A real-time online monitoring and early warning system and method for battery status

Abstract

The present invention discloses a real-time online monitoring and early warning system and method for a battery, which relates to the technical field of battery status monitoring. The system comprises a data processing module, a data analysis module, a comprehensive early warning analysis module, and an anomaly detection module; the data processing module collects various parameter data of the battery in real time through sensors, performs pre-processing, and transmits the data to the data analysis module; the data analysis module extracts features from the processed data and integrates them into a feature data set; the comprehensive early warning analysis module dynamically adjusts the threshold range of each parameter based on the feature data set of the data analysis module and integrates the changes of multiple parameters to calculate the comprehensive failure parameters of the battery and perform graded early warning reminders; the anomaly detection module is used to receive early warning signals and locate the specific location of the anomaly based on the early warning signal analysis, and finally generate a comprehensive failure report.

Description

Technical Field

[0001] The present invention relates to the technical field of battery status monitoring, and in particular to a real-time online status monitoring and early warning system and method for a battery. Background Art

[0002] In key areas such as electricity, communications, transportation, data centers, and new energy vehicles, batteries, as important energy storage devices, have a performance and stability that is directly related to the normal operation and safety of the entire system. However, with the increase of usage time, the performance of batteries will gradually decline, and various abnormal conditions will occur, such as voltage fluctuations, abnormal current, temperature increase, and increased internal resistance. If these abnormal phenomena are not discovered and accurately located in time, they may lead to system performance degradation, frequent failures, and even catastrophic consequences.

[0003] Traditional battery monitoring methods, such as manual inspections and periodic testing, have many limitations. Manual inspections rely on the inspector's experience and subjective judgment, making it difficult to capture subtle abnormal changes. Furthermore, the inspection cycle is long, making it easy to miss the optimal time to address a fault. While periodic testing can provide a more comprehensive performance evaluation, the testing process is cumbersome and requires system interruption. Furthermore, the test data often lags behind the actual operating status and cannot reflect the battery's health status in real time. Summary of the Invention

[0004] The purpose of the present invention is to provide a real-time online monitoring and early warning system and method for a battery to solve the problems raised in the prior art.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A battery real-time status online monitoring and early warning system, the system includes a data processing module, a data analysis module, a comprehensive early warning analysis module, and an anomaly detection module;

[0007] The data processing module collects various parameter data of the battery in real time through sensors and pre-processes and transmits them to the data analysis module, including a data acquisition unit and a data pre-processing unit;

[0008] The data acquisition unit collects various parameter data of the battery in real time through various sensors installed in the battery pack, and transmits the data to the data preprocessing unit through a communication protocol; the data preprocessing unit processes the received data, including cleaning, filtering, and denoising.

[0009] In the above technical solution, the data processing module captures the various parameter data of the battery pack in real time, and performs cleaning, filtering and denoising processing, effectively eliminating noise and interference, ensuring the accuracy and reliability of the data, and laying the foundation for subsequent data analysis.

[0010] Furthermore, the data analysis module extracts characteristic data related to battery performance from the preprocessed data and integrates them into a characteristic data set, including voltage characteristics, current characteristics, temperature characteristics, internal resistance characteristics, and capacity characteristics, ultimately forming a characteristic data set.

[0011] In the above technical solution, the data analysis module extracts characteristic data related to battery performance from the preprocessed data and integrates it into a characteristic data set, highlighting the key factors affecting the health status of the battery and providing data support for subsequent comprehensive early warning analysis.

[0012] Furthermore, the comprehensive warning analysis module dynamically adjusts the threshold range of each parameter based on the characteristic data set of the data analysis module and integrates the changes in multiple parameters to calculate the comprehensive failure parameters of the battery and provide graded warning reminders. It includes a threshold setting unit and a parameter fusion analysis unit; the threshold setting unit dynamically adjusts the threshold range by calculating the mean and standard deviation, takes the characteristic data set of the data analysis module as input, and calculates the mean and standard deviation of each parameter observation value in the characteristic data set respectively;

[0013] First calculate the mean μ of all observations i , according to the formula:

[0014]

[0015] Where N is the number of observations, x i,j represents the i-th observation value of the j-th observation;

[0016] Then calculate the standard deviation σ of each parameter i , according to the formula:

[0017]

[0018] Finally, based on the calculated mean and standard deviation, set the dynamic threshold, combined with the preset weight coefficient, dynamically adjust the threshold of each parameter according to the formula:

[0019] Y i =μ i +kσ i ;

[0020] L i =μ i -kσ i ;

[0021] Where k is the preset weight coefficient, Y i is the previous threshold, L iThe lower threshold is used to obtain the observed values ​​of each parameter in real time. When the real-time observed value of any parameter exceeds the set upper threshold or is lower than the lower threshold, it indicates that the parameter is abnormal. An early warning signal is output to remind and output the abnormal parameter.

[0022] The parameter fusion analysis unit comprehensively considers the changes of multiple parameters and calculates the comprehensive failure risk index H of the battery by weighted summation;

[0023] First calculate the risk value of each parameter according to the formula:

[0024]

[0025] where R i represents the risk value of the i-th parameter, P i Represents the current observation value;

[0026] Combine the risk values ​​of all parameters and use weighted summation to calculate the comprehensive failure risk index H. According to the output of the threshold setting module, adjust the weight of the abnormal parameter when allocating weights. Set a weight adjustment factor α for the abnormal parameter to increase the weight of the abnormal parameter, where α>1. At the same time, on the basis of ensuring that the sum of the weights of all parameters is 1, adjust the weight of the non-abnormal parameter accordingly. Set w i is the weight coefficient of the i-th parameter, I is the indicator function, when the i-th parameter is abnormal, I = 1, otherwise it is 0:

[0027]

[0028] Where y represents the total number of parameters, and m represents the number of abnormal parameters;

[0029] Combined with the weight distribution results of each parameter, the comprehensive failure risk index H is calculated according to the formula:

[0030]

[0031] Set risk thresholds and issue different levels of early warning reminders based on the size of the risk index.

[0032] In the above technical solution, the comprehensive early warning analysis module calculates the mean and standard deviation of each parameter and dynamically adjusts the threshold range based on preset weight coefficients, allowing the early warning system to more flexibly adapt to changes in battery performance. At the same time, a weighted summation method is used to calculate the comprehensive failure risk index, which is combined with risk thresholds for graded early warning, ensuring the timeliness and accuracy of early warnings.

[0033] Furthermore, the model design unit is used to construct a three-dimensional spatial model by physically measuring the battery pack, including the overall length, width, and height dimensions, the dimensions of the individual cells, the spacing between the individual cells, and the installation position and angle of the sensor on the battery pack. The individual cells are placed in the model according to the measured dimensions and positions and each cell is assigned a unique number. A virtual sensor node is placed in the model according to the actual installation position of the sensor on each cell, and a three-dimensional rectangular coordinate system and sensor parameters are set.

[0034] The abnormality locating unit receives the warning signal and abnormal parameter information from the comprehensive warning analysis module, identifies the abnormal sensor node by comparing the real-time sensor data with the data of the virtual sensor node in the model, and locates the corresponding single battery and its corresponding number in the three-dimensional space model;

[0035] The comprehensive output unit is used to generate a comprehensive failure report, which includes abnormal parameters, abnormal locations, comprehensive failure risk index H and its corresponding warning level.

[0036] In the above technical solution, after receiving the early warning signal, the anomaly detection module accurately locates the anomaly through the constructed three-dimensional spatial model. By using the physical measurement data and the precise information of the sensor installation position, it can quickly identify the abnormal sensor node in the three-dimensional space and calculate the specific position of the abnormal single cell in the battery pack, thereby improving the accuracy and efficiency of anomaly detection. Finally, the comprehensive failure report generated by the comprehensive output unit records in detail the abnormal parameters, abnormal location, comprehensive failure risk index and its early warning level, and provides comprehensive abnormal information and analysis results.

[0037] A method for online monitoring and early warning of real-time battery status, comprising the following steps:

[0038] Step S100 collects various parameter data of the battery in real time through various sensors installed in the battery pack and transmits the data to the data preprocessing unit through the communication protocol. The data preprocessing unit processes the received data, including cleaning, filtering, and denoising.

[0039] Step S200 extracts characteristic data related to battery performance from the preprocessed data and integrates them into a characteristic data set, including voltage characteristics, current characteristics, temperature characteristics, internal resistance characteristics, and capacity characteristics, ultimately forming a characteristic data set.

[0040] Step S300 dynamically adjusts the threshold range by calculating the mean and standard deviation, taking the feature data set of the data analysis module as input and calculating the mean and standard deviation of each parameter observation value in the feature data set;

[0041] First calculate the mean μ of all observations i , according to the formula:

[0042]

[0043] Where N is the number of observations, x i,j represents the i-th observation value of the j-th observation;

[0044] Then calculate the standard deviation σ of each parameter i , according to the formula:

[0045]

[0046] Finally, based on the calculated mean and standard deviation, set the dynamic threshold, combined with the preset weight coefficient, dynamically adjust the threshold of each parameter according to the formula:

[0047] Y i =μ i +kσ i ;

[0048] L i =μ i -kσ i ;

[0049] Where k is the preset weight coefficient, Y i is the previous threshold, L i The lower threshold is used to obtain the observed values ​​of each parameter in real time. When the real-time observed value of any parameter exceeds the set upper threshold or is lower than the lower threshold, it indicates that the parameter is abnormal. An early warning signal is output to remind and output the abnormal parameter.

[0050] Comprehensively consider the changes in multiple parameters and use weighted summation to calculate the comprehensive failure risk index H of the battery;

[0051] First calculate the risk value of each parameter according to the formula:

[0052]

[0053] where R i represents the risk value of the i-th parameter, P i Represents the current observation value;

[0054] Combine the risk values ​​of all parameters and use weighted summation to calculate the comprehensive failure risk index H. According to the output of the threshold setting module, adjust the weight of the abnormal parameter when allocating weights. Set a weight adjustment factor α for the abnormal parameter to increase the weight of the abnormal parameter, where α>1. At the same time, on the basis of ensuring that the sum of the weights of all parameters is 1, adjust the weight of the non-abnormal parameter accordingly. Set w iis the weight coefficient of the i-th parameter, I is the indicator function, when the i-th parameter is abnormal, I = 1, otherwise it is 0:

[0055]

[0056] Where y represents the total number of parameters, and m represents the number of abnormal parameters;

[0057] Combined with the weight distribution results of each parameter, the comprehensive failure risk index H is calculated according to the formula:

[0058]

[0059] Set risk thresholds and issue different levels of early warning reminders based on the size of the risk index.

[0060] Step S400 is used to construct a three-dimensional spatial model. By physically measuring the battery pack, including the overall length, width, and height dimensions, the dimensions of the individual cells, the spacing between the individual cells, and the installation position and angle of the sensor on the battery pack, the individual cells are placed in the model according to the measured dimensions and positions and each individual cell is assigned a unique number. On each individual cell, according to the actual installation position of the sensor, a virtual sensor node is placed in the model and a three-dimensional rectangular coordinate system and sensor parameters are set. The comprehensive early warning analysis module receives early warning signals and abnormal parameter information, identifies abnormal sensor nodes by comparing real-time sensor data with the data of the virtual sensor nodes in the model, and locates the corresponding individual cells and their corresponding numbers in the three-dimensional spatial model. Finally, a comprehensive failure report is generated, which includes abnormal parameters, abnormal locations, a comprehensive failure risk index H, and its corresponding early warning level.

[0061] Compared with the prior art, the present invention has the following beneficial effects:

[0062] By adopting dynamic threshold adjustment and parameter fusion analysis technology, the present invention enables the system to flexibly adjust the warning strategy according to the actual performance changes of the battery. Compared with the fixed threshold method, this intelligent warning mechanism can better adapt to the changes in battery performance under different circumstances, and improve the accuracy and reliability of the warning.

[0063] The present invention not only focuses on the abnormality of a single parameter, but also comprehensively evaluates the overall performance of the battery by calculating a comprehensive failure risk index. This comprehensive evaluation method helps to more accurately judge the health status of the battery and provide more comprehensive information support for operation and maintenance decisions.

[0064] By constructing a three-dimensional spatial model, the present invention can locate abnormal locations, which greatly shortens the troubleshooting time and improves operation and maintenance efficiency; at the same time, the rapid response early warning mechanism also provides strong support for timely repair measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 This is a system module diagram of a real-time online monitoring and early warning system for a battery according to the present invention;

[0066] Figure 2 The present invention is a method flow chart of a method for online monitoring and early warning of the real-time status of a battery. DETAILED DESCRIPTION

[0067] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0068] Example: Figure 1-Figure 2 As shown, the present invention provides a technical solution.

[0069] A battery real-time status online monitoring and early warning system, the system includes a data processing module, a data analysis module, a comprehensive early warning analysis module, and an anomaly detection module;

[0070] The data processing module collects various parameter data of the battery in real time through sensors and pre-processes and transmits them to the data analysis module, including a data acquisition unit and a data pre-processing unit;

[0071] The data acquisition unit collects various parameter data of the battery in real time through various sensors installed in the battery pack, and transmits the data to the data preprocessing unit through a communication protocol; the data preprocessing unit processes the received data, including cleaning, filtering, and denoising.

[0072] In the above technical solution, the data processing module captures the various parameter data of the battery pack in real time, and performs cleaning, filtering and denoising processing, effectively eliminating noise and interference, ensuring the accuracy and reliability of the data, and laying the foundation for subsequent data analysis.

[0073] The data analysis module extracts characteristic data related to battery performance from the preprocessed data and integrates them into a characteristic data set, including voltage characteristics, current characteristics, temperature characteristics, internal resistance characteristics, and capacity characteristics, ultimately forming a characteristic data set.

[0074] In the above technical solution, the data analysis module extracts characteristic data related to battery performance from the preprocessed data and integrates it into a characteristic data set, highlighting the key factors affecting the health status of the battery and providing data support for subsequent comprehensive early warning analysis.

[0075] The comprehensive warning analysis module dynamically adjusts the threshold range of each parameter based on the characteristic data set of the data analysis module and integrates the changes in multiple parameters to calculate the comprehensive failure parameters of the battery and provide graded warning reminders. It includes a threshold setting unit and a parameter fusion analysis unit. The threshold setting unit dynamically adjusts the threshold range by calculating the mean and standard deviation. It uses the characteristic data set of the data analysis module as input and calculates the mean and standard deviation of each parameter observation value in the characteristic data set.

[0076] First calculate the mean μ of all observations i , according to the formula:

[0077]

[0078] Where N is the number of observations, x i,j represents the i-th observation value of the j-th observation;

[0079] Then calculate the standard deviation σ of each parameter i , according to the formula:

[0080]

[0081] Finally, based on the calculated mean and standard deviation, set the dynamic threshold, combined with the preset weight coefficient, dynamically adjust the threshold of each parameter according to the formula:

[0082] Y i =μ i +kσ i ;

[0083] L i =μ i -kσ i ;

[0084] Where k is the preset weight coefficient, Y i is the previous threshold, L i The lower threshold is used to obtain the observed values ​​of each parameter in real time. When the real-time observed value of any parameter exceeds the set upper threshold or is lower than the lower threshold, it indicates that the parameter is abnormal. An early warning signal is output to remind and output the abnormal parameter.

[0085] Comprehensively consider the changes in multiple parameters and use weighted summation to calculate the comprehensive failure risk index H of the battery;

[0086] First calculate the risk value of each parameter according to the formula:

[0087]

[0088] where R irepresents the risk value of the i-th parameter, P i Represents the current observation value;

[0089] Combine the risk values ​​of all parameters and use weighted summation to calculate the comprehensive failure risk index H. According to the output of the threshold setting module, adjust the weight of the abnormal parameter when allocating weights. Set a weight adjustment factor α for the abnormal parameter to increase the weight of the abnormal parameter, where α>1. At the same time, on the basis of ensuring that the sum of the weights of all parameters is 1, adjust the weight of the non-abnormal parameter accordingly. Set w i is the weight coefficient of the i-th parameter, I is the indicator function, when the i-th parameter is abnormal, I = 1, otherwise it is 0:

[0090]

[0091] Where y represents the total number of parameters, and m represents the number of abnormal parameters;

[0092] Combined with the weight distribution results of each parameter, the comprehensive failure risk index H is calculated according to the formula:

[0093]

[0094] Set the risk threshold H t , the comprehensive failure risk index H and the risk threshold H t Compare, if H>H t , the early warning mechanism is triggered, and different levels of early warning reminders are issued according to the size of the risk index.

[0095] In the above technical solution, the comprehensive early warning analysis module calculates the mean and standard deviation of each parameter and dynamically adjusts the threshold range based on preset weight coefficients, allowing the early warning system to more flexibly adapt to changes in battery performance. At the same time, a weighted summation method is used to calculate the comprehensive failure risk index, which is combined with risk thresholds for graded early warning, ensuring the timeliness and accuracy of early warnings.

[0096] The model design unit is used to construct a three-dimensional spatial model by physically measuring the battery pack, including the overall length, width, and height dimensions, the dimensions of the individual cells, the spacing between the individual cells, and the installation position and angle of the sensor on the battery pack. The individual cells are placed in the model according to the measured dimensions and positions and each cell is assigned a unique number. A virtual sensor node is placed in the model based on the actual installation position of the sensor on each cell and a three-dimensional rectangular coordinate system and sensor parameters are set.

[0097] The abnormality locating unit receives the warning signal and abnormal parameter information from the comprehensive warning analysis module, identifies the abnormal sensor node by comparing the real-time sensor data with the data of the virtual sensor node in the model, and locates the corresponding single battery and its corresponding number in the three-dimensional space model;

[0098] The comprehensive output unit is used to generate a comprehensive failure report, which includes abnormal parameters, abnormal locations, comprehensive failure risk index H and its corresponding warning level.

[0099] In the above technical solution, after receiving the early warning signal, the anomaly detection module accurately locates the anomaly through the constructed three-dimensional spatial model. By using the physical measurement data and the precise information of the sensor installation position, it can quickly identify the abnormal sensor node in the three-dimensional space and calculate the specific position of the abnormal single cell in the battery pack, thereby improving the accuracy and efficiency of anomaly detection. Finally, the comprehensive failure report generated by the comprehensive output unit records in detail the abnormal parameters, abnormal location, comprehensive failure risk index and its early warning level, and provides comprehensive abnormal information and analysis results.

[0100] A method for online monitoring and early warning of real-time battery status, comprising the following steps:

[0101] Step S100 collects various parameter data of the battery in real time through various sensors installed in the battery pack and transmits the data to the data preprocessing unit through the communication protocol. The data preprocessing unit processes the received data, including cleaning, filtering, and denoising.

[0102] Step S200 extracts characteristic data related to battery performance from the preprocessed data and integrates them into a characteristic data set, including voltage characteristics, current characteristics, temperature characteristics, internal resistance characteristics, and capacity characteristics, ultimately forming a characteristic data set.

[0103] Step S300 dynamically adjusts the threshold range by calculating the mean and standard deviation, taking the feature data set of the data analysis module as input and calculating the mean and standard deviation of each parameter observation value in the feature data set;

[0104] First calculate the mean μ of all observations i , according to the formula:

[0105]

[0106] Where N is the number of observations, x i,j represents the i-th observation value of the j-th observation;

[0107] Then calculate the standard deviation σ of each parameter i , according to the formula:

[0108]

[0109] Finally, based on the calculated mean and standard deviation, set the dynamic threshold, combined with the preset weight coefficient, dynamically adjust the threshold of each parameter according to the formula:

[0110] Y i =μ i +kσ i ;

[0111] L i =μ i -kσ i ;

[0112] Where k is the preset weight coefficient, Y i is the previous threshold, L i The lower threshold is used to obtain the observed values ​​of each parameter in real time. When the real-time observed value of any parameter exceeds the set upper threshold or is lower than the lower threshold, it indicates that the parameter is abnormal. An early warning signal is output to remind and output the abnormal parameter.

[0113] Comprehensively consider the changes in multiple parameters and use weighted summation to calculate the comprehensive failure risk index H of the battery;

[0114] First calculate the risk value of each parameter according to the formula:

[0115]

[0116] where R i represents the risk value of the i-th parameter, P i Represents the current observation value;

[0117] Combine the risk values ​​of all parameters and use weighted summation to calculate the comprehensive failure risk index H. According to the output of the threshold setting module, adjust the weight of the abnormal parameter when allocating weights. Set a weight adjustment factor α for the abnormal parameter to increase the weight of the abnormal parameter, where α>1. At the same time, on the basis of ensuring that the sum of the weights of all parameters is 1, adjust the weight of the non-abnormal parameter accordingly. Set w i is the weight coefficient of the i-th parameter, I is the indicator function, when the i-th parameter is abnormal, I = 1, otherwise it is 0:

[0118]

[0119] Where y represents the total number of parameters, and m represents the number of abnormal parameters;

[0120] Combined with the weight distribution results of each parameter, the comprehensive failure risk index H is calculated according to the formula:

[0121]

[0122] Set risk thresholds and issue different levels of early warning alerts based on the size of the risk index.

[0123] Step S400 is used to construct a three-dimensional spatial model. By physically measuring the battery pack, including the overall length, width, and height dimensions, the dimensions of the individual cells, the spacing between the individual cells, and the installation position and angle of the sensor on the battery pack, the individual cells are placed in the model according to the measured dimensions and positions and each individual cell is assigned a unique number. On each individual cell, according to the actual installation position of the sensor, a virtual sensor node is placed in the model and a three-dimensional rectangular coordinate system and sensor parameters are set. The comprehensive early warning analysis module receives early warning signals and abnormal parameter information, identifies abnormal sensor nodes by comparing real-time sensor data with the data of the virtual sensor nodes in the model, and locates the corresponding individual cells and their corresponding numbers in the three-dimensional spatial model. Finally, a comprehensive failure report is generated, which includes abnormal parameters, abnormal locations, a comprehensive failure risk index H, and its corresponding early warning level.

[0124] In this example, a battery pack is assumed to consist of 10 cells. Each cell is equipped with a sensor to monitor its voltage, current, temperature, internal resistance, and capacity. At a certain point in time, a set of data is collected and preprocessed. The following is a portion of the preprocessed data:

[0125] Number of observations Voltage (V) Temperature (℃) 1 12.5 25.0 2 12.4 25.1 ... ... ... 10 12.3 25.2

[0126] 1. Calculate the mean and standard deviation:

[0127] Voltage: μ = 1 / 10 (12.5 + 12.4 + ... + 12.3) = 12.42σ = 0.05;

[0128] Temperature: μ = 25.08σ = 0.06;

[0129] 2. Set the dynamic threshold and weight coefficient K to 0.5, then:

[0130] Voltage upper threshold Y V =μ+kσ=12.42+0.5×0.05=12.445;

[0131] Voltage lower threshold L V = = 12.42-0.5×0.05 = 12.405;

[0132] Temperature upper threshold Y T = = 25.08 + 0.5 × 0.06 = 25.11;

[0133] The lower limit of temperature L T==25.08 - 0.5×0.06 == 25.05;

[0134] 3. Calculate the risk value:

[0135] The current temperature of a single battery is observed to be 25.15°C, which is within the normal range; the voltage is 12.38V, lower than the lower bound threshold, indicating that the voltage parameter is abnormal. An early warning signal is output for reminder and the voltage parameter is output.

[0136] Calculate the risk value: R v == |12.38 - 12.42| / 0.05 == 0.8;

[0137] R t == |25.1 - 25.08| / 0.06 == 0.33;

[0138] 4. Adjust the weights and calculate the comprehensive failure index H:

[0139] When allocating weights, adjust the weights of abnormal parameters; set a weight adjustment factor α of 1.5 for abnormal parameters, w v == 1.5 / 2 == 0.75, w T == 0.15;

[0140] The comprehensive failure risk index H == 0.75 * 0.8 + 0.15 * 0.33 == 0.65;

[0141] 5. Set the risk threshold for hierarchical early warning:

[0142] Low risk level: H ≤ 0.3, indicating that the battery status is good and no special attention is required;

[0143] Medium - low risk level: 0.3 < H ≤ 0.5, indicating that the battery performance has slightly declined but is still within the acceptable range. It is recommended to conduct regular inspections;

[0144] Medium - high risk level: 0.5 < H ≤ 0.7, indicating that the battery has obvious performance degradation or potential failure risk, and professional personnel need to be arranged immediately for detailed inspection and maintenance;

[0145] High risk level: H > 0.7, indicating that the battery is in a serious failure state and urgent measures need to be taken immediately, such as replacing the battery or starting the backup power supply;

[0146] The calculated comprehensive risk index H == 0.65, which belongs to the medium - high - level warning. The system outputs a prompt "The battery has obvious performance degradation or potential failure risk, and professional personnel need to be arranged immediately for detailed inspection and maintenance";

[0147] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and range of equivalents of the claims are intended to be included therein. Any reference sign in a claim should not be construed as limiting the claim to which it relates.

Claims

1. A battery real-time status online monitoring and early warning system, characterized by: The system includes a data processing module, a data analysis module, a comprehensive early warning analysis module, and an anomaly detection module; The data processing module collects various parameter data of the battery in real time through sensors and pre-processes and transmits them to the data analysis module; The data analysis module extracts features from the processed data and integrates them into a feature data set; The comprehensive early warning analysis module dynamically adjusts the threshold range of each parameter based on the characteristic data set of the data analysis module and integrates the changes of multiple parameters to calculate the comprehensive failure parameters of the battery and provide graded early warning reminders; The comprehensive early warning analysis module includes a threshold setting unit and a parameter fusion analysis unit; the threshold setting unit dynamically adjusts the threshold range by calculating the mean and standard deviation, takes the feature data set of the data analysis module as input, and calculates the mean and standard deviation of each parameter observation value in the feature data set; First calculate the mean μ of all observations i , according to the formula: ； Where N represents the number of observations, represents the i-th observation value of the j-th observation; Then calculate the standard deviation σ of each parameter i , according to the formula: ； Finally, based on the calculated mean and standard deviation, set the dynamic threshold, combined with the preset weight coefficient, dynamically adjust the threshold of each parameter according to the formula: ； ； Where k is the preset weight coefficient, Y i is the previous threshold, L i The lower threshold is used to obtain the observed values ​​of various parameters in real time. When the real-time observed value of any parameter exceeds the set upper threshold or is lower than the lower threshold, it indicates that the parameter is abnormal. An early warning signal is output to remind and output the abnormal parameter. The parameter fusion analysis unit comprehensively considers the changes of multiple parameters and calculates the comprehensive failure risk index H of the battery by weighted summation. First, the risk value of each parameter is calculated according to the formula: ； where R i represents the risk value of the i-th parameter, P i Represents the current observation value; Combine the risk values ​​of all parameters and use weighted summation to calculate the comprehensive failure risk index H. Adjust the weight of abnormal parameters when allocating weights based on the output of the threshold setting module. For abnormal parameters, a weight adjustment factor α is set to increase the weight of abnormal parameters, where α>1. At the same time, on the basis of ensuring that the sum of the weights of all parameters is 1, the weights of non-abnormal parameters are adjusted accordingly; set w i is the weight coefficient of the i-th parameter, I is the indicator function, when the i-th parameter is abnormal, I=1, otherwise it is 0: where R i represents the risk value of the i-th parameter, P i Represents the current observation value; Combine the risk values ​​of all parameters and use weighted summation to calculate the comprehensive failure risk index H. Adjust the weight of abnormal parameters when allocating weights based on the output of the threshold setting module. For abnormal parameters, a weight adjustment factor α is set to increase the weight of abnormal parameters, where α>1. At the same time, on the basis of ensuring that the sum of the weights of all parameters is 1, the weights of non-abnormal parameters are adjusted accordingly; set w i is the weight coefficient of the i-th parameter, I is the indicator function, when the i-th parameter is abnormal, I=1, otherwise it is 0: ； Where y represents the total number of parameters, and m represents the number of abnormal parameters; Combined with the weight distribution results of each parameter, the comprehensive failure risk index H is calculated according to the formula: ； Set risk thresholds and issue different levels of early warning reminders based on the size of the risk index; The anomaly detection module is used to receive warning signals and locate the specific location of the anomaly based on the warning signals, and finally generate a comprehensive failure report; The anomaly detection module includes a model design unit and an anomaly location unit; The model design unit is used to construct a three-dimensional spatial model by physically measuring the battery pack, including the overall length, width, and height dimensions, the dimensions of the individual cells, the spacing between the individual cells, and the installation position and angle of the sensor on the battery pack. The individual cells are placed in the model according to the measured dimensions and positions and each cell is assigned a unique number. A virtual sensor node is placed in the model based on the actual installation position of the sensor on each cell and a three-dimensional rectangular coordinate system and sensor parameters are set. The abnormality locating unit receives the warning signal and abnormal parameter information from the comprehensive warning analysis module, identifies the abnormal sensor node by comparing the real-time sensor data with the data of the virtual sensor node in the model, and locates the corresponding single battery and its corresponding number in the three-dimensional space model; The comprehensive output unit is used to generate a comprehensive failure report, which includes abnormal parameters, abnormal locations, comprehensive failure risk index H and its corresponding warning level.

2. The battery real-time status online monitoring and early warning system according to claim 1 is characterized by: The data processing module includes a data acquisition unit and a data preprocessing unit; The data acquisition unit collects various parameter data of the battery in real time through various sensors installed in the battery pack, and transmits them to the data preprocessing unit through the communication protocol; The data preprocessing unit processes the received data, including cleaning, filtering, and denoising.

3. The battery real-time status online monitoring and early warning system according to claim 1 is characterized by: The data analysis module extracts characteristic data related to battery performance from the preprocessed data and integrates them into a characteristic data set, including voltage characteristics, current characteristics, temperature characteristics, internal resistance characteristics, and capacity characteristics, ultimately forming a characteristic data set.

4. A method for online monitoring and early warning of real-time battery status, characterized by: The method comprises the following steps: Step S100: collecting various parameter data of the battery in real time through sensors, pre-processing and transmitting to the data analysis module; Step S200: extracting features from the processed data and integrating them into a feature data set; Step S300: Based on the characteristic data set of the data analysis module, dynamically adjust the threshold range of each parameter and integrate the changes of multiple parameters to calculate the comprehensive failure parameters of the battery and issue a graded warning reminder; Step S400: receiving an early warning signal and locating the specific location of the abnormality based on the early warning signal analysis, and finally generating a comprehensive failure report; Step S300 dynamically adjusts the threshold range by calculating the mean and standard deviation, taking the feature data set of the data analysis module as input and calculating the mean and standard deviation of each parameter observation value in the feature data set; First calculate the mean μ of all observations i , according to the formula: ； Where N represents the number of observations, represents the i-th observation value of the j-th observation; Then calculate the standard deviation σ of each parameter i , according to the formula: ； Finally, based on the calculated mean and standard deviation, set the dynamic threshold, combined with the preset weight coefficient, dynamically adjust the threshold of each parameter according to the formula: ； ； Where k is the preset weight coefficient, Y i is the previous threshold, L i The lower threshold is used to obtain the observed values ​​of each parameter in real time. When the real-time observed value of any parameter exceeds the set upper threshold or is lower than the lower threshold, it indicates that the parameter is abnormal. An early warning signal is output to remind and output the abnormal parameter. Based on the changes in multiple parameters, the comprehensive failure risk index H of the battery is calculated using weighted summation. First, the risk value of each parameter is calculated according to the formula: ； where R i represents the risk value of the i-th parameter, P i Represents the current observation value; Combine the risk values ​​of all parameters and use weighted summation to calculate the comprehensive failure risk index H. According to the output of the threshold setting module, adjust the weight of the abnormal parameter when allocating weights. Set a weight adjustment factor α for the abnormal parameter to increase the weight of the abnormal parameter, where α>1. At the same time, on the basis of ensuring that the sum of the weights of all parameters is 1, adjust the weight of the non-abnormal parameter accordingly. Set w i is the weight coefficient of the i-th parameter, I is the indicator function, when the i-th parameter is abnormal, I=1, otherwise it is 0: ； Where y represents the total number of parameters, and m represents the number of abnormal parameters; Combined with the weight distribution results of each parameter, the comprehensive failure risk index H is calculated according to the formula: ； Set risk thresholds and issue different levels of early warning reminders based on the size of the risk index; Step S400 is used to construct a three-dimensional spatial model. This involves physically measuring the battery pack, including the overall length, width, and height dimensions, the dimensions of the individual cells, the spacing between the cells, and the sensor installation positions and angles on the battery pack. The cells are then placed in the model according to the measured dimensions and positions, and each cell is assigned a unique number. Virtual sensor nodes are placed in the model based on the actual sensor installation positions on each cell, and a three-dimensional rectangular coordinate system and sensor parameters are set. The comprehensive early warning analysis module receives early warning signals and abnormal parameter information, identifies abnormal sensor nodes by comparing real-time sensor data with the data of virtual sensor nodes in the model, and locates the corresponding single battery and its corresponding number in the three-dimensional space model; finally, a comprehensive failure report is generated, which includes abnormal parameters, abnormal location, comprehensive failure risk index H and its corresponding early warning level.

5. The method for online monitoring and early warning of real-time battery status according to claim 4, characterized in that: The step S100 collects various parameter data of the battery in real time through various sensors installed in the battery pack, and transmits the data to the data preprocessing unit through the communication protocol. The data preprocessing unit processes the received data, including cleaning, filtering, and denoising. The step S200 extracts characteristic data related to battery performance from the preprocessed data and integrates it into a characteristic data set, including voltage characteristics, current characteristics, temperature characteristics, internal resistance characteristics, and capacity characteristics, ultimately forming a characteristic data set.

Images