A method, device, and medium for testing a battery management system safety function

Abstract

The application discloses a kind of battery management system safety function test method, equipment and medium, it is related to battery management system test technical field, including, overcurrent pulse and dynamic voltage disturbance is directionally injected into propagation path thermal map, diffusion proportion coefficient and relay response time are generated;Based on diffusion proportion coefficient and relay response time, calculate dynamic inhibition efficiency value, and according to dynamic inhibition efficiency value dynamic determination ASIL safety level, generate three-dimensional space-time thermal map;According to three-dimensional space-time thermal map, map voltage offset amplitude, superimpose BMS response trajectory, construct failure probability-inhibition efficiency correlation surface, generate digitized test report;The application is fused by multi-source sensing and evidence theory, realizes the dynamic quantification of fault feature, accurately locates high-risk electric core, improves complex working condition adaptability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of battery management system testing, and in particular to a battery management system safety function test method, device and medium. BACKGROUND

[0002] As a core control unit for ensuring battery safety and prolonging battery life, the importance of the battery management system (BMS) is increasingly prominent. Especially in the fields of electric vehicles, energy storage systems and renewable energy, the BMS needs to monitor key parameters such as voltage, current and temperature of the battery in real time, and ensure that the battery operates within a safe range through mechanisms such as overcharge protection, overdischarge protection and short circuit protection. In recent years, BMS testing technology has gradually evolved from traditional physical testing to digitalization and intelligentization. Intelligent sensing systems achieve multi-dimensional perception and dynamic analysis of battery status through the integration of high-precision sensors, edge computing units and adaptive algorithms. For example, based on multi-modal sensor fusion technology, early features of abnormal battery behavior can be accurately captured.

[0003] Traditional testing methods rely on static or preset fault modes, making it difficult to reproduce the actual response behavior of the BMS under complex working conditions. For example, in the case of extreme temperature fluctuations or multi-parameter coupling changes, existing test equipment often cannot accurately simulate real operating conditions, resulting in deviations between test results and actual applications. On the other hand, existing test systems lack real-time data recording and analysis, making it difficult to support dynamic response evaluation of the BMS on a millisecond time scale. For example, the trigger time of the short circuit protection function, the response delay of the overcharge and overdischarge protection, and other key performance indicators need to rely on high-precision timing analysis tools for quantification, but current test equipment still has bottlenecks in sampling frequency and data processing capacity. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a battery management system safety function test method to solve the problems of insufficient simulation accuracy of dynamic multi-parameter coupling working conditions and limited high-precision real-time evaluation capability of traditional BMS safety function testing.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides a battery management system safety function test method, which includes,

[0008] Real-time acquisition of voltage, temperature, acoustic and gas data, and calculation of fault probability value through evidence theory fusion algorithm to dynamically generate high-risk cell coordinates and risk level labels;

[0009] Based on the high-risk cell coordinates and risk level labels, a bionic neural network model is constructed, a fault conduction coefficient matrix is obtained, and a propagation path heat map is generated.

[0010] The overcurrent pulse and dynamic voltage disturbance are injected into the propagation path heat map, and the diffusion proportion coefficient and the relay response time are generated.

[0011] Based on the diffusion proportion coefficient and the relay response time, the dynamic inhibition efficiency value is calculated, and the ASIL safety level is dynamically determined according to the dynamic inhibition efficiency value, and a three-dimensional space-time heat map is generated.

[0012] According to the three-dimensional space-time heat map, the voltage offset amplitude is mapped, the BMS response trajectory is superimposed, the fault probability-inhibition efficiency correlation surface is constructed, and a digital test report is generated.

[0013] As a preferred scheme of the battery management system safety function test method, wherein: the real-time acquisition of voltage, temperature, acoustic and gas data, and the calculation of fault probability value through evidence theory fusion algorithm, dynamic generation of high-risk cell coordinates and risk level labels, the steps are as follows,

[0014] Real-time acquisition of voltage fluctuation variance data, temperature gradient distribution data, acoustic characteristic spectrum data and gas concentration change rate data, and generation of original sensor data stream;

[0015] According to the original sensor data stream, the joint fault probability value is calculated through the evidence theory fusion algorithm, and the probability overrun alarm signal and the characteristic vector peak time marker are dynamically generated;

[0016] Based on the probability overrun alarm signal and the characteristic vector peak time marker, spatial matching positioning is performed, the intersection point of abnormal heat source and sound source is determined, and the risk level is judged according to the scoring rule, and the high-risk cell coordinates and risk level labels are generated.

[0017] As a preferred scheme of the battery management system safety function test method, wherein: based on the high-risk cell coordinates and risk level labels, a bionic neural network model is constructed, and the steps are as follows,

[0018] The high-risk cell coordinates are mapped to the corresponding nodes of the battery pack space topology grid by a dynamic topology mapping method, and the risk level label color coding is marked, and a battery pack space topology grid with risk labels is generated.

[0019] Based on the battery pack space topology grid with risk labels, a multilayer perceptron structure and a biological neuron activation mechanism are used to simulate the fault conduction characteristics between cells, and a bionic neural network model is constructed.

[0020] As a preferred scheme of the battery management system safety function test method of the application, wherein: the fault conduction coefficient matrix is obtained, and a propagation path thermal map is generated, the steps are as follows,

[0021] Based on the bionic neural network model, the reciprocal of the three-dimensional space distance between the cells is combined to determine the conduction intensity value, and the fault conduction coefficient matrix is generated;

[0022] Scan the high-risk conduction path in the conduction coefficient matrix, calculate the fault propagation time delay combined with the time decay factor, and generate the propagation path thermal map.

[0023] As a preferred scheme of the battery management system safety function test method of the application, wherein: the overcurrent pulse and dynamic voltage disturbance are injected into the propagation path thermal map to generate the diffusion proportion coefficient and the relay response time, the steps are as follows,

[0024] Analyze the propagation path thermal map, and screen the cell nodes in the propagation path thermal map whose color depth exceeds the red warning threshold to generate a hierarchical injection strategy list;

[0025] Based on the hierarchical injection strategy list, drive the high-precision power supply control node to inject the overcurrent pulse and dynamic voltage disturbance in stages, and generate an injection operation log file;

[0026] According to the injection operation log file, scan the voltage value of each cell node through the voltage sensor array, count the total number of cells whose voltage deviation exceeds the voltage deviation threshold, and generate the diffusion proportion coefficient;

[0027] Detect the high-voltage relay action signal, capture the electrical characteristic jump point of the relay disconnection action, record the absolute time stamp of the relay action, and generate the relay response time.

[0028] As a preferred scheme of the battery management system safety function test method of the application, wherein: based on the diffusion proportion coefficient and the relay response time, the dynamic suppression efficiency value is calculated, and the ASIL safety level is dynamically determined according to the dynamic suppression efficiency value, and a three-dimensional space-time thermal map is generated, the steps are as follows,

[0029] Based on the diffusion proportion coefficient and the relay response time, perform parameter validity verification and data cleaning to generate a compliant parameter set;

[0030] According to the compliant parameter set, the dynamic suppression efficiency value is calculated, and the ASIL safety level is determined according to the suppression efficiency threshold to generate an ASIL safety level certification file;

[0031] Based on the ASIL safety level certification file, mark the spatial coordinates of the fault injection starting point, draw the fault propagation path trajectory line, and generate a three-dimensional space-time thermal map.

[0032] As a preferred solution of the test method of the battery management system safety function of the application, wherein: according to the three-dimensional space-time thermal diagram, the voltage offset amplitude is mapped, and the BMS response track is superimposed, the steps are as follows,

[0033] The three-dimensional space-time thermal diagram is analyzed, the color scale coding rule is extracted, the thermal diagram space node is scanned, the voltage offset amplitude value of the battery cell is labeled, and a voltage offset amplitude matrix table is generated.

[0034] The BMS response track marker is superimposed at the corresponding space-time coordinate position of the voltage offset amplitude matrix table, and an enhanced thermal diagram data with BMS response marker is generated.

[0035] As a preferred solution of the test method of the battery management system safety function of the application, wherein: the fault probability-inhibition efficiency correlation surface is constructed, and a digital test report is generated, the steps are as follows,

[0036] According to the joint fault probability value and the dynamic inhibition efficiency value, a three-dimensional correlation coordinate system is constructed, the data points in the voltage offset amplitude matrix table are mapped to the three-dimensional correlation coordinate system, and a fault probability-inhibition efficiency correlation surface is generated.

[0037] Based on the fault probability-inhibition efficiency correlation surface, combined with the enhanced thermal diagram data with BMS response marker and the ASIL safety level certification file, a digital test report is packaged and generated.

[0038] In a second aspect, the application provides a computer device comprising a memory and a processor, the memory storing a computer program, wherein: the computer program is executed by the processor to realize any step of the test method of the battery management system safety function according to the first aspect of the application.

[0039] In a third aspect, the application provides a computer readable storage medium having a computer program stored thereon, wherein: the computer program is executed by the processor to realize any step of the test method of the battery management system safety function according to the first aspect of the application.

[0040] The application has the advantages that: through multi-source sensing and evidence theory fusion, fault feature dynamic quantization is realized, high-risk battery cells are accurately positioned, and complex working condition adaptability is improved. Through the construction of a propagation path thermal diagram by a bionic neural network, fault propagation prediction is realized, response evaluation capability is enhanced, protection strategy is optimized, and safety level judgment is supported. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0042] Fig. 1 Flowchart for the test method of the safety function of the battery management system.

[0043] Fig. 2 Flowchart for generating the high-risk cell coordinate and risk level label.

[0044] Fig. 3 Flowchart for constructing the bionic neural network model and generating the propagation path thermogram.

[0045] Fig. 4 Flowchart for injecting the overcurrent pulse and generating the response. DETAILED DESCRIPTION

[0046] In order to make the above objectives, features and advantages of the present application more apparent and comprehensible, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0047] In the following description, a large number of specific details are set forth in order to facilitate a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the connotation of the present application, so the present application is not limited to the specific embodiments disclosed below.

[0048] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0049] Reference Figs. 1-4 For one embodiment of the present application, the embodiment provides a test method of a safety function of a battery management system, comprising the following steps:

[0050] S1: Real-time acquisition of voltage, temperature, acoustic and gas data, and calculation of fault probability value through evidence theory fusion algorithm, dynamic generation of high-risk cell coordinate and risk level label;

[0051] S1.1: Real-time acquisition of voltage fluctuation variance data, temperature gradient distribution data, acoustic feature spectrum data and gas concentration change rate data, generation of original sensor data stream;

[0052] Further, high-precision voltage sensors are deployed to capture voltage fluctuation variance in real time, high-precision infrared temperature arrays are used to scan temperature gradient distribution, high-precision acoustic emission sensors are used to capture acoustic characteristic spectrum, and high-precision electrochemical gas sensors are used to monitor gas concentration change rate; a time-synchronized precision acquisition link is constructed; a overcharge test protocol is executed; high-precision voltage sensors obtain voltage fluctuation variance data, high-precision infrared temperature arrays record temperature gradient distribution data, high-precision acoustic emission sensors capture acoustic characteristic spectrum data, and high-precision electrochemical gas sensors update gas concentration change rate data; after timestamp alignment, raw sensing data stream is generated.

[0053] S1.2: According to the raw sensing data stream, the joint fault probability value is calculated by the evidence theory fusion algorithm, and the probability overrun alarm signal and the characteristic vector peak time marker are dynamically generated;

[0054] Further, the evidence theory fusion algorithm is performed on the raw sensing data stream: weight coefficients are assigned (for example, voltage fluctuation variance value weight 40%, maximum temperature rise rate value weight 30%, acoustic energy integral value weight 20%, and carbon monoxide concentration change slope value weight 10%); Sigmoid function normalization processing is performed on the voltage fluctuation variance value, the maximum temperature rise rate value, the acoustic energy integral value, and the carbon monoxide concentration change slope value, respectively; the joint fault probability value is calculated; when the joint fault probability value exceeds the fault threshold value (an example value of 0.8 is a theoretical safety boundary derived based on the electrochemical characteristics of battery materials and the critical parameters of thermal runaway), the probability overrun alarm signal is generated; the characteristic vector peak time is marked synchronously, and the probability overrun alarm signal and the characteristic vector peak time marker are output.

[0055] The formula for calculating the joint fault probability value is:

[0056] ;

[0057] wherein, represents the fault probability value, represents the voltage characteristic weight, represents the voltage abnormality scoring function, represents the voltage fluctuation variance, represents the temperature characteristic weight, represents the temperature abnormality scoring function, represents the temperature gradient change, represents the acoustic characteristic weight, represents the acoustic abnormality scoring function, represents the acoustic characteristic spectrum, represents the gas characteristic weight, represents the gas abnormality scoring function, represents the gas concentration change rate.

[0058] S1.3: Based on the probability overrun alarm signal and the peak time mark of the feature vector, spatial matching positioning is performed to determine the intersection point of the abnormal heat source and the sound source, and the risk level is judged according to the scoring rules to generate the high-risk battery coordinates and the risk level label.

[0059] Further, based on the probability overrun alarm signal triggering the spatial matching positioning operation, the corresponding time point is marked to extract the acoustic emission sensor array data to obtain the sound source coordinates at the peak time of the feature vector, and the infrared temperature array data is extracted synchronously to trace the heat source coordinates; the sound source coordinates and the heat source coordinates are fused by weighting, and the abnormal heat source and the sound source intersection point coordinates are output; the sound energy intensity value, the temperature rise rate value and the gas concentration change slope value are loaded, and the total risk score is accumulated according to the scoring rules, which is mapped to a five-level risk level label; finally, the high-risk battery coordinates and the risk level label are generated.

[0060] It should be noted that the scoring rules adopt a three-dimensional independent scoring accumulation mechanism: the sound energy intensity value is scored by interval segmentation, the temperature rise rate value is graded by trigger point, and the gas concentration change slope value is graded by gradient; the three scores are superimposed on the basic risk level, and the total score range is mapped to a five-level risk label (for example, a total score of 2 corresponds to L1 level, 3-4 corresponds to L2 level, 5 corresponds to L3 level, 6 corresponds to L4 level, and ≥7 corresponds to L5 level), and a dynamic correction factor is simultaneously bound.

[0061] S2: Based on the high-risk battery coordinates and the risk level label, a bionic neural network model is constructed to obtain a fault conduction coefficient matrix and generate a propagation path thermal map;

[0062] S2.1: Map the high-risk battery coordinates to the corresponding nodes of the battery pack space topology grid by a dynamic topology mapping method, and mark the risk level label color coding to generate a battery pack space topology grid with risk markers;

[0063] Specifically, the battery pack space topology grid is loaded by a dynamic topology mapping method, the high-risk battery coordinates are matched to the nearest grid node, the spatial deviation of the high-risk battery coordinates and the node center is obtained; according to the risk level label, the color coding rule is called to superimpose the pulse marker at the grid node; the battery pack space topology grid is updated in real time to generate a battery pack space topology grid with risk markers.

[0064] It should be noted that the battery pack space topology grid is composed of battery cell node matrix, connection relationship matrix and space attribute matrix, wherein each grid node stores the accurate physical coordinates, electrical connection attributes, thermal conductivity coefficient and mechanical stress parameters of the battery cell, the real-time collected high-risk coordinates are matched to the nearest grid node by a dynamic topology mapping method, and the preset color coding rule and dynamic marker special effect are called according to the risk level label, and finally an enhanced topology structure with space coordinate mapping, risk visualization and conduction path prediction is generated.

[0065] The risk level label color coding adopts a five-level RGBA color mapping mechanism, wherein the L1 level corresponds to a light green semi-transparent label, the L2 level is a blue-green static label, the L3 level enables a yellow breathing flashing special effect, the L4 level triggers a red pulse label, and the L5 level activates a deep red diffusion animation, and each level is synchronously bound to a dynamic label rule.

[0066] S2.2: Based on the risk-labeled battery pack space topology grid, a multi-layer perceptron structure and a biological neuron activation mechanism are used to simulate the fault conduction characteristics between cells, and a bionic neural network model is constructed.

[0067] Specifically, based on the risk-labeled battery pack space topology grid, high-risk cell coordinates and risk level labels are loaded; a multi-layer perceptron structure is constructed; a biological neuron activation mechanism is implemented in the hidden layer; the fault conduction characteristics between cells are simulated, and the bionic neural network model is constructed.

[0068] It should be noted that based on the risk-labeled battery pack space topology grid, a multi-layer perceptron structure is constructed: the topology grid node data is loaded in the input layer, the full connection architecture is configured in the hidden layer, and the output layer is mapped to the target cell conduction intensity value; the physical location of the cell is bound to the neural network node through the spatial topology mapping relationship, realizing the deep coupling of the cell spatial distribution characteristics and the neural network structure.

[0069] The training process is executed in stages: in the data preparation stage, multiple fault scenarios are injected and actual conduction paths are collected as training labels; in the back propagation stage, the neural network weight parameters are adjusted through an optimization algorithm, and the error function is used to quantify the deviation between the predicted conduction path and the actual path; in the dynamic parameter adjustment stage, the neuron activation parameters are adaptively corrected according to the risk level label, and after multiple rounds of iterative training, the matching degree between the output conduction path of the bionic neural network model and the actual fault propagation trajectory reaches the engineering application standard.

[0070] S2.3: Based on the bionic neural network model, the reciprocal of the three-dimensional space distance between cells is determined to generate a conduction intensity value, and a fault conduction coefficient matrix is generated.

[0071] Specifically, based on the bionic neural network model, the three-dimensional space distance between cells is loaded to obtain the reciprocal of the distance; the risk level gain coefficient is superimposed to generate the conduction intensity value; all adjacent node combinations of the battery pack space topology grid are traversed, and the fault conduction coefficient matrix is filled according to the row and column index rule, and finally a fault conduction coefficient matrix containing all the conduction relationships between cells is generated.

[0072] It should be noted that the conduction intensity value is represented by the reciprocal of the three-dimensional Euclidean distance between cells to represent the spatial attenuation characteristics, and the conduction weight and risk level gain coefficient output by the bionic neural network are superimposed, which quantifies the fault propagation ability and is finally used to fill the fault conduction coefficient matrix to achieve precise safety protection.

[0073] S2.4: Scan the high-risk conduction path in the conduction coefficient matrix, calculate the fault propagation delay combined with the time attenuation factor, and generate a propagation path heat map.

[0074] Specifically, scan the high-risk conduction path in the conduction coefficient matrix with a conduction strength value greater than the risk intensity threshold (example value: 0.8, a theoretical critical value determined by reverse deduction based on the physical limit of thermal-electric coupling failure of battery materials), extract the source cell coordinates and target cell coordinates of the high-risk conduction path, and obtain the three-dimensional space distance between cells; calculate the fault propagation delay combined with the time attenuation factor; mark the cell position number on the horizontal axis and the time line on the vertical axis in the two-dimensional coordinate system, and map the high-risk conduction path to the color scale according to the delay value to generate a propagation path heat map.

[0075] The formula for calculating the fault propagation delay is:

[0076] ;

[0077] Among them, represents the fault propagation delay, represents the three-dimensional space distance, represents the conduction coefficient (value range: 0.1-1.0, set according to the physical constraints of spatial geometric layout and electrical connection impedance), represents the material attenuation factor (value range: 0.5-1.2, set according to the difference in intrinsic ionic conductivity of different battery chemical systems), represents the temperature compensation coefficient (value range: 0.8-1.5, set according to the universal influence of temperature on electrochemical reaction kinetics).

[0078] S3: Directly inject the overcurrent pulse and dynamic voltage disturbance into the propagation path heat map to generate a diffusion proportion coefficient and a relay response time;

[0079] S3.1: Analyze the propagation path heat map and select the cell nodes in the propagation path heat map with color scale depth exceeding the red alert threshold to generate a hierarchical injection strategy list;

[0080] Further, analyze the propagation path heat map, scan the color scale depth data of all cell nodes on the time axis; select the cell nodes with color scale depth exceeding the red alert threshold (example value: 0.8, set by measuring the ratio of the maximum ion flow impact that the battery separator can withstand to the minimum insulation resistance in the early stage of thermal runaway); extract the position index, timestamp, and conduction strength value of the exceeding nodes; classify according to the conduction strength value interval to generate a hierarchical injection strategy list containing position-time-level.

[0081] It should be noted that the hierarchical injection strategy list is a set of instructions generated based on the substandard battery cell nodes selected from the propagation path heatmap. It contains three core fields: location index, timestamp, and injection level, and achieves millisecond-level precise fault injection control through structured data format.

[0082] S3.2: Based on the hierarchical injection strategy list, drive the high-precision power control node to inject overcurrent pulses and dynamic voltage disturbances in a hierarchical manner, and generate injection operation log files;

[0083] Furthermore, based on the hierarchical injection strategy list, the high-precision power control node is driven to perform hierarchical injection operations: an overcurrent pulse is injected into the target cell coordinate position at an absolute time (e.g., 4.2 milliseconds, defined by a high-precision synchronous clock source); after a delay time (e.g., 50 milliseconds, the time interval between the overcurrent pulse triggering time and the dynamic voltage disturbance execution time), a dynamic voltage disturbance is applied to adjacent nodes with a conduction coefficient greater than a set value (e.g., 0.85, defined according to the critical decay rate of the inter-cell insulation dielectric breakdown voltage); the high-precision power control node records the injected current value, voltage value, and time deviation in real time; and generates an injection operation log file containing timestamps, coordinate positions, and measured current and voltage values.

[0084] It should be noted that the high-precision power control node, as the core execution node for battery management system safety testing, is essentially a multi-channel programmable power module with dual closed-loop control capabilities for current and voltage, precise spatiotemporal control characteristics, intelligent protection mechanisms, and a data traceability architecture. In node battery pack testing, it achieves high current control accuracy, low timing error rate, and full data integrity. Engineering applications include overcurrent pulse injection and microsecond-level fault sequence playback functions.

[0085] Dynamic voltage perturbation is a dynamic adaptive voltage modulation method that simulates abnormal voltage fluctuations during the battery pack fault propagation process by applying a precise voltage offset to the target cell within a specific time window (e.g., a delay of 50 milliseconds, based on the timing control requirements defined by the fault injection process).

[0086] S3.3: Based on the injection operation log file, scan the voltage values ​​of each cell node through the voltage sensor array, count the total number of cells whose voltage deviation exceeds the voltage deviation threshold, and generate the diffusion ratio coefficient;

[0087] Furthermore, based on the timestamps and coordinate locations recorded in the injection operation log file, a voltage sensor array scanning operation is synchronously initiated: real-time voltage values ​​of each cell node are captured by distributed voltage sensors at a sampling rate of 10 kHz to obtain voltage deviation values; a voltage deviation threshold is set (e.g., 10%, a dynamic critical value set based on the battery electrochemical characteristics and safety risk nodes), the total number of cells with voltage deviation values ​​exceeding the voltage deviation threshold is counted, and a diffusion ratio coefficient is generated.

[0088] S3.4: Detecting the high-voltage relay action signal, capturing the electrical characteristic jump point of relay disconnection action, recording the absolute time stamp of relay action, and generating the relay response time.

[0089] Further, the high-precision current sensor is used to monitor the high-voltage relay coil current value in real time, and when the current value suddenly drops to near zero, it is determined as the action starting point; the differential voltage probe is used to capture the voltage jump characteristics between the relay contacts, and the accurate time when the voltage jump is too large is recorded as the disconnection action point; the GPS synchronous clock source is used to mark the absolute time stamp of the action; the difference value between the fault injection starting time and the action time stamp is obtained, and the relay response time is generated.

[0090] S4: Based on the diffusion proportionality coefficient and the relay response time, the dynamic suppression efficiency value is calculated, and the ASIL safety level is dynamically determined according to the dynamic suppression efficiency value, and a three-dimensional space-time thermal map is generated;

[0091] S4.1: Based on the diffusion proportionality coefficient and the relay response time, parameter validity verification and data cleaning are performed, and a compliant parameter set is generated;

[0092] Specifically, based on the diffusion proportionality coefficient and the relay response time, three parameter verifications are performed: value range verification checks whether the diffusion proportionality coefficient is in the safe interval (example range: [0, 1], the proportion value cannot be negative or exceed 1) and whether the relay response time is within the safe range (example range: [0, 200], defined based on the critical time window of the battery thermal runaway chain reaction development, the critical time window is the physical limit to ensure that electrical isolation is completed before the battery internal temperature reaches the electrolyte ignition point); sign verification confirms that the diffusion proportionality coefficient is non-negative and the response time is positive; physical rationality verification determines whether the product of the diffusion proportionality coefficient and the response time is less than the safety threshold (example value: 1.5, set according to the maximum allowed product limit of fault diffusion energy and relay suppression energy in unit time); for abnormal parameters, trigger data cleaning process: use adjacent node mean replacement or mark as invalid data points and exclude from subsequent calculations; finally generate a compliant parameter set.

[0093] S4.2: According to the compliant parameter set, the dynamic suppression efficiency value is calculated, and the ASIL safety level is determined according to the suppression efficiency threshold, and an ASIL safety level certification file is generated;

[0094] Specifically, based on the diffusion ratio coefficient and the relay response time in the compliance parameter set, the dynamic suppression performance value is calculated; according to the suppression performance threshold judgment rule (for example, performance value > 0.95 is judged as ASIL D level, 0.85-0.95 is judged as ASIL C level), the safety level certification result is output; the test timestamp, compliance parameter value, dynamic suppression performance value and ASIL safety level certification file are generated.

[0095] It should be noted that the suppression performance threshold judgment rule is the core standard for dividing the safety integrity level of the battery management system based on the numerical interval of the dynamic suppression performance value, and the example rule is: when the dynamic suppression performance value is greater than 0.95 and simultaneously satisfies the diffusion ratio coefficient less than 0.03 and the response time less than 100 milliseconds, it is judged as ASIL D level; when the dynamic suppression performance value is in the interval of 0.85 to 0.95 and the diffusion ratio coefficient is less than 0.1, it is judged as ASIL C level; other numerical combinations are classified as ASIL B level or below; and a dynamic compensation mechanism is set, and the final determination result is output to the ASIL safety level certification file in a structured data format.

[0096] The formula for calculating the dynamic suppression performance value is:

[0097] ;

[0098] Among them, represents the dynamic suppression performance value, represents the diffusion ratio coefficient, (value range: [0, 1], the proportion value cannot be negative or exceed 1), represents the time factor, represents the relay response time;

[0099] S4.3: Based on the ASIL safety level certification file, mark the fault injection starting point spatial coordinates, draw the fault propagation path trajectory line, and generate a three-dimensional space-time thermal map.

[0100] Specifically, based on the ASIL safety level certification file, the fault injection starting point spatial coordinates are extracted, and the fault injection starting point spatial coordinates are calibrated in the battery pack three-dimensional node; according to the high-risk path data in the conduction coefficient matrix, the fault propagation path trajectory line is drawn; in the three-dimensional coordinate system, the time dimension is superimposed, and combined with the dynamic suppression performance value, the color scale is mapped according to the dynamic suppression performance value interval, and a three-dimensional space-time thermal map is generated, which integrates spatial coordinates, time series and performance value.

[0101] S5: According to the three-dimensional space-time thermal map, map the voltage offset amplitude, superimpose the BMS response trajectory, construct the fault probability-suppression performance correlation surface, and generate a digital test report.

[0102] S5.1: Analyze the three-dimensional space-time thermal map, extract the color scale encoding rule, scan the space node of the thermal map, label the voltage offset amplitude value of the battery cell, and generate a voltage offset amplitude matrix table;

[0103] Specifically, analyze the three-dimensional space-time thermal map, extract the color scale encoding rule; scan the space node of the thermal map, label the voltage offset amplitude value of the corresponding battery cell; traverse all space nodes and time points, and generate a voltage offset amplitude matrix table mapping the battery cell position and time stamp.

[0104] It should be noted that the color scale encoding rule divides the continuous gradient color band based on the numerical range of the dynamic suppression efficiency value. For example, the L value in the interval [0, 0.3) is mapped to dark blue, indicating suppression failure; the interval [0.3, 0.6) transitions to blue-green, indicating partial suppression; the interval [0.6, 0.85) gradually changes to yellow-orange, indicating effective suppression; the interval [0.85, 1] is displayed as red, indicating close to complete suppression; the color depth is positively correlated with the L value, and the dynamic marking rule is superimposed synchronously, which strictly matches the visualization requirements of the three-dimensional space-time thermal map.

[0105] S5.2: Superimpose BMS response trajectory marker symbols on the corresponding space-time coordinate positions of the voltage offset amplitude matrix table to generate enhanced thermal map data with BMS response markers;

[0106] Specifically, based on the voltage offset amplitude matrix table, superimpose BMS response trajectory marker symbols at specific space-time coordinate positions: relay disconnection action is marked as a red triangle symbol, and fault alarm event is marked as a yellow circle symbol; extract the relay response timestamp and action position, and add marker symbols to the corresponding coordinate nodes in the voltage offset amplitude matrix table; synchronously record the mapping relationship between the marker type and the space-time coordinate; generate enhanced thermal map data containing original voltage offset amplitude data and BMS response marker layers.

[0107] S5.3: According to the joint fault probability value and the dynamic suppression efficiency value, construct a three-dimensional associated coordinate system, map the data points in the voltage offset amplitude matrix table to the three-dimensional associated coordinate system, and generate a fault probability-suppression efficiency correlation surface;

[0108] Specifically, based on the joint fault probability value and the dynamic suppression efficiency value, construct a three-dimensional associated coordinate system: X axis represents the joint fault probability value, Y axis represents the dynamic suppression efficiency value, and Z axis represents the voltage offset amplitude value; extract the data points in the voltage offset amplitude matrix table, and map the data points to the corresponding positions in the three-dimensional associated coordinate system; traverse all data points in the voltage offset amplitude matrix table, and generate a continuous and smooth fault probability-suppression efficiency correlation surface using B-spline surface fitting algorithm.

[0109] S5.4: Based on the failure probability-inhibition efficiency correlation surface, combined with the enhanced heat map data with BMS response markers and ASIL safety level certification documents, encapsulate and generate a digital test report.

[0110] Based on the failure probability-inhibition efficiency correlation surface, combined with the enhanced heat map data with BMS response markers and ASIL safety level certification documents, encapsulate and generate a digital test report.

[0111] The embodiment also provides a computer device suitable for the battery management system safety function test method, which includes a memory and a processor.

[0112] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner. The wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0113] The embodiment also provides a storage medium on which a computer program is stored, the program being executed by a processor to implement the test method for realizing the safety function of the battery management system as proposed in the above embodiment; the storage medium can be realized by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a magnetic disk, or an optical disk.

[0114] To sum up, the application realizes dynamic quantification of fault features, accurate positioning of high-risk battery cells, and improvement of adaptability to complex working conditions by multi-source sensing and evidence theory fusion.

[0115] It should be noted that the above embodiments are only used to illustrate the technical solutions of the application but not limit the application. Although the application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the application can be modified or replaced equivalently without departing from the spirit and scope of the application, and all modifications or equivalent replacements should be included in the scope of the claims of the application.

Claims

1. A method of testing a battery management system safety function, the method comprising: Comprising, Real-time acquisition of voltage, temperature, acoustic and gas data, and calculation of fault probability value through evidence theory fusion algorithm, dynamic generation of high-risk battery coordinate and risk level label; Based on the high-risk battery coordinate and risk level label, a bionic neural network model is constructed to obtain the fault conduction coefficient matrix and generate a propagation path heat map; The overcurrent pulse and dynamic voltage disturbance are injected into the propagation path heat map to generate diffusion proportion coefficient and relay response time; Based on the diffusion proportion coefficient and relay response time, the dynamic inhibition efficiency value is calculated, and the ASIL safety level is dynamically determined according to the dynamic inhibition efficiency value, and a three-dimensional space-time heat map is generated, as follows, Based on the diffusion proportion coefficient and relay response time, perform parameter validity verification and data cleaning to generate a compliant parameter set; According to the compliant parameter set, the dynamic inhibition efficiency value is calculated, and the ASIL safety level is determined according to the inhibition efficiency threshold, and an ASIL safety level certification file is generated; Based on the ASIL safety level certification file, label the fault injection starting point spatial coordinates, draw the fault propagation path trajectory, and generate a three-dimensional space-time heat map; According to the three-dimensional space-time heat map, map the voltage offset amplitude, superimpose the BMS response trajectory, construct the fault probability-inhibition efficiency correlation surface, and generate a digital test report, as follows, Parse the three-dimensional space-time heat map, extract the color scale coding rules, scan the heat map space nodes, label the voltage offset amplitude value of the battery, and generate a voltage offset amplitude matrix table; In the corresponding space-time coordinate position of the voltage offset amplitude matrix table, superimpose the BMS response trajectory marker symbol to generate an enhanced heat map data with BMS response markers; According to the joint fault probability value and dynamic inhibition efficiency value, a three-dimensional correlation coordinate system is constructed, and the data points in the voltage offset amplitude matrix table are mapped to the three-dimensional correlation coordinate system to generate a fault probability-inhibition efficiency correlation surface; Based on the fault probability-inhibition efficiency correlation surface, combined with the enhanced heat map data with BMS response markers and the ASIL safety level certification file, a digital test report is generated.

2. The method of testing battery management system safety functions of claim 1, wherein: The real-time acquisition of voltage, temperature, acoustic and gas data, and the calculation of fault probability value through evidence theory fusion algorithm, dynamic generation of high-risk battery coordinate and risk level label, steps as follows, Real-time acquisition of voltage fluctuation variance data, temperature gradient distribution data, acoustic characteristic spectrum data and gas concentration change rate data, to generate original sensor data stream; According to the original sensor data stream, the joint fault probability value is calculated through the evidence theory fusion algorithm, and the probability over-limit alarm signal and feature vector peak time marker are dynamically generated; Based on the probability over-limit alarm signal and feature vector peak time marker, perform spatial matching positioning to determine the intersection point of abnormal heat source and sound source, and judge the risk level according to the scoring rules to generate high-risk battery coordinate and risk level label.

3. The method of testing battery management system safety functions of claim 2, wherein: The bionic neural network model is constructed based on the high-risk battery coordinate and risk level label, as follows, Map the high-risk battery coordinate to the battery pack space topology grid corresponding node through dynamic topology mapping method, and mark the risk level label color coding to generate a battery pack space topology grid with risk markers; Based on the battery pack space topology grid with risk markers, a multi-layer perceptron structure and a biological neuron activation mechanism are used to simulate the fault conduction characteristics between cells, and a bionic neural network model is constructed.

4. The method of testing battery management system safety functions of claim 3, wherein: The fault conduction coefficient matrix is obtained, and a propagation path heat map is generated, and the steps are as follows, Based on the bionic neural network model, the reciprocal of the three-dimensional spatial distance between cells is combined to determine the conduction intensity value, and a fault conduction coefficient matrix is generated. Scan the high-risk conduction path in the conduction coefficient matrix, combine the time decay factor to calculate the fault propagation delay, and generate a propagation path heat map.

5. The method of testing battery management system safety functions of claim 4, wherein: The overcurrent pulse and dynamic voltage disturbance are injected into the propagation path heat map to generate a diffusion proportion coefficient and a relay response time, and the steps are as follows, Analyze the propagation path heat map and screen the cell nodes in the propagation path heat map whose color depth exceeds the red alert threshold to generate a hierarchical injection strategy list; Based on the hierarchical injection strategy list, drive the high-precision power control node to inject the overcurrent pulse and dynamic voltage disturbance in stages, and generate an injection operation log file; According to the injection operation log file, the voltage values of each cell node are scanned through the voltage sensor array, the total number of cells with voltage deviation exceeding the voltage deviation threshold is counted, and the diffusion proportion coefficient is generated; Detect the high-voltage relay action signal, capture the electrical characteristic jump point of the relay disconnection action, record the absolute time stamp of the relay action, and generate the relay response time. 6.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is characterized in that: The processor executes the computer program to realize the steps of the test method of the battery management system safety function of any one of claims 1-5.

7. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to realize the steps of the test method of the battery management system safety function of any one of claims 1-5.

Images