with the cycle distribution model to obtain the current battery health deviation index.

[0027] It should be noted that in this application, a battery monitoring slave device is used to monitor the response parameters of the battery during the current charge-discharge cycle. The battery monitoring slave device is a front-end data acquisition device used to independently complete the real-time acquisition and preliminary processing of physical signals such as voltage, temperature, and internal resistance in the energy storage battery system, and upload the acquisition results to the upper-level host. The response parameters are a set of dynamic measurement values ​​that characterize the real-time voltage change, temperature change, and internal resistance change of the energy storage battery in response to the external charge-discharge current during the current charge-discharge cycle.

[0028] In practice, firstly, a battery monitoring slave is pre-installed inside each battery module of the energy storage battery. This slave is directly electrically connected to the positive and negative output terminals of the energy storage battery and is also connected via wires to temperature sensors arranged on the surface of each cell. After the energy storage battery enters its current charge / discharge cycle, the slave triggers a sampling action at a pre-set reference frequency, for example, 10 times per second. At each sampling moment, the slave performs three measurement actions sequentially: the first is that the slave measures the potential difference between the positive and negative terminals of each cell through its internal high-impedance sampling circuit, reading the voltage value at that moment; the second is that the slave reads the current temperature value of the surface of each cell through the connected temperature sensors; the third is... The battery monitoring slave device briefly injects an AC test current of known frequency and amplitude into the battery circuit, while simultaneously measuring the AC voltage response across the battery terminals. The internal resistance value of each cell at that moment is obtained by dividing the AC voltage amplitude by the AC current amplitude. The slave device temporarily stores the collected voltage, temperature, and internal resistance values ​​of each cell, along with the current charge / discharge cycle number and the timestamp of the current sampling time, in its internal register. The slave device continuously repeats this acquisition process until the end of the current charge / discharge cycle. During the steady-state phase of the current charge / discharge cycle (excluding the first five minutes after charging begins and the last five minutes before discharging ends), the slave device packages all the voltage, temperature, and internal resistance data collected during the steady-state phase as a complete set of response parameters for the next charge / discharge cycle.

[0029] In some embodiments, the health deviation index of the current battery can be obtained by feature matching between the response parameters and the periodic distribution model using the following steps: Extract the response parameters of the current charge / discharge cycle in the steady state phase. The response parameters include the average voltage, median temperature, and mode of internal resistance of the current charge / discharge cycle. The response parameters of the current charge / discharge cycle are compared with the expected interval of the same cycle number in the cycle distribution model, and the voltage deviation, temperature deviation and internal resistance deviation are calculated respectively to obtain the preliminary deviation value of the current charge / discharge cycle. The current battery health deviation index is determined by the initial deviation values ​​of three adjacent cycles.

[0030] It should be noted that, in this application, the response parameter is a statistical quantity characterizing the actual working state of the energy storage battery in the steady state stage of the current charge-discharge cycle; the preliminary deviation value is a single value that measures the comprehensive degree of deviation of the actual response parameter of the energy storage battery from the expected interval in the cycle distribution model in the current single charge-discharge cycle; and the health deviation index is a comprehensive index used to assess the overall degree of deviation of the health status of the energy storage battery from the normal aging trajectory in the most recent consecutive charge-discharge cycles.

[0031] In practice, firstly, the cycle number of the current charge / discharge cycle is identified, for example, cycle number 50. From all response parameters, the voltage, temperature, and internal resistance values ​​that have entered the steady-state phase are selected, avoiding the first 5 minutes after charging begins and the last 5 minutes before discharging ends. The average voltage value is calculated for all voltage values ​​within the steady-state phase, which is achieved by summing all voltage values ​​and dividing by the total number of voltage sampling points. The median temperature value is calculated for all temperature values ​​collected within the steady-state phase, which is achieved by arranging all temperature values ​​in ascending order and taking the middle value. The mode of internal resistance value is calculated for all internal resistance values ​​collected within the steady-state phase, which is achieved by dividing the internal resistance values ​​in increments of 0.01 milliohms. After dividing the system into several groups, the interval with the most frequent occurrence is identified, and the median value of this interval is taken as the mode of internal resistance. The calculated average voltage, median temperature, and mode of internal resistance are combined as the response parameters for the current charge / discharge cycle. Then, based on the cycle number of the current charge / discharge cycle, for example, cycle number 50, the three expected intervals corresponding to this cycle number are found from the already constructed cycle distribution model: the voltage expected interval, the temperature expected interval, and the internal resistance expected interval. Each expected interval is defined by a lower limit and an upper limit. The actual average voltage, the lower limit of the voltage expectation, and the upper limit of the voltage expectation are taken out, and the voltage deviation is calculated as follows: If the actual average voltage is... The voltage deviation is 0 if the voltage is between the lower and upper expected limits. If the actual average voltage is less than the lower expected limit, the voltage deviation is equal to the lower expected limit minus the actual average voltage, then divided by the lower expected limit. If the actual average voltage is greater than the upper expected limit, the voltage deviation is equal to the actual average voltage minus the upper expected limit, then divided by the upper expected limit. The same logic is used to calculate the temperature deviation and internal resistance deviation, respectively. This involves comparing the median actual temperature with the expected temperature range and the mode of actual internal resistance with the expected internal resistance range, resulting in three deviation values ​​between 0 and 1. The voltage deviation is then multiplied by 0.4, the temperature deviation by 0.3, and the internal resistance... Multiply the deviation by 0.3 and add the three products together to get the initial deviation value for the current charge / discharge cycle. Finally, store the initial deviation values ​​for the three most recent charge / discharge cycles in memory. For example, the initial deviation value for cycle number 50 has been calculated. At the same time, take the initial deviation values ​​for cycle number 49 and cycle number 48. Arrange the initial deviation values ​​of the three adjacent cycles in ascending order of cycle number to form a sequence containing three values. First, calculate the arithmetic mean of these three initial deviation values, and then find the maximum value among these three values. The health deviation index is equal to the average deviation value plus half of the difference between the maximum value and the average deviation value. This gives the current battery health deviation index.

[0032] In step 103, when the health deviation index is greater than the safety threshold of the energy storage battery, the sampling frequency gradient of voltage, temperature and internal resistance is increased to the specified test frequency, and the thermal runaway warning threshold of the battery health state is determined based on the voltage drop rate, temperature rise rate and internal resistance change rate of each cell after each increase in sampling frequency.

[0033] It should be noted that in this application, the specified test frequency includes a first test frequency and a second test frequency, wherein the reference sampling frequency is denoted as the reference frequency, the first test frequency is set to twice the reference frequency, and the second test frequency is set to three times the reference frequency.

[0034] In practice, when the health deviation index exceeds the safety threshold of the energy storage battery, the sampling frequency gradient of voltage, temperature, and internal resistance is increased to a specified test frequency. The safety threshold is a warning line set based on steady-state characteristics, with a default value of 0.5, used to distinguish between normal aging fluctuations of the battery and abnormal deviations that require attention. A health deviation index greater than the safety threshold indicates that the battery's health state has exceeded the baseline fluctuation range described by the periodic distribution model, and the battery is in an early abnormal stage, but has not yet shown signs of thermal runaway. If a lower baseline sampling frequency is continued to be used, it may be impossible to capture transient dynamics such as voltage drops, temperature rises, or internal resistance changes that occur in the abnormal cell within a short period of time. In response, by gradually increasing the sampling frequency to the first and second test frequencies, the time interval between adjacent sampling points can be significantly shortened, making the data acquisition density sufficient to analyze rapidly changing electrochemical processes. On the other hand, using a graded increase rather than jumping directly to the highest frequency can reduce the data processing burden when the anomaly is low, while providing sufficient time resolution for subsequent calculations of the voltage drop rate, temperature rise rate, and internal resistance change rate. Therefore, the gradient increase of the sampling frequency dynamically adjusts the observation accuracy according to the severity of the anomaly, ensuring that potential dynamic deterioration signals can be captured at a higher time density after confirming the presence of steady-state deviation in the battery, thereby calibrating the accurate thermal runaway early warning threshold.

[0035] In some embodiments, the thermal runaway early warning threshold for battery health status is determined based on the voltage drop rate, temperature rise rate, and internal resistance change rate of each cell after each increase in sampling frequency, with reference to... Figure 2 The diagram is a flowchart illustrating the process of determining the thermal runaway warning threshold in some embodiments of this application. In this embodiment, the thermal runaway warning threshold can be determined using the following steps: In step 1031, the sampling frequencies of voltage, temperature and internal resistance are sequentially increased from the reference frequency to the first test frequency and the second test frequency by a fixed multiple. In step 1032, data is continuously collected at each test frequency for a preset time window, and the voltage drop rate, temperature rise rate and internal resistance change rate of each cell within the time window are calculated to obtain the average value of the corresponding rates of all cells. In step 1033, the average voltage drop rate, average temperature rise rate, and average internal resistance change rate at each test frequency are compared with the corresponding rate tolerance upper limit. When any two of the three exceed the tolerance upper limit, the current health deviation index is calibrated as the thermal runaway warning threshold of the battery health status.

[0036] It should be noted that, in this application, the voltage drop rate is a physical quantity that measures how quickly the battery voltage drops per unit time; the temperature rise rate is a physical quantity that measures how quickly the battery temperature rises per unit time; the internal resistance change rate is a physical quantity that measures how quickly the battery internal resistance changes per unit time; the rate tolerance upper limit is a set of pre-set boundary values ​​used to determine whether the voltage drop rate, temperature rise rate, or internal resistance change rate is within a safe range; and the thermal runaway warning threshold is a critical value of a health deviation index used to trigger the thermal runaway alarm program of the energy storage battery.

[0037] In practice, firstly, the sampling frequencies for voltage, temperature, and internal resistance are sequentially increased from the reference frequency by fixed multiples, initially to the first test frequency for data acquisition, and then increased to the second test frequency after a period of continuous acquisition. Next, at each test frequency, a preset time window is set, with a default continuous data acquisition period of 10 seconds. Within this time window, the voltage, temperature, and internal resistance values ​​of each cell at the start and end of the time window are recorded. For the voltage drop rate, the difference between the initial and final voltage values ​​is divided by the length of the time window; for the temperature rise rate, the difference between the final and initial temperature values ​​is divided by the length of the time window; and for the internal resistance change rate, the difference between the final and initial internal resistance values ​​is divided by the length of the time window. After calculating these three rates for all cells in the energy storage battery, the voltage drop rates of all cells are summed and divided by the total number of cells to obtain the average voltage drop rate. The temperature rise rates of all cells are summed and divided by the average number of cells. The average temperature rise rate is obtained by counting the total number of cells. The average internal resistance change rate is obtained by summing the internal resistance change rates of all cells and dividing by the total number of cells. Finally, a rate tolerance upper limit is set for the voltage drop rate, temperature rise rate, and internal resistance change rate. The default tolerance upper limit for the voltage drop rate is 0.5 volts per minute, the tolerance upper limit for the temperature rise rate is 2 degrees Celsius per minute, and the tolerance upper limit for the internal resistance change rate is 0.1 milliohms per minute. The three average values ​​calculated at the first test frequency are compared with their respective rate tolerance upper limits. Then, the three average values ​​calculated at the second test frequency are compared with their respective rate tolerance upper limits. For the comparison results at each test frequency, it is checked whether any two of the three indicators (average voltage drop rate, average temperature rise rate, and average internal resistance change rate) exceed the corresponding rate tolerance upper limit. If any two indicators exceed the tolerance upper limit at the first or second test frequency, the calculated health deviation index is marked as the thermal runaway warning threshold for the battery health status.

[0038] In step 104, when the health deviation index is greater than the thermal runaway warning threshold, the abnormal cell is located by the deviation of the voltage and average voltage of each cell in the energy storage battery, and then the abnormal cell is given a graded alarm.

[0039] It should be noted that in this application, when the health deviation index is greater than the thermal runaway warning threshold, the health deviation index represents the degree of deviation of the current overall state of the battery from the normal aging trajectory, while the thermal runaway warning threshold is a critical value calibrated by measuring the dynamic response rate of the battery after increasing the sampling frequency and comparing it with the tolerance upper limit. This critical value is essentially the discrimination boundary for irreversible abnormal heat generation or internal short circuits and other precursory events of thermal runaway that have already occurred inside the battery. Since the health deviation index is a long-term trend indicator calculated based on steady-state characteristics, while the thermal runaway warning threshold is a short-term drastic change indicator calibrated based on dynamic rate indicators, the two are complementary in a physical sense. Therefore, when the health deviation index exceeds the thermal runaway warning threshold, the battery not only has cumulative aging deviation at the steady-state level, but also has rapid deterioration characteristics at the dynamic level. That is, the abnormality of the battery has entered the thermal runaway incubation stage with a significantly increased rate from the slow capacity decay or internal resistance growth stage. At this time, triggering the deviation positioning and graded alarm at the discharge cell level has sufficient physical basis and temporal redundancy.

[0040] In some embodiments, locating abnormal cells by the deviation between the voltage of individual cells and the average voltage within the energy storage battery can be achieved through the following steps: Calculate the average voltage of all cells in the energy storage battery at the current sampling time, and calculate the absolute deviation and relative deviation percentage of each cell voltage from the average voltage; Cells with a relative deviation percentage greater than the abnormal threshold are marked as suspected abnormal cells, and the spatial distribution clusters formed by all suspected abnormal cells are counted. Abnormal cells are located by means of the spatial distribution clusters and individual absolute deviations.

[0041] It should be noted that, in this application, absolute deviation is a numerical value that measures the difference between the voltage of a single cell and the average voltage of all cells; relative deviation percentage is a percentage that measures the relative degree to which the voltage of a single cell deviates from the average voltage of all cells; anomaly threshold is a pre-set relative deviation percentage threshold used to determine whether a cell belongs to a suspected abnormal cell; suspected abnormal cell is used to mark individual cells that need to be judged as abnormal; spatial distribution cluster is used to describe the spatial distribution pattern of the adjacent area formed by multiple suspected abnormal cells in the physical arrangement of the energy storage battery.

[0042] In practice, firstly, the voltage values ​​of all cells in the energy storage battery at the current sampling time are obtained. Assuming there are M cells in the energy storage battery, these M voltage values ​​are added together and divided by M to obtain the average voltage. Then, for each cell, the absolute deviation and relative deviation percentage are calculated. The absolute deviation is calculated by taking the absolute value of the difference between the cell voltage and the average voltage, regardless of the sign. The relative deviation percentage is calculated by first subtracting the average voltage from the cell voltage, then dividing the difference by the average voltage, and finally multiplying by 100% to obtain a percentage. Next, an anomaly threshold is preset, defaulting to 5%. Each cell in the energy storage battery is iterated through, and the relative deviation percentage of each cell is compared with the anomaly threshold. If the relative deviation percentage of a cell is greater than 5%, the cell is marked as a suspected anomaly cell. After marking all cells, the spatial distribution clusters formed by the suspected anomaly cells are identified based on their physical arrangement within the battery module. The spatial distribution clusters are identified by grouping adjacent cells together. Suspected abnormal cells arranged in a continuous sequence are grouped into the same spatial distribution cluster. For example, in a battery pack of 100 cells arranged in 10 rows and 10 columns, if the cells in the 3rd row and 4th column, the 3rd row and 5th column, and the 4th row and 4th column are all marked as suspected abnormal cells and they are adjacent to each other, then these three cells together form a spatial distribution cluster. Finally, for each identified spatial distribution cluster, the number of suspected abnormal cells contained in the cluster and the absolute deviation value of each cell are counted. The cells in the cluster are sorted in descending order of absolute deviation. The cell with the largest absolute deviation is marked as the core abnormal cell, and the remaining cells with smaller absolute deviations are marked as secondary abnormal cells. At the same time, the shape and range of the spatial distribution cluster are considered. If a spatial distribution cluster covers a large area, such as more than 5 cells arranged in a line or forming a rectangular area of ​​2 rows and 3 columns, then it is determined that there is a diffuse anomaly in the area. The position of the cell with the largest absolute deviation in each spatial distribution cluster and the distribution range of the entire cluster are used as the output result of abnormal cell deviation localization.

[0043] Furthermore, in another aspect of this application, in some embodiments, this application provides a thermal runaway early warning and health protection system for energy storage batteries. This system includes a graded alarm unit, as referenced... Figure 3 The figure is a schematic diagram of the structure of a hierarchical alarm unit according to some embodiments of this application. The hierarchical alarm unit includes: a data acquisition module 201, a processing module 202, and an execution module 203, which are described below: The acquisition module 201 in this application is mainly used to acquire voltage, temperature and internal resistance data of the energy storage battery during continuous charge and discharge cycles, extract steady-state operating characteristics under each charge and discharge cycle, and construct a periodic distribution model of battery health status. Processing module 202, in this application, is used to monitor the response parameters of the battery during the current charge-discharge cycle, and perform feature matching between the response parameters and the cycle distribution model to obtain the current battery health deviation index; It should be noted that the processing module 202 is also used to increase the sampling frequency gradient of voltage, temperature and internal resistance to a specified test frequency when the health deviation index is greater than the safety threshold of the energy storage battery, and to determine the thermal runaway warning threshold of the battery health state based on the voltage drop rate, temperature rise rate and internal resistance change rate of each cell after each increase in sampling frequency. The execution module 203 in this application is mainly used to locate abnormal cells by the deviation of the voltage and average voltage of each cell in the energy storage battery when the health deviation index is greater than the thermal runaway warning threshold, and then to perform graded alarms on each abnormal cell.

[0044] The foregoing has detailed examples of the energy storage battery thermal runaway early warning and health protection system and method provided in the embodiments of this application. It is understood that the corresponding device, in order to achieve the above functions, includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should readily recognize that, in conjunction with the units and algorithm steps of the various examples described in the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0045] In some embodiments, this application also provides a computer device, the computer device including a memory and a processor, the memory for storing a computer program, and the processor for calling and running the computer program from the memory, so that the computer device performs the above-described method for thermal runaway warning and health protection of energy storage batteries.

[0046] In some embodiments, reference Figure 4 The dashed lines in the figure indicate that the unit or module is optional. This figure is a structural schematic diagram of a computer device for implementing a method for early warning and health protection of thermal runaway in energy storage batteries according to an embodiment of this application. The method for early warning and health protection of thermal runaway in energy storage batteries described in the above embodiments can be achieved through… Figure 4 The computer device shown is used to implement this, and the computer device includes at least one processor 301, a memory 302 and at least one communication unit 305. The computer device may be a terminal device, a server or a chip.

[0047] Processor 301 can be a general-purpose processor or a special-purpose processor. For example, processor 301 can be a central processing unit (CPU), which can be used to control computer devices, execute software programs, and process data from software programs. The computer device may also include a communication unit 305 for inputting (receiving) and outputting (transmitting) signals.

[0048] For example, the computer device may be a chip, and the communication unit 305 may be the input and / or output circuit of the chip, or the communication unit 305 may be the communication interface of the chip, which may be a component of a terminal device, network device or other device.

[0049] For example, the computer device may be a terminal device or a server, and the communication unit 305 may be a transceiver of the terminal device or the server, or the communication unit 305 may be a transceiver circuit of the terminal device or the server.

[0050] The computer device may include one or more memories 302 storing a program 304. The program 304 can be executed by a processor 301 to generate instructions 303, causing the processor 301 to execute the method described in the above method embodiments according to the instructions 303. Optionally, the memory 302 may also store data (such as a target audit model). Optionally, the processor 301 may also read data stored in the memory 302, which may be stored at the same storage address as the program 304, or it may be stored at a different storage address than the program 304.

[0051] The processor 301 and memory 302 can be configured separately or integrated together, for example, integrated on the system on chip (SOC) of the terminal device.

[0052] It should be understood that each step of the above method embodiment can be completed by hardware logic circuits or software instructions in the processor 301. The processor 301 can be a CPU, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, such as discrete gate, transistor logic devices, or discrete hardware components.

[0053] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0054] For example, in some embodiments, this application also provides a computer-readable storage medium storing instructions or code that, when executed on a computer, cause the computer to implement the above-described method for thermal runaway warning and health protection of energy storage batteries.

[0055] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0056] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for early warning and health protection of thermal runaway in energy storage batteries, characterized in that, Includes the following steps: The voltage, temperature and internal resistance data of the energy storage battery are collected during continuous charge and discharge cycles. The steady-state operating characteristics under each charge and discharge cycle are extracted, and a periodic distribution model of the battery health state is constructed. Monitor the battery's response parameters during the current charge / discharge cycle, and perform feature matching between the response parameters and the cycle distribution model to obtain the current battery's health deviation index; When the health deviation index is greater than the safety threshold of the energy storage battery, the sampling frequency gradient of voltage, temperature and internal resistance is increased to the specified test frequency. The thermal runaway warning threshold of the battery health status is determined based on the voltage drop rate, temperature rise rate and internal resistance change rate of each cell after each increase in sampling frequency. When the health deviation index is greater than the thermal runaway warning threshold, the abnormal cell is located by the deviation of the voltage and average voltage of each cell in the energy storage battery, and then the abnormal cell is given a graded alarm.

2. The method as described in claim 1, characterized in that, Extracting steady-state operating characteristics under each charge-discharge cycle and constructing a periodic distribution model of battery health status specifically includes: Median filtering is applied to the voltage, temperature, and internal resistance data within each charge-discharge cycle to remove transient fluctuation segments at the beginning and end of the charge-discharge phases, thus obtaining a steady-state data window for the middle segment of each charge-discharge cycle. Within each steady-state data window, the mean voltage, median temperature, and mode of internal resistance are calculated respectively, serving as the steady-state characteristic vector for the corresponding charge-discharge cycle; The steady-state eigenvectors of multiple consecutive charge-discharge cycles are arranged in chronological order, and a periodic distribution model of the battery health state is generated by using a moving average method.

3. The method as described in claim 1, characterized in that, The response parameters are matched with the periodic distribution model to obtain the current battery health deviation index, which specifically includes: Extract the response parameters of the current charge / discharge cycle in the steady state phase. The response parameters include the average voltage, median temperature, and mode of internal resistance of the current charge / discharge cycle. The response parameters of the current charge / discharge cycle are compared with the expected interval of the same cycle number in the cycle distribution model, and the voltage deviation, temperature deviation and internal resistance deviation are calculated respectively to obtain the preliminary deviation value of the current charge / discharge cycle. The current battery health deviation index is determined by the initial deviation values ​​of three adjacent cycles.

4. The method as described in claim 1, characterized in that, The specified test frequencies include a first test frequency and a second test frequency, wherein the reference sampling frequency is denoted as the reference frequency, the first test frequency is set to twice the reference frequency, and the second test frequency is set to three times the reference frequency.

5. The method as described in claim 1, characterized in that, The thermal runaway early warning threshold for battery health status is determined based on the voltage drop rate, temperature rise rate, and internal resistance change rate of each cell after each increase in sampling frequency. Specifically, it includes: The sampling frequencies for voltage, temperature, and internal resistance are sequentially increased from the reference frequency to the first test frequency and the second test frequency by fixed multiples. Data is continuously collected at each test frequency within a preset time window. The voltage drop rate, temperature rise rate, and internal resistance change rate of each cell within the time window are calculated, and the average value of the corresponding rates of all cells is obtained. The average voltage drop rate, average temperature rise rate, and average internal resistance change rate at each test frequency are compared with the corresponding rate tolerance upper limit. When any two of the three exceed the tolerance upper limit, the current health deviation index is calibrated as the thermal runaway warning threshold of the battery health status.

6. The method as described in claim 1, characterized in that, The location of abnormal cells by the deviation between the voltage of each cell and the average voltage within the energy storage battery includes: Calculate the average voltage of all cells in the energy storage battery at the current sampling time, and calculate the absolute deviation and relative deviation percentage of each cell voltage from the average voltage; Cells with a relative deviation percentage greater than the abnormal threshold are marked as suspected abnormal cells, and the spatial distribution clusters formed by all suspected abnormal cells are counted. Abnormal cells are located by means of the spatial distribution clusters and individual absolute deviations.

7. The method as described in claim 1, characterized in that, Use a battery monitoring slave device to monitor the battery's response parameters during the current charge / discharge cycle.

8. A thermal runaway early warning and health protection system for energy storage batteries, the system comprising a graded alarm unit, characterized in that, The hierarchical alarm unit includes: The data acquisition module is used to collect voltage, temperature and internal resistance data of the energy storage battery during continuous charge and discharge cycles, extract steady-state operating characteristics under each charge and discharge cycle, and construct a periodic distribution model of the battery health status. The processing module is used to monitor the response parameters of the battery during the current charge-discharge cycle, perform feature matching between the response parameters and the cycle distribution model, and obtain the current battery health deviation index. The processing module is also used to increase the sampling frequency gradient of voltage, temperature and internal resistance to a specified test frequency when the health deviation index is greater than the safety threshold of the energy storage battery, and determine the thermal runaway warning threshold of the battery health state based on the voltage drop rate, temperature rise rate and internal resistance change rate of each cell after each increase in sampling frequency. The execution module is used to locate abnormal cells by measuring the deviation between the voltage and average voltage of each cell in the energy storage battery when the health deviation index is greater than the thermal runaway warning threshold, and then to issue graded alarms for each abnormal cell.

9. A computer device, characterized in that, The computer device includes a memory and a processor. The memory is used to store computer programs, and the processor is used to call and run the computer programs from the memory, so that the computer device performs the energy storage battery thermal runaway early warning and health protection method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions or code that, when executed on a computer, cause the computer to implement the energy storage battery thermal runaway early warning and health protection method as described in any one of claims 1 to 7.

Images