An intelligent detection system for battery failures in new energy vehicles

Abstract

The present invention relates to the field of battery detection technology, specifically to an intelligent detection system for battery faults of new energy vehicles, which includes a voltage change analysis module, an internal resistance micro-analysis module, a current response monitoring module and a fault diagnosis integrated module. In the present invention, by refining voltage monitoring and internal resistance analysis, the ability to identify battery performance degradation and early faults is significantly improved, anomalies are identified through advanced data processing technology, and detailed fluctuation analysis makes fault warning more accurate, reducing the risk of false alarms and missed alarms. The introduction of fractal theory to analyze internal resistance data can more accurately capture the microscopic changes of battery aging and achieve early detection of aging trends. Combined with pulse current testing and current response data analysis, the evaluation of battery performance stability is further enhanced, ensuring the comprehensiveness and depth of the detection results, and greatly improving the accuracy of battery fault detection and the efficiency of determining the fault type.

Description

Technical Field

[0001] The present invention relates to the technical field of battery detection, and in particular to an intelligent detection system for battery failures of new energy vehicles. Background Art

[0002] The field of battery testing technology focuses on monitoring and evaluating battery performance and health status, usually including real-time monitoring of battery status, battery life prediction, fault diagnosis, and performance optimization. Battery testing technology is essential to ensure safe operation of batteries and extend battery life, especially in electric vehicles and large energy storage systems. With the development of new energy technologies, battery testing technology is also constantly improving to adapt to various battery types, such as lithium-ion batteries, nickel-metal hydride batteries, etc. It can not only monitor battery voltage, current, temperature and other parameters in real time, but also predict the future performance of the battery through data analysis, thereby identifying potential fault problems in advance.

[0003] Among them, the new energy vehicle battery fault intelligent detection system is an intelligent system used in new energy vehicles to specifically detect whether the battery has faults or performance degradation. Its main purpose is to monitor the working status and health of the battery in real time through advanced sensing technology and data analysis algorithms, and to promptly detect and diagnose possible battery failures. It is crucial to ensure the safety, reliability and efficiency of electric vehicles, especially in the context of the popularization of electric vehicles and the increasing maturity of battery technology. An effective battery detection system can significantly improve the vehicle's operation and maintenance efficiency and reduce the risk of unexpected failures.

[0004] Although existing technologies are capable of basic monitoring and prediction of battery status, they are still insufficient in early identification of faults and accurate determination of types. Existing systems rely on relatively simple parameter monitoring and general data analysis models, which limits the depth and timeliness of fault detection. Especially when dealing with non-standardized faults and subtle performance degradations, it is difficult to provide sufficient sensitivity and diagnostic accuracy, resulting in significant performance degradation or safety risks before battery faults are accurately diagnosed. Especially in the high-voltage and high-load operating environment of electric vehicles, late fault diagnosis will lead to increased repair costs and reduced user trust. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an intelligent detection system for battery failures of new energy vehicles.

[0006] In order to achieve the above purpose, the present invention adopts the following technical solution: A new energy vehicle battery fault intelligent detection system includes:

[0007] The voltage change analysis module collects the real-time voltage data of the battery cells of new energy vehicles, records the voltage at consecutive time points, calculates the difference between the voltage at each time point and the reference voltage, extracts the voltage change trend based on the difference calculation result, identifies abnormal fluctuations by comparing with the normal voltage pattern, and generates the battery voltage fluctuation index;

[0008] The internal resistance micro-analysis module regularly measures the internal resistance of the battery based on the battery voltage fluctuation index, performs time series analysis on the internal resistance, calculates the fractal dimension using fractal theory, compares the calculated fractal dimension result with the set threshold value, identifies the signs of battery aging and damage, and forms the battery internal resistance aging characteristics;

[0009] The current response monitoring module applies pulse currents of different amplitudes to the battery according to the battery internal resistance aging characteristics, captures the battery current response data to each pulse in real time, performs battery performance stability analysis based on the current response data and in combination with a preset response mode, and outputs a battery current stability analysis record;

[0010] The fault diagnosis integrated module integrates the battery current stability analysis record, battery internal resistance aging characteristics and battery voltage fluctuation index, conducts in-depth performance evaluation on battery cells, confirms the battery fault type and severity, and outputs comprehensive detection results of new energy vehicle battery faults.

[0011] As a further solution of the present invention, the step of calculating the difference between the voltage and the reference voltage at each time point is:

[0012] Install voltage sensors at the batteries of new energy vehicles to capture the voltage data of the batteries in real time and generate an initial voltage data set;

[0013] Extracting the voltage value of each sampling point from the initial voltage data set, and marking the timestamp of each sampling point accordingly, to obtain a voltage data set with a time tag;

[0014] Define the reference voltage value, combine the voltage data set of the time label, and use the formula: ;

[0015] Calculate the difference between each voltage value and the reference voltage , generate voltage difference data at each time point, where, Representing time point The voltage value, Represents the reference voltage.

[0016] As a further solution of the present invention, the steps of obtaining the battery voltage fluctuation index are:

[0017] Performing trend analysis on the voltage difference data at each time point, smoothing the data, revealing key trends of voltage changes, and obtaining a smoothed voltage trend data set;

[0018] Based on the normal voltage mode, the range and characteristics of normal voltage fluctuation are analyzed and determined, and compared with the smoothed voltage trend data set to detect voltage abnormal points and obtain abnormal voltage fluctuation events;

[0019] Based on the abnormal voltage fluctuation event, the formula is used: ;

[0020] Calculate and output battery voltage fluctuation index ,in, Represents the voltage difference of each abnormal fluctuation, represents the duration of abnormal fluctuations, is the total observation time.

[0021] As a further solution of the present invention, the steps of calculating the fractal dimension are:

[0022] Based on the battery voltage fluctuation index, regularly measure the original internal resistance data of the battery, analyze and determine the correlation between the two, and generate a corrected battery internal resistance value;

[0023] Using the corrected battery internal resistance value, perform time series analysis, draw a trend graph of internal resistance changing over time, capture fluctuation characteristics, and obtain internal resistance time series analysis results;

[0024] Based on the internal resistance time series analysis results, the formula is used: ;

[0025] Calculate and output fractal dimension ,in, Indicates that at a given scale The minimum number of boxes required to cover the data sequence, represents the scale of the box used in fractal analysis, Indicates the degree of variation of the internal resistance time series.

[0026] As a further solution of the present invention, the steps for obtaining the battery internal resistance aging characteristics are:

[0027] Based on the fractal dimension, the threshold D of battery internal resistance aging is set for comparison, using the formula: ;

[0028] Calculate and generate battery aging characteristic indicators , where F represents the fractal dimension and D is the set threshold;

[0029] The battery aging characteristic index is compared with the aging thresholds L1 and L2 of the difference level, using the formula: ;

[0030] Determine the aging state of the battery and generate an aging state classification result A, where a, b, and c are defined as normal, slightly aged, and severely aged, respectively;

[0031] Based on the aging status classification results, the aging status is statistically analyzed, the proportion of each aging status is analyzed, and the battery internal resistance aging characteristics are formed.

[0032] As a further solution of the present invention, the step of capturing the current response data is:

[0033] Using the battery internal resistance aging characteristics, design pulse current experimental schemes with different amplitudes for the battery, confirm that the pulse current covers the full range of battery internal resistance changes, and generate differentiated pulse current schemes;

[0034] According to the differentiated pulse current scheme, a corresponding pulse current is applied to the battery, and the instantaneous response of each pulse current is recorded in real time to obtain a real-time current response record;

[0035] The real-time current response record is analyzed to extract actual response data of the battery to the pulse current of the difference amplitude, thereby obtaining current pulse response data.

[0036] As a further solution of the present invention, the steps of obtaining the battery current stability analysis record are:

[0037] Based on the current pulse response data, extract key performance indicators, including peak current and response time, summarize and analyze the performance indicators, and generate statistical analysis results;

[0038] According to the statistical analysis results, a comparative analysis is performed with the preset current response mode, using the formula, ;

[0039] Calculate the total deviation of the actual observed value from the expected value for each data point , generating performance deviation data, where represents the observed value, represents the expected value, Represents the total number of observed data points, and i represents the index sequence;

[0040] Based on the performance deviation data, the battery's usage environment and historical performance data are comprehensively considered to conduct a long-term evaluation of the battery's stability, predict the battery's performance changes in future time periods, and output a battery current stability analysis record.

[0041] As a further solution of the present invention, the steps for obtaining the comprehensive detection result of the new energy vehicle battery fault are:

[0042] Integrate the battery current stability analysis record, battery internal resistance aging characteristics and battery voltage fluctuation index, integrate data items, perform data standardization processing, and generate comprehensive performance evaluation data;

[0043] Applying multi-dimensional data analysis to the comprehensive performance evaluation data, identifying key influencing factors in battery performance, extracting key health indicators, and obtaining battery health status evaluation indicators;

[0044] According to the battery health status evaluation index, the battery health status evaluation index is matched with the historical fault data, and the formula is used: ;

[0045] Calculate the probability of failure type , and obtain the comprehensive detection results of battery failure, where: Represents the battery performance Key health indicators, represents the base of natural logarithms, Indicates the total number of key health indicators.

[0046] Compared with the prior art, the advantages and positive effects of the present invention are:

[0047] In the present invention, by refining voltage monitoring and internal resistance analysis, the ability to identify battery performance degradation and early faults is significantly improved, anomalies are identified through advanced data processing technology, and detailed fluctuation analysis makes fault warning more accurate, reducing the risk of false alarms and missed alarms. The introduction of fractal theory to analyze internal resistance data can more accurately capture the microscopic changes in battery aging and achieve early detection of aging trends. Combined with pulse current testing and current response data analysis, the evaluation of battery performance stability is further enhanced, ensuring the comprehensiveness and depth of the detection results, and greatly improving the accuracy of battery fault detection and the efficiency of determining the fault type. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is a system flow chart of the present invention;

[0049] Figure 2 A flow chart for calculating the difference between the voltage at each time point and the reference voltage of the present invention;

[0050] Figure 3 A flowchart for obtaining a battery voltage fluctuation index according to the present invention;

[0051] Figure 4 It is a calculation flow chart of the fractal dimension of the present invention;

[0052] Figure 5 The flowchart of obtaining the internal resistance aging characteristics of the battery of the present invention;

[0053] Figure 6is a flow chart for capturing current response data of the present invention;

[0054] Figure 7 A flowchart for obtaining the battery current stability analysis record of the present invention;

[0055] Figure 8 This is a flow chart for obtaining comprehensive detection results of battery failures in new energy vehicles of the present invention. DETAILED DESCRIPTION

[0056] 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.

[0057] 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.

[0058] See also Figure 1 , a new energy vehicle battery fault intelligent detection system includes:

[0059] The voltage change analysis module collects the real-time voltage data of the battery cells of new energy vehicles, records the voltage at consecutive time points, calculates the difference between the voltage at each time point and the reference voltage, extracts the voltage change trend based on the difference calculation result, identifies abnormal fluctuations by comparing with the normal voltage pattern, and generates the battery voltage fluctuation index;

[0060] The internal resistance micro-analysis module measures the battery internal resistance value regularly based on the battery voltage fluctuation index, performs time series analysis on the internal resistance value, calculates the fractal dimension using fractal theory, compares the fractal dimension calculation result with the set threshold, identifies the signs of battery aging and damage, and forms the battery internal resistance aging characteristics;

[0061] The current response monitoring module applies pulse currents of different amplitudes to the battery according to the battery internal resistance aging characteristics, and captures the battery current response data to each pulse in real time. Based on the current response data and the preset response mode, the battery performance stability analysis is performed and the battery current stability analysis record is output;

[0062] The fault diagnosis integrated module integrates battery current stability analysis records, battery internal resistance aging characteristics and battery voltage fluctuation indicators, conducts in-depth performance evaluation of battery cells, confirms the type and severity of battery faults, and outputs comprehensive detection results of new energy vehicle battery faults.

[0063] Battery voltage fluctuation indicators include baseline deviation, peak frequency and number of over-limit times; battery internal resistance aging characteristics include dimensionality index, attenuation rate and aging degree; current stability analysis includes response consistency record, fluctuation index and deviation degree; comprehensive detection results of new energy vehicle battery failures include fault location results, damage level and warning level.

[0064] See also Figure 2 , the calculation steps of the difference between the voltage and the reference voltage at each time point are:

[0065] Install voltage sensors at the batteries of new energy vehicles to capture the voltage data of the batteries in real time and generate an initial voltage data set;

[0066] To install a voltage sensor on a battery cell of a new energy vehicle, you first need to select a suitable sensor location, usually located at the positive and negative ends of the battery, so as to directly capture the voltage signal. During the installation process, pay attention to the good contact between the sensor and the battery to avoid data deviation caused by poor contact. After installation, connect the sensor to the data acquisition system through a real-time data transmission line to ensure real-time data transmission without delay and interference. According to the preset sampling frequency, receive and record voltage data in real time. The data reflects the voltage changes of the battery cell under different working conditions. The generated initial voltage data set includes timestamps and corresponding voltage values, providing basic data support for subsequent data analysis and voltage monitoring.

[0067] Extract the voltage value of each sampling point from the initial voltage data set, and mark the timestamp of each sampling point accordingly to obtain a voltage data set with a time tag;

[0068] The timestamp and voltage value of each sampling point are extracted from the initial voltage data set. The sampling time of each data point is synchronized by a high-precision clock to ensure the accuracy of the time label. The time-series voltage data set excludes abnormal values ​​during the creation process, such as voltage drops or surges, which may be caused by sensor failure or external electromagnetic interference. The voltage values ​​within a reasonable range are screened to ensure the accuracy and reliability of the data set. The processed data set contains accurate time and voltage information, which is of great significance for subsequent voltage trend analysis and anomaly detection. The voltage data set with time labels makes it possible to track every voltage change and provide decision support for the battery management system.

[0069] Define the reference voltage value, combine the voltage data set with the time label, and use the formula: ;

[0070] Calculate the difference between each voltage value and the reference voltage , generate voltage difference data at each time point, where, Representing time point The voltage value, represents the reference voltage;

[0071] Record the voltage at a certain moment as 12.5V. is the preset reference voltage, set to 12V. Substitute the numerical value into the calculation:

[0072] ;

[0073] The results show that at this point in time, the voltage is 0.5V above the reference voltage, indicating a potential battery charge state or load change, which is critical for monitoring battery performance and preventing possible battery problems.

[0074] See also Figure 3 , the steps to obtain the battery voltage fluctuation index are:

[0075] Perform trend analysis on the voltage difference data at each time point, smooth the data, reveal the key trend of voltage change, and obtain a smoothed voltage trend data set;

[0076] When analyzing the voltage difference data set at each time point, the weighted moving average method is used to smooth the data. It not only considers the voltage value at a single time point, but also optimizes the trend representation of the overall data by assigning different weights to the previous and next data points. It is particularly suitable for reducing the impact of short-term noise or temporary voltage jumps, and more clearly reveals the actual change trend of the battery voltage. The weight of each data point is set to gradually decrease according to the time interval between it and the current point, which can make the trend line smoother and closer to the actual battery state change. The result after processing is a smoother voltage trend data set, which provides a reliable basis for identifying long-term performance changes of the battery and ensures the accuracy and practicality of the analysis results.

[0077] Based on the normal voltage mode, the range and characteristics of normal voltage fluctuations are analyzed and determined, and compared with the smoothed voltage trend data set to detect voltage abnormal points and obtain abnormal voltage fluctuation events;

[0078] When comparing the smoothed voltage trend data set with the predefined historical voltage pattern, the parameters of the historical voltage pattern are first determined. The parameters include the normal fluctuation range of voltage, maximum and minimum thresholds, etc., based on the standards obtained from long-term observation data and battery performance tests; then by setting specific thresholds, the system can automatically identify those voltage values ​​that exceed the normal fluctuation range, effectively helping to detect and warn possible battery failures or performance degradation; for each abnormal voltage fluctuation detected, the system will record it and conduct a detailed analysis, including the amplitude, frequency and duration of the fluctuation. The data will then be compared with the battery maintenance records and usage conditions to determine the potential causes of the abnormal fluctuations, providing technicians with detailed data support for further diagnosis and processing.

[0079] Based on abnormal voltage fluctuation events, the formula is used: ;

[0080] Calculate and output battery voltage fluctuation index ,in, Represents the voltage difference of each abnormal fluctuation, represents the duration of abnormal fluctuations, is the total observation time;

[0081] An abnormal fluctuation is , if the duration is seconds, then The total observation time is Seconds, substitute the value into the formula to calculate:

[0082] ;

[0083] The results show that during the observation period, the battery voltage stability was relatively high and the voltage fluctuation index was very low, indicating that the battery was in good working condition, which will help further analyze the battery performance and ensure the battery reliability and long-term stable operation.

[0084] See also Figure 4 , the calculation steps of fractal dimension are:

[0085] Based on the battery voltage fluctuation index, the original internal resistance data of the battery is measured regularly, the correlation between the two is analyzed and determined, and the corrected internal resistance value of the battery is generated;

[0086] The existing battery voltage fluctuation index is called, and on this basis, combined with the measurement data of the battery internal resistance, regression analysis is performed to clarify the relationship between voltage fluctuation and internal resistance. In the regression analysis, the voltage fluctuation index is used as the independent variable, and the internal resistance measurement value is used as the dependent variable. The linear regression model is used to calculate the correlation coefficient and slope between the two, and the influence of voltage change on internal resistance is analyzed. This can not only verify the direct influence of voltage fluctuation on internal resistance, but also provide basic data for battery status assessment, and finally generate a corrected battery internal resistance value for subsequent in-depth analysis and decision-making.

[0087] Use the corrected battery internal resistance value to perform time series analysis, draw a trend chart of internal resistance over time, capture fluctuation characteristics, and obtain internal resistance time series analysis results;

[0088] Use the corrected battery internal resistance value for time series analysis. First, collect the internal resistance measurement data over a period of time to ensure the timeliness and representativeness of the data. Then plot the internal resistance value into a time series graph to help observe the trend of internal resistance over time. In this process, the moving average method is used to smooth the internal resistance data to reduce the impact of short-term fluctuations on the results. The average value within a certain window range is calculated to extract the long-term trend of the internal resistance. In addition, the standard deviation of the internal resistance data is calculated to quantify the amplitude of the internal resistance fluctuation. Through these two statistical parameters, a deeper understanding of the internal resistance fluctuation characteristics can be obtained. Through further data analysis, combined with the historical data of internal resistance, outliers and potential fault signs are identified. The generated time series analysis results not only provide an important basis for battery maintenance, but also lay the foundation for subsequent monitoring and evaluation.

[0089] Based on the internal resistance time series analysis results, the formula is used: ;

[0090] Calculate and output fractal dimension ,in, Indicates that at a given scale The minimum number of boxes required to cover the data sequence, represents the scale of the box used in fractal analysis, Indicates the degree of variation of the internal resistance time series;

[0091] Box size is 0.1 seconds; , that is, the minimum number of boxes required to cover the data sequence at 0.1 second intervals. If 10 data points are collected in 0.1 seconds, then ; , the standard deviation of the internal resistance time series is 0.05ohms. Substitute the value into the formula for calculation:

[0092] ;

[0093] ;

[0094] ;

[0095] ;

[0096] ;

[0097] Compute the fractal dimension:

[0098] ;

[0099] This result shows that the time series of the battery internal resistance has high complexity ( , the value is high), indicating that there is a potential problem with the battery's health.

[0100] See also Figure 5 , the steps to obtain the battery internal resistance aging characteristics are:

[0101] Based on the fractal dimension, the threshold D of battery internal resistance aging is set for comparison, using the formula: ;

[0102] Calculate and generate battery aging characteristic indicators , where F represents the fractal dimension and D is the set threshold;

[0103] The average value of the fractal dimension in the historical data is 1.5, which is set as the threshold D. The fractal dimension F of the current battery is 2.626, and the calculation is:

[0104] ;

[0105] The value of E is 0.45. The indicator will provide data support for identifying and classifying the battery health status, ensuring the timeliness and effectiveness of the battery maintenance plan.

[0106] The battery aging characteristic index is compared with the difference level aging thresholds L1 and L2, using the formula, ;

[0107] Determine the aging state of the battery and generate an aging state classification result A, where a, b, and c are defined as normal, slightly aged, and severely aged, respectively;

[0108] Historical data shows that most batteries have no obvious aging when the E value is below 0.5, and the aging gradually becomes obvious between 0.5 and 0.1.25, and the aging is serious when it exceeds 1.25. So let L1=0.5 and L2=1.25. In order to derive the formula, the current E=1.75, then due to 1.75 1.25, according to the formula, the calculation result A=c, which means that the battery is in a serious aging state, which helps to determine the priority and measures of battery maintenance, and thus optimize the battery life and performance.

[0109] Based on the aging status classification results, statistical analysis is performed on the aging status, the proportion of each type of aging status is analyzed, and the battery internal resistance aging characteristics are formed;

[0110] After obtaining the battery aging status classification results, a detailed statistical analysis is performed on the data to quantify the distribution ratio of batteries at each aging level in the population. First, the aging rating data of all test batteries is collected from the battery management system, which includes the aging status value A automatically downloaded from the real-time monitoring system. Each battery is marked as a (normal), b (slightly aged), or c (severely aged). A quantitative count is performed for each aging state to determine the number of batteries in each state. For example, if there are a total of 10 batteries in the new energy vehicle battery pack, 7 are classified as normal, 2 are slightly aged, and 1 is severely aged, using the formula Calculate the percentage for each category, where is the number of batteries in the corresponding aging state, and N is the total number of batteries tested, that is, 10. In this way, the percentage of each aging category can be obtained, such as 70% normal, 20% slightly aged, and 10% severely aged. The statistical results not only provide a quantitative basis for battery maintenance and replacement, but also help the battery management team identify battery groups that may need further diagnosis or replacement, so as to develop more effective maintenance strategies and optimize battery life.

[0111] See also Figure 6 , the steps to capture the current response data are:

[0112] Using the battery internal resistance aging characteristics, design pulse current experimental schemes with different amplitudes for the battery, confirm that the pulse current covers the full range of battery internal resistance changes, and generate differentiated pulse current schemes;

[0113] Based on the known battery internal resistance aging characteristics, a pulse current test scheme is designed. The scheme specifies in detail the current pulse amplitudes corresponding to different internal resistance stages to ensure that the test covers all key aging stages. By analyzing the historical internal resistance data and aging models, the maximum and minimum values ​​of the pulse current, as well as the duration and interval of the pulse are determined to ensure that each pulse can effectively simulate the behavior of the battery at different aging stages. The designed differentiated pulse current scheme aims to simulate the extreme conditions that the battery may encounter in actual use, as well as its performance under normal use conditions, to generate differentiated pulse current schemes.

[0114] According to the differentiated pulse current scheme, a corresponding pulse current is applied to the battery, and the instantaneous response of each pulse current is recorded in real time to obtain a real-time current response record;

[0115] Based on the differentiated pulse current scheme, a predetermined pulse current is applied to the battery, and a dedicated current measuring instrument is used to record the instantaneous response of the current in detail. The detailed data of each pulse is captured in real time. During this process, special attention is paid to the impact of changes in the internal resistance of the battery on the current response. After each pulse, the current data points at the end are quickly collected. The data points reflect the battery's response to different internal resistance states. Time series analysis is performed on the real-time data, especially the analysis of current peak value, duration and decay rate. After each data collection, the recording device will perform a preliminary screening of the collected data to eliminate outliers caused by measurement errors. The retained valid data is used for further analysis, which not only ensures the accuracy of the data, but also provides a reliable basis for subsequent analysis. The result obtained is an accurately recorded real-time current response record.

[0116] Analyze the real-time current response records, extract the actual response data of the battery to the pulse current of the difference amplitude, and obtain the current pulse response data;

[0117] Through in-depth analysis of real-time current response records, classification and pattern recognition of current response data, focusing on analyzing the differences in current response under different pulse amplitudes, the analysis process takes into account the changing trend of battery internal resistance, current stability and repeatability. Through comprehensive evaluation of the data, changes in battery performance at different aging stages can be identified, and the current response pattern corresponding to each pulse can be accurately captured, which helps to understand the reliability and efficiency of the battery in actual use. Ultimately, detailed current response data of the battery to each pulse is obtained.

[0118] See also Figure 7 , the steps to obtain the battery current stability analysis record are:

[0119] Based on the current pulse response data, key performance indicators are extracted, including peak current and response time, and the performance indicators are summarized and analyzed to generate statistical analysis results;

[0120] Key performance indicators, such as peak current and response time, are extracted from the battery's current response data for each pulse for preliminary analysis. This process involves classifying and summarizing the data through statistical methods to generate a benchmark data model that contains various current response patterns. This model reflects the performance of the battery under different pulse current conditions. This benchmark model not only helps predict the performance trend of the battery, but also provides a basis for further comparative analysis. The results of the statistical analysis can directly map the battery's response characteristics and possible performance attenuation paths, thereby establishing a benchmark data model for each pulse response.

[0121] According to the statistical analysis results, a comparative analysis is performed with the preset current response mode, using the formula, ;

[0122] Calculate the total deviation of the actual observed value from the expected value for each data point , generating performance deviation data, where represents the observed value, represents the expected value, Represents the total number of observed data points, and i represents the index sequence;

[0123] For a set of n=10 data points, the observed value are {2.1, 2.5, 2.9, 3.0, 3.1, 3.3, 3.5, 3.8, 4.0, 4.1} respectively, while the expected values ​​are They are {2.0, 2.5, 3.0, 3.0, 3.0, 3.5, 3.5, 3.8, 4.0, 4.2} respectively. The calculation process is as follows:

[0124] ;

[0125] ;

[0126] ;

[0127] The results show that the total deviation R=0.08, which means that the current response data has a high overall consistency with the expected pattern and a small deviation, indicating that the battery's performance stability is good under the current test conditions, which helps to evaluate the long-term performance and reliability of the battery.

[0128] Based on the performance deviation data, the battery's usage environment and historical performance data are comprehensively considered to evaluate the battery's stability over a long period of time, predict the battery's performance changes in the future, and output the battery current stability analysis record;

[0129] Through the performance deviation data, the data is combined with the battery usage environment and historical performance data to conduct long-term performance evaluation. The process includes not only a detailed analysis of the current deviation data, but also involves trend comparison of historical data, so as to predict the performance changes of the battery in the future time period. The analysis methods include but are not limited to time series analysis and regression analysis, which can reveal the performance fluctuations of the battery in different use stages and evaluate its stability. The final analysis results will be used to form a battery current stability analysis record, providing a scientific basis for battery maintenance and performance optimization in the future time period.

[0130] See also Figure 8 ,The steps to obtain the comprehensive detection results of new energy vehicle battery failure are:

[0131] Integrate battery current stability analysis records, battery internal resistance aging characteristics and battery voltage fluctuation indicators, integrate data items, perform data standardization processing, and generate comprehensive performance evaluation data;

[0132] First, the format of battery current stability analysis records, battery internal resistance aging characteristics, and battery voltage fluctuation indicators are standardized to ensure data consistency and comparability. Data cleaning technology is used to remove outliers and noise data. Then, data fusion technology is used to synthesize these standardized data into a comprehensive performance evaluation data, which will serve as the basis for subsequent analysis. When constructing the comprehensive data, the integrity and representativeness of the data must be considered to ensure that the output results can truly reflect the overall health status of the battery. The generated comprehensive performance evaluation data can provide a scientific basis for battery status evaluation and fault prediction.

[0133] Apply multi-dimensional data analysis to comprehensive performance evaluation data, identify key influencing factors in battery performance, extract key health indicators, and obtain battery health status evaluation indicators;

[0134] Multi-dimensional data analysis is applied to the comprehensive performance evaluation data model. First, statistical analysis methods, such as principal component analysis (PCA), are used to identify the main factors affecting battery performance, which can reduce the dimension of the data while retaining the most critical information. Cluster analysis is then performed to identify different battery health status categories. Cluster analysis helps understand the pattern of battery performance degradation and extract health indicators that are significantly correlated with battery health status. Key health indicators are necessary for the subsequent prediction and evaluation of battery failure types. The battery health status evaluation indicators provide a key basis for subsequent modeling and failure analysis.

[0135] According to the battery health status evaluation index, the battery health status evaluation index is matched with the historical fault data, and the formula is used. ;

[0136] Calculate the probability of failure type , and obtain the comprehensive detection results of battery failure, where: Represents the battery performance Key health indicators, represents the base of natural logarithms, Indicates the total number of key health indicators;

[0137] The following battery health status evaluation indicators are available: , , ,and , Is the base of the natural logarithm (approximately equal to 2.71828).

[0138] Calculate the sum of the exponential parts, that is:

[0139] ;

[0140] Substituting the sum into the exponent and the square root, we get:

[0141] ;

[0142] Substituting into the regression formula, we get:

[0143] ;

[0144] The results show that based on the provided battery health indicators, the probability of the fault type occurring is 62.3%, which is of great significance for predicting the fault type of new energy vehicle batteries and provides a quantitative basis for further fault diagnosis and preventive measures.

[0145] 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. An intelligent detection system for battery failure of new energy vehicles, characterized in that: The system comprises: The voltage change analysis module collects the real-time voltage data of the battery cells of new energy vehicles, records the voltage at consecutive time points, calculates the difference between the voltage at each time point and the reference voltage, extracts the voltage change trend based on the difference calculation result, identifies abnormal fluctuations by comparing with the normal voltage pattern, and generates the battery voltage fluctuation index; The internal resistance micro-analysis module regularly measures the internal resistance of the battery based on the battery voltage fluctuation index, performs time series analysis on the internal resistance, calculates the fractal dimension using fractal theory, compares the calculated fractal dimension result with the set threshold value, identifies the signs of battery aging and damage, and forms the battery internal resistance aging characteristics; The current response monitoring module applies pulse currents of different amplitudes to the battery according to the battery internal resistance aging characteristics, captures the battery current response data to each pulse in real time, performs battery performance stability analysis based on the current response data and in combination with a preset response mode, and outputs a battery current stability analysis record; The fault diagnosis integrated module integrates the battery current stability analysis record, battery internal resistance aging characteristics and battery voltage fluctuation index, conducts in-depth performance evaluation on the battery cells, confirms the type and severity of battery failure, and outputs comprehensive detection results of new energy vehicle battery failures; The calculation steps of the fractal dimension are: Based on the battery voltage fluctuation index, regularly measure the original internal resistance data of the battery, analyze and determine the correlation between the two, and generate a corrected battery internal resistance value; Using the corrected battery internal resistance value, perform time series analysis, draw a trend graph of internal resistance changing over time, capture fluctuation characteristics, and obtain internal resistance time series analysis results; Based on the internal resistance time series analysis results, the formula is used: ; Calculate and output fractal dimension ,in, Indicates that at a given scale The minimum number of boxes required to cover the data sequence, represents the scale of the box used in fractal analysis, Indicates the degree of variation of the internal resistance time series; The steps for capturing the current response data are: Using the battery internal resistance aging characteristics, design pulse current experimental schemes with different amplitudes for the battery, confirm that the pulse current covers the full range of battery internal resistance changes, and generate differentiated pulse current schemes; According to the differentiated pulse current scheme, a corresponding pulse current is applied to the battery, and the instantaneous response of each pulse current is recorded in real time to obtain a real-time current response record; The real-time current response record is analyzed to extract actual response data of the battery to the pulse current of the difference amplitude, thereby obtaining current pulse response data.

2. The intelligent detection system for battery failure of new energy vehicles according to claim 1 is characterized in that: The steps for calculating the difference between the voltage and the reference voltage at each time point are as follows: Install voltage sensors at the batteries of new energy vehicles to capture the voltage data of the batteries in real time and generate an initial voltage data set; Extracting the voltage value of each sampling point from the initial voltage data set, and marking the timestamp of each sampling point accordingly, to obtain a voltage data set with a time tag; Define the reference voltage value, combine the voltage data set of the time label, and use the formula: ; Calculate the difference between each voltage value and the reference voltage , generate voltage difference data at each time point, where, Representing time point The voltage value, Represents the reference voltage.

3. The intelligent detection system for battery failure of new energy vehicles according to claim 2 is characterized in that: The steps for obtaining the battery voltage fluctuation index are as follows: Performing trend analysis on the voltage difference data at each time point, smoothing the data, revealing key trends of voltage changes, and obtaining a smoothed voltage trend data set; Based on the normal voltage mode, the range and characteristics of normal voltage fluctuation are analyzed and determined, and compared with the smoothed voltage trend data set to detect voltage abnormal points and obtain abnormal voltage fluctuation events; Based on the abnormal voltage fluctuation event, the formula is used: ; Calculate and output battery voltage fluctuation index ,in, Represents the voltage difference of each abnormal fluctuation, represents the duration of abnormal fluctuations, is the total observation time.

4. The intelligent detection system for battery failure of new energy vehicles according to claim 1 is characterized in that: The steps for obtaining the battery internal resistance aging characteristics are: Based on the fractal dimension, a threshold D of battery internal resistance aging is set for comparison, using the formula: ; Calculate and generate battery aging characteristic indicators , where F represents the fractal dimension and D is the set threshold; The battery aging characteristic index is compared with the difference level aging thresholds L1 and L2, using the formula, ; Determine the aging state of the battery and generate an aging state classification result A, where a, b, and c are defined as normal, slightly aged, and severely aged, respectively; Based on the aging status classification results, the aging status is statistically analyzed, the proportion of each aging status is analyzed, and the battery internal resistance aging characteristics are formed.

5. The intelligent detection system for battery failure of new energy vehicles according to claim 1 is characterized in that: The steps for obtaining the battery current stability analysis record are: Based on the current pulse response data, extract key performance indicators, including peak current and response time, summarize and analyze the performance indicators, and generate statistical analysis results; According to the statistical analysis results, a comparative analysis is performed with the preset current response mode, using the formula, ; Calculate the total deviation of the actual observed value from the expected value for each data point , generating performance deviation data, where represents the observed value, represents the expected value, Represents the total number of observed data points, and i represents the index sequence; Based on the performance deviation data, the battery's usage environment and historical performance data are comprehensively considered to conduct a long-term evaluation of the battery's stability, predict the battery's performance changes in future time periods, and output a battery current stability analysis record.

6. The intelligent detection system for battery failure of new energy vehicles according to claim 5 is characterized in that: The steps for obtaining the comprehensive detection results of the new energy vehicle battery failure are as follows: Integrate the battery current stability analysis record, battery internal resistance aging characteristics and battery voltage fluctuation index, integrate data items, perform data standardization processing, and generate comprehensive performance evaluation data; Applying multi-dimensional data analysis to the comprehensive performance evaluation data, identifying key influencing factors in battery performance, extracting key health indicators, and obtaining battery health status evaluation indicators; According to the battery health status evaluation index, the battery health status evaluation index is matched with the historical fault data, and the formula is used: ; Calculate the probability of failure type , and obtain the comprehensive detection results of battery failure, where: Represents the battery performance Key health indicators, represents the base of natural logarithms, Indicates the total number of key health indicators.

Images