Electric tricycle lithium battery thermal runaway early warning method and system

Abstract

The invention relates to the technical field of battery safety, in particular to an electric tricycle lithium battery thermal runaway early warning method and system, and the method comprises the steps: collecting various physical state information of a battery module through one or more self-powered wireless sensor nodes disposed in a lithium battery pack; the physical state information, the vehicle working condition information and the charging process information from the intelligent charging pile are uploaded to a cloud analysis platform through the vehicle-mounted communication terminal; the cloud analysis platform performs short-term early warning analysis, long-term trend prediction and user behavior analysis based on the received multi-source data to obtain analysis results respectively, performs weighted fusion calculation on the analysis results, and outputs a comprehensive risk score; and starting a grading early warning mechanism according to the comprehensive risk score, and sending corresponding early warning information or control instructions to a user terminal, a motorcade management platform, an intelligent charging pile and / or a fire rescue system through a cloud analysis platform. Therefore, the problems of high false alarm rate, early warning lag and the like are solved.

Description

Technical Field

[0001] This application relates to the field of battery safety technology, and in particular to a method and system for early warning of thermal runaway of lithium batteries in electric tricycles. Background Technology

[0002] Electric tricycles are an important means of public transportation and a crucial carrier for last-mile logistics, making their safety paramount. Lithium-ion batteries are widely used due to their high energy density, but thermal runaway risk remains their primary safety concern. Currently, safety protection in this field faces the following challenges: Most mainstream solutions only set fixed thresholds for voltage and temperature in the battery management system. However, thermal runaway is a complex chain reaction. When temperature or voltage becomes significantly abnormal, it has often already entered an irreversible stage. The warning window is extremely short, failing to provide sufficient time for personnel evacuation. Therefore, the warning methods are limited and delayed. Fixed thresholds cannot distinguish between normal battery operating conditions (such as after high-current discharge in summer) and actual faults. Vibration and electromagnetic interference can also easily lead to sensor misjudgments. Frequent false alarms reduce user trust, causing users to eventually ignore alarms, resulting in a high false alarm rate. To monitor the internal state, a large number of wired temperature sensors are usually placed inside the battery pack. The numerous wiring harnesses not only increase costs, process complexity, and potential failure points, but more importantly, in extreme cases, damaged cables can become pathways for fire to spread, "bringing fire into the room" and creating new safety hazards.

[0003] Therefore, there is an urgent need in this field for a new solution that provides early warning, high accuracy, and a systematic approach to overcome the safety challenges of lithium batteries in electric tricycles. Summary of the Invention

[0004] This application provides a method and system for early warning of thermal runaway of lithium batteries in electric tricycles, in order to solve problems such as high false alarm rate and delayed early warning.

[0005] The first aspect of this application provides a method for early warning of thermal runaway of lithium battery in electric tricycles, including the following steps: S1, collecting various physical state information of the battery module by one or more self-powered wireless sensor nodes deployed inside the lithium battery pack; S2. Upload the physical status information, vehicle operating condition information, and charging process information from the smart charging pile to the cloud analysis platform via the vehicle-mounted communication terminal. S3. The cloud-based analysis platform performs short-term early warning analysis, long-term trend prediction and user behavior analysis based on the received multi-source data to obtain analysis results, and then performs weighted fusion calculation of the above analysis results to output a comprehensive risk score. S4. Based on the comprehensive risk score, activate the graded early warning mechanism and send corresponding early warning information or control instructions to user terminals, fleet management platforms, smart charging piles and / or fire and rescue systems through the cloud analysis platform.

[0006] Optionally, the various physical state information includes: temperature data, pressure data, vibration data, and gas data; the vehicle operating condition information includes: electrical data, operating status data, and time and status data; and the charging process information includes: charging curve data and charger information data.

[0007] Optionally, the short-term early warning analysis includes: Receive real-time physical status information and vehicle operating condition information, and preprocess and extract features from the information; Determine the battery status and dynamically weight the extracted features based on the battery status; The weighted feature vector is input into the trained short-term early warning model, and the model outputs a real-time risk probability value.

[0008] Optionally, the long-term trend forecast includes: Receive the charging process information and select the voltage and capacity data of the constant current charging stage from the information; The capacity is numerically differentiated with respect to the voltage, and an incremental capacity curve is plotted. Based on the peak value on the incremental capacity curve, feature tracking is performed and compared with the baseline curve to calculate the peak value attenuation amplitude and peak position offset of the feature peak. Based on the above calculation results, the latest battery health status estimate and degradation trend are output through the prediction model.

[0009] Optionally, the user behavior analysis includes: Historical behavior baselines and recent behavior data are established based on historical and real-time vehicle operating condition information and charging process information. Compare the historical behavior baseline with the recent behavior data rows and calculate their statistical deviation. The abnormal pattern is determined based on the statistical deviation. The behavioral analysis model outputs behavioral risk labels and specific behavioral descriptions.

[0010] Optionally, the weighted fusion calculation of the analysis results to output a comprehensive risk score includes: Receive short-term early warning analysis, long-term trend prediction, and user behavior analysis results; The analysis results are assigned weights according to the preset dynamic weighting rules, and the final comprehensive risk score is calculated.

[0011] Optionally, the tiered early warning mechanism includes: Level 1 early warning, Level 2 early warning, and Level 3 early warning, wherein, The first-level early warning is to push maintenance prompt information to the user terminal when the comprehensive risk score is lower than the first threshold. The secondary warning is to send a strong reminder warning to the user terminal and send an instruction to the smart charging pile to limit the subsequent charging power when the comprehensive risk score is between the first threshold and the second threshold. The three-level early warning system sends the highest-level alarm to the user terminal and fleet management platform when the comprehensive risk score is higher than the second threshold. It remotely cuts off the high-voltage power to the vehicle, prohibits all charging piles from charging it, and pushes the early warning information and location information to the fire and rescue system.

[0012] A second aspect of this application provides a lithium battery thermal runaway early warning system for an electric tricycle, comprising: a vehicle-side monitoring unit, a cloud-based analysis platform, and an interaction and execution unit, wherein... The vehicle-side monitoring unit includes a self-powered wireless sensor node and an on-board communication terminal. The self-powered wireless sensor node is used to collect various physical state information of the battery module, and the on-board communication terminal is used for long-distance communication and transmission with the cloud analysis platform. The cloud-based analytics platform is used to run short-term early warning models, long-term prediction models, and behavioral analysis models, outputting a comprehensive risk score and activating a tiered early warning mechanism. The interaction and execution unit includes a user terminal APP, a fleet management platform, a smart charging pile, and an interface for linkage with the fire and rescue system, which is used to send corresponding early warning information or control commands through the interface.

[0013] A third aspect of this application provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being configured to perform a method for early warning of thermal runaway of a lithium battery in an electric tricycle as described in the above embodiments.

[0014] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to implement a method for early warning of thermal runaway of a lithium battery in an electric tricycle as described in the above embodiments.

[0015] Therefore, this application has at least the following beneficial effects: This application's embodiments utilize one or more self-powered wireless sensors deployed within a lithium battery pack to form a self-powered wireless sensor network. This fundamentally eliminates the short-circuit risk associated with additional wiring within the battery pack, offering simple installation, extremely high safety, and significant novelty and practicality. By deeply fusing and analyzing physical state information, vehicle operating condition information, and charging process information from smart charging piles in the cloud, a three-dimensional, multi-dimensional evaluation system is constructed, resulting in a significantly higher accuracy and earlier warning rate compared to traditional methods. Short-term warning models, long-term prediction models, and behavioral analysis models are deployed in the cloud, with the vehicle only needing to perform basic data collection and communication, greatly reducing the cost and complexity of a single-vehicle BMS and enabling large-scale deployment. The system not only provides "early warnings" but also achieves "active intervention" through linkage with charging piles and vehicle controllers, forming a complete "prediction-early warning-protection" closed loop, greatly improving the system's effectiveness. This solves technical problems such as high false alarm rates and delayed warnings.

[0016] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0017] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart illustrating a thermal runaway warning method for a lithium battery in an electric tricycle, according to an embodiment of this application. Figure 2 This is a schematic diagram of the structure of a lithium battery thermal runaway early warning system for an electric tricycle according to an embodiment of this application; Figure 3 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application. Detailed Implementation

[0018] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0019] The following description, with reference to the accompanying drawings, illustrates an embodiment of a method and system for early warning of thermal runaway of a lithium battery in an electric tricycle. Addressing the issues of high false alarm rate and delayed warning mentioned in the background section, this application provides a method for early warning of thermal runaway of a lithium battery in an electric tricycle. This method utilizes one or more self-powered wireless sensors deployed inside the lithium battery pack to form a self-powered wireless sensor network, fundamentally eliminating the short-circuit risk caused by additional wiring within the battery pack. It is easy to install, highly safe, and possesses novelty and practicality. By deeply fusing and analyzing physical state information, vehicle operating condition information, and charging process information from smart charging piles in the cloud, a three-dimensional, multi-dimensional evaluation system is constructed, resulting in a warning accuracy and early warning level far exceeding traditional methods. By deploying short-term warning models, long-term prediction models, and behavioral analysis models in the cloud, the vehicle only needs to complete basic data collection and communication, significantly reducing the cost and complexity of a single-vehicle BMS and enabling large-scale deployment. The system not only provides "early warning" but also achieves "active intervention" through linkage with charging piles and vehicle controllers, forming a complete "prediction-early warning-protection" closed loop, greatly improving the system's effectiveness. This solved the technical problems of high false alarm rate and delayed early warning.

[0020] Specifically, Figure 1 This is a flowchart illustrating a method for early warning of thermal runaway of lithium batteries in electric tricycles, provided in an embodiment of this application.

[0021] like Figure 1 As shown, the method for early warning of thermal runaway of lithium battery in electric tricycles includes the following steps: In step S1, various physical state information of the battery module is collected by one or more self-powered wireless sensor nodes deployed inside the lithium battery pack.

[0022] The information includes various physical state data such as temperature, pressure, vibration, and gas; vehicle operating condition data such as electrical data, operating status data, and time and status data; and charging process information such as charging curve data and charger information data.

[0023] Specifically, temperature data includes the temperature value of a single battery cell or battery module, the maximum temperature value, the temperature gradient, and the rate of temperature change; pressure data includes the internal air pressure or pressure value and the rate of pressure change; vibration data includes vibration acceleration and vibration spectrum; and gas data includes the concentration of a specific gas.

[0024] It should be noted that physical state information refers to raw data reflecting the battery's own physical characteristics, directly collected by sensors deployed inside the battery pack. This information is the most direct early signal of thermal runaway. Temperature data directly reflects heat generation; internal short circuits, overcharging, and over-discharging can all cause abnormal temperature increases, and rapid temperature rise is a hallmark of a thermal runaway chain reaction. Pressure data is detected because side reactions (such as electrolyte decomposition) in the battery generate gas, leading to battery bulging and increased internal pressure; a sudden increase in pressure is an important precursor to thermal runaway. Vibration data can diagnose changes in mechanical structure and identify abuse conditions, as faults such as internal short circuits can cause slight changes in the internal structure, altering its vibration response characteristics.

[0025] Specifically, electrical data includes: total output voltage of the battery pack, charging and discharging current, remaining battery capacity, and battery health status; operational status data includes: vehicle speed; mileage; gear information and accelerator pedal opening; time and status data includes: timestamps, i.e., the exact time the data was generated, and vehicle status indicators, such as stationary, driving, charging, and off.

[0026] It should be noted that vehicle operating condition information refers to data obtained through the vehicle controller that reflects the vehicle's operating status and the battery's electrical status. This information is used to determine the battery's operating environment. Electrical data can determine the battery's basic electrical state. Overvoltage, overcurrent, and over-discharge are common causes of thermal runaway. The battery's remaining charge state and battery health state are the basic background for risk assessment. Operating status data is used to determine the battery load. High-current discharge (such as rapid acceleration or hill climbing) will cause normal temperature rise. The system must distinguish this from abnormal temperature rise to avoid false alarms. Time and status data provide a time reference and status background for data analysis. Under different states, the system adopts different analysis strategies and alarm thresholds.

[0027] Specifically, the charging curve data includes: voltage time-series data, current time-series data, and capacity data; the charging information includes the charging pile ID, set charging power, set charging current, and charging start / stop time. Among them, the voltage time-series data is the curve of voltage changing over time throughout the entire charging process; the current time-series data is the curve of current changing over time throughout the entire charging process; and the capacity data is the amount of electricity already charged.

[0028] It's important to note that charging process information specifically refers to high-precision, high-sampling-rate time-series data collected by smart charging stations or on-board chargers during vehicle charging. This data is used for in-depth analysis of battery health degradation trends. Charging curve data serves as the raw data for incremental capacity analysis. By analyzing the differential of voltage with respect to capacity during the constant-current charging phase, characteristic peaks reflecting changes in the battery's internal electrochemical performance can be obtained, enabling precise assessment of health and degradation trends. Charger information is used to identify charging stations and determine the compliance of charging practices. Overall, charging process information is used for "regular in-depth health checks" of the battery, providing insights into aging trends at the electrochemical level and enabling long-term risk prediction.

[0029] In step S2, the physical status information, vehicle operating condition information, and charging process information from the smart charging pile are uploaded to the cloud analysis platform through the vehicle communication terminal.

[0030] In step S3, the cloud-based analytics platform performs short-term early warning analysis, long-term trend prediction, and user behavior analysis based on the received multi-source data to obtain analysis results. The platform then performs weighted fusion calculations on the above analysis results to output a comprehensive risk score.

[0031] The short-term early warning analysis includes: receiving real-time physical state information and vehicle operating condition information, preprocessing the information and extracting features; determining the battery status and dynamically weighting the extracted features according to the battery status; inputting the weighted feature vector into the trained short-term early warning model, and the model outputs a real-time risk probability value.

[0032] Specifically, the goal of short-term early warning analysis is to determine whether the battery is entering a chain reaction of thermal runaway within the second to minute timeframe. The specific process involves: receiving real-time physical state information and vehicle operating condition information, and aligning the received multi-sensor data with timestamps, handling lost and outlier values; calculating key features from the raw data, including temporal features, spatial features, frequency domain features, and statistical features. Temporal features include calculating the rate of temperature change and the rate of pressure change; spatial features include calculating the maximum temperature difference (temperature gradient) between different monitoring points within the battery pack; frequency domain features involve performing a fast Fourier transform on the vibration signal and analyzing its spectrum; and statistical features include calculating the mean and variance within a sliding time window.

[0033] The system determines the current state of the battery (stationary, driving, charging, or just finished a high-current discharge). Based on the determined state, the extracted features are dynamically weighted. For example, when the vehicle is stationary, any slight temperature rise or pressure change is highly suspicious, and the weights of the temperature change rate and pressure change rate are set to the highest, with the thresholds set to be extremely sensitive. When the battery has just finished a high-current discharge, it is normal for the overall battery temperature to be high. The system pays more attention to the temperature gradient (whether there is local overheating) and the rate of temperature decrease (whether the heat dissipation is normal). At this time, the threshold for absolute temperature is temporarily relaxed, but the weight of gradient features is increased.

[0034] The weighted feature vector is input into a pre-trained short-term early warning model, and the model outputs a real-time risk probability value, representing the possibility of thermal runaway in the current state.

[0035] Understandably, short-term early warning analysis offers strong real-time capabilities, enabling it to detect sudden faults and buy valuable time for personnel evacuation. Short-term early warning analysis can achieve second-level response, directly addressing the chain reaction of thermal runaway. The short-term early warning analysis model is the only one capable of responding to ongoing thermal runaway (such as a sudden and drastic temperature rise caused by an internal short circuit) within seconds or minutes. It focuses on "what is happening right now," serving as the last and most critical technical line of defense against disaster. This model integrates multi-dimensional information such as temperature change rate, pressure change rate, temperature gradient, and vibration spectrum, combined with vehicle status (such as whether a high-current discharge has just ended) for dynamic threshold judgment, effectively distinguishing between "normal operating condition heat generation" and "abnormal fault heat generation," with accuracy far exceeding single-parameter early warning. Due to the use of a distributed sensor array, when an alarm occurs, the system can roughly pinpoint which module or area first experienced an anomaly, providing crucial information for subsequent maintenance and emergency response.

[0036] The long-term trend prediction includes: receiving charging process information and selecting voltage and capacity data from the constant current charging stage; numerically differentiating the capacity with respect to the voltage and plotting an incremental capacity curve; performing feature tracking based on the peak value on the incremental capacity curve and comparing it with a reference curve to calculate the characteristic peak attenuation magnitude and characteristic peak position offset; and outputting the latest battery health estimate and degradation trend through a prediction model based on the above calculation results.

[0037] Specifically, the goal of long-term trend prediction is to predict the degradation trend of battery health and identify potential risks on a daily to weekly scale. The process involves receiving charging process information and selecting voltage and capacity data from the constant-current charging phase. Constant-current charging ensures data consistency and facilitates analysis. The capacity (Q) in the data is numerically differentiated with respect to voltage (V), i.e., dQ / dV is calculated, and a dQ / dV-V curve, i.e., the incremental capacity curve, is plotted. Each peak on this curve corresponds to a specific electrochemical reaction within the battery.

[0038] Characteristic peaks on the curve are identified and compared with the baseline curve at its new state to accurately calculate the peak value decay magnitude and peak position shift. The peak value decay magnitude corresponds to the loss of active material, while the peak position shift corresponds to the increase in internal resistance and the degradation of reaction kinetics. Based on these changes, an algorithmic model calculates the latest and more accurate battery health value after this charge. Finally, a predictive model outputs the latest battery health estimate and degradation trend.

[0039] Understandably, long-term trend prediction, with its strong foresight, can identify potential slow-deterioration risks and enable predictive maintenance. Thermal runaway is not always sudden; many are the result of long-term, slow battery performance degradation until it reaches a critical point. Long-term trend prediction models, through in-depth analysis of charging data (such as incremental capacity analysis), can detect trends weeks or months in advance when battery performance shows slight degradation but has not yet exhibited external fault characteristics, issuing an early warning of "declining battery health." This represents a leap from "post-event warnings" to "pre-event predictions." Its output SOH (State of Health) estimate is an extremely valuable indicator, allowing users or fleet managers to clearly understand the battery's remaining lifespan and potential risk level, providing a scientific basis for decisions such as battery replacement and secondary use. Because the analysis object is a relatively stable charging curve, its results are not affected by complex operating conditions such as vibrations during driving or instantaneous high currents, ensuring stable and reliable analysis results.

[0040] The user behavior analysis includes: establishing historical behavior baselines and recent behavior data based on historical and real-time vehicle operating condition information and charging process information; comparing the historical behavior baselines and recent behavior data to calculate their statistical deviation; determining abnormal patterns based on the statistical deviation; and outputting behavior risk labels and specific behavior descriptions through the behavior analysis model.

[0041] The process involves weighted fusion of analysis results to output a comprehensive risk score, which includes receiving short-term early warning analysis, long-term trend prediction, and user behavior analysis results; assigning weights to the above analysis results according to preset dynamic weighting rules; and calculating the final comprehensive risk score.

[0042] Specifically, the goal of user behavior analysis is to identify users' poor usage habits, which are indirect causes of accelerated battery aging or thermal runaway. The specific process is as follows: Based on historical and real-time vehicle operating condition information and charging process information, a historical behavior baseline and recent behavior data are established. The historical behavior baseline is the statistical distribution of data such as typical charging power, common charging duration, typical depth of discharge, and average vehicle speed of the user / vehicle over a past period (e.g., 30 days); the recent behavior data is the charging power, charging duration, depth of discharge, and other data from the most recent one or several times.

[0043] The historical behavioral baseline and recent behavioral data are compared to calculate the statistical deviation, which is the statistical distance between the recent behavioral data points and the historical behavioral baseline. The higher the calculated deviation, the more abnormal the recent behavior. The deviation is a quantified deviation score.

[0044] The calculated deviation score is compared with preset thresholds. If the deviation score is greater than the first threshold, it is marked as suspicious; if the deviation score is greater than the second threshold, it is marked as abnormal. The first threshold is less than the second threshold. Simultaneously, the system examines which feature dimensions or dimensions caused the high deviation, thus determining the specific abnormal pattern. Finally, the behavioral analysis model outputs a behavioral risk label and a detailed behavioral description.

[0045] Understandably, user behavior analysis, focusing on the root causes, can correct usage behaviors that lead to battery damage, eliminating risks at the source. The vast majority of battery safety incidents stem from misuse, such as using substandard chargers, over-discharging, and prolonged storage in a state of low charge. This model does not directly monitor the battery, but rather how the "person" uses it. It can accurately identify these bad habits and issue corrective instructions before the battery is damaged, preventing it from entering a dangerous state at its source. The system establishes an independent behavior profile for each user, with personalized criteria for judging "normal" and "abnormal." Different measurement scales are used for cautious and careless users, making the solution highly adaptable and accurate. Correcting user behavior through algorithms is a low-cost yet highly efficient safety measure, achieving low-cost, high-efficiency risk intervention compared to hardware upgrades.

[0046] In step S4, based on the comprehensive risk score, a graded early warning mechanism is activated, and corresponding early warning information or control instructions are sent to user terminals, fleet management platforms, smart charging piles and / or fire and rescue systems through the cloud analysis platform.

[0047] The tiered early warning mechanism includes three levels: Level 1, Level 2, and Level 3. Level 1 warning sends a maintenance reminder to the user terminal when the comprehensive risk score is below the first threshold. Level 2 warning sends a strong reminder warning to the user terminal and sends instructions to the smart charging pile to limit subsequent charging power when the comprehensive risk score is between the first and second thresholds. Level 3 warning sends the highest level alarm to the user terminal and the fleet management platform when the comprehensive risk score is above the second threshold, remotely cuts off the vehicle's high-voltage power, prohibits all charging piles from charging the vehicle, and pushes the warning information and location information to the fire and rescue system.

[0048] In some embodiments, Lao Wang is a deliveryman driving an electric tricycle equipped with an intelligent early warning system. The battery has been used for two years, and the cloud records its state of health (SOH) at 78%, indicating slight aging but still within the normal range of use.

[0049] One evening, after finishing work, Lao Wang parked his vehicle in the warehouse to charge it. The charging station was charging normally.

[0050] The system analyzed the charging curve using a long-term trend prediction model and found that the State of Harmony (SOH) decreased slightly from 78% to 77.5%, showing a stable trend with no significant risk. However, a user behavior analysis model revealed that Mr. Wang's discharge depth reached 90% (almost completely draining the battery), far exceeding his usual baseline of 70%, thus marking his behavior as "suspicious." The cloud-based decision center calculated a comprehensive risk score, concluding that the long-term risk was low, the short-term risk was 0, but the behavioral risk was amplified, resulting in a comprehensive risk score of 35 (assuming the first threshold is set at 40).

[0051] The system triggered a Level 1 warning, and Mr. Wang received a push notification on his mobile app: "[Maintenance Tip] Your battery has discharged deeply this time. It is recommended to avoid frequently draining the battery completely to extend its lifespan." Mr. Wang glanced at the notification but didn't pay much attention to it, and the vehicle continued charging.

[0052] The next morning, Lao Wang prepared to go to work and found that his vehicle was fully charged. He started the vehicle and began delivery. The system, through a short-term early warning analysis model, detected a slight anomaly in the temperature change rate in the middle of the battery pack, and that the temperature at one point was about 5°C higher than the surrounding area (abnormal temperature gradient); the long-term trend prediction model and the user behavior analysis model remained unchanged. The cloud-based decision center recalculated the score, concluding that the probability of short-term risk had increased, and that battery aging amplified the weight of short-term risk, resulting in a comprehensive risk score of 65 points (between the first threshold of 40 points and the second threshold of 80 points).

[0053] Triggering a level-two warning, Lao Wang's app suddenly emitted a rapid alarm and a pop-up window: "Warning! Abnormal battery temperature. Please stop immediately at a safe location to check!" Simultaneously, the cloud sent instructions to all charging stations to add the vehicle's ID to the "restricted charging list." The instructions stipulated that the maximum charging current for this vehicle on its next charge must not exceed 5A (standard charging is 20A). This measure aims to prevent high-power charging from potentially exacerbating existing battery hazards. Lao Wang became alert upon hearing the alarm, but seeing that the vehicle seemed to be operating normally and being in a hurry to deliver packages, he ignored the warning and continued driving.

[0054] Another hour passed, and Lao Wang was driving. The short-term early warning analysis model detected that the temperature at the previous anomaly point began to rise sharply, and the pressure sensor simultaneously detected a rapid increase in internal pressure. Multiple sensors cross-verified and confirmed that the thermal runaway chain reaction had begun; the comprehensive risk score instantly soared to 92 points (far exceeding the second threshold of 80 points).

[0055] Upon triggering a Level 3 warning, the cloud platform executes the following series of operations within seconds: The user terminal and fleet management platform send the highest-level audible and visual alarms to Mr. Wang and the courier company's fleet manager's apps, clearly stating: "Critical! Battery thermal runaway danger! Please evacuate immediately!" The vehicle controller remotely sends a command to forcibly disconnect the vehicle's high-voltage power. The vehicle may suddenly stop or slide to the roadside to ensure personnel safety and prevent further deterioration of the electrical system. The charging network broadcasts a command to all smart charging stations in the area, completely prohibiting charging of the vehicle to prevent reconnection to power in a faulty state. The fire and rescue system automatically pushes the warning information, the vehicle's precise GPS location, battery type (e.g., ternary lithium), license plate number, and owner's contact information to the 119 command center via a data interface. The rescue center can immediately dispatch fire trucks and, having been informed in advance, prepare for handling the lithium battery fire.

[0056] Understandably, this tiered mechanism ensures that response measures are strictly matched to the level of risk, avoiding both overreacting (frequent false alarms desensitize users) and missing crucial opportunities (missed reports lead to accidents). Level 1 warnings provide educational and preventative intervention through gentle reminders; Level 2 warnings proactively intervene through strong warnings and restrictions, attempting to prevent the situation from escalating; Level 3 warnings, when danger is unavoidable, involve the system taking over control, prioritizing personnel safety, and coordinating with social emergency response forces to minimize losses.

[0057] The lithium battery thermal runaway early warning method for electric tricycles proposed in this application eliminates the short-circuit risk caused by additional wiring inside the battery pack by deploying one or more self-powered wireless sensors inside the battery pack to form a self-powered wireless sensor network. This method is simple to install, highly safe, and novel and practical. It deeply integrates and analyzes physical state information, vehicle operating condition information, and charging process information from smart charging piles in the cloud, constructing a three-dimensional, multi-dimensional evaluation system. The early warning accuracy and early warning level far exceed traditional methods. By deploying short-term early warning models, long-term prediction models, and behavioral analysis models in the cloud, the vehicle only needs to complete basic data collection and communication, significantly reducing the cost and complexity of a single-vehicle BMS and enabling large-scale deployment. The system not only provides early warning but also achieves proactive intervention through linkage with charging piles and vehicle controllers, forming a complete "prediction-early warning-protection" closed loop, greatly improving the system's effectiveness. This solves the technical problems of high false alarm rate and delayed early warning.

[0058] Next, referring to the accompanying drawings, an electric tricycle lithium battery thermal runaway early warning system is described according to an embodiment of this application.

[0059] Figure 2 This is a schematic diagram of the structure of an electric tricycle lithium battery thermal runaway early warning system according to an embodiment of this application.

[0060] like Figure 2 As shown, the electric tricycle lithium battery thermal runaway early warning system 10 includes: a vehicle-side monitoring unit 100, a cloud-based analysis platform 200, and an interaction and execution unit 300.

[0061] The vehicle-side monitoring unit 100 includes a self-powered wireless sensor node and an on-board communication terminal. The self-powered wireless sensor node is used to collect various physical state information of the battery module, and the on-board communication terminal is used for long-distance communication and transmission with the cloud analysis platform. The cloud analysis platform 200 is used to run short-term early warning models, long-term prediction models, and behavioral analysis models, output a comprehensive risk score, and activate a graded early warning mechanism. The interaction and execution unit includes a user terminal APP, a fleet management platform, a smart charging pile, and an interface for linkage with the fire and rescue system, which is used to send corresponding early warning information or control commands through the interface.

[0062] It should be noted that the foregoing explanation of the embodiment of the thermal runaway warning method for lithium batteries of electric tricycles also applies to the thermal runaway warning system for lithium batteries of electric tricycles in this embodiment, and will not be repeated here.

[0063] The lithium battery thermal runaway early warning system for electric tricycles proposed in this application eliminates the short-circuit risk caused by additional wiring inside the battery pack by deploying one or more self-powered wireless sensors within the battery pack, forming a self-powered wireless sensor network. This simplifies installation, provides high safety, and demonstrates novelty and practicality. The system deeply integrates and analyzes physical state information, vehicle operating condition information, and charging process information from smart charging piles in the cloud, constructing a three-dimensional, multi-dimensional evaluation system. The early warning accuracy and timing far exceed traditional methods. By deploying short-term early warning models, long-term prediction models, and behavioral analysis models in the cloud, the vehicle only needs to perform basic data collection and communication, significantly reducing the cost and complexity of a single-vehicle BMS and enabling large-scale deployment. The system not only provides early warning but also achieves proactive intervention through linkage with charging piles and vehicle controllers, forming a complete "prediction-early warning-protection" closed loop, greatly improving the system's effectiveness. This solves technical problems such as high false alarm rates and delayed early warnings.

[0064] Figure 3 This is a schematic diagram of the structure of a vehicle provided in an embodiment of this application. The electronic device may include: The memory 301, the processor 302, and the computer program stored on the memory 301 and capable of running on the processor 302.

[0065] When the processor 302 executes the program, it implements the method for early warning of thermal runaway of lithium battery in electric tricycle provided in the above embodiment.

[0066] Furthermore, the vehicle also includes: Communication interface 303 is used for communication between memory 301 and processor 302.

[0067] The memory 301 is used to store computer programs that can run on the processor 302.

[0068] The memory 301 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0069] If the memory 301, processor 302, and communication interface 303 are implemented independently, then the communication interface 303, memory 301, and processor 302 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 3 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0070] Optionally, in a specific implementation, if the memory 301, processor 302, and communication interface 303 are integrated on a single chip, then the memory 301, processor 302, and communication interface 303 can communicate with each other through an internal interface.

[0071] Processor 302 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0072] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the above-described method for early warning of thermal runaway of lithium batteries in electric tricycles.

[0073] This application also provides a computer program product, which stores a computer program that, when executed by a processor, implements the above-described method for early warning of thermal runaway of lithium batteries in electric tricycles.

[0074] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0075] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0076] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0077] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0078] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

Claims

1. A method for early warning of thermal runaway of lithium batteries in electric tricycles, characterized in that, Includes the following steps: S1. Collect various physical state information of the battery module by deploying one or more self-powered wireless sensor nodes inside the lithium battery pack; S2. Upload the physical status information, vehicle operating condition information, and charging process information from the smart charging pile to the cloud analysis platform via the vehicle-mounted communication terminal. S3. The cloud-based analysis platform performs short-term early warning analysis, long-term trend prediction, and user behavior analysis based on the received multi-source data to obtain analysis results, and then performs weighted fusion calculation on the above analysis results to output a comprehensive risk score. S4. Based on the comprehensive risk score, activate the graded early warning mechanism and send corresponding early warning information or control instructions to user terminals, fleet management platforms, smart charging piles and / or fire and rescue systems through the cloud analysis platform.

2. The method for early warning of thermal runaway of lithium battery in an electric tricycle according to claim 1, characterized in that, The various physical state information includes: temperature data, pressure data, vibration data, and gas data; the vehicle operating condition information includes: electrical data, operating status data, and time and status data; the charging process information includes: charging curve data and charger information data.

3. The method for early warning of thermal runaway of lithium battery in electric tricycles according to claim 1, characterized in that, The short-term early warning analysis includes: Receive real-time physical status information and vehicle operating condition information, and preprocess and extract features from the information; Determine the battery status and dynamically weight the extracted features based on the battery status; The weighted feature vector is input into the trained short-term early warning model, and the model outputs a real-time risk probability value.

4. The method for early warning of thermal runaway of lithium battery in electric tricycles according to claim 1, characterized in that, The long-term trend forecast includes: Receive the charging process information and select the voltage and capacity data of the constant current charging stage from the information; The capacity is numerically differentiated with respect to the voltage, and an incremental capacity curve is plotted. Based on the peak value on the incremental capacity curve, feature tracking is performed and compared with the baseline curve to calculate the peak value attenuation amplitude and peak position offset of the feature peak. Based on the above calculation results, the latest battery health status estimate and degradation trend are output through the prediction model.

5. The method for early warning of thermal runaway of lithium battery in electric tricycles according to claim 1, characterized in that, The user behavior analysis includes: Historical behavior baselines and recent behavior data are established based on historical and real-time vehicle operating condition information and charging process information. Compare the historical behavior baseline with the recent behavior data rows and calculate their statistical deviation. The abnormal pattern is determined based on the statistical deviation. The behavioral analysis model outputs behavioral risk labels and specific behavioral descriptions.

6. The method for early warning of thermal runaway of lithium battery in an electric tricycle according to claim 1, characterized in that, The weighted fusion calculation of the above analysis results to output a comprehensive risk score includes: Receive short-term early warning analysis, long-term trend prediction, and user behavior analysis results; The analysis results are assigned weights according to the preset dynamic weighting rules, and the final comprehensive risk score is calculated.

7. The method for early warning of thermal runaway of lithium battery in an electric tricycle according to claim 1, characterized in that, The tiered early warning mechanism includes: Level 1 early warning, Level 2 early warning, and Level 3 early warning, wherein... The first-level early warning is to push maintenance prompt information to the user terminal when the comprehensive risk score is lower than the first threshold. The secondary warning is to send a strong reminder warning to the user terminal and send an instruction to the smart charging pile to limit the subsequent charging power when the comprehensive risk score is between the first threshold and the second threshold. The three-level early warning system sends the highest-level alarm to the user terminal and fleet management platform when the comprehensive risk score is higher than the second threshold. It remotely cuts off the high-voltage power to the vehicle, prohibits all charging piles from charging it, and pushes the early warning information and location information to the fire and rescue system.

8. A thermal runaway early warning system for lithium batteries in electric tricycles according to any one of claims 1-7, characterized in that, include: The system comprises a vehicle-side monitoring unit, a cloud-based analytics platform, and an interaction and execution unit. The vehicle-side monitoring unit includes a self-powered wireless sensor node and an on-board communication terminal. The self-powered wireless sensor node is used to collect various physical state information of the battery module, and the on-board communication terminal is used for long-distance communication and transmission with the cloud analysis platform. The cloud-based analytics platform is used to run short-term early warning models, long-term prediction models, and behavioral analysis models, outputting a comprehensive risk score and activating a tiered early warning mechanism. The interaction and execution unit includes a user terminal APP, a fleet management platform, a smart charging pile, and an interface for linkage with the fire and rescue system, which is used to send corresponding early warning information or control commands through the interface.

9. An electronic device, characterized in that, include: The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement a method for early warning of thermal runaway of a lithium battery for an electric tricycle as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement a method for early warning of thermal runaway of lithium battery in electric tricycles as described in any one of claims 1-7.

Images