A new energy vehicle battery fault diagnosis system

Abstract

The present invention relates to the field of battery fault diagnosis technology, specifically to a new energy vehicle battery fault diagnosis system, the system comprising: a battery operating status monitoring module, an abnormal data detection module, a fault type identification module, and a fault battery pack positioning module. In the present invention, by calculating the correlation coefficient between battery operating parameters and monitoring the key operating data of the battery in real time, accurate monitoring and evaluation of the battery status is achieved, and by comparing the key features of various abnormal data with preset fault mode features, accurate identification of the fault type and clear marking of the fault location in the battery pack are achieved, and various problems such as battery performance degradation, imbalance and internal short circuit are discovered in time, simplifying the fault location and repair work of the maintenance team, reducing the impact on vehicle operation, effectively extending the service life of the battery, improving the reliability of the battery, optimizing the energy utilization efficiency, and optimizing the maintenance cost.

Description

Technical Field

[0001] The present invention relates to the technical field of battery fault diagnosis, and in particular to a new energy vehicle battery fault diagnosis system. Background Art

[0002] The field of battery fault diagnosis technology involves detecting and analyzing the performance and health status of new energy vehicle batteries, which is used to predict and prevent vehicle battery failures. By monitoring multiple key parameters such as battery voltage, current, temperature, and internal resistance, and using machine learning and artificial intelligence, it can detect signs of battery performance degradation or damage at an early stage, prevent serious battery failures, diagnose the causes of battery performance degradation, predict the remaining service life of the battery, and improve the safety and reliability of the vehicle. It aims to improve battery life and safety, optimize energy utilization efficiency, optimize maintenance costs, and formulate battery recycling strategies.

[0003] Among them, a new energy vehicle battery fault diagnosis system is used to identify and analyze potential problems of new energy vehicle batteries, including detecting and diagnosing faults or performance degradation in battery packs. By integrating sensing technology and data processing algorithms, it collects battery operation data in real time, analyzes the health of the battery, and detects and warns of potential fault problems at an early stage, including detecting battery performance degradation, battery imbalance, and internal battery short circuits. It provides data support for maintenance and replacement, helps vehicle manufacturers and maintenance teams optimize maintenance processes, ensures safe operation of vehicles, and provides car owners with a safe and efficient use experience.

[0004] Traditional automotive battery fault diagnosis technology does not respond promptly or diagnose accurately enough when dealing with sudden or initial battery failure problems. There are delays in data processing and fault warning under high load and extreme operating conditions. It is difficult to accurately diagnose in the early stages of battery performance degradation. The lack of in-depth analysis of internal changes in the battery leads to battery failures being ignored, which accelerates the degradation of battery performance without timely maintenance. It is difficult to prevent and handle failures in advance through real-time data analysis, and effective intervention can only be carried out after the battery has obviously failed, which increases the risk of safety accidents and maintenance costs, affecting the user experience and vehicle reliability. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a new energy vehicle battery fault diagnosis system.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme: A new energy vehicle battery fault diagnosis system comprises:

[0007] The battery operation status monitoring module analyzes the relationship between the battery load and various operation data under ideal conditions according to the specifications of the vehicle battery, and calculates the correlation coefficient between the load data and various operation data. Combined with sensors, it collects the operation data of multiple battery cells in real time, evaluates and records the operation status of the battery pack, and generates battery status monitoring data;

[0008] The abnormal data detection module uses the battery status monitoring data to construct and analyze the topological data structure in the battery parameter data, identify structural changes and persistent abnormal characteristics, set abnormal detection thresholds, and analyze the operating data to generate abnormal data point detection results;

[0009] The fault type identification module performs time-frequency decomposition on the abnormal signal and extracts multiple key features through the abnormal data point detection results, identifies the fault type by comparing the multiple key features with the preset fault mode features, and generates a fault type analysis result;

[0010] The faulty battery pack locating module identifies the battery pack of the battery unit involved in the fault according to the fault type analysis result, marks the number and location information of the faulty battery pack, and generates fault location information.

[0011] As a further solution of the present invention, the battery status monitoring data includes correlation coefficient calculation results, operation data monitoring records, and battery status evaluation information; the abnormal data point detection results include temperature abnormality indicators, voltage abnormality indicators, and charge and discharge rate abnormality indicators; the fault type analysis results are specifically overheating fault identification results, overcharging fault identification information, and aging fault identification records; the fault location information includes the fault battery number, the fault battery location, and the battery pack fault category data set.

[0012] As a further solution of the present invention, the step of obtaining the correlation coefficient is specifically:

[0013] Based on the specifications of the car battery, the relationship between the battery load and the operating temperature, voltage, and charge and discharge rate under ideal conditions is analyzed to obtain the ideal operating parameter data set;

[0014] Based on the ideal operating parameter data set, the current, voltage, and temperature data of multiple battery cells in actual operation are monitored and recorded in real time by sensors to generate a real-time battery monitoring data set;

[0015] Based on the real-time battery monitoring data set, by formula;

[0016] ;

[0017] Calculate the correlation coefficient between load data and various operating data;

[0018] in, represents the correlation coefficient, Indicates Observed values ​​of battery load data, Indicates Observed values ​​of battery operation data, represents the average value of the battery load data series, represents the average value of the battery operation data series, represents the total number of data points, Represents the index, is the summation symbol.

[0019] As a further solution of the present invention, the step of acquiring the battery status monitoring data is specifically as follows:

[0020] Based on the correlation coefficient, extract multiple standard specification parameters of the target vehicle battery and monitoring values ​​under actual operating conditions, including operating current, voltage, and temperature difference data, to obtain performance deviation extraction data;

[0021] Extract data based on the performance deviation, using the formula:

[0022] ;

[0023] Calculate the operating status score and evaluate the operating status of the battery pack;

[0024] in, Score the battery pack's health status, For the The current value of each battery cell, is the standard value of current, For the The voltage value of each battery cell, is the standard value of voltage, For the The temperature value of each battery cell, is the standard value of the temperature under the target operating state, is the weight factor, is the index of the battery cell;

[0025] Based on the battery pack operating status, battery status monitoring data is generated by evaluating and recording the operating status of multiple battery packs in real time.

[0026] As a further solution of the present invention, the step of obtaining the abnormality detection threshold is specifically as follows:

[0027] Utilizing the battery status monitoring data, extracting key performance parameter data of multiple batteries under operating conditions, including voltage, current, and temperature, and combining the interrelationships and dependencies between battery parameters to construct a topological model, reflect the performance and health of the battery in real time, and generate a topological data structure;

[0028] Based on the topological data structure, the interaction between battery parameters is analyzed, and structural changes, including abnormal fluctuations in current and voltage, and abnormal temperature rise, are identified to obtain structural changes and persistent abnormal characteristics;

[0029] The structural changes and persistent abnormal characteristics mentioned above use threshold analysis and anomaly detection techniques to identify structural changes that deviate significantly from normal performance;

[0030] Based on the structural changes and persistent abnormal characteristics, the formula:

[0031] ;

[0032] Calculate anomaly detection thresholds;

[0033] in, is the anomaly detection threshold, is the adjustment factor, is the weighting factor used to enhance the influence of the median, is the median of the persistence values, is the weighting factor used to adjust the effect of standard deviation, is the quantitative value of a single data feature, is the mean of the data eigenvalues, is the total number of data feature values.

[0034] As a further solution of the present invention, the step of obtaining the abnormal data point detection result is specifically:

[0035] Based on the anomaly detection threshold, a data point marking list is obtained by comparing a plurality of data points in the topological structure with the anomaly detection threshold;

[0036] Based on the data point tag list, by performing duration and repeatability analysis on a plurality of tagged data points, evaluating the duration and frequency of occurrence of data points exceeding a threshold, identifying the systematicity and repeatability of abnormal tags, and generating a data point verification record;

[0037] Verify the record based on the data point, by the formula:

[0038] ;

[0039] Calculate the anomaly index of each abnormal event and generate abnormal data point detection results;

[0040] in, is the abnormal index of the abnormal event, For any time The data point value of For Exceeding the threshold When part, , is the time point when the abnormality starts and ends, is the anomaly detection threshold, Represents a small time interval.

[0041] As a further solution of the present invention, the steps of obtaining the multiple key features are specifically as follows:

[0042] Based on the abnormal data point detection results, the formula is:

[0043] ;

[0044] Calculate frequency domain data;

[0045] in, is the frequency domain data, is the time series index, is the total number of data points, is the data point value in the time series, is the base of natural logarithms, is a plural unit, is the circumference of a circle, is the frequency index, It is the complex exponential part used to convert the time signal to the frequency domain in Fourier transform;

[0046] Based on the frequency domain data, the amplitude of each frequency component is analyzed, the energy value of each frequency component is compared, the energy peak frequency component is identified, and multiple key features are extracted.

[0047] As a further solution of the present invention, the steps of obtaining the fault type analysis result are specifically as follows:

[0048] Based on the multiple key features, extract key feature data of multiple known failure modes to generate a comparative feature data set;

[0049] Based on the comparative feature data set, using multiple key features extracted from abnormal signals, the formula:

[0050] ;

[0051] Calculate similarity scores;

[0052] in, represents the similarity score, is the first of the key feature vectors extracted from the abnormal signal elements, is the first key feature vector of the preset failure mode elements, is the dimension of the feature vector, is the index in the feature vector;

[0053] Based on the similarity score, by calculating the similarity between the target key feature and multiple known fault modes, the fault mode of the target abnormal signal is identified and a fault type analysis result is generated.

[0054] As a further solution of the present invention, the step of acquiring the fault location information is specifically as follows:

[0055] Based on the fault type analysis result, identifying the battery pack involving the faulty battery cell, and generating a faulty battery pack mark;

[0056] Based on the faulty battery pack mark, obtaining the location information of the battery cells involved in the fault in the battery pack, including the row number and column number of each battery cell involved in the fault, to provide spatial positioning data for fault location;

[0057] Based on the spatial positioning data, the formula:

[0058] ;

[0059] Calculate the linear index of the faulty battery cell to obtain the fault location information;

[0060] in, is the linear index of the fault location, is the row number where the battery cell is located, is the number of cells in each row, It is the column number of the battery cell.

[0061] Compared with the prior art, the advantages and positive effects of the present invention are:

[0062] In the present invention, by calculating the correlation coefficient between battery operating parameters and monitoring the key operating data of the battery in real time, accurate monitoring and evaluation of the battery status is achieved. By comparing the key features of various abnormal data with preset fault mode features, accurate identification of fault types and clear marking of fault locations in the battery pack are achieved, and various problems such as battery performance degradation, imbalance and internal short circuit are discovered in time, simplifying the maintenance team's fault location and repair work, reducing the impact on vehicle operation, effectively extending the battery life, improving battery reliability, optimizing energy utilization efficiency, and optimizing maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 is a system flow chart of the present invention;

[0064] Figure 2 The flowchart of calculating the correlation coefficient of the present invention is as follows;

[0065] Figure 3 The flowchart of obtaining battery status monitoring data of the present invention;

[0066] Figure 4 A flow chart of setting anomaly detection threshold value of the present invention;

[0067] Figure 5 This is a flow chart of obtaining abnormal data point detection results of the present invention;

[0068] Figure 6 A flowchart for extracting various key features of the present invention;

[0069] Figure 7 A flow chart of obtaining a fault type analysis result of the present invention;

[0070] Figure 8 This is a flow chart of identifying fault location information of the present invention. DETAILED DESCRIPTION

[0071] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0072] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0073] See also Figure 1 , a new energy vehicle battery fault diagnosis system includes:

[0074] The battery operation status monitoring module analyzes the relationship between the battery load and various operation data under ideal conditions according to the specifications of the vehicle battery, and calculates the correlation coefficient between the load data and various operation data. Combined with sensors, it collects the operation data of multiple battery cells in real time, evaluates and records the operation status of the battery pack, and generates battery status monitoring data;

[0075] The abnormal data detection module uses the battery status monitoring data to construct and analyze the topological data structure in the battery parameter data, identify structural changes and persistent abnormal characteristics, set abnormal detection thresholds, and analyze the operating data to generate abnormal data point detection results;

[0076] The fault type identification module performs time-frequency decomposition on the abnormal signal and extracts multiple key features through the abnormal data point detection results, identifies the fault type by comparing the multiple key features with the preset fault mode features, and generates a fault type analysis result;

[0077] The faulty battery pack locating module identifies the battery pack of the battery unit involved in the fault according to the fault type analysis result, marks the number and location information of the faulty battery pack, and generates fault location information.

[0078] The battery status monitoring data includes the correlation coefficient calculation results, operation data monitoring records, and battery status evaluation information. The abnormal data point detection results include temperature abnormality indicators, voltage abnormality indicators, and charge and discharge rate abnormality indicators. The fault type analysis results are specifically overheating fault identification results, overcharging fault identification information, and aging fault identification records. The fault location information includes the fault battery number, fault battery location, and battery pack fault category data set.

[0079] See also Figure 2 , the steps to obtain the correlation coefficient are as follows:

[0080] Based on the specifications of the car battery, the relationship between the battery load and the operating temperature, voltage, and charge and discharge rate under ideal conditions is analyzed to obtain the ideal operating parameter data set;

[0081] Based on the specification parameters of automotive batteries, the standard charge and discharge rate, maximum load capacity, recommended operating voltage and temperature range of each battery are extracted from manufacturer documents and battery performance test results. The parameters are used to establish a theoretical model to simulate the behavior of the battery under different operating conditions, such as the voltage and current changes of the battery at full load under standard ambient temperature, the temperature increase and the impact on battery life during charging and discharging, and the relationship curves and expressions between the parameters under ideal conditions are determined to obtain an ideal operating parameter data set, which contains all the key operating parameters of the battery under optimized operating conditions.

[0082] Based on the ideal operating parameter data set, the current, voltage, and temperature data of multiple battery cells in actual operation are monitored and recorded in real time through sensors to generate a real-time battery monitoring data set;

[0083] Based on the ideal operating parameter data set, the sensor array in the battery management system is configured. The sensors can capture the current, voltage and temperature data of the battery cells in actual operation in real time. The data collection frequency and accuracy of each sensor are set according to the predetermined parameters in the ideal data set to ensure the validity and accuracy of the data. Through the continuous collection of real-time data, a data recording system is used, including timestamps and corresponding operating conditions, to generate a real-time battery monitoring data set. The data set reflects the performance of the battery in different actual usage scenarios. By comparing the ideal parameter data set, any deviation from the predetermined operating conditions can be identified to ensure that the battery operates in an environment with maximum safety and efficiency.

[0084] Based on the real-time battery monitoring data set, through the formula;

[0085] ;

[0086] Calculate the correlation coefficient between load data and various operating data;

[0087] in, Represents the correlation coefficient, reflecting the strength of the linear relationship between battery load and operating data. Indicates Observed values ​​of battery load data, Indicates Observed values ​​of battery operation data, represents the average value of the battery load data series, represents the average value of the battery operation data series, represents the total number of data points, Represents an index, used to identify specific data points, and a summation symbol Indicates accumulation of subsequent expressions;

[0088] formula:

[0089] ;

[0090] Parameter meaning and acquisition method:

[0091] and : Battery load and operation data series, collected by real-time monitoring system;

[0092] , : No. 1 in the sequence Battery load and operation data of each data point is directly obtained from the real-time data acquisition system;

[0093] , : and The mean of a data series is obtained by calculating the arithmetic mean of these data points;

[0094] : The total number of data points, obtained by the number of data points recorded by the data acquisition system;

[0095] Calculation example:

[0096] Set the battery load and operating data to be: , , , , ;

[0097] Substitute into the formula to calculate:

[0098] ;

[0099] ;

[0100] ;

[0101] ;

[0102] Calculation results It shows that there is a high positive correlation between battery load and operating data. The numerical value shows that as the battery load changes, the operating data also shows a similar trend change.

[0103] See also Figure 3 , the specific steps for obtaining battery status monitoring data are:

[0104] Based on the correlation coefficient, multiple standard specification parameters of the target vehicle battery and the monitoring values ​​under the actual operating state are extracted, including operating current, voltage, and temperature difference data, to obtain performance deviation extraction data;

[0105] Based on the correlation coefficient, a data comparison and analysis is performed on each automotive battery from the ideal and actual operating parameter data sets, including selecting the current, voltage and temperature data corresponding to the ideal operating parameter data set from the real-time battery monitoring data set, and using statistical analysis software to calculate the correlation coefficient between multiple parameters, evaluate the interdependence and influence, and identify monitoring values ​​that exceed the standard specification parameters according to the pre-set standard deviation and mean value range, such as abnormal current increase or voltage drop. The deviation data will be extracted and marked as performance anomalies to obtain performance deviation extraction data, which reflects the difference between the performance of the battery under actual operating conditions and the ideal state.

[0106] Extract data based on performance deviation, using the formula:

[0107] ;

[0108] Calculate the operating status score and evaluate the operating status of the battery pack;

[0109] in, Score the operating status of the battery pack, reflecting the degree of deviation of the battery pack from the ideal state. For the The current value of each battery cell is obtained through real-time current monitoring data. is the standard value of current, provided by the battery manufacturer or set based on safety regulations. For the The voltage value of each battery cell is obtained through real-time voltage monitoring data. is the standard value of voltage, provided by the manufacturer or set based on specifications, For the The temperature value of each battery cell is obtained through real-time temperature monitoring data. is the standard value of the temperature under the target operating state, is a weighting factor used to adjust the impact of temperature deviation in the score, is the index of the battery cell;

[0110] formula:

[0111] ;

[0112] Parameter meaning and acquisition method:

[0113] : The current measurement value is obtained in real time through the current sensor in the battery management system;

[0114] : The standard value of current, usually provided by the battery manufacturer, used to set the standard current range for normal operation of the battery;

[0115] : The current voltage measurement value is obtained in real time through the voltage sensor in the battery management system;

[0116] : The standard value of voltage, provided by the battery manufacturer, indicating the standard voltage for normal operation of the battery;

[0117] : Current temperature measurement value, obtained in real time through the temperature sensor in the battery pack;

[0118] : The standard value of the temperature under the target operating state, which is set according to the battery manufacturer's specifications or empirical data, indicating the ideal temperature of the battery under the target operating state;

[0119] : The weight factor of temperature deviation is set based on past experience and the impact of battery sensitivity on temperature changes;

[0120] Calculation example:

[0121] Setting current data , current standard value , voltage data , voltage standard value , temperature data , temperature standard value , ;

[0122] Substitute the set value into the formula for calculation:

[0123] Calculate the current deviation ratio:

[0124] ;

[0125] ;

[0126] ;

[0127] Calculate the voltage deviation ratio:

[0128] ;

[0129] ;

[0130] ;

[0131] Calculate the temperature deviation ratio:

[0132] ;

[0133] ;

[0134] ;

[0135] Aggregate deviations and calculate scores :

[0136] ;

[0137] ;

[0138] Explanation of calculation results:

[0139] score It indicates that the battery condition is relatively good, close to the ideal state, and the total deviation is small. The calculation process is used to evaluate the overall health status of the battery pack and help guide maintenance and operation decisions.

[0140] Based on the operating status of the battery pack, the battery status monitoring data is generated by real-time evaluation and recording of the operating status of multiple battery packs;

[0141] Based on the operating status of the battery pack, the battery management system is configured to track and record the key performance parameters of each battery pack in real time, such as current, voltage, temperature, and the overall efficiency and health status of the battery pack. By deploying sensors and data acquisition equipment, the target parameters are monitored, and the data is automatically collected and stored in the central database at preset time intervals. Real-time data stream analysis and anomaly detection algorithms are implemented to identify performance degradation or potential failures. Through this continuous monitoring and real-time evaluation, battery status monitoring data is generated. The data set provides instant and accurate information for battery operation and maintenance, supporting decision-making and preventive maintenance measures.

[0142] See also Figure 4 , the specific steps for obtaining the anomaly detection threshold are:

[0143] Using battery status monitoring data, extract key performance parameter data of various batteries under operating conditions, including voltage, current, and temperature. Combined with the interrelationships and dependencies between battery parameters, build a topological model to reflect the performance and health of the battery in real time and generate a topological data structure.

[0144] Using battery status monitoring data, we extract data on a variety of key battery performance parameters, including voltage, current, and temperature. We use data analysis tools and algorithms to evaluate the relationships and dependencies between target parameters. We apply graph theory and network science methods to build a topological model that represents the performance and health status of the battery pack. The model dynamically represents the relationship between battery cells and between the cells and the overall performance. Through this topological model, the performance status and health status of each battery are reflected in real time, and a topological data structure is generated. The data structure records the structural information and node performance of the battery network, providing a basis for performance analysis and health monitoring.

[0145] Based on the topological data structure, the interaction between battery parameters is analyzed and structural changes are identified, including abnormal fluctuations in current and voltage, abnormal temperature rise, and structural changes and persistent abnormal characteristics are obtained;

[0146] Based on the topological data structure, network analysis tools and algorithms are used to analyze the interactions between battery parameters and monitor changes in the topological structure, including abnormal fluctuations in current and voltage, abnormal temperature increases and other structural changes. Threshold analysis and anomaly detection technology are used to identify structural changes that deviate significantly from normal performance. The changes indicate battery performance degradation or potential failures. Through continuous monitoring and data analysis, structural changes and persistent abnormal characteristics are obtained. The characteristics reflect key issues and health risks in the battery system, providing decision support for formulating maintenance strategies and optimizing battery use.

[0147] Based on structural changes and persistent abnormal characteristics, the formula:

[0148] ;

[0149] Calculate anomaly detection thresholds;

[0150] in, is the anomaly detection threshold, which is used to identify abnormal states in the data. is the adjustment factor, set according to the threshold sensitivity requirements of the system, is a weight factor used to enhance the influence of the median, which is set according to the data distribution characteristics. is the median of the persistence value, obtained from the quantitative statistical results of the data characteristics. is a weight factor used to adjust the impact of the standard deviation, set according to the variability of the data, is the quantitative value of a single data feature, obtained from the data structure feature set. is the mean of the data eigenvalues, obtained by calculating the average value of the data eigenvalue set. is the total number of data eigenvalues;

[0151] Detailed explanation of formula and calculation process

[0152] formula:

[0153] ;

[0154] Parameter meaning and acquisition method:

[0155] : Anomaly detection threshold, used to identify abnormal states in the data;

[0156] : Adjustment factor, which is set according to the threshold sensitivity requirements of the system and affects the final size of the threshold;

[0157] : Weight factor, which strengthens the influence of the median and is set based on the central tendency of the analytical data;

[0158] : The median of the persistence value, obtained by calculating the median of the data feature set;

[0159] : Weight factor, which adjusts the effect of standard deviation and is set based on the variability of the data;

[0160] : The quantitative value of a single data feature, obtained from the data structure feature set;

[0161] : The mean of the data eigenvalues, obtained by calculating the average value of the data eigenvalue set;

[0162] : The total number of data feature values, that is, the number of data structure features obtained by analysis;

[0163] Calculation example:

[0164] Setting the persistence value for , is 20, is 20, is 1.0, is 1.0, is 1.5, 5, calculate the anomaly detection threshold :

[0165] ;

[0166] ;

[0167] ;

[0168] ;

[0169] ;

[0170] ;

[0171] result Used to identify potential abnormal data points in the battery management system. The threshold is set based on the median of the topological features and the variability calculated from these features, ensuring that important topological changes can be sensitively detected and improving the system's monitoring and early warning capabilities.

[0172] See also Figure 5 , the specific steps for obtaining the abnormal data point detection results are:

[0173] Based on the anomaly detection threshold, a data point labeling list is obtained by comparing multiple data points in the topological structure with the anomaly detection threshold;

[0174] Based on the anomaly detection threshold, multiple data points such as current, voltage, temperature, etc. in the topological structure are compared with the set anomaly threshold. Each data point is scanned by data analysis software to check whether it exceeds the threshold. For data points that exceed the threshold, they are marked and added to the data point marking list. The list records potential abnormal data points and related attributes, including timestamps, parameter values, and location information, and screens key data points that need to be analyzed.

[0175] Based on the data point tag list, by performing duration and repeatability analysis on multiple tagged data points, evaluate the duration and frequency of data points exceeding the threshold, identify the systematicity and repeatability of abnormal tags, and generate data point verification records;

[0176] Based on the data point marking list, the duration and repeatability analysis is performed on each marked data point in the list. Statistical tools are used to calculate the specific duration of each abnormal data point exceeding the threshold and the frequency of occurrence during the monitoring period, which helps to evaluate whether the abnormal nature of each marked data point is systematic and repeatable. If a data point exceeds the threshold multiple times and lasts for a certain period of time, it is considered a systematic anomaly. The analysis results are summarized to generate a data point verification record, which explains the behavior pattern and potential impact of each abnormal data point, providing a basis for subsequent fault diagnosis and system maintenance.

[0177] Verify records based on data points, using the formula:

[0178] ;

[0179] Calculate the anomaly index of each abnormal event and generate abnormal data point detection results;

[0180] in, is the abnormality index of the abnormal event, which is used to quantify the severity and persistence of each abnormal event. For any time The data point value is obtained from the real-time monitoring data stream, representing the voltage, current or temperature of the battery. For Exceeding the threshold The part of the time indicates the degree of deviation. , is the time point when the abnormality starts and ends, is the anomaly detection threshold, which is used to determine whether a data point is abnormal. Represents a small time interval;

[0181] formula:

[0182] ;

[0183] Parameter meaning and acquisition method:

[0184] : At any time The data point value represents the voltage, current or temperature of the battery; it is obtained through the real-time monitoring system, which continuously monitors the working status of the battery and records the data value at each time point;

[0185] : Preset anomaly detection threshold, determined from system settings, usually set based on historical data analysis and battery performance standards; used to define when a data point is abnormal, usually set to a safe upper or lower limit for battery operating parameters;

[0186] , : The time when the anomaly starts and ends, by continuously monitoring the first and last data points in the data that exceed the threshold The time determination reflects the duration of the abnormal event;

[0187] :if Exceed , then the value is , otherwise it is 0, which reflects the degree to which the data point exceeds the threshold, and is calculated by real-time data comparison;

[0188] Calculation example:

[0189] Setting battery temperature threshold Spend, exist Seconds to The data for seconds are Degree, calculation :

[0190] ;

[0191] ;

[0192] Result interpretation:

[0193] Calculation results It shows that within 10 seconds, the cumulative deviation of the battery temperature above the preset threshold is 35. The value reflects the severity and persistence of the abnormal event. The calculation process is used to determine whether there is a significant thermal abnormality in the battery performance and to guide the implementation of maintenance and safety measures.

[0194] See also Figure 6 , the steps for obtaining multiple key features are as follows:

[0195] Based on the abnormal data point detection results, the formula is:

[0196] ;

[0197] Calculate frequency domain data;

[0198] in, It is frequency domain data, which means the signal is in frequency The frequency domain performance of To sum the symbols, calculate the total contribution of the frequency domain components, It is a time series index, representing the position in a discrete time series and used to index a specific time point. is the total number of data points, indicating the total number of data points for Fourier transform, used to normalize the scale of frequency domain conversion. is the data point value in the time series, representing the time point The signal value of is the base of natural logarithms, is a plural unit, is a mathematical constant representing the ratio of pi, is a frequency index used to represent a specific frequency component in the analysis. It is the complex exponential part used in Fourier transform to convert time signal to frequency domain, describing the behavior of signal in frequency domain;

[0199] formula:

[0200] ;

[0201] Parameter meaning and acquisition method:

[0202] : The total number of data points is obtained by choosing a small value suitable for simplifying the example; here it is set to 8, which is a FFT size commonly used in examples and teaching;

[0203] : Data point value in the time series, each point is assumed to be 1, simplifying the calculation and simulating equal amplitude signal sources;

[0204] : Time series index, from 0 to , used to iterate over each data point in the Fourier transform;

[0205] : Frequency index, set to 1, represents the analysis of the first non-zero frequency component;

[0206] : Complex exponential function, used for frequency domain transformation of signals, the frequency component of each data point is considered during calculation;

[0207] Calculation example:

[0208] set up , ,all ,calculate :

[0209] ;

[0210] The specific calculation process is as follows:

[0211] :

[0212] ;

[0213] :

[0214] ;

[0215] :

[0216] ;

[0217] :

[0218] ;

[0219] :

[0220] ;

[0221] :

[0222] ;

[0223] :

[0224] ;

[0225] :

[0226] ;

[0227] To add up the values:

[0228] ;

[0229] Formula calculation results:

[0230] Calculation results is 0, which means that at the frequency In the component, all sine and cosine components cancel each other out. The result reflects the phase and amplitude balance of the frequency component in the signal, indicating that the signal has no significant amplitude at this frequency. The calculation process is used to analyze the contribution and characteristics of each frequency component in the periodic signal.

[0231] Based on frequency domain data, by analyzing the amplitude of each frequency component, comparing the energy value of each frequency component, identifying the energy peak frequency component, and extracting multiple key features;

[0232] Based on frequency domain data, Fourier transform is applied to convert time series data into frequency domain representation, the amplitude of each frequency component is analyzed, and the energy value of each frequency component is calculated through spectrum analysis tools, including evaluating the energy distribution of the entire spectrum and identifying the energy peak frequency component. For each significant energy peak, the relevant amplitude and phase information is analyzed, and key features such as main frequency, energy density and spectral asymmetry are extracted from the frequency components. The features reflect the basic characteristics and potential dynamic changes of the signal, and a set of feature sets that describe the key attributes of the signal are obtained, which provides the necessary information for understanding the behavior and state of the signal source.

[0233] See also Figure 7 , the specific steps for obtaining the fault type analysis results are:

[0234] Based on multiple key features, extract key feature data of multiple known failure modes and generate comparative feature data sets;

[0235] Based on multiple key features, by accessing the maintenance database and fault records, key feature data matching the known fault modes are extracted, including frequency characteristics, energy peaks and multiple spectrum indicators for identifying specific fault modes from historical fault data. The data are compared with the key features obtained from the current analysis. Through the data analysis software, the similarity between the characteristics of the current signal and the characteristics of the known fault mode is checked to determine the corresponding fault types that may exist in the signal and generate a comparative feature data set. The data set includes detailed comparison results between the features extracted from the current operating state and the historical fault features, providing an evidence-based analysis basis for fault diagnosis.

[0236] Based on the contrast feature data set, using multiple key features extracted from abnormal signals, through the formula:

[0237] ;

[0238] Calculate similarity scores;

[0239] in, Represents the similarity score, reflecting the similarity between two sets of feature vectors. is the first of the key feature vectors extracted from the abnormal signal elements, obtained through time-frequency analysis, reflecting the specific frequency components of the abnormal signal. is the first key feature vector of the preset failure mode elements, obtained from the fault database, describing the frequency response of known fault types, is the dimension of the feature vector, which is set according to the actual number of features and is used to define the range of the summation operation. is the index in the feature vector;

[0240] formula:

[0241] ;

[0242] Parameter meaning and acquisition method:

[0243] : Similarity score, reflecting the similarity between two sets of feature vectors;

[0244] : Summation symbol, used to calculate the index From 1 to The sum of all items of ;

[0245] : The first key feature vector extracted from the abnormal signal elements, obtained through actual time-frequency analysis;

[0246] :The first key feature vector of the preset failure mode elements, obtained by accessing the fault database;

[0247] : The dimension of the eigenvector, which determines the range of the summation operation and is usually set to the length of the eigenvector;

[0248] Calculation example:

[0249] Set feature vector data , , , calculate the similarity between the target key features and the known failure modes:

[0250] ;

[0251] ;

[0252] ;

[0253] ;

[0254] ;

[0255] ;

[0256] result It is about 0.802, indicating that there is a high similarity between the two sets of feature vectors. The value shows that, according to the given feature vectors, the abnormal signal and the preset fault mode are statistically similar, indicating that there may be matching fault types;

[0257] Based on the similarity score, the similarity between the target key features and multiple known fault modes is calculated to identify the fault mode of the target abnormal signal and generate the fault type analysis result;

[0258] Based on the similarity score, statistical methods and machine learning algorithms are used to calculate the similarity between the target key features and multiple known fault modes, including comparing the feature data set currently extracted from the equipment or system with the feature set stored in the fault mode database. Through similarity scoring models such as cosine similarity or Euclidean distance, the degree of match between each fault mode and the target signal is evaluated, and the fault mode that best matches the target abnormal signal is identified. The results illustrate the similarity score of each fault mode and highlight the fault type with the highest matching score, generating fault type analysis results. The results provide clear guidance for the maintenance team, pointing out the possible causes of the fault and related repair measures, supporting the rapid diagnosis and effective resolution of the fault.

[0259] See also Figure 8 , the specific steps for obtaining fault location information are:

[0260] Based on the fault type analysis result, a battery pack involving a faulty battery cell is identified, and a faulty battery pack tag is generated;

[0261] Based on the fault type analysis results, the fault identification function in the battery management system is used to scan the operating data of the entire battery pack to identify fault modes that match the analysis results, including specific fault codes and descriptions obtained from the fault type analysis. By analyzing the performance data of each unit in the battery pack, it is identified which battery cells display abnormal behaviors similar to the identified fault modes, and the battery cells are marked as faulty battery cells. A faulty battery pack mark is generated, which includes the identification information and fault types of all battery cells involved in the fault, providing guidance for maintenance activities.

[0262] Based on the faulty battery pack mark, the location information of the battery cells involved in the fault in the battery pack is obtained, including the row number and column number of each faulty cell, to provide spatial positioning data for fault location;

[0263] Based on the faulty battery pack marking, the battery management system accesses the detailed configuration information of the battery pack and obtains the precise location information of the battery cells involved in the fault, including the row and column numbers of each battery cell marked as faulty in the battery pack. The system integrates the location information into a spatial positioning data set, which provides the technical team with the necessary spatial coordinates to help locate the faulty battery cells.

[0264] Based on spatial positioning data, through the formula:

[0265] ;

[0266] Calculate the linear index of the faulty battery cell to obtain the fault location information;

[0267] in, is the linear index of the fault location, used to identify the location of the faulty battery, is the row number where the battery cell is located, is the number of cells in each row, is the column number of the battery cell;

[0268] formula:

[0269] ;

[0270] Parameter meaning and acquisition method:

[0271] : The linear index of the fault location, calculated and used to specifically identify the location of the faulty battery;

[0272] The row number of the battery cell is obtained from the position data of the battery management system, which reflects the row position of the battery in the battery pack;

[0273] : The number of battery cells in each row of the battery pack, obtained from the layout configuration of the battery pack and used to calculate the position within the row;

[0274] : The column number of the battery cell, which is also obtained from the position data of the battery management system, indicating the specific position of the battery in the row;

[0275] Calculation example:

[0276] Set line number , represents the battery cells in the third row, the number of battery cells in each row , indicating that there are 10 battery cells in each row, and the column numbers , represents the battery cell located in the fourth column, and substitute the set value into the formula for calculation:

[0277] ;

[0278] ;

[0279] ;

[0280] Calculated fault location index It indicates that the faulty battery cell is located at the 34th position of the battery pack. The value reflects the specific fault location determined by the row and column numbers. The formula is used to locate the faulty unit and help the maintenance team handle the faulty battery.

[0281] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A new energy vehicle battery fault diagnosis system, characterized in that: The system comprises: The battery operation status monitoring module analyzes the relationship between the battery load and various operation data under ideal conditions according to the specifications of the vehicle battery, and calculates the correlation coefficient between the load data and various operation data. Combined with sensors, it collects the operation data of multiple battery cells in real time, evaluates and records the operation status of the battery pack, and generates battery status monitoring data; The abnormal data detection module uses the battery status monitoring data to construct and analyze the topological data structure in the battery parameter data, identify structural changes and persistent abnormal characteristics, set abnormal detection thresholds, and analyze the operating data to generate abnormal data point detection results; The steps for obtaining the abnormality detection threshold are specifically as follows: Utilizing the battery status monitoring data, extracting key performance parameter data of multiple batteries under operating conditions, including voltage, current, and temperature, and combining the interrelationships and dependencies between battery parameters to construct a topological model, reflect the performance and health of the battery in real time, and generate a topological data structure; Based on the topological data structure, the interaction between battery parameters is analyzed, and structural changes, including abnormal fluctuations in current and voltage, and abnormal temperature rise, are identified to obtain structural changes and persistent abnormal characteristics; The structural changes and persistent abnormal characteristics described above use threshold analysis and anomaly detection techniques to identify structural changes that deviate significantly from normal performance; Based on the structural changes and persistent abnormal characteristics, the formula: ； Calculate anomaly detection thresholds; in, is the anomaly detection threshold, is the adjustment factor, is the weighting factor used to enhance the influence of the median, is the median of the persistence values, is the weighting factor used to adjust the effect of standard deviation, is the quantitative value of a single data feature, is the mean of the data eigenvalues, is the total number of data eigenvalues; The fault type identification module performs time-frequency decomposition on the abnormal signal and extracts multiple key features through the abnormal data point detection results, identifies the fault type by comparing the multiple key features with the preset fault mode features, and generates a fault type analysis result; The faulty battery pack locating module identifies the battery pack of the battery unit involved in the fault according to the fault type analysis result, marks the number and location information of the faulty battery pack, and generates fault location information.

2. The new energy vehicle battery fault diagnosis system according to claim 1 is characterized in that: The battery status monitoring data includes the correlation coefficient calculation results, operation data monitoring records, and battery status evaluation information; the abnormal data point detection results include temperature abnormality indicators, voltage abnormality indicators, and charge and discharge rate abnormality indicators; the fault type analysis results are specifically overheating fault identification results, overcharging fault identification information, and aging fault identification records; the fault location information includes the fault battery number, fault battery location, and battery pack fault category data set.

3. The new energy vehicle battery fault diagnosis system according to claim 1, characterized in that: The steps for obtaining the correlation coefficient are specifically as follows: Based on the specifications of the car battery, the relationship between the battery load and the operating temperature, voltage, and charge and discharge rate under ideal conditions is analyzed to obtain the ideal operating parameter data set; Based on the ideal operating parameter data set, the current, voltage, and temperature data of multiple battery cells in actual operation are monitored and recorded in real time by sensors to generate a real-time battery monitoring data set; Based on the real-time battery monitoring data set, by formula; ； Calculate the correlation coefficient between load data and various operating data; in, represents the correlation coefficient, Indicates Observed values ​​of battery load data, Indicates Observed values ​​of battery operation data, represents the average value of the battery load data series, represents the average value of the battery operation data series, represents the total number of data points, Represents the index, is the summation symbol.

4. The new energy vehicle battery fault diagnosis system according to claim 3 is characterized in that: The steps for acquiring the battery status monitoring data are specifically as follows: Based on the correlation coefficient, extract multiple standard specification parameters of the target vehicle battery and monitoring values ​​under actual operating conditions, including operating current, voltage, and temperature difference data, to obtain performance deviation extraction data; Extract data based on the performance deviation, using the formula: ； Calculate the operating status score and evaluate the operating status of the battery pack; in, Score the battery pack's health status, For the The current value of each battery cell, is the standard value of current, For the The voltage value of each battery cell, is the standard value of voltage, For the The temperature value of each battery cell, is the standard value of the temperature under the target operating state, is the weight factor, is the index of the battery cell; Based on the battery pack operating status, battery status monitoring data is generated by evaluating and recording the operating status of multiple battery packs in real time.

5. The new energy vehicle battery fault diagnosis system according to claim 1, characterized in that: The steps for obtaining the abnormal data point detection result are specifically as follows: Based on the anomaly detection threshold, a data point label list is obtained by comparing a plurality of data points in the topological structure with the anomaly detection threshold; Based on the data point tag list, by performing duration and repeatability analysis on a plurality of tagged data points, evaluating the duration and frequency of occurrence of data points exceeding a threshold, identifying the systematicity and repeatability of abnormal tags, and generating a data point verification record; Verify the record based on the data point, by the formula: ； Calculate the anomaly index of each abnormal event and generate abnormal data point detection results; in, is the abnormal index of the abnormal event, For any time The data point value of For Exceeding the threshold When part, , is the time point when the abnormality starts and ends, is the anomaly detection threshold, Represents a small time interval.

6. The new energy vehicle battery fault diagnosis system according to claim 1, characterized in that: The steps for obtaining the multiple key features are specifically as follows: Based on the abnormal data point detection results, the formula is: ； Calculate frequency domain data; in, is the frequency domain data, is the time series index, is the total number of data points, is the data point value in the time series, is the base of natural logarithms, is a plural unit, is the circumference of a circle, is the frequency index, It is the complex exponential part used to convert the time signal to the frequency domain in Fourier transform; Based on the frequency domain data, the amplitude of each frequency component is analyzed, the energy value of each frequency component is compared, the energy peak frequency component is identified, and multiple key features are extracted.

7. The new energy vehicle battery fault diagnosis system according to claim 6, characterized in that: The steps for obtaining the fault type analysis result are specifically as follows: Based on the multiple key features, extract key feature data of multiple known failure modes to generate a comparative feature data set; Based on the comparative feature data set, using multiple key features extracted from abnormal signals, the formula: ； Calculate similarity scores; in, represents the similarity score, is the first of the key feature vectors extracted from the abnormal signal elements, is the first key feature vector of the preset failure mode elements, is the dimension of the feature vector, is the index in the feature vector; Based on the similarity score, by calculating the similarity between the target key feature and multiple known fault modes, the fault mode of the target abnormal signal is identified and a fault type analysis result is generated.

8. The new energy vehicle battery fault diagnosis system according to claim 1, characterized in that: The steps for obtaining the fault location information are specifically as follows: Based on the fault type analysis result, identifying the battery pack involving the faulty battery cell, and generating a faulty battery pack mark; Based on the faulty battery pack mark, obtaining the location information of the battery cells involved in the fault in the battery pack, including the row number and column number of each battery cell involved in the fault, to provide spatial positioning data for fault location; Based on the spatial positioning data, the formula: ； Calculate the linear index of the faulty battery cell to obtain the fault location information; in, is the linear index of the fault location, is the row number where the battery cell is located, is the number of cells in each row, It is the column number of the battery cell.

Images