Battery module health early detection method and device through pressure online detection

Abstract

This invention discloses a method and apparatus for early detection of battery module health through online pressure detection, relating to the field of battery monitoring technology. The method includes: deploying a thin-film pressure sensor in the battery module to collect multi-dimensional data in real time; transmitting the data to a BMS module via 485 communication, and generating a current health status index for the battery cell after preprocessing and feature extraction; using the average value of the cell's cycle data as a benchmark health status index, and comparing it to obtain the deviation value of the health status index; matching graded early warning information based on the deviation value, and implementing linked response control of the battery module accordingly. This invention solves the technical problem that traditional BMS-based battery health status assessments based on voltage, current, temperature, and internal resistance suffer from detection lag and cannot capture the expansion force changes in the early stages of cell aging, thus making it difficult to achieve real-time online assessment and early warning of battery health status. It achieves the technical effect of improving the timeliness of battery module health status assessment and the ability to provide safety warnings.

Description

Technical Field

[0001] This invention relates to the field of battery monitoring technology, and more specifically to a method and apparatus for early detection of battery module health through online pressure detection. Background Technology

[0002] In the grand wave of global energy transition, the energy storage industry is playing an increasingly crucial role. Lithium-ion batteries, due to their high energy density, long cycle life, and low self-discharge rate, have been widely used in new energy vehicles, energy storage power stations, and consumer electronics. With the deepening of the "dual carbon" goal, new energy storage technologies are being given a new mission. Energy storage systems must not only meet the needs of power supply and demand balance and renewable energy consumption, but also achieve higher standards in safety and reliability. However, during long-term charge-discharge cycles, lithium-ion batteries inevitably experience performance degradation and aging, which can lead to serious safety accidents such as thermal runaway. Several typical safety accidents have occurred in the global energy storage industry, and the safety situation is becoming increasingly severe. Traditional battery management systems mainly rely on collecting electrical parameters such as voltage, current, and temperature from the battery cell assembly to assess the cell's health status. This monitoring method can only passively prevent the risk of thermal runaway caused by cell defects due to over-temperature, over-current, and over-voltage, but it cannot accurately assess the early health status of the battery. Faced with the aforementioned severe safety situation and technological limitations, the traditional BMS voltage / current / temperature monitoring system can no longer meet the high safety standards required for energy storage. This has made the shift from traditional "passive fire protection" to "active prevention" an industry consensus, based on the increasingly urgent need for early warning of battery health status and full life cycle safety management. Summary of the Invention

[0003] This application provides a method and device for early detection of battery module health through online pressure detection. It solves the technical problem that traditional BMS (Battery Management System) assessment of battery health status based on voltage, current, temperature, and internal resistance has detection lag and cannot capture the expansion force change characteristics in the early stage of cell aging, thus making it difficult to achieve real-time online assessment and early warning of battery health status. It achieves the technical effect of utilizing the characteristic that the change of cell expansion force precedes capacity decay and internal resistance increase, realizing early online detection of battery health status without occupying the internal space of the module, and improving the timeliness of battery module health status assessment and safety warning capability.

[0004] The first aspect of this application provides a method for early detection of battery module health via online pressure detection, the method comprising: Deploy thin film pressure sensors on the battery module to collect multi-dimensional battery data streams in real time. The multi-dimensional battery data streams include expansion force data during the battery charging and discharging process, as well as battery voltage, current, internal resistance and temperature data; transmit the multi-dimensional battery data streams to the BMS module via 485 communication for preprocessing and feature extraction to obtain battery characteristic parameters, and convert the battery characteristic parameters into the current health state indicator of the cell; take the average value of the cell cycle data as the reference health state indicator, compare the current health state indicator of the cell with the reference health state indicator to obtain the deviation value of the health state indicator; match and output graded warning information according to the deviation value of the health state indicator, and perform linkage response control on the battery module based on the graded warning information.

[0005] In a second aspect of the present application, a battery module health early detection device through online pressure detection is provided, the device comprising: Data acquisition module: deploy thin film pressure sensors on the battery module to collect multi-dimensional battery data streams in real time, where the multi-dimensional battery data streams include expansion force data during the battery charging and discharging process, as well as battery voltage, current, internal resistance and temperature data; feature conversion module: transmit the multi-dimensional battery data streams to the BMS module via 485 communication for preprocessing and feature extraction to obtain battery characteristic parameters, and convert the battery characteristic parameters into the current health state indicator of the cell; indicator comparison module: take the average value of the cell cycle data as the reference health state indicator, compare the current health state indicator of the cell with the reference health state indicator to obtain the deviation value of the health state indicator; graded warning module: match and output graded warning information according to the deviation value of the health state indicator, and perform linkage response control on the battery module based on the graded warning information.

[0006] The one or more technical solutions provided in the present application at least have the following technical effects or advantages: Firstly, thin film pressure sensors are arranged inside the battery module to continuously monitor the expansion force of the cell during the charging and discharging process, and synchronously acquire operating data such as voltage, current, internal resistance and temperature. The collected data is transmitted to the BMS for processing and feature analysis, so as to comprehensively form a health state indicator that can characterize the health degree of the cell. Then, the historical average data of the cell in a certain cycle stage is taken as the health reference, and the current health state is compared with the reference state to judge the change degree of the cell health state. Finally, different warning levels are divided according to the deviation degree, and corresponding linkage control measures such as monitoring, maintenance or deactivation are adopted, so as to realize online assessment of the battery health state, early identification of abnormalities and graded safety management. Brief Description of the Drawings

[0007] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 This is a schematic flowchart of a method for early detection of battery module health via online pressure detection, provided in an embodiment of this application.

[0009] Figure 2 This is a schematic diagram of the structure of a battery module health early detection device that uses online pressure detection, as provided in an embodiment of this application.

[0010] Figure labeling: Data acquisition module 11, feature transformation module 12, indicator comparison module 13, graded early warning module 14. Detailed Implementation

[0011] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0012] Example 1, as Figure 1 As shown, this application provides a method for early detection of battery module health through online pressure detection, the method including: A thin-film pressure sensor is deployed on the battery module to collect multi-dimensional data streams of the battery in real time. The multi-dimensional data streams of the battery include expansion force data, battery voltage, current, internal resistance and temperature data during the battery charging and discharging process.

[0013] In this embodiment, during the battery module assembly process, a thin-film pressure sensor is placed between the battery cell and the module end plate, enabling effective contact between the sensor and the force-bearing end face of the cell. When the cell expands or contracts during charging or discharging, the force generated by the cell is transmitted to the thin-film pressure sensor. The sensor acquires the corresponding expansion force data in real time based on the force change. Simultaneously, the battery module's acquisition system synchronously acquires the voltage, current, internal resistance, and temperature data of each cell during operation. The expansion force data is then correlated with the aforementioned electrical and thermal parameters according to the acquisition sequence, forming a multi-dimensional battery data stream that reflects the current operating and mechanical state of the cell, providing fundamental data for subsequent battery health status analysis.

[0014] Furthermore, deploying a thin-film pressure sensor on the battery module includes: The battery module consists of a lower housing, a top cover, a module, a liquid cooling system, a fire suppression system, connecting copper and aluminum busbars, a PDU, a BMS system, and connectors. The module itself consists of a cell, an end plate, a steel strip, cushioning foam, and a data acquisition harness assembly. A thin-film pressure sensor is deployed between the cell and the module end plate, and the data acquisition harness assembly of the battery module is connected to the thin-film pressure sensor.

[0015] Preferably, the battery module structure is first arranged, including a lower casing, upper cover, module, liquid cooling system, fire suppression system, connecting copper and aluminum busbars, PDU, BMS system, and connectors. The module's internal components include battery cells, end plates, steel strips, cushioning foam, and a data acquisition harness assembly. These structures collectively constitute the basic hardware environment for battery operation, status acquisition, and safety control. During module assembly, a thin-film pressure sensor is placed between the battery cell and the module end plate, positioning it to detect changes in force at the cell's end. When the battery cell expands or contracts during charge / discharge cycles, the expansion force is transmitted to the module end plate and simultaneously acts on the thin-film pressure sensor located between them, enabling the sensor to detect changes in cell expansion force in real time. Subsequently, the thin-film pressure sensor is connected to the data acquisition harness assembly in the battery module, allowing the expansion force information detected by the sensor to enter the battery's data acquisition channel. In addition to collecting expansion force data, the data acquisition harness assembly also simultaneously collects operating parameters of the battery cell, such as voltage, current, internal resistance, and temperature, and transmits the collected information to the BMS module. This combines the newly added mechanical state information with the existing electrical and temperature parameters, providing a multi-dimensional data foundation for subsequent battery cell health status analysis.

[0016] The battery multidimensional data stream is transmitted to the BMS module via 485 communication for preprocessing and feature extraction to obtain battery feature parameters, which are then converted into indicators of the current health status of the battery cell.

[0017] In one embodiment, after acquiring multi-dimensional battery data such as cell expansion force, voltage, current, internal resistance, and temperature, the data is transmitted to the BMS module via RS-485 communication. The BMS module correlates the received data of different types according to the data acquisition time, filters out abnormal data caused by communication fluctuations or instantaneous sensor disturbances, and smooths continuously acquired data to reduce the interference of data noise on subsequent health status assessments, thereby forming effective data that reflects the actual operating status of the cell. Subsequently, the BMS module extracts feature information related to changes in the cell's health status from the processed data. For example, it extracts the pressure change amplitude and trend from the expansion force change curve and combines this with voltage, internal resistance, and temperature changes under the same operating stage to form battery characteristic parameters characterizing the cell's current state. The obtained characteristic parameters are recorded according to the corresponding cell and cycle period for subsequent continuous status tracking. Subsequently, the BMS module corrects the extracted battery feature parameters accordingly, so that the feature data obtained under different operating conditions have a unified basis for comparison. Then, different types of feature parameters are converted into corresponding health characterization values, and the correlation between various features and cell aging status is integrated to form a current health status index that characterizes the current health of the cell. This health status index can transform the original multi-dimensional operating data into quantitative results that are easy to compare and judge, providing a data basis for subsequent comparison with the benchmark health status and implementation of graded early warning.

[0018] Furthermore, the battery multidimensional data stream is transmitted to the BMS module via 485 communication for preprocessing and feature extraction to obtain battery feature parameters, including: The battery multidimensional data stream is transmitted to the BMS module via 485 communication for data cleaning to obtain a usable battery multidimensional data stream; a low-pass filtering algorithm is used to filter the usable battery multidimensional data stream to obtain a standard battery multidimensional data stream; battery feature parameters corresponding to the usable battery multidimensional data stream are extracted and synchronously saved locally and in the cloud.

[0019] Preferably, during battery module operation, the data acquisition harness assembly continuously acquires expansion force data collected by the thin-film pressure sensor, as well as voltage, current, internal resistance, and temperature data corresponding to the battery cells. This data is then transmitted to the BMS module via 485 communication according to a preset data frame format. Each data frame includes at least the cell identifier, data type, sampling time, and corresponding sampled value, enabling the BMS module to correlate the expansion force, voltage, current, internal resistance, and temperature data obtained by the same cell within the same sampling period based on the cell identifier and sampling time. After receiving the multidimensional data stream from the battery, the BMS module performs data cleaning on the received data. Specifically, the data is sorted and time-aligned according to the cell identifier and timestamp in the data frame. The system then sequentially checks for communication verification errors, missing data, duplicate sampling, exceeding the preset effective range, and abnormal abrupt changes between adjacent sampling points. For data frames with communication verification errors or that cannot be identified, the corresponding data is marked as invalid and discarded. For duplicate data, one valid sample value is retained. For a single missing sampling point appearing within a short period, interpolation compensation is performed based on the valid data before and after the missing sampling point. For continuously missing data segments exceeding the preset allowable missing length, the data segment is marked as invalid and not included in the current feature calculation. If any sample value of expansion force, voltage, current, internal resistance, or temperature exceeds the preset effective range of the corresponding sensor or cell, the sample value is identified as out-of-bounds data. If the change in adjacent sampling points exceeds the preset maximum change rate of the corresponding parameter, the corresponding sampling point is identified as a suspected abnormal abrupt change point, and a judgment is made based on the consecutive sampling values ​​before and after it. When the abnormal change exists only in a single sampling point and subsequent sampling values ​​recover to the original trend, it is discarded as interference data or compensated using adjacent valid data. After completing the above processing, retain the data that has passed the validity check and completed time alignment to form a multidimensional data stream of usable batteries.

[0020] Subsequently, low-pass filtering is performed on each continuously sampled parameter in the available battery multidimensional data stream to suppress high-frequency noise generated by the thin-film pressure sensor, voltage acquisition circuit, current acquisition circuit, and temperature acquisition circuit. In one feasible approach, a first-order low-pass filtering algorithm is used. Let the original sampled value of a certain type at the k-th sampling time be x(k), and the corresponding filtered output value be y(k). Then, the current filtered output is calculated according to y(k) = αx(k) + (1-α)y(k-1), where 0 < α < 1, α is the filtering coefficient, and y(k-1) is the filtering result of the previous sampling time. When performing filtering for the first time, y(0) = x(0). The filtering coefficient α can be predetermined based on the actual sampling period and the range of high-frequency interference to be filtered out. The filtering coefficient is adjusted accordingly when the sampling period or sensor response characteristics change. The BMS module continuously filters the time series of expansion force, voltage, current, internal resistance, and temperature according to the above method to suppress random jitter and transient interference, while retaining low-frequency trend information generated during cell charging and discharging with changes in SOC and cycle number, thus forming a standard battery multidimensional data stream. Then, the BMS module extracts features from the cleaned data using the cell number and charge / discharge cycle number as indices. For pressure data, at least the peak expansion force, cycle change rate, and interval expansion difference corresponding to the preset SOC interval are extracted. The peak expansion force is the maximum effective pressure value in the current cycle pressure sequence. Let the expansion force feature value of the current nth cycle be P. n The characteristic value corresponding to the previous iteration is P. n-1 The rate of cyclic change can be determined according to (P). n -P n-1 ) / P n-1 The calculation is used to characterize the degree of change in cell expansion characteristics with the number of cycles; the interval expansion difference is determined by selecting two preset SOC positions or SOC intervals corresponding to the pressure characteristic value P. a and P b And according to ΔP=|P b -P a The calculations characterize the extent of cell expansion during a single charge-discharge cycle. For voltage, current, internal resistance, and temperature data, corresponding effective feature values ​​are extracted according to the same cycle period or the same SOC range. This ensures that each feature is correlated with its corresponding pressure feature in terms of cell number, cycle number, and data acquisition conditions, thus forming a set of battery feature parameters including pressure, voltage, current, internal resistance, and temperature characteristics.

[0021] After feature extraction, the BMS module establishes a corresponding data record for each group of battery feature parameters. This data record is associated with at least the cell number, cycle count, acquisition time, and corresponding feature value. The battery feature parameters are written to the BMS local storage unit, and simultaneously uploaded to cloud storage via the battery system's data communication link. This allows the local end to meet the needs of real-time health status calculation and early warning control, while the cloud can store historical feature data formed during long-term cell cycling. This provides a historical data foundation for establishing health benchmarks, analyzing feature change trends, and comparing cell health status at different cycling stages. Through this process, the impact of communication anomalies, invalid sampling, and high-frequency noise on subsequent health status judgments can be reduced, and the continuously collected raw multidimensional data can be converted into effective feature parameters that reflect the cell's operating status and cycling change trends, thereby improving the stability and reliability of subsequent cell health status index calculations.

[0022] Furthermore, converting the battery characteristic parameters into indicators of the current health status of the cell includes: The battery characteristic parameters are corrected under operating conditions to obtain standard battery characteristic parameters; each type of characteristic parameter in the battery characteristic parameters is mapped to a single-dimensional health status to generate multi-dimensional health sub-values; based on the multi-dimensional health sub-values, a weighted fusion transformation is performed to determine the current health status index of the battery cell.

[0023] Preferably, after the module completes the preprocessing and feature extraction of the battery's multidimensional data, it constructs a mapping table corresponding to cell temperature, SOC, and expansion force. This mapping table is then used to perform operating condition correction and compensation on the battery characteristic parameters to obtain standard battery characteristic parameters. Subsequently, single-dimensional health status mapping is performed on the characteristic parameters of each type of standard battery. That is, the pressure characteristic parameter in the battery characteristic parameters is normalized and mapped to obtain multi-dimensional health sub-values. Then, a preset weight allocation rule is used to perform weighted fusion transformation on the obtained multi-dimensional health sub-values ​​to calculate the comprehensive health sub-value of the pressure dimension as the current health status index of the cell. After completing the above calculations, the BMS module establishes and stores the obtained current health status index of the cell with the corresponding cell number, cycle number, and collection time for subsequent comparison with the benchmark health status index, thereby improving the stability, comparability, and ability to identify early abnormal changes in the cell health status evaluation.

[0024] Furthermore, the battery characteristic parameters are corrected for operating conditions to obtain standard battery characteristic parameters, including: A cell temperature-SOC-expansion force mapping table is constructed; the cell temperature-SOC-expansion force mapping table is used to perform operating condition correction and compensation on the battery characteristic parameters to obtain standard battery characteristic parameters.

[0025] Optionally, to reduce the impact of cell temperature and state of charge (SOC) changes on expansion force characteristics, a cell temperature-SOC-expansion force mapping table is pre-constructed before evaluating cell health. First, a normal cell with the same model, capacity, and structure as the cells in the battery module to be monitored is selected as a benchmark sample. The benchmark sample is subjected to cyclic testing under preset normal charge and discharge conditions, and cell temperature, SOC, and expansion force data detected by a thin-film pressure sensor are collected simultaneously during the test. Multiple temperature nodes or temperature ranges are set according to the actual allowable operating temperature range of the cell, and multiple SOC nodes or SOC ranges are set according to SOC from low to high. Multiple operating condition calibration units are formed by combining temperature nodes and SOC nodes. For each operating condition calibration unit, the battery cell is charged or discharged under the corresponding temperature conditions. When the battery cell reaches the corresponding SOC node, the expansion force data at that moment or within that SOC range is acquired. The expansion force data obtained from multiple cycles or multiple normal battery cells under the same operating condition are statistically processed, and the average value, median value, or average value after outlier removal is determined as the normal expansion force reference value corresponding to that temperature-SOC operating condition node. This yields multiple ternary correspondences (T...). i S j P ij ), where T i Let S represent the i-th temperature node. j Let P represent the j-th SOC node. ij This indicates that a normal battery cell operates at temperature T. i and SOC is S j The corresponding reference expansion force under the given conditions. The reference expansion force P is analyzed along two dimensions: temperature and SOC. ij The data is arranged to form a cell temperature-SOC-expansion force mapping table, which is pre-stored in the BMS module. For temperature-SOC combination nodes where data is not directly obtained during testing, interpolation can be performed using the expansion force reference values ​​of adjacent calibrated nodes. For example, when the current temperature T is at T... i With T i+1 Between, the current SOC value S is located at S j With S j+1 When the time interval is between these intervals, bilinear interpolation can be performed based on the reference expansion forces corresponding to the four adjacent working condition nodes to obtain the reference value P of the normal expansion force under the current working condition. ref (T,S), thus enabling the mapping table to cover the continuously changing temperature and SOC operating range of the battery cell.

[0026] After the mapping table is constructed, during the actual operation of the battery module, the BMS module acquires the target cell's current temperature T, current SOC value S, and corresponding actual expansion force data P in real time. Using the current temperature T and current SOC value S as query indexes, it searches for the corresponding normal expansion force reference value in the temperature-SOC-expansion force mapping table. When T and S match the calibration node in the mapping table, the corresponding reference expansion force P is directly read. ref (T,S), when T or S is located between adjacent calibration nodes, interpolation calculation is performed using adjacent nodes to obtain the reference expansion force P under the current working condition. ref (T,S). Subsequently, a uniform standard temperature T0 and standard SOC value S0 are pre-selected from the mapping table as the standard comparison condition, and the normal expansion force reference value P corresponding to this standard condition is read. ref (T0, S0), and then compensate for the actual expansion force based on the difference between the reference expansion force under the current working condition and the reference expansion force under the standard working condition. In one feasible approach, a differential compensation method is used, according to P std =PP ref (T,S)+P ref The equivalent expansion force P under standard working conditions is calculated from (T0,S0). std Where P is the actual expansion force collected under the current working conditions, P ref (T,S) represents the normal reference expansion force corresponding to the current temperature and SOC, P ref (T0, S0) represents the reference expansion force corresponding to the standard operating condition. Therefore, when a cell experiences normal expansion force changes only due to temperature increases, decreases, or SOC changes, the changes caused by these operating conditions can be offset during the compensation process, while retaining the additional expansion changes relative to the normal cell reference state. For pressure characteristic parameters such as peak expansion force, cyclic change rate, and interval expansion difference, continuous expansion force data can be compensated point-by-point according to the above method. Then, based on the compensated standard expansion force sequence, the peak expansion force, cyclic change rate, and interval expansion difference are re-extracted to obtain standard pressure characteristic parameters. For non-pressure characteristics such as voltage, current, internal resistance, and temperature, their effective characteristic values ​​obtained after the aforementioned preprocessing can be retained and combined with the pressure characteristics corrected for operating conditions to form a standard battery characteristic parameter set. The BMS module establishes a correlation between the corrected standard battery characteristic parameters and the corresponding cell number, sampling time, cycle number, temperature, and SOC for subsequent single-dimensional health status mapping and multi-dimensional health sub-value weighted fusion. Through the above process, the impact of normal expansion force fluctuations caused by different temperatures and SOC operating conditions on the health status assessment can be reduced, and the pressure characteristics collected under different operating conditions can be compared under a unified reference condition, thereby improving the comparability of pressure characteristics and the stability and accuracy of cell health status evaluation results.

[0027] Furthermore, the specific steps for generating the comprehensive health sub-value of the stress dimension include: A preset weight allocation rule is established, specifically a weight allocation coefficient for the peak expansion force, the rate of change of the cycle, and the difference in expansion within the interval. The pressure characteristic parameters in the battery characteristic parameters are normalized and mapped to obtain health sub-values ​​of multiple pressure characteristic dimensions. The health sub-values ​​of multiple pressure characteristic dimensions are weighted and fused according to the weight allocation rule to obtain a comprehensive health sub-value of the pressure dimension.

[0028] Optionally, after obtaining the standard battery characteristic parameters corrected for operating conditions, pressure characteristic parameters for characterizing the mechanical health of the cell are extracted from the standard battery characteristic parameters. These include the peak expansion force, the rate of change of cycle, and the interval expansion difference. The peak expansion force is used to characterize the maximum expansion force generated by the cell during a single charge-discharge cycle. Let P be the peak expansion force corresponding to the nth cycle. max,n The rate of change of pressure is used to characterize how quickly pressure characteristics change with increasing cycle number; for example, it can be expressed according to R... n =(P max,n -P max,n-1 ) / (P max,n-1 +ε) calculates the cyclic change rate of the nth cycle, where P max,n-1 The peak expansion force of the previous cycle is given by ε, which is a preset positive number to avoid a denominator of zero. The interval expansion difference is used to characterize the degree of expansion force change between two preset SOC positions or SOC intervals in the same charge-discharge cycle. Let P be the standard expansion force corresponding to the first preset SOC position. a The standard expansion force corresponding to the second preset SOC position is P. b Then it can be calculated according to ΔP=|P b -P a | Obtain the interval expansion difference for the current cycle. After obtaining the above three types of pressure characteristics, a weighting allocation rule is pre-established: a first weighting coefficient w1 is set for the peak expansion force, a second weighting coefficient w2 is set for the cycle change rate, and a third weighting coefficient w3 is set for the interval expansion difference, satisfying w1+w2+w3=1. These weighting coefficients can be determined using cycle calibration data of cells of the same model before the battery is put into actual operation. Specifically, the degree of change of the above three types of pressure characteristics relative to the normal health state at different cycle stages can be statistically analyzed, and the weight of each pressure characteristic can be determined according to its sensitivity to changes in cell aging. The more sensitive a pressure characteristic is to changes in cell health state, the greater its weight is assigned. For cells of the same model and under the same application conditions, the determined weighting coefficients are stored in the BMS module and directly called during actual health state calculation.

[0029] Subsequently, the peak expansion force, cyclic change rate, and interval expansion difference were normalized to eliminate the influence caused by the differences in dimensions and numerical ranges among the three types of pressure characteristics. Specifically, based on pre-selected health baseline cyclic data, baseline values ​​for the three types of pressure characteristics were determined, denoted as P0, R0, and ΔP0, respectively. Furthermore, the maximum allowable deviation D corresponding to each pressure characteristic was determined based on cyclic calibration data of the same type of battery cell. 1,max D 2,max and D 3,max For the current battery cell, the normalized deviations of the three types of pressure characteristics relative to the reference values ​​are calculated respectively. Among them, the normalized deviation of the peak expansion force can be expressed as D1=|P max -P0| / (|P0|+ε), the normalized deviation of the cyclic rate of change can be expressed as D2=|R-R0| / (|R0|+ε), and the normalized deviation of the interval expansion difference can be expressed as D3=|Δ P -ΔP0| / (|ΔP0|+ε). Then, each normalized deviation is converted into a health sub-value within a unified range. In one feasible approach, the range of the health sub-value is set to 0~100, where 100 indicates that the corresponding pressure characteristic is close to the healthy baseline state, and 0 indicates that the corresponding pressure characteristic has reached or exceeded the preset maximum abnormal deviation. The peak expansion force health sub-value can then be calculated according to H1=100×max[0, 1-D1 / D...]. 1,max The calculation is performed, and the healthy sub-value of the cyclic change rate can be calculated according to H2=100×max[0,1-D2 / D]. 2,max The calculation is performed, and the healthy sub-value of the interval expansion difference can be calculated according to H3=100×max[0,1-D3 / D]. 3,max Calculations are performed to determine the health sub-value. When the deviation between the current pressure characteristic and the health baseline value is small, the corresponding health sub-value is close to 100. As the deviation of the pressure characteristic increases, the corresponding health sub-value gradually decreases, resulting in the health sub-values ​​H1 (peak expansion force), H2 (cycle change rate), and H3 (interval expansion difference), forming a set of health sub-values ​​across multiple pressure characteristic dimensions. Finally, the BMS module performs weighted fusion of the health sub-values ​​across multiple pressure characteristic dimensions according to pre-stored weight allocation rules, based on H... P =w1×H1+w2×H2+w3×H3 to calculate the comprehensive health sub-value H of the stress dimension. P , where H P It also falls within the unified health evaluation range of 0-100. When the peak expansion force, cycle change rate, and range expansion difference of the battery cell are all close to the health baseline state, the comprehensive health sub-value H of the pressure dimension is... P Maintaining a high level, when one or more of the stress characteristics continuously deviate from the healthy baseline state, the corresponding health sub-value decreases and is reflected in H through the corresponding weight. PIn this process, the BMS (Battery Management System) correlates the obtained pressure dimension comprehensive health sub-value with the cell number, cycle count, and acquisition time, and uses this as the pressure dimension input for subsequent comprehensive fusion with other characteristic dimensions such as voltage, current, internal resistance, and temperature. Through this process, multiple pressure characteristics with different dimensions and ranges of variation can be uniformly converted into comparable health evaluation quantities, while simultaneously considering the absolute degree of cell expansion, cyclic evolution trends, and expansion change characteristics within a specific SOC (State of Charge) range. This improves the comprehensive characterization ability of pressure information for changes in the early health status of the cell.

[0030] The average value of the cell cycle data is used as a benchmark health status indicator. The current health status indicator of the cell is compared with the benchmark health status indicator to obtain the deviation value of the health status indicator.

[0031] In one embodiment, after obtaining the current health status index of the battery cell, the BMS module calls the historical cycle health status data of the corresponding battery cell to establish a benchmark health status index for comparison. During the preset benchmark cycle phase after the battery module is put into operation, the battery cell health status index corresponding to each effective charge-discharge cycle is continuously recorded, and a correspondence is established between the battery cell number, cycle number, and health status index. The preset benchmark cycle phase is preferably a cycle phase in the early stage of the battery cell's lifespan where the health status is relatively stable. For example, the first 500 cycles can be selected as the benchmark data range, and the number of benchmark cycles can be adjusted according to the battery cell model, cycle life, operating conditions, and calibration test results. Let N effective cycles be obtained in the benchmark cycle phase, and the health status index corresponding to the i-th effective cycle be H. i The BMS module then calculates the average of the health status indicators corresponding to N valid cycles to obtain the baseline health status indicator H. base The baseline health status index is associated with the corresponding cell number and stored locally in the BMS and in the cloud, serving as a reference benchmark for the cell's subsequent cycle health status changes. During normal operation after the battery enters the baseline cycle stage, the BMS calculates the current health status index H of the cell based on the aforementioned multi-dimensional feature fusion process after each effective cycle or upon reaching the preset health status evaluation time. cur And read the baseline health status index H corresponding to the cell. base The current health status indicators are compared with the baseline health status indicators to determine the degree of change in the current cell's health status relative to its previous state. Subsequently, the deviation value D of the health status indicators is calculated using a relative deviation method. H D H =|H cur -H base | / (|H baseThe formula is |+ε)×100%, where ε is a preset positive number to avoid the baseline health status index being zero. When the health status index adopts a non-zero health score range, ε can take a sufficiently small positive number. The BMS associates and stores the obtained health status index deviation value with the cell number, current cycle number, calculation time, and current health status index, and uses the deviation value as a direct input parameter for subsequent graded early warning judgment. Through the above process, an individualized health reference benchmark can be established using the stable cycle state of the cell in the early stage of its own life, and the degree of decline of the cell's health status can be characterized by the relative change between the current state and the historical benchmark. This reduces the impact of the initial consistency differences of different cells on the health judgment results and provides a quantitative basis for subsequent graded early warning and linkage control based on the health status index deviation value.

[0032] Based on the deviation value of the health status index, a graded early warning information is output, and the battery module is controlled in a coordinated manner based on the graded early warning information.

[0033] In one embodiment, after obtaining the health status index deviation value of the target cell, the BMS module uses the health status index deviation value as a graded early warning judgment parameter, and calls a pre-stored early warning threshold table to perform interval matching on the current deviation value to determine the corresponding health risk level of the target cell, including Level 1, Level 2, Level 3, and Level 4 early warnings. Under Level 1 early warning, the BMS maintains the normal operation of the battery module and continuously monitors and records the cell's expansion force, voltage, current, internal resistance, temperature, and health status index deviation value. Under Level 2 early warning, the BMS continues to maintain the operation of the battery module while continuously tracking the health status index and its deviation trend in subsequent cycles of the target cell, and records the Level 2 early warning status, corresponding cell number, current health status index, health status index deviation value, cycle number, and early warning generation. The time is associated with the storage to determine if the deviation value continues to increase. Under a Level 3 warning, the BMS marks the target cell as an abnormal cell requiring maintenance and generates maintenance prompts to instruct the battery module containing the target cell to be inspected and maintained. At the same time, the Level 3 warning information is transmitted to the upper-level control terminal of the battery system, enabling maintenance personnel to arrange corresponding maintenance based on the warning information. Under a Level 4 warning, the BMS generates a stop-use command and executes the stop-use control of the battery module or battery module through the control link with the power distribution unit (PDU), causing the target battery module to stop participating in normal charging and discharging. Simultaneously, the BMS can transmit the Level 4 warning status to the fire protection system, causing the fire protection system to enter the linkage response state corresponding to the higher-level warning, so that corresponding safety protection can be implemented when an abnormal state that meets the fire triggering conditions is subsequently detected.

[0034] For the specific matching process of warning levels, the BMS compares the deviation values ​​of health status indicators from low to high with the aforementioned warning thresholds. When the current deviation value crosses the threshold of an adjacent level, the warning status is updated to the corresponding higher level, and the corresponding linkage control command is updated simultaneously. For example, if the deviation value of a cell's health status indicator in the previous cycle was 25%, and it increases to 35% in the current cycle, the BMS updates its warning status from Level 2 to Level 3, and switches from continuous key monitoring status to maintenance reminder status. If the subsequent deviation value further exceeds 40%, the Level 3 warning is upgraded to Level 4, and a shutdown control is executed. To avoid frequent switching of warning levels due to single data fluctuations near the threshold, the health status indicator deviation value is required to meet the threshold conditions of a higher level for a preset number of consecutive times before the warning upgrade is executed. After each generation or update of the warning level, the BMS forms a warning record with the warning level, corresponding cell number, health status indicator deviation value, trigger time, number of cycles, and corresponding linkage response results, and saves it locally. It can also be simultaneously uploaded to the cloud to form a record of abnormal health changes throughout the target cell's entire life cycle. Therefore, by dividing the deviation values ​​of health status indicators into multiple risk ranges and configuring progressive response strategies such as continuous monitoring, key attention, maintenance reminders, and shutdown respectively, the degree of abnormal battery health status can be correlated with the actual control measures. When the degree of abnormality is low, unnecessary downtime can be avoided, and when the health status continues to deteriorate, intervention measures can be upgraded in a timely manner. This achieves closed-loop management from online health status monitoring, anomaly identification, graded early warning to safety linkage control, thereby improving the timeliness of handling early health anomalies of battery modules and operational safety.

[0035] Furthermore, the tiered early warning information is as follows: when the deviation value of the health status indicator is less than 10%, it is a Level 1 warning, requiring only continuous monitoring; when the deviation value of the health status indicator is in the range of 10% to 30%, it is a Level 2 warning, requiring continued close monitoring; when the deviation value of the health status indicator is in the range of 30% to 40%, it is a Level 3 warning, prompting the scheduling of maintenance; and when the deviation value of the health status indicator is greater than 40%, it is a Level 4 warning, directly triggering a stop-use command.

[0036] Preferably, when the deviation value of the health status index is less than 10%, it indicates that the current health status of the cell has changed little compared to the baseline state. In this case, it is classified as a Level 1 warning, and operational data such as expansion force, voltage, current, internal resistance, and temperature continue to be collected to continuously track the subsequent trend of the health status index. When the deviation value of the health status index reaches the range of 10% to 30%, it indicates that the health status of the cell has deviated significantly from the baseline state. This is classified as a Level 2 warning. At this time, the BMS marks the corresponding cell as a key focus object, and while maintaining the operation of the battery module, it continuously monitors the changes in the health status of the cell to determine whether the deviation has a further increasing trend. When the deviation value of the health status index reaches the range of 30% to 40%, it indicates that there is a significant difference between the current state of the cell and the baseline health status, and the cell may show a significant aging or abnormal expansion trend. This is classified as a Level 3 warning. At this time, the BMS generates a maintenance prompt message, lists the corresponding cell or battery module as a maintenance object, and prompts maintenance personnel to arrange inspection so that maintenance can be carried out before the abnormality escalates further. When the deviation value of the health status indicator exceeds 40%, it indicates that the health status of the battery cell has significantly deviated from the normal benchmark, and continued operation may pose a high safety risk. This is classified into four levels of warning. At this time, the BMS directly generates a stop-use command and can send corresponding control commands to execution units such as the PDU to restrict or stop the charging and discharging operation of the relevant battery module, preventing the abnormal state from continuing to develop. Through the above-mentioned classification judgment, a correspondence can be established between the degree of deviation of the health status indicator and different handling measures such as continuous monitoring, key attention, maintenance scheduling, and stop-use, so as to achieve layered safety management from early attention to high-risk shutdown.

[0037] In summary, the embodiments of this application have at least the following technical effects: Deploy a thin film pressure sensor on the battery module to collect multi-dimensional battery data streams in real time. The multi-dimensional battery data streams include expansion force data, battery voltage, current, internal resistance and temperature data during the battery charging and discharging process; transmit the multi-dimensional battery data streams to the BMS module via 485 communication for preprocessing and feature extraction to obtain battery characteristic parameters, and convert the battery characteristic parameters into the current health state indicator of the cell; take the average value of the cell cycle data as the reference health state indicator, compare the current health state indicator of the cell with the reference health state indicator to obtain the deviation value of the health state indicator; match and output graded warning information according to the deviation value of the health state indicator, and perform linkage response control on the battery module based on the graded warning information. It solves the technical problem that the traditional BMS battery health state assessment based on voltage, current, temperature and internal resistance has detection lag, cannot capture the expansion force change characteristics in the early stage of cell aging, and thus is difficult to realize real-time online assessment and early warning of battery health state, and achieves the technical effect of utilizing the characteristic that the cell expansion force change precedes capacity attenuation and internal resistance increase, realizing early online detection of battery health state without occupying the internal space of the module, and improving the timeliness of battery module health state assessment and safety warning capability.

[0038] Embodiment 2 is based on the same inventive concept as the aforementioned embodiment of the method for early online detection of battery module health through pressure online detection, as Figure 2 shown, the present application provides a device for early online detection of battery module health through pressure online detection, wherein the device comprises: Data acquisition module 11: deploy a thin film pressure sensor on the battery module to collect multi-dimensional battery data streams in real time, the multi-dimensional battery data streams include expansion force data, battery voltage, current, internal resistance and temperature data during the battery charging and discharging process; Feature conversion module 12: transmit the multi-dimensional battery data streams to the BMS module via 485 communication for preprocessing and feature extraction to obtain battery characteristic parameters, and convert the battery characteristic parameters into the current health state indicator of the cell; Indicator comparison module 13: take the average value of the cell cycle data as the reference health state indicator, compare the current health state indicator of the cell with the reference health state indicator to obtain the deviation value of the health state indicator; Graded warning module 14: match and output graded warning information according to the deviation value of the health state indicator, and perform linkage response control on the battery module based on the graded warning information.

[0039] Further, the data acquisition module 11 is configured to execute the following method: The battery module consists of a lower housing, a top cover, a module, a liquid cooling system, a fire suppression system, connecting copper and aluminum busbars, a PDU, a BMS system, and connectors. The module itself consists of a cell, an end plate, a steel strip, cushioning foam, and a data acquisition harness assembly. A thin-film pressure sensor is deployed between the cell and the module end plate, and the data acquisition harness assembly of the battery module is connected to the thin-film pressure sensor.

[0040] Furthermore, the feature conversion module 12 is used to perform the following method: The battery multidimensional data stream is transmitted to the BMS module via 485 communication for data cleaning to obtain a usable battery multidimensional data stream; a low-pass filtering algorithm is used to filter the usable battery multidimensional data stream to obtain a standard battery multidimensional data stream; battery feature parameters corresponding to the usable battery multidimensional data stream are extracted and synchronously saved locally and in the cloud.

[0041] Furthermore, the feature conversion module 12 is used to perform the following method: The battery characteristic parameters are corrected under operating conditions to obtain standard battery characteristic parameters; 5. Single-dimensional health status mapping is performed on each type of characteristic parameter in the battery characteristic parameters to generate multi-dimensional health sub-values; Based on the multi-dimensional health sub-values, a weighted fusion transformation is performed to determine the current health status index of the cell.

[0042] Furthermore, the feature conversion module 12 is used to perform the following method: A cell temperature-SOC-expansion force mapping table is constructed; the cell temperature-SOC-expansion force mapping table is used to perform operating condition correction and compensation on the battery characteristic parameters to obtain standard battery characteristic parameters.

[0043] Furthermore, the feature conversion module 12 is used to perform the following method: A preset weight allocation rule is established, specifically a weight allocation coefficient for the peak expansion force, the rate of change of the cycle, and the difference in expansion within the interval. The pressure characteristic parameters in the battery characteristic parameters are normalized and mapped to obtain health sub-values ​​of multiple pressure characteristic dimensions. The health sub-values ​​of multiple pressure characteristic dimensions are weighted and fused according to the weight allocation rule to obtain a comprehensive health sub-value of the pressure dimension.

[0044] Furthermore, the graded early warning module 14 is used to perform the following method: The tiered early warning information is as follows: when the deviation value of the health status indicator is less than 10%, it is a Level 1 warning, requiring only continuous monitoring; when the deviation value of the health status indicator is in the range of 10% to 30%, it is a Level 2 warning, requiring continued close monitoring; when the deviation value of the health status indicator is in the range of 30% to 40%, it is a Level 3 warning, prompting the scheduling of maintenance; and when the deviation value of the health status indicator is greater than 40%, it is a Level 4 warning, directly triggering a stop-use command.

[0045] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for early detection of battery module health via online pressure detection, characterized in that, The method includes: A thin-film pressure sensor is deployed on the battery module to collect multi-dimensional data streams of the battery in real time. The multi-dimensional data streams of the battery include expansion force data, battery voltage, current, internal resistance and temperature data during the battery charging and discharging process. The battery multidimensional data stream is transmitted to the BMS module via 485 communication for preprocessing and feature extraction to obtain battery feature parameters, which are then converted into the current health status index of the battery cell. The average value of the cell cycle data is used as a benchmark health status indicator. The current health status indicator of the cell is compared with the benchmark health status indicator to obtain the deviation value of the health status indicator. Based on the deviation value of the health status index, a graded early warning information is output, and the battery module is controlled in a coordinated manner based on the graded early warning information.

2. The method for early detection of battery module health via online pressure detection as described in claim 1, characterized in that, Deploying a thin-film pressure sensor on the battery module includes: The battery module consists of a lower housing, a top cover, a module, a liquid cooling system, a fire protection system, connecting copper and aluminum busbars, a PDU, a BMS system, and connectors. The module itself consists of a cell, an end plate, a steel strip, buffer foam, and a data acquisition harness assembly. A thin-film pressure sensor is deployed between the battery cell and the module end plate of the battery module, wherein the battery module's data acquisition harness assembly is connected to the thin-film pressure sensor.

3. The method for early detection of battery module health via online pressure detection as described in claim 1, characterized in that, The battery multidimensional data stream is transmitted to the BMS module via 485 communication for preprocessing and feature extraction to obtain battery feature parameters, including: The battery multidimensional data stream is transmitted to the BMS module via 485 communication for data cleaning to obtain a usable battery multidimensional data stream. A low-pass filtering algorithm is used to filter the available battery multidimensional data stream to obtain a standard battery multidimensional data stream. Extract the battery feature parameters corresponding to the multidimensional data stream of the available battery, and synchronously save the battery feature parameters locally and in the cloud.

4. The method for early detection of battery module health via online pressure detection as described in claim 1, characterized in that, Converting the battery characteristic parameters into indicators of the current health status of the cell includes: The battery characteristic parameters are corrected for operating conditions to obtain standard battery characteristic parameters; A single-dimensional health status mapping is performed on each type of feature parameter in the battery feature parameters to generate multi-feature-dimensional health sub-values; The current health status index of the battery cell is determined by weighted fusion transformation based on the multi-feature dimension health sub-values.

5. The method for early detection of battery module health via online pressure detection as described in claim 4, characterized in that, The battery characteristic parameters are corrected for operating conditions to obtain standard battery characteristic parameters, including: Construct a mapping table corresponding to cell temperature, state of charge (SOC), and expansion force; The battery characteristic parameters are corrected and compensated for under operating conditions using the cell temperature-SOC-expansion force mapping table to obtain standard battery characteristic parameters.

6. The method for early detection of battery module health via online pressure detection as described in claim 4, characterized in that, The specific steps for generating the comprehensive health sub-value of the stress dimension include: A preset weight allocation rule is defined, which specifically includes weight allocation coefficients for peak expansion force, cyclic change rate, and interval expansion difference. Normalize and map the pressure feature parameters in the battery feature parameters to obtain health sub-values ​​of multiple pressure feature dimensions; The health sub-values ​​of the multiple stress feature dimensions are weighted and fused according to the weight allocation rules to obtain the comprehensive health sub-value of the stress dimension.

7. The method for early detection of battery module health via online pressure detection as described in claim 1, characterized in that, The tiered early warning information is as follows: when the deviation value of the health status indicator is less than 10%, it is a Level 1 warning, requiring only continuous monitoring; when the deviation value of the health status indicator is in the range of 10% to 30%, it is a Level 2 warning, requiring continued close monitoring; when the deviation value of the health status indicator is in the range of 30% to 40%, it is a Level 3 warning, prompting the scheduling of maintenance; and when the deviation value of the health status indicator is greater than 40%, it is a Level 4 warning, directly triggering a stop-use command.

8. A battery module health early detection device based on online pressure detection, characterized in that, The apparatus for implementing the early health detection method for battery modules via online pressure detection as described in any one of claims 1-7 comprises: Data acquisition module: A thin-film pressure sensor is deployed on the battery module to collect multi-dimensional data streams of the battery in real time. The multi-dimensional data streams of the battery include expansion force data, battery voltage, current, internal resistance and temperature data during the battery charging and discharging process. Feature conversion module: Transmits the multidimensional data stream of the battery to the BMS module via 485 communication for preprocessing and feature extraction to obtain battery feature parameters, and converts the battery feature parameters into the current health status index of the cell; Indicator comparison module: Based on the average value of the cell cycle data, a baseline health status indicator is used. The current health status indicator of the cell is compared with the baseline health status indicator to obtain the deviation value of the health status indicator. Graded early warning module: Based on the deviation value of the health status index, it matches and outputs graded early warning information, and performs linkage response control on the battery module based on the graded early warning information.

Images