Power battery test data abnormity early warning method and system

Abstract

The invention relates to the technical field of lithium battery manufacturing, in particular to a power battery test data abnormity early warning method and system, and the method comprises the following steps: collecting the operation parameters of test equipment and the test data of a tested power battery in a power battery test process in real time; based on battery state information in the test data, inputting the test data into a preset virtual battery model; calculating a relative deviation ratio between the actually measured voltage of the measured power battery and the analog voltage value; if it is judged that the data exception is not caused by the battery problem, an equipment fault diagnosis step is executed, and fault diagnosis and grade judgment are carried out by comprehensively applying at least two of the following judgment bases in the step: corresponding early warning operation is executed according to the fault grade judged in the fault diagnosis step. The reliability of test data and the credibility of the test process are remarkably improved, and battery performance misjudgment caused by equipment problems is avoided.

Description

Technical Field

[0001] This invention belongs to the field of lithium battery technology, specifically relating to a method and system for early warning of abnormal test data in power batteries. Background Technology

[0002] With the rapid development of the new energy vehicle industry, the performance and safety of power batteries, as core components, directly affect the overall vehicle quality. Therefore, the power battery testing process is crucial. During power battery testing, various indicators such as battery capacity, charge / discharge efficiency, cycle life, and high / low temperature performance are measured using testing equipment, generating a large amount of test data. This data is a key basis for evaluating battery performance, optimizing battery design, and ensuring product quality.

[0003] However, existing power battery testing processes often encounter two types of data problems caused by equipment: First, hardware failures (such as poor sensor contact or data acquisition module lag) or software anomalies (such as data transmission protocol errors or storage module failures) lead to abnormal data recording, manifested as missing data, data jumps, or data duplication. If these problems are not detected in time, they will result in incomplete test data, affecting subsequent battery performance evaluation. Second, the decline in equipment accuracy after long-term use (such as current sensor drift, voltage measurement module aging, or temperature control accuracy deviation) causes test data to deviate from the true value beyond the allowable range, i.e., inaccurate data. These problems are insidious and difficult to detect in the short term, easily leading to unqualified batteries entering the market or high-quality batteries being misjudged, causing economic losses and safety risks to enterprises.

[0004] Currently, the processing of power battery test data in the industry mainly focuses on data analysis and verification after the test, lacking real-time monitoring and early warning mechanisms. Some companies use manual periodic checks of equipment status, which is not only inefficient but also fails to cover real-time data anomalies during the testing process. A few systems with basic early warning functions can only alarm for single data indicators (such as voltage out of range), unable to distinguish whether the data anomaly is caused by a problem with the battery itself or the equipment, and even less able to provide early warnings for data inaccuracies caused by decreased equipment precision. Therefore, there is an urgent need for a technical solution that can identify equipment-related data problems in real time and provide early warnings to overcome the shortcomings of existing technologies. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for early warning of abnormal power battery test data, which can monitor power battery test data in real time, intelligently distinguish between "battery abnormality" and "equipment abnormality", and provide early warning of equipment accuracy decline.

[0006] To achieve the above objectives, the present invention provides the following technical solution: This invention provides a method for early warning of abnormal test data in power batteries, comprising the following steps: S1. Real-time acquisition of the operating parameters of the testing equipment and the test data of the power battery under test during the power battery testing process; S2. Based on the battery status information in the test data, input it into the preset virtual battery model to calculate the simulated voltage value under the current operating condition; S3. Calculate the relative deviation rate between the measured voltage and the simulated voltage value of the tested power battery, and determine whether the data abnormality is caused by a battery problem based on whether the relative deviation rate exceeds a preset threshold corresponding to the battery type. S4. If the data anomaly is determined not to be caused by a battery problem, then the equipment fault diagnosis procedure is executed. This procedure comprehensively applies at least two of the following judgment criteria to diagnose the fault and determine its severity: (a) Basic fault mode determination based on data record characteristics, wherein the basic fault mode includes rule-based identification of at least one of data missing, data duplication and data jump; (b) Matching and judging based on the historical fault database of equipment historical fault information; (c) Based on the comparison of calibration data between the actual measured data of the equipment and the calibration reference data, the accuracy of the equipment is identified; S5. Based on the fault level determined in the fault diagnosis steps, execute the corresponding early warning operation.

[0007] Preferably, in step S2, the battery state information includes the battery's initial SOC, real-time current, and temperature; the virtual battery model is constructed based on an advanced equivalent circuit model and an SOC correction algorithm.

[0008] Preferably, in step S3, the preset threshold is configured differently according to the battery type, including a basic threshold, a charge / discharge switching threshold, a multi-scenario superposition threshold, and a continuous threshold determination abnormal time.

[0009] Preferably, the battery type includes lithium iron phosphate batteries and ternary lithium batteries; for lithium iron phosphate batteries, the basic threshold is 1%; for ternary lithium batteries, the basic threshold is 1.5%.

[0010] Preferably, the basic fault mode determination specifically includes: determining the temporal continuity of data recording based on a preset sampling frequency to identify data missing or duplicate data; and / or calculating the parameter change rate of adjacent data points and comparing it with a preset jump threshold to identify data jumps.

[0011] Preferably, the calibration data comparison and judgment specifically includes: periodically comparing the measured data of the test equipment under standard operating conditions with the benchmark data obtained by the high-precision calibration equipment to establish a health baseline for equipment accuracy; in real-time testing, by comparing the deviation trend between the current measured data and the health baseline, it is determined whether the equipment accuracy has decayed and exceeded the confidence threshold.

[0012] Preferably, the fault level determined in step S4 includes at least a reminder level and an alarm level; wherein, the reminder level corresponds to anomalies that do not affect core testing services and can be recorded through logs; the alarm level corresponds to anomalies that have affected or may affect the validity of test data and require immediate intervention.

[0013] Preferably, in step S5, for reminder-level warnings, system log marking and / or periodic summary notifications are executed; for alarm-level warnings, real-time platform pop-ups and / or instant messaging tool pushes are executed.

[0014] This invention also discloses a power battery test data anomaly early warning system for implementing the method, comprising: The data acquisition system is used to collect measured data in real time during the testing process of power batteries; A data analysis system, connected to the data acquisition system, has a built-in virtual battery model for receiving the measured data and calculating the battery simulation data; The fault diagnosis system is connected to the data analysis system and is used to compare the measured data with the simulated data, and to determine whether the root cause of the data anomaly is the test equipment based on preset rules, and to determine the anomaly level. A fault alarm system, connected to the fault judgment system, is used to perform corresponding early warning operations based on the anomaly level.

[0015] Preferably, the data acquisition system includes a power battery testing cabinet and an industrial control computer; the functions of the data analysis system, fault diagnosis system and fault alarm system are achieved collaboratively by a conversion protocol module, a database module, a gateway and a data processing system connected in sequence.

[0016] The beneficial effects of this invention are as follows: By introducing a virtual battery model to generate a theoretical voltage benchmark, this invention effectively distinguishes whether data anomalies originate from battery-related issues or test equipment malfunctions. Combined with a multi-dimensional comprehensive judgment mechanism based on rule-based fault mode diagnosis, historical fault database matching, and equipment accuracy degradation trend analysis, it achieves real-time and accurate identification and early warning of data recording anomalies and equipment accuracy degradation during power battery testing. This significantly improves the reliability of test data and the credibility of the testing process, avoiding misjudgments of battery performance due to equipment problems; it enables early prediction of implicit equipment accuracy degradation, transforming passive inspection into proactive maintenance and extending the effective service life of the equipment; and through a tiered early warning mechanism, it optimizes the operation and maintenance response process, improves testing efficiency and automation levels, and reduces manual inspection costs and safety risks. Attached Figure Description

[0017] Figure 1 This is a functional block diagram of a power battery test data anomaly early warning system provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the hardware composition of a power battery test data anomaly early warning system provided in an embodiment of the present invention; Figure 3 This is a flowchart of a power battery test data anomaly early warning method provided in an embodiment of the present invention.

[0018] Explanation of the labels in the diagram: 1: Data acquisition system; 2: Data analysis system; 3: Fault diagnosis system; 4: Fault alarm system; 11: Power battery test cabinet; 12: Industrial control computer; 13: Protocol conversion module; 14: Database module; 15: Gateway; 16: Data processing system. Detailed Implementation

[0019] The technical solution of the present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that the following embodiments are only used to explain the present invention and do not constitute a limitation on the scope of protection of the present invention.

[0020] like Figure 2 As shown, the early warning system of the present invention mainly comprises four parts that work together: a data acquisition system 1, a data analysis system 2, a fault judgment system 3, and a fault alarm system 4.

[0021] Its specific hardware implementation architecture is as follows Figure 3 As shown, it includes: Power Battery Test Cabinet 11: Used to apply charging and discharging conditions to the power battery under test, and has built-in sensors for voltage, current, temperature, etc., to collect the equipment's own operating parameters (such as total voltage and total current) and battery test data.

[0022] Industrial computer 12: Connected to test cabinet 11, it receives the raw data uploaded by the test cabinet 11. Test cabinet 11 and industrial computer 12 together constitute the data acquisition system 1.

[0023] Protocol conversion module 13: Used to convert data from the industrial computer 12 from device proprietary protocols (such as CAN bus data) to standard communication protocols (such as Modbus / TCP / IP) to facilitate network transmission and system integration.

[0024] Database module 14: Preferably using a MySQL database to store the converted measured data, preset virtual battery model data, historical calibration baseline data of the equipment, and the accumulated historical fault case library.

[0025] Gateway 15: Responsible for secure communication and data routing between modules within the system.

[0026] Data Processing System 16: As the core analysis engine of the system, it carries the logical functions of Data Analysis System 2, Fault Judgment System 3, and Fault Alarm System 4. It calls data from Database Module 14, performs model calculations and rule judgments, and generates early warning instructions.

[0027] The conversion protocol module 13, database module 14, gateway 15 and data processing system 16 work together to achieve data analysis, fault diagnosis and alarm functions.

[0028] Data analysis system 2 is built on MySQL (database) and is configured with virtual battery model data (including lithium iron phosphate and ternary lithium batteries, etc.). After the received data is formatted and cleaned, the obtained battery information (initial SOC, real-time current and temperature) is input into the virtual battery model in time. The virtual model takes "initial SOC + real-time current + temperature" as input and calculates "normal voltage value (simulated voltage) under the current operating condition" based on a preset algorithm (advanced equivalent circuit model + SOC correction), which serves as the benchmark for judging whether the measured voltage is abnormal. The difference is quantified by the relative deviation rate between the actual battery voltage and the simulated voltage. Relative deviation rate ΔU% = (U 实测 -U 模拟 ) / U 实测

[0029] If the data exceeds the threshold range, it is considered to be a data problem caused by a battery issue.

[0030] The fault diagnosis system 3 uses basic fault modes, historical fault databases and calibration data comparison as its core, and uses different methods to identify equipment data problems that correspond to different levels of early warning.

[0031] The basic failure modes include data errors caused by software problems, such as data loss, data duplication, and data jumps. The main approach is to deconstruct the equipment's recorded data through the database and, in conjunction with the characteristics that data should follow, such as temporal continuity, temporal uniqueness, and logical rationality, construct quantitative analysis rules to achieve automatic screening and location of abnormal data. For example: (1) The data recording time must be unique and continuous (e.g., a fixed sampling frequency of 10Hz, i.e., one data point every 100ms), as the basis for identifying missing, duplicate, or abrupt data. (2) Use Python to query the database to see if there is empty data or if previously normal data has become empty data; First, define the normal range for each parameter, and then mark the parts that exceed the range.

[0032] (3) Determine the voltage jump range by querying the database using Python, such as ensuring that the voltage jump does not exceed 10% every 100ms. The specific steps are as follows: Data grouping and preprocessing: Monitoring data is grouped according to the unique device identifier (device_id) (for scenarios involving parallel monitoring of multiple devices).

[0033] The time interval calculation is used to determine the time difference between two adjacent voltage data points from the same device, and is used to determine the temporal continuity of data acquisition.

[0034] The voltage change rate is calculated based on the voltage values ​​of adjacent data. The relative voltage change rate is calculated (formula: |current voltage - previous voltage| / previous voltage × 100%), which quantifies the voltage fluctuation amplitude.

[0035] Abnormal voltage jump determination: When the time interval is within the preset valid range (e.g., 90-110ms, which can be adjusted according to the actual scenario) and the voltage change rate exceeds the threshold (e.g., 1%), it is determined to be an abnormal voltage jump.

[0036] The historical fault database consists of the fault table that comes with the device initially and updates for problems that occur during operation, such as data errors caused by sudden hardware failures, such as sensor hardware damage, poor contact of the data acquisition card, communication abnormalities, and abnormal power supply voltage. It is combined with the data of the device's built-in sensors and fault combinations for comprehensive judgment. Calibration data comparison serves as a predictive tool for judging slow hardware failure. It involves periodically comparing the data with the calibrated data using standard charge and discharge methods. By setting a health baseline and confidence threshold, it determines the degree of degradation in the device's accuracy, thereby achieving the purpose of early warning.

[0037] After periodic calibration, the equipment's accuracy should meet the accuracy requirements specified in its parameters. For example, if the equipment's voltage accuracy is 0.05%FS and its maximum voltage range is 1000V, during calibration, when the required output voltage is 1000V, its actual output should be within ±1000V*1.005%, i.e., within the range of 999.5V to 1000.5V. During actual operation, the target voltage and current data need to be recorded and compared with the actual voltage and current data for trend analysis. If the accuracy deviation meets the trend prediction for three consecutive times and exceeds the accuracy range, the equipment accuracy degradation is considered to be excessive.

[0038] Based on the above data, the following fault levels and early warning methods are defined:

[0039] Example 1: Lithium Iron Phosphate Power Battery R&D Testing Scenario – Data Missing Anomaly Detection I. Scene Background Cycle life testing of a batch of lithium iron phosphate power batteries (model LFP-50Ah) requires continuous collection of individual cell voltage, current, and temperature data during the charging and discharging process (sampling frequency 10Hz, one data point every 100ms). In this scenario, occasional communication interruptions between the test cabinet and the industrial control computer can easily lead to data loss.

[0040] II. Implementation Resources Core resources: This invention's early warning system (including 1 power battery testing cabinet, 1 industrial control computer, MySQL database, conversion protocol module, gateway, and data processing system), 3 sets of lithium iron phosphate test samples (initial SOC 80%), and communication status monitoring software. III. Implementation Steps 1. System Deployment and Parameter Configuration Hardware connection: Connect the test cabinet (including voltage / current / temperature sensors) to the industrial control computer via CAN bus to form a data acquisition system; connect the conversion protocol module, database, gateway and data processing system in series to ensure that the data can be retrieved in real time after being transmitted to the database via Modbus protocol.

[0041] Rule setting: Configure "data missing judgment rule" in the data processing system - based on the 10Hz sampling frequency, set "data missing is judged when the time interval between two adjacent data acquisitions is >110ms (allowing ±10ms error); at the same time, configure the warning level: a reminder level warning is triggered when a single device loses a single piece of data (missing time ≤300ms), and an alarm level warning is triggered when there are 3 or more consecutive missing data.

[0042] 2. Data Acquisition and Anomaly Simulation Initiate the cycle life test. For the first 2 hours of normal equipment operation, data acquisition is continuous and without loss. A complete data record is generated in the database every 100ms (e.g., 10:00:00.000 records voltage 3.21V, 10:00:00.100 records voltage 3.20V).

[0043] Artificial communication interruption: At 2 hours and 5 minutes into the test, disconnect the CAN bus connector between the test cabinet and the industrial control computer for 1 second (corresponding to the time period during which 10 data points should be collected), and then reconnect it to simulate the "data loss caused by a momentary communication interruption" in a real-world scenario.

[0044] 3. Anomaly Detection and Early Warning Execution The data processing system verifies the temporal continuity of data in the database in real time: when it detects that there are only two data points, 10:02:05.000 (voltage 3.19V) and 10:02:05.900 (voltage 3.18V), within the period from 10:02:05.000 to 10:02:05.900, with no data in the eight 100ms intervals in between, it is determined as "eight consecutive data missing", which meets the alarm-level warning triggering conditions.

[0045] The fault alarm system immediately issued a warning: a red pop-up window appeared on the platform (displaying "Device ID: TC-001, 8 consecutive data entries were missing during the period of 10:02:05, it is recommended to check the communication connection"), and was simultaneously pushed to the test team group via DingTalk; team members arrived at the site within 5 minutes and found that the CAN bus connector was loose. After tightening it, data acquisition returned to normal.

[0046] IV. Implementation Results After data loss occurs, the system completes anomaly identification within 100ms and triggers an alert within 2 seconds, enabling timely detection of data anomalies.

[0047] Example 2: Ternary Lithium-ion Power Battery R&D Testing Scenario – Implementation of Equipment Precision Attenuation Anomaly Detection I. Scene Background For high-rate charge-discharge performance testing of a ternary lithium-ion battery (model NCM-70Ah), precise data acquisition of individual cell voltage and current at 1C, 2C, and 3C rates is required using a test cabinet to analyze the battery's charge-discharge efficiency and cycle stability. In this scenario, the voltage measurement module of the test cabinet is prone to current sensor drift and voltage sampling accuracy degradation due to prolonged high-frequency use (average daily testing time of 12 hours).

[0048] II. Implementation Resources Core resources: This invention includes an early warning system (containing 2 dedicated high-rate test cabinets, 1 industrial control computer, MySQL database, conversion protocol module, gateway, and data processing system), 5 sets of novel ternary lithium test samples (initial SOC 90%), a high-precision standard voltage calibrator (accuracy ±0.005% FS, used for equipment accuracy benchmark calibration), and high-rate charge / discharge control software. III. Implementation Steps 1. System Deployment and Accuracy Baseline Setting Hardware and software configuration: Two high-rate test cabinets (TC-01 and TC-02) are connected to an industrial control computer via a CAN bus to form a data acquisition system; a ternary lithium virtual battery model (adapted to 70Ah capacity, with a preset voltage calculation algorithm under high-rate conditions) is imported into the MySQL database; the data processing system is configured with a "precision attenuation judgment rule"—the voltage value measured by the standard calibrator is the "healthy baseline". If the measured voltage of the test cabinet deviates from the baseline for three consecutive times beyond the allowable range (equipment voltage accuracy 0.05% FS, maximum voltage 4.2V, allowable deviation ±0.0021V), it is judged as exceeding the accuracy attenuation limit.

[0049] Initial accuracy calibration: Before testing, use a high-precision standard voltage calibrator to calibrate the target voltage points (e.g., 3.0V, 3.6V, 4.2V) of the two test cabinets at 1C, 2C, and 3C rates. Record the baseline of "target voltage - measured voltage" at each rate (e.g., TC-01 at 2C rate and target voltage of 3.6V, the measured value is 3.6008V, the deviation is 0.0008V, which meets the accuracy requirements).

[0050] 2. Research and development testing and real-time accuracy monitoring Initiating high-rate charge-discharge testing: According to the R&D plan, 5 groups of ternary lithium samples were sequentially subjected to charge-discharge cycle tests at 1C (70A), 2C (140A), and 3C (210A) rates, with each group undergoing 10 cycles. The test cabinet collected individual cell voltage and current data every 50ms, which were transmitted to the database via an industrial control computer and a conversion protocol module. The data processing system compared the measured voltage with the initial calibration baseline in real time and generated accuracy deviation trend curves at each rate.

[0051] Deviation Anomaly Detection: When the test reached the 8th cycle (2C rate), the data processing system detected that the voltage deviation of TC-01 showed a continuous upward trend: when the target voltage was 3.6V, the first measured value was 3.6015V (deviation 0.0015V), the second was 3.6020V (deviation 0.0020V), and the third was 3.6023V (deviation 0.0023V). The deviation exceeded the allowable range of ±0.0021V for three consecutive times, triggering the accuracy decay warning judgment logic.

[0052] 3. Anomaly Confirmation and Early Warning Response Handling The system automatically retrieved historical accuracy data from TC-01 and found that during the past 10 days of testing, the voltage deviation at the 2C rate gradually increased from the initial 0.0008V to 0.0023V, which is consistent with the characteristics of "slow hardware failure after long-term use of equipment". It was determined to be an alarm-level warning and was immediately pushed to the testing team through a platform pop-up window, and a warning email (including deviation trend chart and test data during the period of abnormality) was sent at the same time.

[0053] The TC-01 test was suspended and recalibrated using a standard calibrator: it was confirmed that the voltage measurement module of the test cabinet was drifting due to high frequency and high current impact. After replacing it with a high-precision measurement module of the same model, it was recalibrated (the actual measured value was 3.6009V at 2C rate and target value of 3.6V), and the equipment accuracy was restored.

[0054] IV. Implementation Results Reliability of R&D data: This implementation uses real-time early warning to detect the accuracy degradation of one test cabinet in a timely manner, avoiding R&D misjudgments caused by data deviation (such as mistakenly treating equipment accuracy problems as insufficient high-rate performance of batteries).

[0055] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for early warning of abnormal test data in power batteries, characterized in that, Includes the following steps: S1. Real-time acquisition of the operating parameters of the testing equipment and the test data of the power battery under test during the power battery testing process; S2. Based on the battery status information in the test data, input it into the preset virtual battery model to calculate the simulated voltage value under the current operating condition; S3. Calculate the relative deviation rate between the measured voltage and the simulated voltage value of the tested power battery, and determine whether the data abnormality is caused by a battery problem based on whether the relative deviation rate exceeds a preset threshold corresponding to the battery type. S4. If the data anomaly is determined not to be caused by a battery problem, then the equipment fault diagnosis procedure is executed. This procedure comprehensively applies at least two of the following judgment criteria to diagnose the fault and determine its severity: (a) Basic fault mode determination based on data record characteristics, wherein the basic fault mode includes rule-based identification of at least one of data missing, data duplication and data jump; (b) Matching and judging based on the historical fault database of equipment historical fault information; (c) Based on the comparison of calibration data between the actual measured data of the equipment and the calibration reference data, the accuracy of the equipment is identified; S5. Based on the fault level determined in the fault diagnosis steps, execute the corresponding early warning operation.

2. The method for early warning of abnormal power battery test data according to claim 1, characterized in that, In step S2, the battery state information includes the battery's initial SOC, real-time current, and temperature; the virtual battery model is constructed based on an advanced equivalent circuit model and an SOC correction algorithm.

3. The method for early warning of abnormal power battery test data according to claim 1, characterized in that, In step S3, the preset threshold is configured differently according to the battery type, including a basic threshold, a charge / discharge switching threshold, a multi-scenario superposition threshold, and a continuous threshold exceeding the abnormal time.

4. The method for early warning of abnormal power battery test data according to claim 3, characterized in that, The battery types include lithium iron phosphate batteries and ternary lithium batteries; for lithium iron phosphate batteries, the basic threshold is 1%; for ternary lithium batteries, the basic threshold is 1.5%.

5. The method for early warning of abnormal power battery test data according to claim 1, characterized in that, The basic fault mode determination specifically includes: determining the temporal continuity of data recording based on a preset sampling frequency to identify data missing or duplicate data; and / or calculating the parameter change rate of adjacent data points and comparing it with a preset jump threshold to identify data jumps.

6. The method for early warning of abnormal power battery test data according to claim 1, characterized in that, The calibration data comparison and judgment specifically includes: periodically comparing the measured data of the test equipment under standard operating conditions with the benchmark data obtained by the high-precision calibration equipment to establish a health baseline for equipment accuracy; in real-time testing, by comparing the deviation trend between the current measured data and the health baseline, it is determined whether the equipment accuracy has decayed and exceeded the confidence threshold.

7. The method for early warning of abnormal power battery test data according to claim 1, characterized in that, The fault level determined in step S4 includes at least the alert level and the alarm level; wherein, the alert level corresponds to anomalies that do not affect core test services and can be recorded through logs; the alarm level corresponds to anomalies that have affected or may affect the validity of test data and require immediate intervention.

8. The method for early warning of abnormal power battery test data according to claim 7, characterized in that, In step S5, for reminder-level alerts, system log marking and / or periodic summary notifications are executed; for alarm-level alerts, real-time platform pop-ups and / or instant messaging tool pushes are executed.

9. A power battery test data anomaly early warning system, used to implement the method as described in any one of claims 1-8, characterized in that, include: The data acquisition system (1) is used to collect measured data in real time during the power battery testing process; The data analysis system (2) is connected to the data acquisition system (1) and has a built-in virtual battery model for receiving the measured data and calculating the battery simulation data. The fault judgment system (3) is connected to the data analysis system (2) and is used to compare the measured data with the simulated data, and to determine whether the root cause of the data anomaly is the test equipment based on preset rules, and to determine the anomaly level. The fault alarm system (4) is connected to the fault judgment system (3) and is used to perform corresponding early warning operations according to the abnormality level.

10. The system according to claim 9, characterized in that, The data acquisition system (1) includes a power battery test cabinet (11) and an industrial control computer (12); the functions of the data analysis system (2), the fault judgment system (3) and the fault alarm system (4) are realized by the sequentially connected conversion protocol module (13), database module (14), gateway (15) and data processing system (16).

Images