A safety monitoring system and method for new energy vehicle batteries

Abstract

The present invention discloses a safety monitoring system and method for new energy vehicle batteries, including a conventional vibration monitoring module, an impact capture module, a cross-validation module, a cumulative impact module, an early warning module, and a storage and learning module. The conventional vibration monitoring module obtains a vibration feature baseline through multi-scale analysis, and the impact capture module captures instantaneous impacts and extracts features such as peak value, duration, and energy concentration. If it is determined that there is a loosening risk, the cross-validation module further collects electromagnetic field, acoustic and temperature information and comprehensively evaluates whether substantial loosening occurs; if there is no loosening, the cumulative impact module calculates the remaining impact value and accumulates it into the historical impact parameter to capture the fatigue effect caused by the accumulation of multiple small impacts. The storage and learning module records each impact and evaluation result and optimizes the threshold through a self-learning algorithm, thereby improving the detection accuracy and safety of the loosening risk of the battery fixing structure.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy vehicles, and in particular to a safety monitoring system and method for new energy vehicle batteries. Background Art

[0002] With the increasing popularity and expansion of new energy vehicles, power batteries have become a critical component affecting vehicle safety and performance. Existing new energy vehicles typically secure the battery pack to the chassis or body support frame using multiple bolts or bracket structures. However, under various operating conditions, such as long-term driving, frequent acceleration and deceleration, and repeated bumps and impacts, the battery mounting structure may be subjected to various forms of vibration and impact.

[0003] If fatigue or loosening of connecting components is not detected in advance, the following technical risks may arise: Under repeated impact loads or road bumps, some bolts, brackets, or clips may experience preload loss and loosening, resulting in uneven load bearing and increased vibration amplitude of the battery pack during vehicle operation. This not only affects the NVH characteristics of the entire vehicle, but may also cause stress concentration and potential cracks in the battery housing, frame structure, and even chassis. If the battery module and connection parts become loose, in addition to mechanical structural instability, it may also cause poor contact or increased contact resistance in local electrical connections, further leading to local overheating, noise interference, or accelerated component aging, posing a hidden danger to the overall safety of the vehicle. Conventional vehicle vibration monitoring methods mostly only obtain overall acceleration or simple diagnostic signals. It is difficult to accurately distinguish whether the loosening critical point is approaching when multiple small impacts accumulate or multiple medium impacts are experienced in a short period of time. Traditional single-threshold judgment modes are prone to omissions or misjudgments in extreme road conditions (such as potholes and concentrated speed bumps). At the same time, if risk assessment relies solely on a single physical signal (such as vibration acceleration), it will not be possible to effectively capture trace signs of loosening in a complex vibration background.

[0004] Based on the above status quo and problems, it is necessary to develop a more refined and real-time risk monitoring system for new energy vehicle battery fixing structures. Summary of the Invention

[0005] The purpose of the present invention is to provide a safety monitoring system for new energy vehicle batteries, which has the ability to predict the risk of loosening of the battery fixing structure earlier and more accurately, ensuring that the vehicle maintains high safety and durability in a long-term, multi-operating environment.

[0006] The above technical objectives of the present invention are achieved through the following technical solutions:

[0007] A safety monitoring system for new energy vehicle batteries, comprising:

[0008] A conventional vibration monitoring module, which is used to collect background vibration data at the battery fixing structure during vehicle driving and perform trend analysis on the background vibration data;

[0009] An impact capture module is used to capture instantaneous impacts and, when a transient impact signal is detected, extract transient vibration data within a preset period before and after the instantaneous impact signal is generated, and combine the transient vibration data with the background vibration trend to make a preliminary judgment on the loosening risk;

[0010] A cross-validation module, comprising a cross-collection submodule and a comprehensive judgment submodule. The cross-collection submodule is configured to collect cross-judgment data at the battery fixing structure when the preliminary judgment result of the loosening risk is that there is a loosening risk. The comprehensive judgment submodule comprehensively judges whether substantial loosening occurs based on the cross-judgment data.

[0011] A cumulative impact module, which is used to calculate the residual impact value when the comprehensive judgment result is that no substantial loosening has occurred, and convert the residual impact value into a cumulative impact parameter. The cumulative impact parameter is used to add the impact of the superposition effect caused by multiple historical shocks when making a preliminary judgment on the subsequent loosening risk;

[0012] An early warning module, configured to trigger an early warning signal when a comprehensive judgment indicates that substantial loosening has occurred or when the cumulative impact parameter exceeds a substantial risk threshold;

[0013] The storage and learning module is used to store the background vibration data, instantaneous impact signals, comprehensive judgment results, and the evolution of cumulative impact parameters when each instantaneous impact signal is generated, and to optimize the actual risk threshold through data analysis or self-learning algorithm to obtain a new actual risk threshold.

[0014] By adopting the above-mentioned technical solution, through the organic combination of conventional vibration monitoring modules, impact capture modules, cross-validation modules, cumulative impact modules, early warning modules, and storage and learning modules, the system can realize real-time detection and comprehensive assessment of the loosening risk of the battery fixing structure under various working conditions; it can not only capture instantaneous impacts, but also record and manage the cumulative impact of multiple light and moderate impacts, thereby improving the detection rate and judgment accuracy of potential loosening faults of the battery fixing structure, and ensuring the safe operation of new energy vehicles.

[0015] It is further provided that: the conventional vibration monitoring module includes a multi-scale vibration analysis unit and a characteristic baseline generation unit;

[0016] The multi-scale vibration analysis unit collects background vibration data during driving in segments and uses wavelet transform or short-time Fourier transform to decompose and obtain low-frequency stationary vibration components and medium- and high-frequency random vibration components, and extracts the root mean square value and energy distribution index of each component;

[0017] The characteristic baseline generation unit calculates the current vibration level based on the extracted root mean square value and energy distribution index of each component and vehicle driving information to form a vibration characteristic baseline for preliminary judgment. The vibration characteristic baseline is updated over time and is used to characterize the vehicle driving status.

[0018] By adopting the above technical solution, the background vibration data is segmented and collected and decomposed by wavelet / short-time Fourier through the multi-scale vibration analysis unit, which can accurately distinguish the vibration components of different frequency bands. The characteristic baseline generation unit can obtain the "vibration characteristic baseline" reflecting the current vehicle driving vibration level in real time based on the vehicle driving information and the decomposed root mean square value, energy distribution and other indicators, so that the subsequent impact capture or risk judgment link can adaptively adjust the risk judgment threshold based on this baseline, reduce false alarms and missed alarms, and take into account changes under different road conditions.

[0019] Further configuration: the impact capture module includes an acceleration sensor unit and a short-time energy detection unit,

[0020] The acceleration sensor unit acquires the transient vibration data of the battery fixing structure in real time at a preset high sampling frequency, generates a transient impact signal when a sudden change in impact energy is detected, and stores the transient vibration data at that moment and within a preset period before and after that moment in a cache;

[0021] The short-time energy detection unit performs short-window segmentation processing or wavelet local analysis on the stored transient vibration data, extracts the impact peak, duration and energy concentration as key feature information, and compares them with the vibration feature baseline provided by the conventional vibration monitoring module to comprehensively evaluate the loosening risk caused by the transient impact.

[0022] By adopting the above technical solution, the acceleration sensor unit acquires transient vibration data in real time at a high sampling frequency to avoid missing the details of rapid impact. The short-time energy detection unit performs short-window segmentation or wavelet local analysis on the cached data, and can accurately extract the impact peak value, duration and energy concentration. After comparison with the "vibration characteristic baseline" provided by the conventional vibration monitoring module, a comprehensive evaluation is performed, enabling the system to more accurately determine the loosening risk caused by transient impact under different driving conditions.

[0023] Further configuration: the comparison with the vibration characteristic baseline provided by the conventional vibration monitoring module to comprehensively evaluate the loosening risk caused by the transient impact specifically includes:

[0024] Obtain the vibration characteristic baseline of the current vehicle from the conventional vibration monitoring module, and set the corresponding impact risk threshold according to the level of the vibration characteristic baseline;

[0025] The extracted key feature information is compared with the impact risk threshold at this level. If it exceeds the impact risk threshold, it is judged that there is a loosening risk. If it does not exceed the impact risk threshold, it is judged that there is no loosening risk.

[0026] By adopting the above technical solution, by comparing the extracted key feature information with the impact risk threshold corresponding to the current vibration feature baseline level, the preliminary risk quantification of a single impact event can be achieved. If the impact exceeds the corresponding threshold, it is judged that "there is a loosening risk". If it does not exceed it, it passes safely. This simplifies the preliminary judgment logic of the instantaneous impact link and makes the evaluation steps more efficient. It adapts to different vibration baseline levels of the vehicle, allowing the system to make targeted risk threshold adjustments under violent bumpy or relatively stable working conditions.

[0027] Further configuration: the cross-validation module includes an electromagnetic interference detection unit, an acoustic monitoring unit, and a temperature sensing unit; when the impact capture module determines that there is a loosening risk, the electromagnetic interference detection unit obtains electromagnetic field change information at the battery fixing structure, the acoustic monitoring unit obtains friction or impact sound signals generated at the fixing structure, and the temperature sensing unit detects abnormal heat signals at the fixing structure;

[0028] Whether substantial loosening has occurred is determined comprehensively based on abnormal heat signals, sound signals and electromagnetic field change information.

[0029] By adopting the above technical solutions, the electromagnetic interference detection unit, acoustic monitoring unit and temperature sensing unit can capture multi-dimensional physical abnormality information (such as micro-arcing, friction impact sound, local temperature rise, etc.), significantly reducing the missed alarms / false alarms caused by a single sensing path. When the impact capture module determines that "there is a risk of loosening", it further obtains multiple physical signals through the cross-acquisition submodule and combines the fusion algorithm of the comprehensive judgment submodule to more accurately confirm whether actual loosening has occurred, thereby improving the overall system's ability to identify signs of minor or intermittent faults and enhancing the safety monitoring level of the battery fixing structure of new energy vehicles.

[0030] It is further configured that: the cumulative impact module includes an impact calculation unit and a cumulative calculation unit, the impact calculation unit is used to calculate the residual impact value according to the transient vibration data, the vibration characteristic baseline and the vehicle driving information;

[0031] The cumulative calculation unit is used to store and update the cumulative influence parameter, and to update the cumulative influence parameter according to the residual influence value to obtain a new cumulative influence parameter.

[0032] By adopting the above technical solution, the impact calculation unit calculates the "residual impact value" based on transient vibration data, vibration characteristic baseline and vehicle driving information, and quantifies the potential impact that may occur after a single impact but has not yet caused substantial loosening. The cumulative calculation unit accumulates or updates the "residual impact value" obtained from each calculation into a "cumulative impact parameter", which can capture the structural fatigue risk brought about by the continuous superposition of multiple small or medium impacts, enhance the system's ability to identify the process of "quantitative change to qualitative change", and make up for the traditional limitation of only caring about a single impact.

[0033] Further configuration: the cumulative impact module also includes an attenuation unit, which obtains a vibration characteristic baseline, and when the vibration characteristic baseline indicates that the vehicle is in a stable driving state, the cumulative impact parameter is attenuated and updated according to the attenuation parameter at each attenuation period to obtain a new cumulative impact parameter.

[0034] By adopting the above technical solution and introducing the "attenuation unit", the cumulative impact parameters can be periodically attenuated when the vehicle is in a stable driving state for a long time, which helps to prevent the cumulative parameters from growing infinitely due to multiple impacts in the early stage and causing misjudgment. The vibration characteristic baseline is combined with the attenuation unit, so that the system can adaptively attenuate the cumulative impact parameters after judging the actual operating conditions of the vehicle, taking into account the different scenarios of "occasional impact" and "long-term stability" during dynamic driving, retaining a reasonable historical impact "memory" while avoiding excessive sensitivity, thereby improving the stability and user experience of the entire monitoring system.

[0035] Further configuration: the early warning module is connected to the vehicle control system or battery management system (BMS), and when substantial looseness occurs or the cumulative impact parameters exceed the substantial risk threshold, the vehicle power limitation or safety protection strategy is triggered.

[0036] By adopting the above technical solution, the early warning module is interconnected with the vehicle control system (VCU) or battery management system (BMS), and the vehicle power limitation or safety protection is automatically triggered when it is confirmed to be "substantially loose" or the cumulative impact parameters exceed the risk threshold. The reaction time between the fault alarm and the vehicle safety response is greatly shortened, and more serious safety hazards caused by looseness are prevented in advance. When the system detects a risk, power reduction or graded protection measures are taken in time, which helps to ensure the safety of the vehicle's power battery and the entire vehicle.

[0037] It is further configured that the instantaneous impact signal generated when the impact energy suddenly changes includes a single large-amplitude transient vibration data and transient vibration data with an amplitude lower than a single judgment threshold that is continuously detected multiple times within a preset time window. The transient vibration data with an amplitude lower than a single judgment threshold that is continuously detected multiple times within the preset time window belong to the same instantaneous impact signal.

[0038] By adopting the above technical solution, "a single large-amplitude transient impact" and "multiple consecutive impacts slightly below the single threshold within a short time window" are both regarded as "the same instantaneous impact" and are uniformly captured and judged, avoiding the omission of multiple small impacts. The definition of impact events is expanded, allowing the system to identify the same impact caused by "multiple consecutive medium impacts" under complex road conditions, improving the robustness of the system and having more flexible impact recognition capabilities for long-band bumps or continuous speed bumps.

[0039] Another object of the present invention is to provide a safety monitoring method for new energy vehicle batteries, which has the advantages of predicting the risk of loosening of the battery fixing structure earlier and more accurately, ensuring that the vehicle maintains high safety and durability in a long-term, multi-operating environment.

[0040] The above technical objectives of the present invention are achieved through the following technical solutions:

[0041] A safety monitoring method for a new energy vehicle battery is applied to the safety monitoring system for a new energy vehicle battery described above, and the method comprises the following steps:

[0042] Background vibration monitoring: Utilizes conventional vibration monitoring modules to perform segmented acquisition and multi-scale analysis of background vibration data during vehicle driving, obtaining low-frequency steady vibration components and medium- and high-frequency random vibration components, and forming a vibration characteristic baseline to characterize the current driving state.

[0043] Transient shock capture: When a sudden acceleration or energy change is detected, the system records transient vibration data at the moment of the change and within a preset period before and after the change. Key features such as peak value, duration, and energy concentration extracted are compared with the vibration feature baseline to make a preliminary assessment of loosening risk.

[0044] Multi-physical signal cross-validation: When the preliminary judgment indicates a loosening risk, electromagnetic interference signals, acoustic monitoring signals, and temperature signals are acquired separately, and indicators such as thermal anomalies, electromagnetic field changes, and friction and impact sounds are comprehensively analyzed to confirm whether actual loosening has occurred;

[0045] Cumulative impact management: If the comprehensive judgment result is that no substantial loosening has occurred, the residual impact value is calculated based on the intensity of the transient impact, vehicle driving information, and vibration characteristic baseline, and is accumulated to the cumulative impact parameter. After the vehicle is in stable driving or the tightening operation is completed, the cumulative impact parameter is attenuated or reset;

[0046] Risk warning or tightening: If the comprehensive judgment results indicate that substantial loosening has occurred, or when the cumulative impact parameters exceed the preset substantial risk threshold, the warning module is triggered or a control signal is output to the vehicle control system to implement vehicle power limitation or safety protection strategies;

[0047] Storage and learning: The instantaneous impact signals, background vibration data, comprehensive judgment results and the evolution of cumulative impact parameters are stored in the storage and learning module. The actual risk threshold is dynamically optimized through data analysis or self-learning algorithms to continuously update the risk assessment accuracy during subsequent vehicle driving.

[0048] In summary, the present invention has the following beneficial effects:

[0049] The dual detection method of combining conventional vibration monitoring and high-sampling rate impact capture can not only continuously monitor the low-frequency / medium-frequency vibration characteristics of the vehicle during normal driving, but also instantly capture acceleration mutations and energy concentration when sudden impact occurs, thereby improving the efficiency of capturing signs of loosening of the battery fixing structure.

[0050] After an initial assessment, if a loosening risk is suspected, multiple physical signals, such as electromagnetic, acoustic, and temperature, can be introduced for cross-checking, significantly reducing false positives or missed negatives caused by a single path. This mechanism ensures the system can accurately identify substantial loosening risks even in complex environments.

[0051] To address the potential for multiple small impacts to evolve into major loosening failures over time, the system introduces a "residual impact value" and "cumulative impact parameter" to manage risk from quantitative to qualitative changes. Combined with the vehicle's driving status or the decay / reset mechanism after tightening operations, the system can detect potential hazards while preventing excessive misjudgments due to the unlimited accumulation of early impact information.

[0052] The vibration baseline level under vehicle driving conditions is dynamically associated with the impact judgment threshold, and the impact risk threshold is reasonably adjusted according to the vehicle's vibration state (smooth / bumpy) at different times to avoid frequent false alarms under bumpy road conditions and missed reporting of small but critical impact signals under smooth road conditions. When the system confirms that "substantial looseness" or "cumulative risk" exceeds the limit, it can be linked with the vehicle control unit (VCU) or BMS to promptly trigger power limitation or safety protection strategies, quickly reducing the possibility of escalating danger and ensuring the safety of the vehicle and occupants.

[0053] During the operation of the system, a large number of data samples of "instantaneous impact - cross-validation - cumulative impact - loosening conclusion" are collected, and the risk judgment threshold and impact feature extraction strategy are continuously updated and optimized through learning algorithms or big data analysis; it achieves a high degree of adaptability to different vehicle models, road conditions, and usage environments, laying the foundation for improving the long-term safety of the fixed structure of new energy vehicle batteries. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is an overall flow chart of the embodiment. DETAILED DESCRIPTION

[0055] The present invention will be further described in detail below with reference to the accompanying drawings.

[0056] Example:

[0057] like Figure 1 As shown, a safety monitoring system for new energy vehicle batteries includes:

[0058] A conventional vibration monitoring module, which is used to collect background vibration data at the battery fixing structure during vehicle driving and perform trend analysis on the background vibration data;

[0059] An impact capture module is used to capture instantaneous impacts and, when a transient impact signal is detected, extract transient vibration data within a preset period before and after the instantaneous impact signal is generated, and combine the transient vibration data with the background vibration trend to make a preliminary judgment on the loosening risk;

[0060] A cross-validation module, comprising a cross-collection submodule and a comprehensive judgment submodule. The cross-collection submodule is configured to collect cross-judgment data at the battery fixing structure when the preliminary judgment result of the loosening risk is that there is a loosening risk. The comprehensive judgment submodule comprehensively judges whether substantial loosening occurs based on the cross-judgment data.

[0061] A cumulative impact module, which is used to calculate the residual impact value when the comprehensive judgment result is that no substantial loosening has occurred, and convert the residual impact value into a cumulative impact parameter. The cumulative impact parameter is used to add the impact of the superposition effect caused by multiple historical shocks when making a preliminary judgment on the subsequent loosening risk;

[0062] An early warning module, configured to trigger an early warning signal when a comprehensive judgment indicates that substantial loosening has occurred or when the cumulative impact parameter exceeds a substantial risk threshold;

[0063] The storage and learning module is used to store the background vibration data, instantaneous impact signals, comprehensive judgment results, and the evolution of cumulative impact parameters when each instantaneous impact signal is generated, and to optimize the actual risk threshold through data analysis or self-learning algorithm to obtain a new actual risk threshold.

[0064] When a vehicle travels on a relatively smooth road, vibration amplitude is often small or fluctuates periodically. If the vehicle is well secured, it's difficult to distinguish potential loosening based solely on these vibration data. However, when the vehicle passes over potholes, speed bumps, or encounters an unexpected collision, it often generates one or a series of large-amplitude transient shock waveforms. If there are already hidden loosening trends, the shock will amplify the risk. Background vibration can reflect the degree and frequency distribution of the turbulence experienced by the vehicle over a period of time. The characteristics of transient impacts (amplitude, energy peak, and impact duration) will produce very different consequences depending on the "background vibration state." Therefore, the risk of the same impact intensity in the two backgrounds of "very stable" and "continuous turbulence" is different. Once the "background + impact" combined analysis preliminarily determines that "loosening is possible," other physical signals such as electromagnetic, acoustic, and temperature are quickly tested. If certain anomalies (such as a sharp increase in electromagnetic scattering, increased temperature at the bolt, and acoustic resonance after impact) are consistent with signs of loosening, the accuracy of the judgment can be greatly improved. If there are no obvious anomalies in multiple physical signals, it means that the impact has not yet caused actual loosening, or only a slight change, which is still within the safe range. However, the system will fully record this data to facilitate differentiation during subsequent algorithm analysis and model training.

[0065] Through the organic combination of conventional vibration monitoring modules, impact capture modules, cross-validation modules, cumulative impact modules, early warning modules, and storage and learning modules, the system can achieve real-time detection and comprehensive assessment of the risk of loosening of the battery fixing structure under various working conditions; it can not only capture instantaneous impacts, but also record and manage the cumulative impact of multiple light and moderate impacts, thereby improving the detection rate and judgment accuracy of potential loosening faults of the battery fixing structure, and ensuring the safe operation of new energy vehicles.

[0066] The conventional vibration monitoring module includes a multi-scale vibration analysis unit and a characteristic baseline generation unit;

[0067] The multi-scale vibration analysis unit collects background vibration data during driving in segments and uses wavelet transform or short-time Fourier transform to decompose and obtain low-frequency stationary vibration components and medium- and high-frequency random vibration components, and extracts the root mean square value and energy distribution index of each component;

[0068] The characteristic baseline generation unit calculates the current vibration level based on the extracted root mean square value and energy distribution index of each component and vehicle driving information to form a vibration characteristic baseline for preliminary judgment. The vibration characteristic baseline is updated over time and is used to characterize the vehicle driving status.

[0069] Multi-scale vibration analysis unit, acquisition cycle: When the vehicle is driving normally, the background vibration signal at the battery fixture is acquired at set time intervals (such as 1 to 5 seconds). Decomposition method: Perform wavelet transform or short-time Fourier transform (STFT) on the background vibration signal to separate the vibration components in the low-frequency band (such as 0-20Hz) and the medium- and high-frequency band (such as 20-200Hz). Index extraction: Extract key indicators such as the root mean square value RMS, peak-to-peak value Pk-Pk, and frequency domain energy concentration for each decomposition segment to measure the vibration conditions of the vehicle under smooth driving, slight bumps, and large bumps.

[0070] During the driving process, a vibration signal x(t) is obtained in each sampling period, with a length of N; it is recorded as a discrete sequence {x1, x2, ..., xN}. The wavelet transform or short-time Fourier transform (STFT) is used to decompose the signal into a low-frequency (stationary) part and a medium- and high-frequency (random) part, which are recorded as X LOW (t) and X HIGH (t), in this embodiment, short-time Fourier transform is used, and the specific formula is:

[0071]

[0072] Where h(.) is the window function; summing or integrating within a specific frequency band can obtain the local energy distribution in different frequency bands.

[0073] The characteristic baseline generation unit calculates vibration levels by applying the index values ​​of low-frequency and medium-high-frequency components to a pre-established vibration level classification model (e.g., based on statistical thresholds or machine learning classifiers) to output a "basic vibration level" or "vibration intensity score" reflecting the vehicle's current road conditions. It also generates a baseline based on the basic vibration level, combined with vehicle driving information such as vehicle speed and suspension system status, to form a "vibration characteristic baseline" that updates over time. For reference, when a sudden transient shock occurs, if the current vibration baseline is in the "stable" range, even relatively small shock peaks may pose a higher risk of loosening. If the baseline is already in the "medium or high turbulence" state, the risk judgment threshold corresponding to the same shock peak will be appropriately increased to reduce misjudgments.

[0074] The RMS and peak-to-peak values ​​of the low-frequency band, as well as the energy distribution of the mid- and high-frequency bands, are combined with vehicle speed v, acceleration and other driving data and input into the pre-designed scoring model. A simple linear scoring function S is defined:

[0075] S = αRMS low +βE high +γv

[0076] Among them, α, β, and Y are weights obtained through experience or learning.

[0077] S is calculated for each sampling period and compared with the historical distribution to obtain the current vibration level (such as "stable", "moderate bumpy" or "strong bumpy"). This level is recorded as the "characteristic baseline" B(t) and updated over time. It is used to correct the impact risk threshold or distinguish working conditions when the impact capture module is called.

[0078] The multi-scale vibration analysis unit collects background vibration data in segments and performs wavelet / short-time Fourier decomposition, which can accurately distinguish vibration components in different frequency bands. The characteristic baseline generation unit can obtain a "vibration characteristic baseline" reflecting the current vehicle driving vibration level in real time based on vehicle driving information and decomposed root mean square value, energy distribution and other indicators. Subsequent impact capture or risk assessment links can adaptively adjust the risk assessment threshold based on this baseline, reducing false alarms and missed alarms while taking into account changes under different road conditions.

[0079] The impact capture module includes an acceleration sensor unit and a short-time energy detection unit.

[0080] The acceleration sensor unit acquires the transient vibration data of the battery fixing structure in real time at a preset high sampling frequency, generates a transient impact signal when a sudden change in impact energy is detected, and stores the transient vibration data at that moment and within a preset period before and after that moment in a cache;

[0081] The short-time energy detection unit performs short-window segmentation processing or wavelet local analysis on the stored transient vibration data, extracts the impact peak, duration and energy concentration as key feature information, and compares them with the vibration feature baseline provided by the conventional vibration monitoring module to comprehensively evaluate the loosening risk caused by the transient impact.

[0082] Sensor placement: Install high-sensitivity accelerometers around the battery fixture or connecting bolts to ensure that transient, high-frequency impact vibrations can be captured. Sampling frequency setting: Sampling frequency f s Higher than conventional vibration monitoring (hundreds of Hz and above), to avoid missing the waveform details of transient impacts. Real-time data streaming: The continuously collected acceleration data is packaged into short-term data frames (such as tens of milliseconds) and transmitted to the subsequent short-term energy detection unit.

[0083] Energy mutation detection: Calculate the energy of the acceleration signal within the sliding time window, such as short-time energy ST-Energy or transient intensity index based on wavelet coefficients; when the energy exceeds the preset threshold (or the energy difference between adjacent windows exceeds the limit), it is considered that a "shock mutation" is detected.

[0084] or E ψ (n)=∑ k |(a k , b n )|2

[0085] Where x(n) represents the discrete acceleration signal, M is the analysis window length, and W X is the wavelet transform coefficient.

[0086] At the moment of detecting the mutation and within several milliseconds (or sampling windows) before and after it, the system caches the vibration data for subsequent feature extraction. The cached impact segment data is subjected to feature analysis such as peak value, level change rate, main frequency band energy ratio, etc. to identify the impact intensity and duration. The instantaneous impact signal generated when the impact energy suddenly changes includes a single large-amplitude transient vibration data and transient vibration data with an amplitude lower than a single judgment threshold that is continuously detected multiple times within a preset time window. The transient vibration data with an amplitude lower than a single judgment threshold that is continuously detected multiple times within the preset time window belong to the same instantaneous impact signal.

[0087] The vibration baseline (smooth / moderate bumps / strong bumps) of the current vehicle driving is obtained from the conventional vibration monitoring module, and the corresponding impact risk threshold is set according to the baseline level. The extracted transient impact amplitude, impact energy peak, etc. are compared with the threshold under this level; if the threshold is exceeded, a preliminary conclusion of "possible loosening risk" is output, and a trigger signal is sent to the cross-validation module.

[0088] The acceleration sensor unit acquires transient vibration data in real time at a high sampling frequency to avoid missing details of rapid impacts. The short-time energy detection unit performs short-window segmentation or wavelet local analysis on the cached data, accurately extracting the impact peak value, duration, and energy concentration. After comparison with the "vibration characteristic baseline" provided by the conventional vibration monitoring module, a comprehensive evaluation is performed, enabling the system to more accurately determine the loosening risk caused by transient impacts under different driving conditions.

[0089] The comparison with the vibration characteristic baseline provided by the conventional vibration monitoring module to comprehensively evaluate the loosening risk caused by the transient impact specifically includes:

[0090] Obtain the vibration characteristic baseline of the current vehicle from the conventional vibration monitoring module, and set the corresponding impact risk threshold according to the level of the vibration characteristic baseline;

[0091] The extracted key feature information is compared with the impact risk threshold at this level. If it exceeds the impact risk threshold, it is judged that there is a loosening risk. If it does not exceed the impact risk threshold, it is judged that there is no loosening risk.

[0092] By comparing the extracted key feature information with the impact risk threshold corresponding to the current vibration feature baseline level, the preliminary risk quantification of a single impact event can be achieved. If the impact exceeds the corresponding threshold, it is judged that "there is a loosening risk". If it does not exceed it, it passes safely. This simplifies the preliminary judgment logic of the instantaneous impact link and makes the evaluation steps more efficient. It adapts to different vibration baseline levels of the vehicle and allows the system to make targeted risk threshold adjustments under violent bumpy or relatively stable working conditions.

[0093] The cumulative impact module includes an impact calculation unit and a cumulative calculation unit, wherein the impact calculation unit is used to calculate the residual impact value according to the transient vibration data, the vibration characteristic baseline and the vehicle driving information;

[0094] The cumulative calculation unit is used to store and update the cumulative influence parameter, and to update the cumulative influence parameter according to the residual influence value to obtain a new cumulative influence parameter.

[0095] The cumulative impact module also includes an attenuation unit, which obtains a vibration characteristic baseline and, when the vibration characteristic baseline indicates that the vehicle is in a stable driving state, updates the cumulative impact parameter according to the attenuation parameter at every attenuation period to obtain a new cumulative impact parameter.

[0096] Calculation of residual impact value:

[0097] ΔI=f(P,B(t),v,…)

[0098] Where v represents the vehicle speed or related parameters, and f is an empirical or model function. The residual impact value is accumulated into the cumulative impact parameter to complete the update of the cumulative impact parameter.

[0099] Due to the existence of the cumulative impact module, before each comparison of key feature information with the impact risk, the cumulative impact parameter is converted into a threshold correction parameter through calculation. The impact risk threshold obtained based on the vibration feature baseline level is corrected using the threshold correction parameter to obtain a new impact risk threshold, and the key feature information is compared with the corrected impact risk threshold. The cumulative impact parameter and the converted threshold correction parameter are inversely proportional, that is, the larger the cumulative impact parameter, the smaller the threshold correction parameter. The impact risk threshold after correction by the threshold correction parameter also becomes smaller accordingly, indicating that the impact limit that the vehicle can withstand has been reduced due to the accumulation of multiple previous impacts. Due to the involvement of the cumulative impact parameter, the cumulative impact of multiple previous impacts is taken into account in each preliminary judgment, ensuring the accuracy of the preliminary judgment.

[0100] The impact calculation unit calculates the "residual impact value" based on transient vibration data, vibration characteristic baselines, and vehicle driving information. This quantifies the potential impact that may occur after a single impact without causing actual loosening. The cumulative calculation unit accumulates or updates the "residual impact value" obtained from each calculation into a "cumulative impact parameter." This can capture the structural fatigue risk brought about by the continuous superposition of multiple small or medium impacts, improving the system's ability to identify the process of "quantitative change to qualitative change," and overcoming the traditional limitation of only focusing on single impacts. The introduction of the "attenuation unit" allows for periodic attenuation of the cumulative impact parameter when the vehicle is in a stable driving state for a long period of time. This helps prevent the cumulative parameter from growing infinitely due to multiple early impacts, leading to misjudgment. The combination of the vibration characteristic baseline and the attenuation unit allows the system to adaptively attenuate the cumulative impact parameter after determining the actual vehicle operating conditions. This takes into account the different scenarios of "occasional impact" and "long-term stability" during dynamic driving, retaining a reasonable "memory" of historical impacts while avoiding oversensitivity, thereby improving the stability and user experience of the entire monitoring system.

[0101] The cross-validation module includes an electromagnetic interference detection unit, an acoustic monitoring unit, and a temperature sensing unit. When the impact capture module determines that there is a risk of loosening, the electromagnetic interference detection unit obtains electromagnetic field change information at the battery fixing structure, the acoustic monitoring unit obtains friction or impact sound signals generated by the fixing structure, and the temperature sensing unit detects abnormal heat signals at the fixing structure.

[0102] Whether substantial loosening has occurred is determined comprehensively based on abnormal heat signals, sound signals and electromagnetic field change information.

[0103] Electromagnetic interference detection unit, detection principle: When the battery connection components are loose, micro-arcing or poor contact may occur at the connection contacts, which in turn generates abnormal electromagnetic radiation or local current fluctuations around the metal interface; install a miniature electromagnetic induction probe or high-frequency current sensor around the battery connection position, and use a bandpass or narrowband filter to remove the electromagnetic interference of the conventional vehicle electronic system, and extract the characteristics of abnormal pulses or noise (such as peak amplitude, pulse repetition rate, etc.); when the detected electromagnetic interference characteristics exceed the preset threshold, an "electromagnetic anomaly" signal is output.

[0104] Acoustic monitoring unit: loose connection parts will produce abnormal friction, knocking or whistling sounds after being subjected to vibration or impact; one or more miniature microphones or acceleration acoustic sensors are arranged near the battery module or connection parts; use short-time Fourier transform (STFT) or wavelet-based acoustic analysis algorithm to compare the collected sound data with the acoustic fingerprint under normal working conditions; if a sharp impact sound or abnormal frequency domain energy distribution is detected, and it matches the loose feature library (such as the friction signal in the ultrasonic segment), an "acoustic anomaly" signal is output.

[0105] Temperature sensing unit, detection principle: Loose connections often lead to increased local contact resistance or heat generated by impact and collision, causing local temperature rise. Temperature sensors (thermocouples, RTDs, or infrared micrometers) are mounted on key bolts or brackets to periodically or continuously monitor local temperature. This temperature is compared with a typical temperature baseline (determined by vehicle power and ambient temperature). If the temperature rises significantly within a short period of time, it is identified as a "temperature anomaly" event.

[0106] To reduce false positives from a single signal, the detection results of the three sub-units can be combined through weighted scoring to define the total score:

[0107] S=w1A′ EM +w2E′ a +w3ΔT′

[0108] Among them, w1, w2, and w3 are weight coefficients obtained through experience or training.

[0109] If the calculated total score S is greater than the preset score threshold, the multi-physical signal synthesis is judged to be abnormal, confirming the existence of a substantial loosening risk.

[0110] The electromagnetic interference detection unit, acoustic monitoring unit and temperature sensing unit can capture multi-dimensional physical abnormality information (such as micro-arcing, friction impact sound, local temperature rise, etc.), significantly reducing the missed alarms / false alarms caused by a single sensing path. When the impact capture module determines that "there is a risk of loosening", it further obtains multiple physical signals through the cross-acquisition sub-module and combines the fusion algorithm of the comprehensive judgment sub-module to more accurately confirm whether actual loosening has occurred, thereby improving the overall system's ability to identify signs of minor or intermittent faults and enhancing the safety monitoring level of the battery fixing structure of new energy vehicles.

[0111] The early warning module is connected to the vehicle control system or battery management system (BMS), and triggers vehicle power limitation or safety protection strategy when substantial looseness occurs or the cumulative impact parameters exceed the substantial risk threshold.

[0112] By interconnecting the early warning module with the vehicle control unit (VCU) or battery management system (BMS), the vehicle power limit or safety protection is automatically triggered when "substantial looseness" is confirmed or the cumulative impact parameters exceed the risk threshold. This greatly shortens the reaction time between fault alarm and vehicle safety response, and takes precautions in advance against more serious safety hazards that may be caused by looseness. When the system detects a risk, it takes power reduction or graded protection measures in a timely manner, which helps to ensure the safety of the vehicle's power battery and the entire vehicle.

[0113] Another object of the present invention is to provide a safety monitoring method for new energy vehicle batteries, which has the advantages of predicting the risk of loosening of the battery fixing structure earlier and more accurately, ensuring that the vehicle maintains high safety and durability in a long-term, multi-operating environment.

[0114] A safety monitoring method for a new energy vehicle battery is applied to the safety monitoring system for a new energy vehicle battery described above, and the method comprises the following steps:

[0115] Background vibration monitoring: Utilizes conventional vibration monitoring modules to perform segmented acquisition and multi-scale analysis of background vibration data during vehicle driving, obtaining low-frequency steady vibration components and medium- and high-frequency random vibration components, and forming a vibration characteristic baseline to characterize the current driving state.

[0116] Transient shock capture: When a sudden acceleration or energy change is detected, the system records transient vibration data at the moment of the change and within a preset period before and after the change. Key features such as peak value, duration, and energy concentration extracted are compared with the vibration feature baseline to make a preliminary assessment of loosening risk.

[0117] Multi-physical signal cross-validation: When the preliminary judgment indicates a loosening risk, electromagnetic interference signals, acoustic monitoring signals, and temperature signals are acquired separately, and indicators such as thermal anomalies, electromagnetic field changes, and friction and impact sounds are comprehensively analyzed to confirm whether actual loosening has occurred;

[0118] Cumulative impact management: If the comprehensive judgment result is that no substantial loosening has occurred, the residual impact value is calculated based on the intensity of the transient impact, vehicle driving information, and vibration characteristic baseline, and is accumulated to the cumulative impact parameter. After the vehicle is in stable driving or the tightening operation is completed, the cumulative impact parameter is attenuated or reset;

[0119] Risk warning or tightening: If the comprehensive judgment results indicate that substantial loosening has occurred, or when the cumulative impact parameters exceed the preset substantial risk threshold, the warning module is triggered or a control signal is output to the vehicle control system to implement vehicle power limitation or safety protection strategies;

[0120] Storage and learning: The instantaneous impact signals, background vibration data, comprehensive judgment results and the evolution of cumulative impact parameters are stored in the storage and learning module. The actual risk threshold is dynamically optimized through data analysis or self-learning algorithms to continuously update the risk assessment accuracy during subsequent vehicle driving.

[0121] In summary, the present invention has the following beneficial effects:

[0122] The dual detection method of combining conventional vibration monitoring and high-sampling rate impact capture can not only continuously monitor the low-frequency / medium-frequency vibration characteristics of the vehicle during normal driving, but also instantly capture acceleration mutations and energy concentration when sudden impact occurs, thereby improving the efficiency of capturing signs of loosening of the battery fixing structure.

[0123] After an initial assessment, if a loosening risk is suspected, multiple physical signals, such as electromagnetic, acoustic, and temperature, can be introduced for cross-checking, significantly reducing false positives or missed negatives caused by a single path. This mechanism ensures the system can accurately identify substantial loosening risks even in complex environments.

[0124] To address the potential for multiple small impacts to evolve into major loosening failures over time, the system introduces a "residual impact value" and "cumulative impact parameter" to manage risk from quantitative to qualitative changes. Combined with the vehicle's driving status or the decay / reset mechanism after tightening operations, the system can detect potential hazards while preventing excessive misjudgments due to the unlimited accumulation of early impact information.

[0125] The vibration baseline level under vehicle driving conditions is dynamically associated with the impact judgment threshold, and the impact risk threshold is reasonably adjusted according to the vehicle's vibration state (smooth / bumpy) at different times to avoid frequent false alarms under bumpy road conditions and missed reporting of small but critical impact signals under smooth road conditions. When the system confirms that "substantial looseness" or "cumulative risk" exceeds the limit, it can be linked with the vehicle control unit (VCU) or BMS to promptly trigger power limitation or safety protection strategies, quickly reducing the possibility of escalating danger and ensuring the safety of the vehicle and occupants.

[0126] During the operation of the system, a large number of data samples of "instantaneous impact - cross-validation - cumulative impact - loosening conclusion" are collected, and the risk judgment threshold and impact feature extraction strategy are continuously updated and optimized through learning algorithms or big data analysis; it achieves a high degree of adaptability to different vehicle models, road conditions, and usage environments, laying the foundation for improving the long-term safety of the fixed structure of new energy vehicle batteries.

[0127] The above-described embodiments do not constitute a limitation on the scope of protection of this technical solution. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the above-described embodiments shall be included in the scope of protection of this technical solution.

Claims

1. A safety monitoring system for new energy vehicle batteries, characterized in that: include: A conventional vibration monitoring module, which is used to collect background vibration data at the battery fixing structure during vehicle driving and perform trend analysis on the background vibration data; An impact capture module is used to capture instantaneous impacts and, when a transient impact signal is detected, extract transient vibration data within a preset period before and after the instantaneous impact signal is generated, and combine the transient vibration data with the background vibration trend to make a preliminary judgment on the loosening risk; A cross-validation module, comprising a cross-collection submodule and a comprehensive judgment submodule. The cross-collection submodule is configured to collect cross-judgment data at the battery fixing structure when the preliminary judgment result of the loosening risk is that there is a loosening risk. The comprehensive judgment submodule comprehensively judges whether substantial loosening occurs based on the cross-judgment data. A cumulative impact module, which is used to calculate the residual impact value when the comprehensive judgment result is that no substantial loosening has occurred, and convert the residual impact value into a cumulative impact parameter. The cumulative impact parameter is used to add the impact of the superposition effect caused by multiple historical shocks when making a preliminary judgment on the subsequent loosening risk; An early warning module, configured to trigger an early warning signal when a comprehensive judgment indicates that substantial loosening has occurred or when cumulative impact parameters exceed a substantial risk threshold; A storage and learning module, which is used to store background vibration data, instantaneous impact signals, comprehensive judgment results, and the evolution of cumulative impact parameters when each instantaneous impact signal is generated, and to optimize the actual risk threshold through data analysis or self-learning algorithm to obtain a new actual risk threshold; The conventional vibration monitoring module includes a multi-scale vibration analysis unit and a characteristic baseline generation unit; The multi-scale vibration analysis unit collects background vibration data during driving in segments and uses wavelet transform or short-time Fourier transform to decompose and obtain low-frequency stationary vibration components and medium- and high-frequency random vibration components, and extracts the root mean square value and energy distribution index of each component; The characteristic baseline generation unit calculates the current vibration level based on the extracted root mean square value and energy distribution index of each component and the vehicle driving information to form a vibration characteristic baseline for preliminary judgment. The vibration characteristic baseline is updated over time and is used to characterize the vehicle driving state; The impact capture module includes an acceleration sensor unit and a short-time energy detection unit. The acceleration sensor unit acquires the transient vibration data of the battery fixing structure in real time at a preset high sampling frequency, generates a transient impact signal when a sudden change in impact energy is detected, and stores the transient vibration data at that moment and within a preset period before and after that moment in a cache; The short-time energy detection unit performs short-window segmentation processing or wavelet local analysis on the stored transient vibration data, extracts the impact peak value, duration and energy concentration as key feature information, and compares it with the vibration feature baseline provided by the conventional vibration monitoring module to comprehensively evaluate the loosening risk caused by the transient impact; The instantaneous impact signal generated when the impact energy suddenly changes includes a single large-amplitude transient vibration data and transient vibration data with an amplitude lower than a single judgment threshold that is continuously detected multiple times within a preset time window. The transient vibration data with an amplitude lower than a single judgment threshold that is continuously detected multiple times within the preset time window belong to the same instantaneous impact signal.

2. A safety monitoring system for new energy vehicle batteries according to claim 1, characterized in that: The comparison with the vibration characteristic baseline provided by the conventional vibration monitoring module to comprehensively evaluate the loosening risk caused by the transient impact specifically includes: Obtain the vibration characteristic baseline of the current vehicle from the conventional vibration monitoring module, and set the corresponding impact risk threshold according to the level of the vibration characteristic baseline; The extracted key feature information is compared with the impact risk threshold at this level. If it exceeds the impact risk threshold, it is judged that there is a loosening risk. If it does not exceed the impact risk threshold, it is judged that there is no loosening risk.

3. A safety monitoring system for new energy vehicle batteries according to claim 1, characterized in that: The cross-validation module includes an electromagnetic interference detection unit, an acoustic monitoring unit, and a temperature sensing unit. When the impact capture module determines that there is a risk of loosening, the electromagnetic interference detection unit obtains electromagnetic field change information at the battery fixing structure, the acoustic monitoring unit obtains friction or impact sound signals generated by the fixing structure, and the temperature sensing unit detects abnormal heat signals at the fixing structure. Whether substantial loosening has occurred is determined comprehensively based on abnormal heat signals, sound signals and electromagnetic field change information.

4. A safety monitoring system for new energy vehicle batteries according to claim 2, characterized in that: The cumulative impact module includes an impact calculation unit and a cumulative calculation unit, wherein the impact calculation unit is used to calculate the residual impact value according to the transient vibration data, the vibration characteristic baseline and the vehicle driving information; The cumulative calculation unit is used to store and update the cumulative influence parameter, and to update the cumulative influence parameter according to the residual influence value to obtain a new cumulative influence parameter.

5. A safety monitoring system for new energy vehicle batteries according to claim 4, characterized in that: The cumulative impact module also includes an attenuation unit, which obtains a vibration characteristic baseline and, when the vibration characteristic baseline indicates that the vehicle is in a stable driving state, updates the cumulative impact parameter according to the attenuation parameter at every attenuation period to obtain a new cumulative impact parameter.

6. A safety monitoring system for new energy vehicle batteries according to claim 1, characterized in that: The early warning module is connected to the vehicle control system or battery management system (BMS), and triggers vehicle power limitation or safety protection strategy when substantial looseness occurs or the cumulative impact parameters exceed the substantial risk threshold.

7. A safety monitoring method for a new energy vehicle battery, applied to a safety monitoring system for a new energy vehicle battery according to any one of claims 1 to 6, characterized in that: The method comprises the following steps: Background vibration monitoring: Utilizes conventional vibration monitoring modules to perform segmented acquisition and multi-scale analysis of background vibration data during vehicle driving, obtaining low-frequency steady vibration components and medium- and high-frequency random vibration components, and forming a vibration characteristic baseline to characterize the current driving state. Transient shock capture: When a sudden acceleration or energy change is detected, the system records transient vibration data at the moment of the change and within a preset period before and after the change. The extracted key features of peak value, duration, and energy concentration are compared with the vibration feature baseline to make a preliminary assessment of loosening risk. Multi-physical signal cross-validation: When the preliminary judgment result indicates that there is a risk of loosening, electromagnetic interference signals, acoustic monitoring signals, and temperature signals are obtained separately, and thermal anomalies, electromagnetic field changes, and friction and impact sound indicators are comprehensively analyzed to confirm whether actual loosening has occurred; Cumulative impact management: If the comprehensive judgment result is that no substantial loosening has occurred, the residual impact value is calculated based on the intensity of the transient impact, vehicle driving information, and vibration characteristic baseline, and is accumulated to the cumulative impact parameter. After the vehicle is in stable driving or the tightening operation is completed, the cumulative impact parameter is attenuated or reset; Risk warning or tightening: If the comprehensive judgment results indicate that substantial loosening has occurred, or when the cumulative impact parameters exceed the preset substantial risk threshold, the warning module is triggered or a control signal is output to the vehicle control system to implement vehicle power limitation or safety protection strategies; Storage and learning: The instantaneous impact signals, background vibration data, comprehensive judgment results and the evolution of cumulative impact parameters are stored in the storage and learning module. The actual risk threshold is dynamically optimized through data analysis or self-learning algorithms to continuously update the risk assessment accuracy during subsequent vehicle driving.

Images