A method and system for detecting faults in a battery cell
By collecting battery cell parameter data, using generalized principal component analysis and machine learning to build a fault detection model, the real-time and automation problems of hybrid circuit chip fault detection in the existing technology are solved, and accurate monitoring and dynamic adjustment of battery cell status are achieved. , improves the accuracy and efficiency of fault detection.
Patent Information
- Application Number
- CN202411018131.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-29
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2044-07-29
AI Technical Summary
The prior art has real-time limitations in online fault detection of hybrid circuit chips, cannot cover all possible fault conditions, low degree of automation and lack dynamic adjustment capabilities.
By regularly collecting battery cell parameter data, using generalized principal component analysis to extract battery cell status characteristics, setting the DC internal resistance calculation cycle, automatically triggering the detection process, combining machine learning to build a fault detection model, dynamically adjust the detection cycle, generate a detection report and take Corresponding measures.
It improves the identification accuracy and automation of fault detection, reduces maintenance costs, and ensures the healthy state of the battery cells of the tower base station and the stable operation of the system.
Smart Images

Figure CN118534334B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of electronic circuits, and in particular to a method and system for detecting faults of a battery cell. Background Art
[0002] As an indispensable energy storage device in modern society, batteries are widely used in mobile devices, electric vehicles, energy storage systems and other fields. With the rapid development of these fields, the demand for high-performance and high-safety batteries is increasing. However, due to the characteristics of the battery itself and the uncertainty of the use environment, battery abnormalities occur from time to time, which has prompted the continuous development of battery fault detection technology.
[0003] Battery failure not only affects the normal operation of the equipment, but may also cause safety accidents and economic losses. For example, the attenuation of battery capacity will affect the endurance of the equipment; the increase of battery internal resistance will reduce the discharge performance; abnormal battery temperature may cause thermal runaway safety issues; abnormal battery voltage will affect the normal operation of the equipment or even damage the equipment. Therefore, it is of great significance to detect battery faults. Traditional methods mainly rely on monitoring and analyzing battery parameters, such as measuring capacity, internal resistance, temperature and voltage parameters, and comparing them with standard values. However, this method can often only detect obvious abnormalities, and it is difficult to detect more hidden abnormalities.
[0004] For example, the patent with the authorization announcement number CN108241114B discloses an online fault detection method, device and fault detection circuit for a hybrid circuit chip, including: in the normal operation of the device where the hybrid circuit chip is located, real-time acquisition of parameter values of preset fault-sensitive parameters of the hybrid circuit chip; feature extraction of the acquired parameter values of each fault-sensitive parameter to obtain the current waveform features of each fault-sensitive parameter; matching the obtained current waveform features of each fault-sensitive parameter with the correspondence between the preset waveform features and the fault mode to obtain the fault mode corresponding to the current waveform features; and determining and outputting the fault diagnosis result according to the obtained fault mode. By using the embodiment of the present invention, the purpose of online detection of whether a hybrid circuit chip has a fault is achieved.
[0005] For example, the patent with the authorization announcement number CN217360196U discloses a fault detection device for an electronic chip, including: a detection device body and a mounting member fixedly mounted on the detection device body, the mounting member is provided with a mounting groove, two groups of mounting plates are slidably mounted on both sides of the mounting member, and a support assembly for fixing the chip body is fixedly mounted inside the mounting groove; a plurality of groups of detection grooves are provided on one end face of the mounting plate, and a plurality of groups of connection seats for connecting with the detection pen are staggeredly installed on the mounting plate, and a connection sleeve for inserting the pins is fixedly installed inside the plurality of detection grooves, and each group of the connection sleeves is connected to each group of connection seats through each group of connecting wires. Through the above structure, the technical solution can conveniently lead out the pins of the chip, and can conveniently insert and fix the detection pen during the detection process, which can make the detection process more convenient.
[0006] The above existing technologies all have the following problems: 1) The online fault detection method of hybrid circuit chips is limited in real time due to the need to process multiple fault-sensitive parameters and complex feature extraction processes; 2) The correspondence between the pre-set waveform features and the fault mode may not cover all possible fault conditions; 3) The degree of automation is low and lacks dynamic adjustment capabilities. Summary of the invention
[0007] In view of the shortcomings of the prior art, the present invention proposes a method and system for fault detection of battery cells, which collects battery cell parameter data regularly, and uses generalized principal component analysis to extract battery cell status characteristics after preprocessing; sets a DC internal resistance calculation cycle to automatically trigger a detection process, including charging, discharging tests and internal resistance calculation; calculates the maximum distortion coefficient of the battery cell internal resistance, and compares it with a preset threshold to achieve fault warning and protection; combines machine learning to build a fault detection model to improve recognition accuracy; automatically generates a detection report, takes measures according to the fault level and dynamically adjusts the detection cycle; effectively monitors the health status of the battery cells of the tower base station, ensures the stable operation of the system, and reduces maintenance costs.
[0008] To achieve the above object, the present invention provides the following technical solutions:
[0009] A method for detecting a fault of a battery cell, comprising:
[0010] Step S1: collecting the cell parameter data in the tower base station battery system and preprocessing it, and extracting the cell state feature quantity using the generalized principal component analysis method based on the preprocessed cell parameter data;
[0011] Step S2: Setting the DC internal resistance calculation cycle, the battery management system monitors the current time in real time, and if the set DC internal resistance calculation cycle is reached, the battery cell detection process is automatically triggered, otherwise, it continues to wait until the next cycle;
[0012] Step S3: After the detection process is started, the battery management system sends a charging instruction to the DC / DC module through the SCI interface, starts charging the battery, and monitors the battery charging status. If it is detected that the battery is fully charged, the battery output voltage is adjusted to the current busbar voltage level, and a short-time discharge test is performed;
[0013] Step S4: During the charging and short-time discharging process, the battery management system collects the battery pack current and battery cell voltage in real time, and uses the collected data to calculate the internal resistance value of each battery cell, construct the battery overall internal resistance array, and analyze the internal resistance array at the same time, and obtain the maximum distortion coefficient of the battery cell internal resistance by calculating the maximum value and median of the DC internal resistance;
[0014] Step S5: setting an alarm threshold and a protection threshold, comparing the maximum distortion coefficient of the cell internal resistance with the preset alarm threshold and protection threshold, judging whether it is necessary to issue an alarm or perform a protection operation according to the comparison result, and using a machine learning algorithm in combination with the cell state feature quantity to construct a cell fault detection model;
[0015] Step S6: According to the recognition result of the battery cell fault detection model, a detection report is automatically generated. If a battery cell fault is detected, an alarm message is issued or a protection operation is performed according to the fault level, and the operation and maintenance personnel are notified to handle it in time. According to the health status of the battery cell, the DC internal resistance detection cycle is dynamically adjusted.
[0016] Specifically, the judgment formula for reaching the DC internal resistance calculation cycle in step S2 is: , then the calculation cycle is reached, where e represents the error, Indicates the current time. Indicates the last DC internal resistance calculation time, Indicates the set DC internal resistance calculation cycle.
[0017] Specifically, the specific steps of step S4 include:
[0018] S4.1: Calculate the internal resistance of each battery cell based on the collected data. The formula is: ;in, Indicates n The internal resistance of each battery cell, Indicates the first n The voltage of each battery cell, Indicates the first n The voltage of each battery cell, represents the temperature correction factor, represents the voltage recovery correction factor, Indicates the battery pack current collected during the charging process. Indicates the battery pack current collected when charging is suspended. and represents the empirical coefficient, Indicates the change in battery temperature during charging pause. Indicates the change in battery pack current before and after charging is suspended;
[0019] S4.2: Store the internal resistance value of each battery cell into an array to form the overall internal resistance array of the battery Specifically, the specific steps of step S4 also include:
[0020] S4.3: Calculate the maximum and median statistics of the battery's overall internal resistance array, analyze the internal resistance distribution, and evaluate the health of the current battery cell based on the internal resistance distribution results. The formula is: ;
[0021] in, Indicates the maximum value of the battery's overall internal resistance array, Indicates n The correction function of each battery cell for temperature and battery aging, Indicates n The temperature of each battery cell, Indicates n The battery aging degree of each battery cell, Represents the median of the battery as a whole, Indicates remainder;
[0022] S4.4: Based on , Statistics, calculate the maximum distortion coefficient of the battery cell internal resistance, and output the internal resistance array, statistics, and distortion coefficient results to the BMS display.
[0023] Specifically, the calculation formula for the maximum distortion coefficient of the cell internal resistance in S4.4 is: ;in, Indicates the maximum distortion coefficient of the cell internal resistance, and represents the weight coefficient, Indicates the current focal length. represents the basic focal length, Indicates perspective.
[0024] Specifically, the specific steps in step S5 include: S5.1: Setting the alarm threshold and protection threshold , Will and Make comparisons;
[0025] like , then DC internal resistance protection is generated; if , a DC internal resistance abnormality alarm is generated, and an improvement strategy is used to increase the detection cycle. The improvement strategy is expressed as: ,in, Indicates the next round of calculation cycle, and indicates the shrinkage coefficient.
[0026] Specifically, the specific steps in step S5 also include:
[0027] S5.2: Based on the acquired cell state feature quantity, a cell fault detection model is trained using a machine learning algorithm, and the performance of the cell fault detection model is evaluated using a cross-validation method. The formula is: ; Where P represents the failure probability, represents the machine learning model, U represents voltage, I represents current, and D represents the overall internal resistance array of the battery. t Indicates battery temperature;
[0028] S5.3: Adjust the parameters of the cell fault detection model according to the test results, optimize the performance of the cell fault detection model, and introduce regularization to prevent overfitting;
[0029] S5.4: Deploy the trained cell fault detection model to the actual system, receive the cell status data in real time and perform fault detection.
[0030] A battery cell fault detection system includes: a data acquisition module, a trigger module, a charging module, an internal resistance analysis module, a fault detection module, and an update module;
[0031] The data acquisition module is used to collect data from the battery module;
[0032] The trigger module is used to monitor whether the parameter data of the battery cell exceeds the normal range, and to start the entire detection process when it exceeds the normal range; the charging module is used to charge the battery cell to a specific power level or perform a charging cycle test before performing fault detection;
[0033] The internal resistance analysis module is used to measure and analyze the internal resistance of the battery cell;
[0034] The fault detection module is used to comprehensively analyze the information obtained from the data acquisition module, the trigger module and the internal resistance analysis module, and to evaluate and analyze the state of the battery cell using a pattern recognition algorithm and a preset standard or threshold, so as to determine whether the battery cell has a fault and the type and degree of the fault. Once an abnormality is found, the module will trigger an alarm or take preventive measures;
[0035] The update module is responsible for updating system software and firmware.
[0036] Specifically, the internal resistance analysis module includes: an internal resistance calculation unit and an internal resistance analysis unit;
[0037] The internal resistance calculation unit is used to calculate the internal resistance value of the battery cell using the collected voltage and current data;
[0038] The internal resistance analysis unit is used to analyze the internal resistance value to determine whether the health of the battery cell is good or not, and whether there is any fault.
[0039] Specifically, the fault detection module includes: a model building unit and a judgment unit;
[0040] The model building unit is used to build a fault detection model based on historical data and expert knowledge;
[0041] The judgment unit is used to compare the real-time data with the fault detection model to judge whether the battery cell has a fault.
[0042] Compared with the prior art, the present invention has the following beneficial effects:
[0043] 1. The present invention proposes a battery cell fault detection system, and optimizes and improves the architecture, operation steps and processes. The system has the advantages of simple process, low investment and operation costs, and low production work costs.
[0044] 2. The present invention proposes a method for fault detection of battery cells. By collecting the voltage, current and temperature parameter data of the battery cells in real time and performing preprocessing and feature extraction, the actual state of the battery cells can be reflected more accurately. The generalized principal component analysis method is used to extract the characteristic quantity of the battery cell state, which is helpful to eliminate redundant information and improve the recognition accuracy of the subsequent fault detection model. A DC internal resistance calculation cycle is set, and the battery cell detection process is automatically triggered to realize the automation of detection, which not only reduces manual intervention and improves detection efficiency, but also avoids errors that may be caused by human factors.
[0045] 3. The present invention proposes a method for fault detection of battery cells. During the charging and short-time discharging process, the battery pack current and battery cell voltage are collected in real time, and the internal resistance value of each battery cell is calculated to construct an overall internal resistance array of the battery. By analyzing the internal resistance array and calculating the maximum distortion coefficient of the battery cell internal resistance, the abnormal conditions of the battery cells can be comprehensively detected, including but not limited to the fault modes of increased internal resistance and capacity attenuation. A battery cell fault detection model is constructed using a machine learning algorithm, which can continuously learn and optimize the recognition algorithm to improve the intelligence level of fault detection. At the same time, fault identification is performed in combination with the characteristic quantity of the battery cell state, so that the type and degree of the fault can be more accurately determined. The DC internal resistance detection cycle is dynamically adjusted according to the health status of the battery cell, so as to realize continuous monitoring and adaptive adjustment of the battery cell state, which helps to reduce the detection cost and improve the detection efficiency while ensuring the detection accuracy. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] Figure 1 A schematic diagram of a method for detecting a fault of a battery cell according to the present invention;
[0047] Figure 2 This is a principle flow chart of a method for detecting a fault in a battery cell according to the present invention;
[0048] Figure 3 A flow chart of calculation of internal resistance of a battery cell in a method for detecting a battery cell fault according to the present invention;
[0049] Figure 4 This is an architecture diagram of a battery cell fault detection system according to the present invention. DETAILED DESCRIPTION
[0050] Example 1: Please refer to Figure 1-Figure 2 , an embodiment of the present invention provides: a method for detecting a fault of a battery cell, comprising the following steps:
[0051] Step S1: collecting the cell parameter data in the tower base station battery system and preprocessing it, and extracting the cell state feature quantity using the generalized principal component analysis method based on the preprocessed cell parameter data;
[0052] Among them, the battery cell parameter data includes voltage, current, and temperature; the generalized principal component analysis method is the existing technology content in this field, is not the creative solution of this application, and will not be elaborated here.
[0053] Step S2: Setting the DC internal resistance calculation cycle, the battery management system monitors the current time in real time, and if the set DC internal resistance calculation cycle is reached, the battery cell detection process is automatically triggered, otherwise, it continues to wait until the next cycle;
[0054] Furthermore, the specific steps of step S2 include:
[0055] (1) Set the DC internal resistance calculation cycle: In the configuration interface of the battery management system, set the DC internal resistance calculation cycle. This cycle can be determined based on the characteristics of the battery cell, usage scenarios, and actual needs, such as once a week, once a day, or once an hour.
[0056] (2) Real-time monitoring of current time: The battery management system needs to integrate a real-time clock function so that it can accurately obtain current time information;
[0057] (3) Determine whether the calculation cycle has been reached: The control unit in the battery management system will periodically check the difference between the current time and the last DC internal resistance calculation time. If the difference reaches or exceeds the set calculation cycle, it is considered that the DC internal resistance calculation needs to be performed;
[0058] (4) Triggering the cell detection process: Once it is determined that the calculation cycle has been reached, the battery management system will automatically trigger the cell detection process, which may include multiple steps such as data collection, preprocessing, feature extraction, and internal resistance calculation;
[0059] (5) Continue waiting or enter the next cycle: If the current time has not reached the calculation cycle, the battery management system will continue to wait until the next cycle arrives and make another judgment.
[0060] Step S3: After the detection process is started, the battery management system sends a charging instruction to the DC / DC module through the SCI interface, starts charging the battery, and monitors the battery charging status. If it is detected that the battery is fully charged, the battery output voltage is adjusted to the current busbar voltage level, and a short-time discharge test is performed;
[0061] Furthermore, the specific steps of step S3 include:
[0062] (1) Start the test process: When the vehicle or equipment starts the charging process, the battery management system first performs a self-check to ensure that all sensors and communication interfaces are working properly. After confirming that the charging conditions are met, such as the charging gun is connected and the battery temperature is within a safe range, the battery management system is ready to send a charging command;
[0063] (2) Sending charging instructions: The battery management system sends charging instructions to the DC / DC module through the SCI communication interface, including target charging current and voltage parameters. The DC / DC module adjusts its working state according to the received instructions and starts charging the battery.
[0064] (3) Monitoring battery charging status: The battery management system continuously monitors the battery’s voltage, current, and temperature parameters to ensure that the charging process is carried out within a safe range. It also uses built-in algorithms to calculate the battery’s state of charge and health status to evaluate the battery’s charging progress and performance.
[0065] (4) Detecting whether the battery is fully charged: When the battery SOC reaches the preset full charge threshold, the battery management system determines that the battery is fully charged. It can also determine whether the battery is fully charged by monitoring whether the battery voltage is stable within a fixed range.
[0066] (5) Adjusting battery output voltage: After the battery is fully charged, the battery management system adjusts the battery output voltage to match the current bus voltage level to reduce energy loss and potential safety risks;
[0067] (6) Perform a short-time discharge test: In order to verify the charging effect and performance of the battery, the battery management system will perform a short-time discharge test. During the test, the battery management system controls the battery to discharge at a certain current for a period of time, and then re-evaluates the battery's state of charge and health status.
[0068] Step S4: During the charging and short-time discharging process, the battery management system collects the battery pack current and battery cell voltage in real time, and uses the collected data to calculate the internal resistance value of each battery cell, construct the battery overall internal resistance array, and analyze the internal resistance array at the same time, and obtain the maximum distortion coefficient of the battery cell internal resistance by calculating the maximum value and median of the DC internal resistance;
[0069] Step S5: setting an alarm threshold and a protection threshold, comparing the maximum distortion coefficient of the cell internal resistance with the preset alarm threshold and protection threshold, judging whether it is necessary to issue an alarm or perform a protection operation according to the comparison result, and using a machine learning algorithm in combination with the cell state feature quantity to construct a cell fault detection model;
[0070] Step S6: According to the recognition result of the battery cell fault detection model, a detection report is automatically generated. If a battery cell fault is detected, an alarm message is issued or a protection operation is performed according to the fault level, and the operation and maintenance personnel are notified to handle it in time. According to the health status of the battery cell, the DC internal resistance detection cycle is dynamically adjusted.
[0071] Example 2: Please refer to Figure 3 In this embodiment, the judgment formula for reaching the DC internal resistance calculation cycle in step S2 is: , then the calculation cycle is reached, where e represents the error, Indicates the current time. Indicates the last DC internal resistance calculation time, Indicates the set DC internal resistance calculation cycle.
[0072] The specific steps of step S4 include:
[0073] S4.1: Calculate the internal resistance of each battery cell based on the collected data. The formula is: ;in, Indicates n The internal resistance of each battery cell, Indicates the first n The voltage of each battery cell is in V. Indicates the first n The voltage of each battery cell is in V. Represents the temperature correction factor, which is used to compensate for the effect of temperature on internal resistance. It can be a temperature function based on experimental data or a constant related to the current battery temperature. It represents the voltage recovery correction factor, which is used to take into account the voltage recovery phenomenon that may occur during the pause period, that is, the voltage may rise slightly due to the weakening of the polarization effect. It can be determined based on the specific battery chemical characteristics and experimental data. Indicates the battery pack current collected during the charging process, in A. Indicates the battery pack current collected when charging is suspended, in A. and Represents empirical coefficients, which are used to adjust the influence of temperature and internal current changes on the internal resistance calculation. These coefficients need to be calibrated through experimental data. Indicates the change in battery temperature during charging pause, which can be obtained through the temperature sensor, in units of , Indicates the change in battery pack current before and after charging is suspended, in A;
[0074] S4.2: Store the internal resistance value of each battery cell into an array to form the overall internal resistance array of the battery .
[0075] S4.3: Calculate the maximum and median statistics of the battery's overall internal resistance array, analyze the internal resistance distribution, and evaluate the health of the current battery cell based on the internal resistance distribution results. The formula is: ;in, Indicates the maximum value of the battery's overall internal resistance array, Indicates n The correction function of the temperature and battery aging of each battery cell is used to adjust the i The DC internal resistance of a battery cell at a given temperature and aging degree, Indicates n The temperature of each battery cell, Indicates n The battery aging degree of each battery cell, Represents the median of the battery as a whole, Indicates remainder;
[0076] Further, The calculation formula is: ;
[0077] in, represents the normalization constant, represents the temperature influence coefficient, represents the reference temperature, A coefficient representing the effect of battery aging, Indicates the capacity attenuation percentage; S4.4: According to , Statistics, calculate the maximum distortion coefficient of the battery cell internal resistance, and output the internal resistance array, statistics, and distortion coefficient results to the BMS display.
[0078] The calculation formula for the maximum distortion coefficient of the cell internal resistance in S4.4 is: ;
[0079] in, Indicates the maximum distortion coefficient of the cell internal resistance, and Represents the weight coefficient, which is used to adjust the influence of different factors on the distortion coefficient. It is obtained by fitting experimental data. Indicates the current focal length. Indicates the base focal length, scale factor It reflects the effect of focal length change on distortion. Represents the viewing angle in degrees or radians.
[0080] The specific steps in step S5 include:
[0081] S5.1: Setting alarm thresholds and protection threshold , Will and Make comparisons;
[0082] like , then DC internal resistance protection is generated;
[0083] like , a DC internal resistance abnormality alarm is generated, and an improvement strategy is used to increase the detection cycle. The improvement strategy is expressed as: ,in, Indicates the next round of calculation cycle, and indicates the shrinkage coefficient.
[0084] S5.2: Based on the acquired cell state feature quantity, a cell fault detection model is trained using a machine learning algorithm, and the performance of the cell fault detection model is evaluated using a cross-validation method. The formula is: ;
[0085] Where P represents the failure probability, represents the machine learning model, U represents voltage, I represents current, and D represents the overall internal resistance array of the battery. t Indicates battery temperature;
[0086] S5.3: Adjust the parameters of the cell fault detection model according to the test results, optimize the performance of the cell fault detection model, and introduce regularization to prevent overfitting;
[0087] S5.4: Deploy the trained cell fault detection model to the actual system, receive the cell status data in real time and perform fault detection.
[0088] Example 3: Please refer to Figure 4 Another embodiment of the present invention provides a battery cell fault detection system, comprising:
[0089] Data acquisition module, trigger module, charging module, internal resistance analysis module, fault detection module, update module;
[0090] A data acquisition module, used to collect data from the battery module, including the voltage, current and temperature parameters of each battery cell;
[0091] The trigger module is used to monitor whether the parameter data of the battery cell exceeds the normal range, and start the entire detection process when it exceeds the normal range. For example, it activates other modules for data collection or analysis at a predetermined time, a specific event, such as the end of the battery charge and discharge cycle, or when an abnormality is detected. It can also control the start and end of the test sequence so that each test step is executed in a predetermined order;
[0092] A charging module, used to charge the battery to a specific charge level or perform a charging cycle test before performing a fault detection;
[0093] Internal resistance analysis module, used to measure and analyze the internal resistance of the battery cell. High internal resistance indicates battery aging and internal short circuit fault;
[0094] The fault detection module is used to comprehensively analyze the information obtained from the data acquisition module, the trigger module and the internal resistance analysis module, and to evaluate and analyze the status of the battery cell using the pattern recognition algorithm and the preset standard or threshold, so as to determine whether the battery cell has a fault and the type and degree of the fault, including overheating, voltage imbalance, internal resistance increase, and capacity attenuation. Once an abnormality is found, the module will trigger an alarm or take preventive measures;
[0095] The update module is responsible for updating the system software and firmware to ensure that the fault detection system can adapt to changes in battery performance, identify new fault modes, and improve the performance and accuracy of the overall system, including downloading the latest detection algorithms, calibration data or safety protocols.
[0096] The data acquisition module includes: a data acquisition unit, a preprocessing unit, a feature extraction unit, a control unit, and a storage unit;
[0097] The data acquisition unit is responsible for collecting real-time parameter data from the battery cell;
[0098] A preprocessing unit is used to perform preliminary processing on the collected parameter data, such as denoising and filtering, to improve data quality;
[0099] A feature extraction unit, used to extract feature information for fault detection from the preprocessed parameter data;
[0100] The control unit is responsible for the process control of the entire data acquisition module to ensure the accuracy and timeliness of parameter data acquisition;
[0101] The storage unit is used to store the collected and processed parameter data.
[0102] The trigger module includes: a cycle setting unit and a time monitoring unit;
[0103] The cycle setting unit is used by the user to set the cycle of fault detection, such as weekly, daily, and hourly;
[0104] The time monitoring unit is used to monitor the current time in real time. When the set detection cycle is reached, the data collection and fault detection process is triggered.
[0105] The charging module includes: a charging control unit, a charging status monitoring unit, and a discharge test unit;
[0106] Charging control unit, used to control charging current and voltage according to the current state of the battery cell and charging requirements to achieve intelligent charging;
[0107] The charging status monitoring unit is used to monitor the status of the battery cell during the charging process in real time, such as voltage, current, and temperature, to ensure the safety of the charging process;
[0108] Discharge test unit, used to perform discharge test on battery cells to evaluate their performance and health.
[0109] The internal resistance analysis module includes: an internal resistance calculation unit and an internal resistance analysis unit;
[0110] An internal resistance calculation unit, used to calculate the internal resistance value of the battery cell using the collected voltage and current data;
[0111] The internal resistance analysis unit is used to analyze the internal resistance value to determine whether the health of the battery cell is good or whether there is any fault.
[0112] The fault detection module includes: a threshold setting unit, a model building unit, a judgment unit, and an alarm output unit;
[0113] A threshold setting unit is used to allow the user to set the thresholds of various fault detection parameters, such as the upper and lower limits of voltage, current, and internal resistance;
[0114] A model building unit, used to build a fault detection model based on historical data and expert knowledge to improve the accuracy and reliability of fault detection;
[0115] A judgment unit, used to compare the real-time data with the fault detection model to determine whether the battery cell has a fault;
[0116] The alarm output unit is used to output alarm information in time when a fault is detected in the battery cell, and notify relevant personnel to handle it.
[0117] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in the field may also change, modify, replace and modify the above-mentioned embodiments without departing from the purpose and scope of protection of the present invention, and all of these are within the protection of the present invention.
Claims
1. A method for detecting a fault in a battery cell, characterized in that: include: Step S1: collecting the cell parameter data in the tower base station battery system and preprocessing it, and extracting the cell state feature quantity using the generalized principal component analysis method based on the preprocessed cell parameter data; Step S2: Setting the DC internal resistance calculation cycle, the battery management system monitors the current time in real time, and if the set DC internal resistance calculation cycle is reached, the battery cell detection process is automatically triggered, otherwise, it continues to wait until the next cycle; Step S3: After the detection process is started, the battery management system sends a charging instruction to the DC / DC module through the SCI interface, starts charging the battery, and monitors the battery charging status. If it is detected that the battery is fully charged, the battery output voltage is adjusted to the current busbar voltage level, and a short-time discharge test is performed; Step S4: During the charging and short-time discharging process, the battery management system collects the battery pack current and battery cell voltage in real time, and uses the collected data to calculate the internal resistance value of each battery cell, construct the battery overall internal resistance array, and analyze the internal resistance array at the same time, and obtain the maximum distortion coefficient of the battery cell internal resistance by calculating the maximum value and median of the DC internal resistance; Step S5: setting an alarm threshold and a protection threshold, comparing the maximum distortion coefficient of the cell internal resistance with the preset alarm threshold and protection threshold, judging whether it is necessary to issue an alarm or perform a protection operation according to the comparison result, and using a machine learning algorithm in combination with the cell state feature quantity to construct a cell fault detection model; Step S6: According to the recognition result of the battery cell fault detection model, a detection report is automatically generated. If a battery cell fault is detected, an alarm message is issued or a protection operation is performed according to the fault level, and the operation and maintenance personnel are notified to handle it in time. According to the health status of the battery cell, the DC internal resistance detection cycle is dynamically adjusted.
2. A method for detecting a fault in a battery cell according to claim 1, characterized in that: The judgment formula for reaching the DC internal resistance calculation cycle in step S2 is: , then the calculation cycle is reached, where e represents the error, Indicates the current time. Indicates the last DC internal resistance calculation time, Indicates the set DC internal resistance calculation cycle.
3. A method for detecting a fault in a battery cell as claimed in claim 2, characterized in that: The specific steps of step S4 include: S4.1: calculating the internal resistance of each battery cell according to the collected data, the formula is: ; in, Indicates n The internal resistance of each battery cell, Indicates the first n The voltage of each battery cell, Indicates the first n The voltage of each battery cell, represents the temperature correction factor, represents the voltage recovery correction factor, Indicates the battery pack current collected during the charging process. Indicates the battery pack current collected when charging is suspended. and represents the empirical coefficient, Indicates the change in battery temperature during charging pause. Indicates the change in battery pack current before and after charging is suspended; S4.2: Store the internal resistance value of each battery cell into an array to form the overall internal resistance array of the battery .
4. A method for detecting a fault in a battery cell as claimed in claim 3, characterized in that: The specific steps of step S4 also include: S4.3: Calculate the maximum and median statistics of the battery's overall internal resistance array, analyze the internal resistance distribution, and evaluate the health of the current battery cell based on the internal resistance distribution results. The formula is: ; in, Indicates the maximum value of the battery's overall internal resistance array, Indicates n The correction function of each battery cell for temperature and battery aging, Indicates n The temperature of each battery cell, Indicates n The battery aging degree of each battery cell, Represents the median of the battery as a whole, Indicates remainder; S4.4: Based on , Statistics, calculate the maximum distortion coefficient of the battery cell internal resistance, and output the internal resistance array, statistics, and distortion coefficient results to the BMS display.
5. A method for detecting a fault in a battery cell as claimed in claim 4, characterized in that: The calculation formula of the maximum distortion coefficient of the cell internal resistance in S4.4 is: ; in, Indicates the maximum distortion coefficient of the cell internal resistance, and represents the weight coefficient, Indicates the current focal length. represents the basic focal length, Indicates perspective.
6. A method for detecting a fault in a battery cell as claimed in claim 5, characterized in that: The specific steps in step S5 include: S5.1: Setting the alarm threshold and protection threshold , Will and Make comparisons; like , then DC internal resistance protection is generated; if , a DC internal resistance abnormality alarm is generated, and an improvement strategy is used to increase the detection cycle. The improvement strategy is expressed as: ,in, Indicates the next round of calculation cycle, Represents the shrinkage coefficient.
7. A method for detecting a fault in a battery cell as claimed in claim 5, characterized in that: The specific steps in step S5 also include: S5.2: Based on the acquired cell state feature quantity, a cell fault detection model is trained using a machine learning algorithm, and the performance of the cell fault detection model is evaluated using a cross-validation method. The formula is: ; Where P represents the failure probability, represents the machine learning model, U represents voltage, I represents current, and D represents the overall internal resistance array of the battery. t Indicates battery temperature; S5.3: Adjust the parameters of the cell fault detection model according to the test results, optimize the performance of the cell fault detection model, and introduce regularization to prevent overfitting; S5.4: Deploy the trained cell fault detection model to the actual system, receive the cell status data in real time and perform fault detection.
8. A battery cell fault detection system, used to implement a battery cell fault detection method according to any one of claims 1 to 7, characterized in that: include: Data acquisition module, trigger module, charging module, internal resistance analysis module, fault detection module, update module; The data acquisition module is used to collect data from the battery module; The trigger module is used to monitor whether the parameter data of the battery cell exceeds the normal range, and to start the entire detection process when it exceeds the normal range; The charging module is used to charge the battery cell to a specific power level or perform a charging cycle test before performing fault detection; The internal resistance analysis module is used to measure and analyze the internal resistance of the battery cell; The fault detection module is used to comprehensively analyze the information obtained from the data acquisition module, the trigger module and the internal resistance analysis module, and to evaluate and analyze the state of the battery cell using a pattern recognition algorithm and a preset standard or threshold, so as to determine whether the battery cell has a fault and the type and degree of the fault. Once an abnormality is found, the module will trigger an alarm or take preventive measures; The update module is responsible for updating system software and firmware.
9. A battery cell fault detection system as claimed in claim 8, characterized in that: The internal resistance analysis module includes: an internal resistance calculation unit and an internal resistance analysis unit; The internal resistance calculation unit is used to calculate the internal resistance value of the battery cell using the collected voltage and current data; The internal resistance analysis unit is used to analyze the internal resistance value to determine whether the health of the battery cell is good or not, and whether there is any fault.
10. A battery cell fault detection system as claimed in claim 9, characterized in that: The fault detection module includes: a model building unit and a judgment unit; The model building unit is used to build a fault detection model based on historical data and expert knowledge; The judgment unit is used to compare the real-time data with the fault detection model to judge whether the battery cell has a fault.
Citation Information
Patent Citations
Online fault detection method, device and fault detection circuit for hybrid circuit chips
CN108241114B
Fault detection equipment for electronic chip
CN217360196U
Cable joint partial discharge on-line monitoring device and method, and electric power system
CN116679177A
Storage battery health prediction method and system based on group end convergence method
CN117289162A
D-WPT system and method for obtaining energy of high-voltage transmission line
CN117914010A
