Thermal runaway fault monitoring and early warning method for new energy automobile battery

Abstract

The invention provides a thermal runaway fault monitoring and early warning method for a new energy automobile battery, and relates to the technical field of fault monitoring, and the method comprises the steps: collecting a battery voltage and current fluctuation curve of a preset monitoring time zone when an automobile battery is started, and carrying out the interaction with a first temperature sensor to obtain an environment characteristic temperature of the preset monitoring time zone; performing positive sample analysis based on the voltage and current fluctuation curve and the environment characteristic temperature to obtain a battery temperature reference curve, and interacting with a second temperature sensor to obtain a battery temperature monitoring curve of a preset monitoring time zone, when the curve deviation coefficient of the battery temperature reference curve and the battery temperature monitoring curve is larger than or equal to the curve deviation coefficient threshold value, a thermal runaway fault early warning signal is generated, and the problems that when the automobile battery is monitored, the response time is slow, the early warning precision is low, and a thermal runaway event cannot be effectively prevented in time are solved; the effects of improving the comprehensiveness, timeliness and accuracy of battery thermal runaway monitoring are achieved.

Description

Technical Field

[0001] This invention relates to the field of fault monitoring technology, and more specifically to a method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries. Background Technology

[0002] In the new energy vehicle industry, battery thermal runaway monitoring is a critical safety measure. Thermal runaway refers to the uncontrolled chemical reaction that occurs in a battery under the influence of overcharging, over-discharging, internal short circuits, or external temperatures, causing a rapid rise in battery temperature and potentially leading to explosion or fire. Existing battery thermal runaway control technologies mainly rely on periodic temperature monitoring and voltage and current state detection. While these technologies provide basic battery state information, they often suffer from delayed response and insufficient fault prediction. For example, existing technologies typically only detect anomalies when the battery has already begun to overheat, exhibiting significant lag. This is usually too late to prevent thermal runaway events, leading to safety risks. Furthermore, traditional methods often rely solely on temperature or voltage / current data, resulting in inaccurate battery state detection and false alarms. Additionally, the lack of effective data analysis and prediction makes it difficult to accurately predict the risk of battery thermal runaway and provide effective early warnings.

[0003] In summary, existing technologies for monitoring thermal runaway in automotive batteries often suffer from technical problems such as insufficient comprehensive data utilization, slow response time, and low early warning accuracy, making it difficult to prevent thermal runaway events in a timely and effective manner. Summary of the Invention

[0004] This application provides a method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries, which addresses the technical problems of existing technologies in monitoring thermal runaway of automotive batteries, such as insufficient comprehensive data utilization, slow response time, and low early warning accuracy, making it difficult to prevent thermal runaway events in a timely and effective manner.

[0005] The method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries provided in this application includes: When the battery of a new energy vehicle is started, the battery voltage fluctuation curve and battery current fluctuation curve are collected in a preset monitoring time zone, wherein the preset monitoring time zone is a time interval obtained by subtracting a preset time step from the current time; the system interacts with a first temperature sensor to obtain the ambient characteristic temperature of the preset monitoring time zone; positive sample analysis is performed based on the battery voltage fluctuation curve, the battery current fluctuation curve and the ambient characteristic temperature to obtain the battery temperature reference curve; the system interacts with a second temperature sensor to obtain the battery temperature monitoring curve of the preset monitoring time zone; when the curve deviation coefficient between the battery temperature reference curve and the battery temperature monitoring curve is greater than or equal to the curve deviation coefficient threshold, a thermal runaway fault warning signal is generated and sent to the user terminal.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: The method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries provided in this application involves collecting battery voltage fluctuation curves and battery current fluctuation curves within a preset monitoring time zone when the new energy vehicle battery is started. The preset monitoring time zone is a time interval obtained by subtracting a preset time step from the current moment. The method interacts with a first temperature sensor to obtain the ambient characteristic temperature of the preset monitoring time zone. Based on the battery voltage fluctuation curve, the battery current fluctuation curve, and the ambient characteristic temperature, positive sample analysis is performed to obtain a battery temperature reference curve. The method also interacts with a second temperature sensor to obtain the battery temperature monitoring curve of the preset monitoring time zone. When the curve deviation coefficient between the battery temperature reference curve and the battery temperature monitoring curve is greater than or equal to a curve deviation coefficient threshold, a thermal runaway fault early warning signal is generated and sent to the user terminal. This method solves the technical problems of insufficient data utilization, slow response time, and low early warning accuracy in existing technologies for monitoring thermal runaway in automotive batteries, which make it difficult to prevent thermal runaway events in a timely and effective manner. It achieves the technical effect of improving the comprehensiveness, timeliness, and accuracy of battery thermal runaway monitoring, and significantly enhancing the early warning capability for potential battery faults. Attached Figure Description

[0007] Figure 1 This application provides a schematic flowchart of a method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries.

[0008] Figure 2 This application provides a schematic diagram of the positive sample analysis process in a method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries.

[0009] Figure 3 This is a schematic diagram of the structure of an exemplary electronic device of this application.

[0010] Explanation of reference numerals in the attached drawings: Processor 31, Memory 32, Input device 33, Output device 34. Detailed Implementation

[0011] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0012] Examples, such as Figure 1 As shown, this application provides a method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries, the method comprising: When the battery of a new energy vehicle is started, the battery voltage fluctuation curve and battery current fluctuation curve are collected in a preset monitoring time zone. The preset monitoring time zone is a time interval obtained by pushing a preset time step forward from the current time.

[0013] In one specific embodiment, when the new energy vehicle battery starts up, the battery voltage fluctuation curve and battery current fluctuation curve are first collected in real time. Specifically, this collection step is performed for a preset monitoring time zone, which is obtained by calculating a preset time step backward from the current moment. The preset time step refers to a time interval determined based on the battery's charge and discharge characteristics, historical data, and actual operating environment, with the aim of capturing the voltage and current fluctuations of the battery under typical operating conditions.

[0014] Based on this, the collected voltage and current data are processed, specifically including recording the voltage and current changes of the battery within the specified time zone. Specifically, the battery voltage fluctuation curve is constructed using neighborhood hierarchical clustering analysis of the battery voltage time-series information, while the battery current fluctuation curve is constructed using a similar clustering analysis method. This clustering analysis compares voltage or current data at adjacent time points; when the deviation of these data exceeds a predetermined threshold, the data from these time points are merged to form a more coherent and representative fluctuation curve. For example, if the voltage change from time t1 to t2 remains stable within a certain range within the preset monitoring time zone, but the voltage suddenly rises or falls and exceeds the preset voltage deviation threshold from time t2 to t3, then t2 and t3 will be considered critical change points requiring special monitoring. This processing not only improves data processing efficiency but also enhances the monitoring system's sensitivity and response speed to potential battery anomalies. Through these steps, the voltage and current status of new energy vehicle batteries during actual operation can be effectively monitored, providing reliable data support for subsequent thermal runaway fault early warning.

[0015] Interact with the first temperature sensor to obtain the ambient characteristic temperature of the preset monitoring time zone.

[0016] Furthermore, the system interacts with a first temperature sensor to obtain the ambient characteristic temperature within a preset monitoring time zone. The first temperature sensor is a device specifically designed to monitor and record temperature changes in a specific environment or device. In this invention, the primary function of the first temperature sensor is to acquire temperature data around the battery of a new energy vehicle or within its operating environment. This sensor is typically installed near the battery module or other critical parts of the vehicle to ensure accurate monitoring of the ambient temperature affecting battery performance and safety. The ambient characteristic temperature refers to the statistical characteristic value of the temperature data recorded within the preset monitoring time zone within the battery's operating environment. This data is crucial for understanding the environmental factors contributing to battery thermal runaway.

[0017] Specifically, in this step, a data communication connection is first established with a first temperature sensor installed inside the vehicle or near the battery module. This connection allows for the real-time collection of temperature data from the sensor, typically recorded in time-series format, covering the entire preset monitoring time zone. The collected temperature time-series information is then further processed, including calculating statistical parameters such as variance and mean. For example, if the ambient temperature time-series information shows a gradual increase in temperature from 8 AM to 1 PM within the preset monitoring time zone, followed by a period of stability, the system will calculate the average temperature during this time period and set it as the characteristic ambient temperature for that time zone. Furthermore, variance statistics can be used to assess the volatility of temperature changes. A smaller variance indicates more stable temperature changes and higher accuracy of the characteristic ambient temperature; conversely, a larger variance suggests that the potential impact of temperature instability on battery performance needs to be considered.

[0018] The above steps provide the system with the temperature background of the battery operating environment, which helps to more accurately assess the risk of thermal runaway in subsequent analyses and enables comprehensive thermal runaway early warning analysis by combining battery voltage and current data. By comprehensively considering the ambient temperature and the battery's own electrical performance, it is possible to more effectively predict and warn of potential thermal runaway faults, thereby enhancing the safety of new energy vehicles.

[0019] Positive sample analysis is performed based on the battery voltage fluctuation curve, the battery current fluctuation curve, and the ambient characteristic temperature to obtain the battery temperature baseline curve.

[0020] Optionally, the positive sample analysis refers to using data samples collected under normal operating conditions to build a model or benchmark. These samples represent the typical behavior of the battery when no fault occurs. First, voltage and current fluctuation curves of the battery within a specific monitoring time zone are collected. These curves are obtained through neighborhood hierarchical clustering analysis of the time-series information of the battery voltage and current. Simultaneously, the ambient characteristic temperature is obtained from a first temperature sensor. This temperature is obtained through variance statistics and mean analysis of the ambient temperature data. Next, a positive sample dataset is constructed. This dataset consists of data collected during normal battery operation, including but not limited to battery voltage and current fluctuation data and corresponding ambient temperatures. This data reflects the battery's performance and response under fault-free conditions. Then, a predictive model is trained using this positive sample data. This model is typically a Long Short-Term Memory (LSTM) neural network model, capable of learning the relationship between voltage, current, and temperature data and predicting the normal trend of battery temperature changes. In this way, the model generates a benchmark curve for battery temperature, which characterizes the expected battery temperature change under given voltage and current conditions. The obtained battery temperature baseline curve will be used to compare with the battery temperature curve monitored in real time. Any behavior that deviates significantly from the baseline curve may indicate that the battery is entering a state of thermal runaway, thereby triggering the warning system.

[0021] Through the above steps, not only can a temperature prediction model of the battery under healthy conditions be accurately established, but also behaviors that deviate from normal operating mode can be monitored and identified in real time, thereby providing early warning of potential thermal runaway events and increasing response time redundancy to ensure the safety of the vehicle and passengers.

[0022] Interact with the second temperature sensor to obtain the battery temperature monitoring curve for the preset monitoring time zone.

[0023] For example, the second temperature sensor is typically installed at a critical location on the battery module to ensure accurate measurement of the battery's actual operating temperature. This sensor is connected to the vehicle's monitoring system or central processing unit (CPU) via a data transmission interface to ensure real-time data transmission. When the new energy vehicle starts and enters the preset monitoring time zone, the second temperature sensor begins to collect battery temperature data in real time. This data is recorded in a time series format, with each data point representing the battery temperature at a specific point in time. The temperature data collected from the second temperature sensor is used to construct a battery temperature monitoring curve, which reflects the trend and pattern of battery temperature changes throughout the entire time zone from the start to the end of monitoring. Furthermore, the collected temperature data undergoes filtering and noise reduction processing to eliminate accidental fluctuations caused by sensor errors or external environmental interference, ensuring the accuracy and reliability of the battery temperature monitoring curve. For instance, if the battery temperature suddenly rises above the normal operating range during a certain period within the monitoring time zone, this will create a significant peak on the battery temperature monitoring curve. The system compares this information with the battery temperature baseline curve to identify any potential risk of thermal runaway.

[0024] By following the steps above, the battery temperature status can be monitored in real time, thus providing an effective early warning before thermal runaway events occur, enhancing the safety and reliability of new energy vehicles. This process is an important supplement to the thermal management strategy in the battery management system, helping to extend battery life and ensure user driving safety.

[0025] When the curve deviation coefficient between the battery temperature reference curve and the battery temperature monitoring curve is greater than or equal to the curve deviation coefficient threshold, a thermal runaway fault early warning signal is generated and sent to the user terminal.

[0026] Specifically, during vehicle operation, the real-time monitored battery temperature curve is continuously compared with a pre-established battery temperature baseline curve based on a model. The battery temperature baseline curve represents the expected temperature performance of the battery under normal operating conditions, while the battery temperature monitoring curve reflects the current actual temperature situation. Further, a deviation coefficient is calculated. This deviation coefficient is derived by calculating the temperature difference between the two curves over the entire monitoring time zone. This involves quantifying the temperature difference at corresponding points on the two curves and calculating a deviation coefficient according to a preset pattern. The deviation coefficient threshold is set based on battery design parameters, historical data analysis, and safety standards. When the calculated deviation coefficient reaches or exceeds this threshold, the battery is considered to be at risk of thermal runaway. Once the deviation coefficient exceeds the threshold, a thermal runaway fault warning signal is immediately generated. This signal includes key information such as the fault type, current battery status, and suggested countermeasures, and is subsequently sent to the user's device, such as a smartphone or in-vehicle information system. This communication can be completed via wireless network, Bluetooth, or other available communication technologies. For example, if during monitoring, the battery temperature rises rapidly due to overcharging or excessively high external temperature, and the deviation coefficient from the baseline curve increases rapidly and exceeds the set threshold, the system will automatically trigger an early warning to prompt the user to take safety measures such as cooling down or stopping the vehicle for inspection.

[0027] Through the aforementioned monitoring and early warning mechanisms, drivers or maintenance personnel can be effectively alerted in advance of potential thermal runaway of the battery, thereby taking timely countermeasures to avoid potential safety accidents and ensure the safety of the vehicle and passengers. Furthermore, the automation and intelligence of this process greatly improve the efficiency and reliability of the new energy vehicle battery management system.

[0028] Furthermore, when the new energy vehicle battery starts, the battery voltage fluctuation curve and battery current fluctuation curve are collected in a preset monitoring time zone, including: Obtain battery voltage timing information and battery current timing information for the preset monitoring time zone.

[0029] Neighborhood hierarchical clustering analysis is performed on the battery voltage time-series information to construct the battery voltage fluctuation curve.

[0030] The battery current time-series information is subjected to neighborhood hierarchical clustering analysis to construct the battery current fluctuation curve.

[0031] Furthermore, at the moment the new energy vehicle battery starts, real-time data on battery voltage and current are collected by sensors installed on the battery module. The battery voltage and current time-series information refers to the sequence of battery voltage and current values ​​recorded chronologically within a preset monitoring time zone. The preset monitoring time zone is set based on a reasonable period after battery startup to ensure sufficient data collection, allowing for subsequent data analysis. The collected voltage and current time-series information is then processed through neighborhood hierarchical clustering analysis. This analysis method groups data points based on the similarity of their adjacent values, thereby identifying natural clusters in the data. For battery voltage time-series information, the system identifies data points that are temporally adjacent and have similar voltage changes, and clusters them into one cluster; similarly, battery current time-series information is also clustered according to similar current changes. After the clustering analysis is completed, the data generated by each cluster is used to construct battery voltage fluctuation curves and battery current fluctuation curves. These fluctuation curves show the changing trends of battery voltage and current within the preset monitoring time zone and are key indicators for identifying the battery's operating status. For example, if the collected data shows that the voltage drops rapidly and then rises quickly in a short period of time, cluster analysis may group this series of changes into a cluster to characterize possible abnormal battery discharge or recharge phenomena. Such fluctuation curves are crucial for the early diagnosis of possible battery failures.

[0032] Through the above steps, not only can the voltage and current changes of the battery be accurately recorded and analyzed, but cluster analysis can also be used to gain a deeper understanding of the battery's operating mode and potential problems. This information provides strong data support for ensuring the safe operation of the battery and timely fault warning.

[0033] Furthermore, neighborhood hierarchical clustering analysis is performed on the battery voltage time-series information to construct the battery voltage fluctuation curve, including: The battery voltage timing information is obtained by first time-domain battery voltage and second time-domain battery voltage, wherein the first time domain and the second time domain are adjacent time domains.

[0034] When the battery voltage deviation between the first time-domain battery voltage and the second time-domain battery voltage is greater than or equal to the voltage deviation threshold, the first time domain and the second time domain are merged to obtain a third time domain. The battery voltage in the first time domain and the battery voltage in the second time domain are then averaged to obtain the third time-domain battery voltage.

[0035] The analysis is repeated, and when the voltage deviation of any neighborhood is less than the voltage deviation threshold, the battery voltage clustering time sequence information is output.

[0036] The battery voltage fluctuation curve is constructed based on the battery voltage clustering time series information.

[0037] Specifically, the system first collects battery voltage time-series information within a preset monitoring time zone. This information records the changes in battery voltage in chronological order, with each data point representing a voltage reading in a specific time domain. The time domain is typically defined as the time interval between battery voltage measurements, such as per second or per minute. During neighborhood hierarchical clustering analysis, the system compares battery voltage values ​​in adjacent time domains (i.e., the first and second time domains). If the deviation between the voltage values ​​in two adjacent time domains exceeds a preset voltage deviation threshold, it indicates a significant voltage change between these two time domains, which could be an indicator of battery performance variation. When the voltage deviation between the first and second time domains exceeds the threshold, these two time domains are merged into a new time domain (the third time domain). At this point, the system calculates the average of the battery voltages in the first and second time domains to represent the voltage value in the newly merged time domain. This averaging analysis helps smooth the voltage data and reduce the impact of random errors. Next, the system continues to perform iterative analysis on all voltage data, examining neighboring time domains in each analysis, until the entire time-series data has been fully processed. In each step, adjacent time domains are kept separate only if their voltage deviations do not exceed a threshold, thus preserving more detailed voltage change information. Finally, a battery voltage fluctuation curve is constructed based on the processed battery voltage clustering time series information. This curve reflects in detail the voltage stability and fluctuation characteristics of the battery within the monitoring time zone.

[0038] By following the steps above, the real-time changes and trends of battery voltage can be monitored more accurately, providing key data for the battery management system to enable more effective battery health monitoring and fault diagnosis. This analysis process not only improves the accuracy of battery monitoring but also helps to detect and prevent battery-related faults in a timely manner, thereby improving battery operating efficiency and extending its service life.

[0039] Furthermore, interacting with the first temperature sensor to obtain the environmental characteristic temperature of the preset monitoring time zone includes: Interact with the first temperature sensor to obtain the environmental monitoring temperature time sequence information of the preset monitoring time zone.

[0040] The variance statistics of the environmental monitoring temperature time series information are performed to obtain the environmental monitoring temperature dispersion coefficient.

[0041] When the environmental monitoring temperature dispersion coefficient is less than or equal to the dispersion coefficient threshold, the mean value of the environmental monitoring temperature time series information is set as the environmental characteristic temperature.

[0042] When the environmental monitoring temperature dispersion coefficient is greater than the dispersion coefficient threshold, the maximum value of the concentrated environmental monitoring temperature of the environmental monitoring temperature time series information is obtained and set as the environmental characteristic temperature.

[0043] For example, when a new energy vehicle starts, the first temperature sensor begins to collect real-time ambient temperature time-series information within a preset monitoring time zone. This preset monitoring time zone refers to a specific time period set according to the vehicle's operating environment and a predetermined monitoring plan, with the aim of capturing continuous temperature data within that time zone. Then, variance statistics are performed on the collected ambient temperature time-series information to assess the degree of temperature data fluctuation and obtain the environmental monitoring temperature dispersion coefficient. This dispersion coefficient is a quantitative indicator of the degree of dispersion of temperature data, reflecting the stability of the ambient temperature. A smaller variance indicates more consistent temperature changes, while a larger variance indicates greater temperature fluctuations. Furthermore, different strategies are adopted to determine the environmental characteristic temperature based on the magnitude of the dispersion coefficient. If the dispersion coefficient is less than or equal to a set threshold, it indicates that the temperature data is relatively stable. In this case, the system calculates the mean of all temperature data and sets this mean as the environmental characteristic temperature, representing the average environmental conditions within the monitoring time zone. Conversely, if the dispersion coefficient is greater than the threshold, it indicates greater temperature fluctuations. In this case, the system selects the maximum value in the temperature time-series information as the environmental characteristic temperature, reflecting extreme environmental conditions. For example, if the temperature gradually rises from 25°C in the morning to 35°C in the afternoon during a high-temperature day in summer, but the change is relatively stable, the system may take the average temperature (about 30°C) of this period as the environmental characteristic temperature; while if the temperature suddenly rises from 25°C to 40°C in a short period of time due to some reason (such as air conditioning failure) on the same day, the system will take 40°C as the environmental characteristic temperature to indicate the highest environmental temperature during that period.

[0044] The above steps enable effective monitoring and evaluation of temperature changes in the operating environment of new energy vehicles, providing important environmental parameters for the battery management system, thereby helping to optimize battery performance and prevent potential thermal runaway risks.

[0045] Furthermore, the centralized environmental monitoring temperature for obtaining the time-series information of the environmental monitoring temperature includes: The first environmental monitoring temperature for obtaining the environmental monitoring temperature timing information.

[0046] The set of temperature deviations between the environmental monitoring temperature time series information and the first environmental monitoring temperature is calculated by iterating through the information.

[0047] The temperature deviations are sorted from smallest to largest in the set of temperature deviations, and the average value is calculated to obtain the first local outlier parameter, which is then added to the set of local outlier parameters.

[0048] Calculate the mean of the local outlier parameters of the local outlier parameter set.

[0049] The first concentration factor is obtained by calculating the ratio of the mean of the local outlier parameter to the first local outlier parameter.

[0050] When the first concentration factor is greater than or equal to the concentration factor threshold, the first environmental monitoring temperature is added to the centralized environmental monitoring temperature.

[0051] Specifically, the system first selects the first environmental monitoring temperature from the collected environmental monitoring temperature time series information. This step marks the beginning of the data processing flow; the first environmental monitoring temperature is usually the first temperature record in the time series data. The entire environmental monitoring temperature time series information is traversed, and the difference between each data point and the first environmental monitoring temperature is calculated, forming a temperature deviation set. This set will be used for subsequent statistical analysis to help determine the temperature distribution in the data. Next, half of the temperature deviations are selected from the temperature deviation set in ascending order, and the mean of these deviations is calculated to obtain the first local outlier parameter. This parameter reflects a local stability or anomaly in the temperature data. Then, the system summarizes all local outlier parameters and calculates the mean of this set, i.e., the mean of the local outlier parameters. Finally, the ratio of the first local outlier parameter to the mean of the local outlier parameters is compared to calculate the first concentration factor. This factor is an indicator of the importance of a certain temperature value in the overall data. If the first concentration factor is greater than or equal to the preset concentration factor threshold, it indicates that the first environmental monitoring temperature has high representativeness or concentration in the entire temperature series. Therefore, it is added to the concentrated environmental monitoring temperature, and this temperature value will be regarded as the key data point that best reflects the actual thermal condition of the environment.

[0052] Through the above steps, not only can key temperature points in the environment be accurately identified, but more scientific data support can also be provided for temperature management and thermal runaway early warning systems for new energy vehicle batteries, thereby improving battery safety and operational efficiency.

[0053] Furthermore, when the curve deviation coefficient between the battery temperature reference curve and the battery temperature monitoring curve is greater than or equal to a curve deviation coefficient threshold, a thermal runaway fault early warning signal is generated and sent to the user, including: Unify the coordinate systems of the battery temperature reference curve and the battery temperature monitoring curve, and interconnect the curve heads and curve tails of the battery temperature reference curve and the battery temperature monitoring curve to obtain the graphic area parameters.

[0054] Set the area parameter of the graphic as the curve deviation coefficient.

[0055] Furthermore, the coordinate systems of the battery temperature reference curve and the battery temperature monitoring curve are unified to ensure that the two curves are compared under the same standard. Coordinate system one is used to eliminate comparison biases caused by different data measurement units or scales, thereby ensuring the accuracy of subsequent analysis. After coordinate system one, the system will perform curve head interconnection and curve tail interconnection operations. Specifically, the starting points and ending points of the two curves are connected by straight lines to form a closed figure. Then, the area of ​​the closed figure enclosed by the two curves and the connecting line is calculated, and this area is defined as the figure area parameter. The figure area parameter is then set as the curve deviation coefficient, which is a quantitative indicator that measures the difference between the two curves. Its magnitude directly reflects the degree of deviation between the reference temperature and the measured temperature. By comparing this coefficient with a preset threshold, the system can determine whether the battery is at risk of potential thermal runaway. When the curve deviation coefficient exceeds the threshold, the system will immediately generate a thermal runaway fault warning signal and send this signal to the user. The warning signal includes key fault information, suggested countermeasures, and the time and location of the fault, enabling the user to take timely measures to prevent or mitigate the impact of thermal runaway. By following the steps above, the battery temperature status can be effectively monitored, potential risks of thermal runaway can be detected and warned in a timely manner, and the safety and reliability of battery use can be significantly improved.

[0056] Furthermore, such as Figure 2 As shown, a positive sample analysis is performed based on the battery voltage fluctuation curve, the battery current fluctuation curve, and the ambient characteristic temperature to obtain the battery temperature baseline curve, including: Collect historical operating data of positive sample batteries for new energy vehicle battery models. The historical operating data of batteries includes ambient temperature records, battery voltage fluctuation time series records, battery current fluctuation time series records, and battery temperature time series records.

[0057] The historical operating data of the positive sample batteries is cropped according to the preset time step to obtain several positive sample battery operating sample data.

[0058] Using the battery temperature time-series recorded data as supervision, and the ambient temperature recorded data, the battery voltage fluctuation time-series recorded data, and the battery current fluctuation time-series recorded data as input, a long short-term memory neural network is trained by retrieving several positive sample battery operation sample data to generate a battery time-series temperature prediction model.

[0059] The battery voltage fluctuation curve, the battery current fluctuation curve, and the ambient characteristic temperature are processed according to the battery time-series temperature prediction model to obtain battery temperature prediction time-series information and construct the battery temperature reference curve.

[0060] Specifically, positive sample data is extracted from the historical operating data of new energy vehicle batteries. This data comes from batteries operating under normal conditions and contains rich information, including ambient temperature records, battery voltage fluctuation time-series records, battery current fluctuation time-series records, and battery temperature time-series records. This data provides the system with a complete historical perspective of the battery's operating status. To make the data processing more timely, the historical data of positive sample batteries is cropped according to a preset time step. This means dividing the entire historical data into several hourly slices, each containing a complete set of battery operating information. These cropped data segments are the positive sample battery operating sample data. Next, using battery temperature time-series records as a supervision signal, and using ambient temperature data, battery voltage fluctuation data, and battery current fluctuation data as inputs, a Long Short-Term Memory (LSTM) neural network is trained. LSTM is a neural network model capable of processing time-series data and effectively capturing long-term dependencies in time-series data. During training, the LSTM model learns the temperature change patterns of the battery and generates a time-series model for predicting battery temperature. By using a trained LSTM model, predictive time-series information about battery temperature can be generated based on real-time battery voltage fluctuation curves, battery current fluctuation curves, and ambient temperature characteristics. This information represents the expected temperature change of the battery under current operating conditions. Based on this predictive data, a battery temperature baseline curve is constructed, reflecting the normal temperature change trend of the battery under corresponding environmental conditions. For example, when the ambient temperature is 30°C, the voltage fluctuation is small but the current increases, the system can predict the rate of temperature rise of the battery and generate a baseline curve as a reference for the battery's health status. Through the above steps, the accuracy of the battery temperature baseline curve is ensured, enabling better monitoring and early warning of the battery during actual use.

[0061] Through the technical solutions of the above embodiments, the method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries provided in this application has the following technical effects: 1. By collecting battery voltage and current fluctuation data in real time within a preset monitoring time zone and performing neighborhood hierarchical clustering analysis to construct battery voltage and current fluctuation curves, more detailed monitoring of battery status is possible. This allows for the capture of minute changes in battery performance, which may be precursors to early thermal runaway. Furthermore, by combining this with dynamic assessment of ambient temperature, conditions that may lead to thermal runaway can be identified earlier.

[0062] 2. By interacting with a temperature sensor, the ambient temperature within the monitoring time zone is obtained. Furthermore, the characteristic ambient temperature is determined by calculating the dispersion coefficient and lumped factor of the ambient temperature. This method takes into account the fluctuation and extreme values ​​of the ambient temperature, and more comprehensively reflects the thermal condition of the battery operating environment.

[0063] 3. Train an LSTM network using historical operating data to predict battery temperature changes. This introduction of machine learning methods improves prediction accuracy and the system's adaptability, making the battery temperature baseline curve based on real-time data more accurate, thus enhancing prediction accuracy and the ability to predict future conditions.

[0064] 4. By comparing the real-time monitored battery temperature curve with the predicted baseline temperature curve, temperature deviations exceeding the normal range can be identified in a timely manner. The graphical area parameter is used as the curve deviation coefficient to intuitively represent the degree of difference between the two curves. This can issue an early warning before the risk of thermal runaway reaches the critical point, providing sufficient time for taking countermeasures.

[0065] Example 2, Figure 3 This is a schematic diagram of the structure of an electronic device provided in Embodiment 2 of the present invention, showing a block diagram of an exemplary electronic device suitable for implementing the embodiments of the present invention. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality or scope of the embodiments of the present invention. Figure 3 As shown, the electronic device includes a processor 31, a memory 32, an input device 33, and an output device 34; the number of processors 31 in the electronic device can be one or more. Figure 3 Taking a processor 31 as an example, the processor 31, memory 32, input device 33, and output device 34 in an electronic device can be connected via a bus or other means. Figure 3 Taking the example of a connection between China and Israel via a bus.

[0066] The memory 32, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the thermal runaway fault monitoring and early warning method for new energy vehicle batteries in this embodiment of the invention. The processor 31 executes various functional applications and data processing of the electronic device by running the software programs, instructions, and modules stored in the memory 32, thereby realizing the aforementioned thermal runaway fault monitoring and early warning method for new energy vehicle batteries.

[0067] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0068] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0069] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries, characterized in that, include: When the battery of a new energy vehicle is started, the battery voltage fluctuation curve and battery current fluctuation curve are collected in a preset monitoring time zone. The preset monitoring time zone is a time interval obtained by pushing a preset time step forward from the current time. Interact with the first temperature sensor to obtain the ambient characteristic temperature of the preset monitoring time zone; Based on the battery voltage fluctuation curve, the battery current fluctuation curve, and the ambient characteristic temperature, a positive sample analysis is performed to obtain the battery temperature baseline curve. Interact with the second temperature sensor to obtain the battery temperature monitoring curve for the preset monitoring time zone; When the curve deviation coefficient between the battery temperature reference curve and the battery temperature monitoring curve is greater than or equal to the curve deviation coefficient threshold, a thermal runaway fault early warning signal is generated and sent to the user terminal.

2. The method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries as described in claim 1, characterized in that, When the battery of a new energy vehicle starts, the battery voltage fluctuation curve and battery current fluctuation curve are collected in a preset monitoring time zone, including: Obtain the battery voltage timing information and battery current timing information of the preset monitoring time zone; Neighborhood hierarchical clustering analysis is performed on the battery voltage time-series information to construct the battery voltage fluctuation curve; The battery current time-series information is subjected to neighborhood hierarchical clustering analysis to construct the battery current fluctuation curve.

3. The method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries as described in claim 2, characterized in that, The battery voltage time-series information is subjected to neighborhood hierarchical clustering analysis to construct the battery voltage fluctuation curve, including: The battery voltage timing information is obtained by first time-domain battery voltage and second time-domain battery voltage, wherein the first time domain and the second time domain are adjacent time domains; When the battery voltage deviation between the first time domain battery voltage and the second time domain battery voltage is greater than or equal to the voltage deviation threshold, the first time domain and the second time domain are merged to obtain a third time domain, and the average value analysis of the first time domain battery voltage and the second time domain battery voltage is performed to obtain the third time domain battery voltage. The cyclic analysis is performed, and when the voltage deviation of any neighborhood is less than the voltage deviation threshold, the battery voltage clustering time sequence information is output. The battery voltage fluctuation curve is constructed based on the battery voltage clustering time series information.

4. The method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries as described in claim 1, characterized in that, Interacting with a first temperature sensor to obtain the environmental characteristic temperature of the preset monitoring time zone includes: Interact with the first temperature sensor to obtain the environmental monitoring temperature time sequence information of the preset monitoring time zone; Variance statistics were performed on the time-series information of the environmental monitoring temperature to obtain the coefficient of variation of the environmental monitoring temperature. When the environmental monitoring temperature dispersion coefficient is less than or equal to the dispersion coefficient threshold, the mean value of the environmental monitoring temperature time series information is set as the environmental characteristic temperature; When the environmental monitoring temperature dispersion coefficient is greater than the dispersion coefficient threshold, the maximum value of the concentrated environmental monitoring temperature of the environmental monitoring temperature time series information is obtained and set as the environmental characteristic temperature.

5. The method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries as described in claim 4, characterized in that, The centralized environmental monitoring temperature for obtaining the time-series information of the environmental monitoring temperature includes: The first environmental monitoring temperature for which the environmental monitoring temperature time-series information is obtained; Calculate the set of temperature deviations between the environmental monitoring temperature time series information and the first environmental monitoring temperature; The temperature deviations are sorted from the set of temperature deviations in ascending order, and the average value is calculated to obtain the first local outlier parameter, which is then added to the set of local outlier parameters. Calculate the mean of the local outlier parameters in the set of local outlier parameters; Calculate the ratio of the mean of the local outlier parameter to the first local outlier parameter to obtain the first concentration factor; When the first concentration factor is greater than or equal to the concentration factor threshold, the first environmental monitoring temperature is added to the centralized environmental monitoring temperature.

6. The method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries as described in claim 1, characterized in that, When the curve deviation coefficient between the battery temperature reference curve and the battery temperature monitoring curve is greater than or equal to the curve deviation coefficient threshold, a thermal runaway fault early warning signal is generated and sent to the user, including: Unify the coordinate system of the battery temperature reference curve and the battery temperature monitoring curve, and interconnect the curve heads and curve tails of the battery temperature reference curve and the battery temperature monitoring curve to obtain the graphic area parameters. Set the area parameter of the graphic as the curve deviation coefficient.

7. The method for monitoring and early warning of thermal runaway faults in new energy vehicle batteries as described in claim 1, characterized in that, Based on the battery voltage fluctuation curve, the battery current fluctuation curve, and the ambient characteristic temperature, a positive sample analysis is performed to obtain the battery temperature baseline curve, including: Collect historical operating data of positive sample batteries for new energy vehicle battery models. The historical operating data of the batteries includes ambient temperature record data, battery voltage fluctuation time series record data, battery current fluctuation time series record data, and battery temperature time series record data. The historical operating data of the positive sample battery is cropped according to the preset time step to obtain several positive sample battery operating sample data. Using the battery temperature time-series recording data as supervision, and the ambient temperature recording data, the battery voltage fluctuation time-series recording data, and the battery current fluctuation time-series recording data as input, the long short-term memory neural network is trained by retrieving the several positive sample battery operation sample data to generate a battery time-series temperature prediction model. The battery voltage fluctuation curve, the battery current fluctuation curve, and the ambient characteristic temperature are processed according to the battery time-series temperature prediction model to obtain battery temperature prediction time-series information and construct the battery temperature reference curve.

8. An electronic device, characterized in that, The electronic device includes: Memory, used to store executable instructions; The processor, when executing executable instructions stored in the memory, implements the thermal runaway fault monitoring and early warning method for new energy vehicle batteries as described in any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the thermal runaway fault monitoring and early warning method for new energy vehicle batteries as described in any one of claims 1-7.

Images