Method for determining power sensitivity of post-blast behavior of power battery

Abstract

This invention relates to the field of power battery technology and discloses a method for determining the power sensitivity of the post-vacuuming buffer behavior of a power battery. The method includes: acquiring temperature time-series data, the time of valve injection, and the time of thermal runaway occurrence of several power batteries under different heating power conditions; determining the post-vacuuming buffer analysis time interval; determining the total net energy accumulation from the heating start time to the thermal runaway occurrence time, and determining the net energy accumulation between buffer zones within the post-vacuuming buffer analysis time interval; generating power scale parameters based on the corresponding heating power, total net energy accumulation, and net energy accumulation between buffer zones; determining the global statistical dispersion; and determining the sensitivity level of the power battery's post-vacuuming buffer behavior to changes in heating power. This application achieves a quantitative determination of the degree to which the post-vacuuming buffer behavior is affected by heating power.

Description

Technical Field

[0001] This invention relates to the field of power battery technology, and more specifically, to a method for determining the power sensitivity of the post-injection valve buffer behavior of a power battery. Background Technology

[0002] Power batteries typically undergo temperature rise, valve ejection, and potential thermal runaway processes under external heating or thermal abuse conditions. Existing research has been able to obtain energy input information of batteries under thermal abuse conditions through energy balance or thermal response models. However, between the occurrence of valve ejection and the triggering of thermal runaway, multiple processes simultaneously occur within the battery system, including enhanced heat dissipation and exothermic internal reactions. These processes exhibit complex coupling behavior and are highly sensitive to applied heating power.

[0003] Especially for high-capacity power batteries, the post-vacuum stage may exhibit significantly different buffering behaviors under different heating power conditions. For example, it may continuously buffer without thermal runaway, delay the triggering of thermal runaway, or occur almost simultaneously with the vacuum stage and thermal runaway. Existing technologies mostly rely on a single temperature threshold or total energy threshold, making it difficult to transform the complex coupling process from the vacuum stage to the thermal runaway stage into a comparable and reproducible engineering judgment criterion.

[0004] Therefore, it is necessary to design a power sensitivity determination method for the post-valve buffer behavior of a power battery to solve the problems existing in the current technology. Summary of the Invention

[0005] In view of this, the present invention proposes a power sensitivity determination method for the post-vacuum behavior of a power battery, aiming to solve the problem of the difficulty in transforming the complex coupling process from the valve to the thermal runaway stage into a comparable and reproducible engineering judgment criterion.

[0006] This invention proposes a method for determining the power sensitivity of the post-injection valve buffer behavior of a power battery, comprising: Cell detection data of several power batteries under different heating power conditions are obtained. The cell detection data of each power battery includes temperature time series data, valve injection time and thermal runaway time. The temperature time series data is then processed by moving average filtering. Based on the timing of the valve ejection and thermal runaway, the post-valve buffer analysis time interval is determined; based on the ambient temperature, equivalent heat dissipation coefficient, and the filtered temperature time series data, the total net energy accumulation from the heating start time to the thermal runaway time is determined, as well as the net energy accumulation between buffer zones within the post-valve buffer analysis time interval; based on the corresponding heating power, the total net energy accumulation, and the net energy accumulation between buffer zones, power scale parameters are generated. Outlier detection and removal are performed on the power scale parameters to generate an effective power scale parameter set; the global statistical dispersion is determined based on the effective power scale parameter set. Based on the global statistical dispersion, the sensitivity level of the buffer behavior of the power battery injection valve to changes in heating power is determined.

[0007] Furthermore, when determining the post-spray valve buffer analysis time interval, it includes: The continuous time window between the start point and the end point, taking the time when the valve occurs as the starting point and the time when thermal runaway occurs as the ending point, is the post-valve buffer analysis time interval.

[0008] Furthermore, when determining the total net energy accumulation from the start of heating to the occurrence of thermal runaway, the following is included: Based on the temperature time series data, the integral value of the instantaneous difference between heating power and heat dissipation power in the time domain from the start of heating to the occurrence of thermal runaway is determined as the total net energy accumulation.

[0009] Furthermore, determining the net energy accumulation between buffer zones within the post-spray valve buffer analysis time interval includes: Based on the temperature time series data, the integral value of the instantaneous difference between heating power and heat dissipation power in the time domain within the buffer analysis time interval after the spray valve is determined as the net energy accumulation in the buffer zone.

[0010] Furthermore, when generating power scale parameters, the following are included: The power scale parameter is generated by calculating the ratio of the product of the total net energy accumulation and the corresponding heating power to the net energy accumulation between the buffer zones.

[0011] Furthermore, when acquiring cell test data for several power batteries under different heating power conditions, the following are included: Standardized heating of the same type of power battery cells is performed under several preset heating power conditions; Collect temperature time-series data; determine the timing of the valve release based on voltage signal abrupt change characteristics and gas release detection signals; determine the timing of thermal runaway based on temperature step rise characteristics and voltage zeroing characteristics.

[0012] Furthermore, determining the global statistical dispersion includes: Based on the effective power scale parameter set, the standard deviation of all the power scale parameters within the effective power scale parameter set is determined, and the standard deviation is used as the global statistical dispersion.

[0013] Furthermore, when determining the sensitivity level of the buffer behavior after the power battery injection valve to changes in heating power, the following includes: The statistical dispersion is compared with the dispersion threshold; When the statistical dispersion is less than or equal to the dispersion threshold, it is determined that the post-spray valve buffer behavior is insensitive to changes in heating power. When the statistical dispersion is greater than the dispersion threshold, it is determined that the post-buffering behavior of the spray valve is sensitive to changes in heating power.

[0014] Furthermore, it also includes: The heating power range is divided into multiple continuous power sub-intervals, and the local statistical dispersion of the effective power scale parameter in each power sub-interval is determined to generate a dispersion distribution sequence.

[0015] Furthermore, it also includes: Based on the discrete distribution sequence, the difference value of the local statistical discreteness corresponding to adjacent power sub-intervals is determined. When the difference value is greater than the difference change threshold, it is determined that the post-vacuum behavior of the corresponding power boundary region is sensitive to the change of heating power.

[0016] Compared with existing technologies, the advantages of this invention are as follows: By constructing power scale parameters and quantifying their statistical dispersion, an objective and quantifiable determination of the degree to which the buffering behavior after the nozzle is affected by heating power is achieved. Based on the precise definition of the buffering analysis interval at the nozzle and thermal runaway moments, and by integrating ambient temperature and equivalent heat dissipation coefficient to physically correct the temperature time-series data, the accuracy of net energy accumulation calculation is improved. By normalizing and correlating the heating power, total net energy accumulation, and net energy accumulation between buffer zones, the generated power scale parameters remove the absolute energy differences in a single experiment, highlighting the relative influence mechanism of power changes on buffering behavior. Through intelligent outlier removal and global statistical dispersion analysis of the parameter set, sensitivity levels are distinguished; high dispersion indicates that the buffering behavior is highly sensitive to power fluctuations, requiring strengthened power fluctuation suppression and dynamic adjustment of early warning thresholds in thermal management; low dispersion reflects stable buffering behavior, simplifying thermal runaway prevention strategies. Attached Figure Description

[0017] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart of a method for determining the power sensitivity of the post-injection valve buffer behavior of a power battery, provided in an embodiment of the present invention. Detailed Implementation

[0018] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0019] In some embodiments of this application, see Figure 1 As shown, a power sensitivity determination method for the post-injection valve buffer behavior of a power battery is proposed, including: The cell test data of several power batteries under different heating power conditions were obtained. The cell test data of each power battery includes temperature time series data, the time of valve injection and the time of thermal runaway. The temperature time series data were then processed by moving average filtering. Based on the timing of the valve ejection and thermal runaway, the buffer analysis time interval after the valve ejection is determined; based on the ambient temperature, equivalent heat dissipation coefficient, and filtered temperature time series data, the total net energy accumulation from the start of heating to the occurrence of thermal runaway is determined, as well as the net energy accumulation between buffer zones within the buffer analysis time interval after the valve ejection; based on the corresponding heating power, total net energy accumulation, and net energy accumulation between buffer zones, power scale parameters are generated. Outlier detection and removal are performed on the power scale parameters to generate an effective power scale parameter set; based on the effective power scale parameter set, the global statistical dispersion is determined. Based on the global statistical dispersion, the sensitivity level of the buffer behavior of the power battery injection valve to changes in heating power is determined.

[0020] Specifically, heating tests were conducted on battery cells of the same model using multiple preset heating powers. Continuous data sequences of cell temperature changes over time were simultaneously collected. The timing of the injection valve occurrence was precisely calibrated based on voltage signal abrupt changes and gas release detection signals. Simultaneously, the timing of thermal runaway was determined by combining the temperature step rise phenomenon and voltage zero-return characteristics. The raw temperature data was smoothed using a moving average filtering method to effectively suppress sensor noise and environmental interference. Subsequently, the analysis time interval for the post-injection valve buffering behavior was strictly defined, starting from the time of injection valve occurrence and ending at the time of thermal runaway. Based on the principle of thermal balance, and combining ambient temperature, the equivalent heat capacity and equivalent heat dissipation coefficient of the battery cell (which can be obtained by heating and calibrating a standard aluminum block under the same experimental conditions, or by fitting and calculating the temperature data of a battery cell that has not experienced thermal runaway), the instantaneous difference between heating power and heat dissipation power is integrated over time. This calculates the total net energy accumulation from the start of heating to the moment thermal runaway occurs, as well as the net energy accumulation within the buffer zone during the post-vacuum analysis period. This energy calculation process also supports the use of other engineering-validated thermal models or energy estimation methods. Furthermore, a power-scale parameter is constructed, calculated by multiplying the total net energy accumulation by the corresponding heating power and then dividing by the net energy accumulation within the buffer zone. A smaller parameter value indicates a higher proportion of accumulated energy in the post-vacuum buffer stage and a relatively longer buffer time window, thus quantitatively characterizing the intensity of the impact of applied heating power on buffer behavior. Outlier detection and removal were performed on the power scale parameters obtained from multiple sets of experiments to form a set of effective parameters, and the standard deviation was calculated as the global statistical dispersion: a smaller standard deviation indicates that the buffering behavior changes smoothly within the test power range and is not sensitive to heating power; a larger standard deviation indicates that the buffering behavior changes significantly with power and is highly sensitive to heating power.

[0021] Understandably, by accurately quantifying the net energy accumulation process through a thermal balance model, the influence of environmental heat dissipation interference and measurement noise is eliminated, thereby improving the physical reliability and experimental repeatability of parameter calculations. The constructed power-scale parameters achieve a normalized correlation between energy accumulation characteristics and heating power, eliminating differences in the scale of a single experiment and focusing on the essential impact of power variables on the buffering mechanism. The parameter values ​​reflect the length of the buffer window. Based on a dual-layer judgment logic of global statistical dispersion and local dispersion distribution, not only are material-level differences distinguished, but the critical power range in which the buffering behavior causes abrupt changes in power response is also identified. This provides a basis for refined segmented optimization of the thermal management system, such as dynamic threshold setting and power fluctuation suppression.

[0022] In some embodiments of this application, determining the post-spray valve buffer analysis time interval includes: The starting point is the moment when the spray valve occurs, and the ending point is the moment when thermal runaway occurs. The continuous time window between the starting point and the ending point is the post-spray valve buffer analysis time interval.

[0023] Specifically, key time points are precisely calibrated based on experimental monitoring data: the occurrence of the valve ejection is confirmed through dual verification using voltage signal abrupt change characteristics (such as a voltage drop exceeding a preset threshold) and gas release detection signals; the occurrence of thermal runaway is determined comprehensively based on the step-like increase in temperature data (temperature change rate consistently exceeding five degrees Celsius per second) and the voltage signal returning to zero. Subsequently, the verified valve ejection occurrence time is used as the starting point of the analysis interval, and the occurrence of thermal runaway is used as the ending point. The continuous and uninterrupted time period between the starting and ending points is clearly defined as the post-valve ejection buffer analysis time interval. This interval strictly corresponds to the complete thermal buffering stage from the opening of the battery safety valve to the outbreak of the violent thermal runaway reaction.

[0024] Understandably, by accurately defining the post-spray valve buffer analysis time interval through multi-feature cross-validation, the objectivity of interval calibration is improved; by clearly defining the continuous time window from the occurrence of the spray valve to the start of thermal runaway, interference from irrelevant stages is isolated, avoiding the calculation deviation of net energy accumulation caused by the ambiguity of interval definition.

[0025] In some embodiments of this application, determining the total net energy accumulation from the start of heating to the occurrence of thermal runaway includes: Based on temperature time-series data, the integral value of the instantaneous difference between heating power and heat dissipation power in the time domain from the start of heating to the occurrence of thermal runaway is determined as the net energy accumulation over the entire process.

[0026] In some embodiments of this application, determining the net energy accumulation between buffer zones within the post-spray valve buffer analysis time interval includes: Based on the temperature time series data, the integral value of the instantaneous difference between heating power and heat dissipation power in the time domain within the buffer analysis time interval after the spray valve is determined as the net energy accumulation between the buffer zones.

[0027] The expression for the instantaneous difference between heating power and heat dissipation power is as follows: ; Where mc is the equivalent heat capacity of the battery cell; dT / dt is the rate of change of the battery cell temperature with respect to time; P is the external heating power; h is the equivalent total heat dissipation coefficient between the battery cell and the environment; T is the real-time temperature of the battery cell; T0 is the ambient temperature; and t is time.

[0028] The expression for the integral value in the time domain is: ; Where E_net(t) is the net energy accumulation of the battery cell from the start of heating (t=0) to time t; P is the external heating power; h is the equivalent total heat dissipation coefficient between the battery cell and the environment; T is the real-time temperature of the battery cell; and T0 is the ambient temperature.

[0029] Specifically, key thermophysical parameters are obtained through standardized calibration: Under the same experimental environment and heating power conditions, a standard aluminum block is heated and its temperature time-series data is recorded. Based on the thermal balance relationship, the equivalent heat capacity of the battery cell and the equivalent total heat dissipation coefficient between the battery cell and the environment are calculated. This parameter can also be obtained by fitting and calibrating the temperature data of the same model of battery cell that has not experienced thermal runaway during the heating stage. The ambient temperature is monitored and recorded in real time by a high-precision temperature and humidity sensor. Subsequently, based on the temperature time-series data after moving average filtering, the rate of change of battery cell temperature with respect to time is calculated. According to the thermal balance principle, the instantaneous difference between heating power and heat dissipation power is characterized by the product of the equivalent heat capacity of the battery cell and the rate of temperature change. At the same time, this difference is also equal to the external heating power minus the product of the equivalent heat dissipation coefficient and the difference between the real-time temperature of the battery cell and the ambient temperature. Perform time-domain numerical integration (using trapezoidal integration) on the instantaneous difference between the heating start time and the thermal runaway occurrence time. The result of the integration is the net energy accumulation over the entire process. Similarly, perform time-domain numerical integration on the instantaneous difference within the buffer analysis time interval after the spray valve. The result of the integration is the net energy accumulation between the buffer zones.

[0030] Understandably, the equivalent heat capacity and heat dissipation coefficient obtained through calibration accurately quantify the impact of environmental heat dissipation, eliminating the systematic bias caused by neglecting heat dissipation in the traditional simple cumulative heat dissipation method; the numerical integration process is based on filtered measured temperature data, which improves the physical reliability and experimental repeatability of the calculation results; the independent and accurate calculation of energy throughout the process and between the buffer provides a highly reliable input for the construction of power scale parameters, making the buffer behavior analysis closely fit the physical mechanism of thermal runaway evolution, and avoiding sensitivity misjudgments caused by energy calculation distortion.

[0031] In some embodiments of this application, the generation of power scale parameters includes: The power scale parameter is generated by calculating the ratio of the product of the total net energy accumulation and the corresponding heating power to the net energy accumulation between the buffer zones.

[0032] The expression for the power scale parameter is as follows: ; in, P * represents power scale parameters; E 0 For the accumulation of energy throughout the heating process; P This refers to the external heating power. E 1 represents the energy accumulation in the analysis region.

[0033] Specifically, the process calls upon the precisely calculated cumulative net energy throughout the heating process, the external heating power value under the corresponding experimental conditions, and the cumulative net energy in the buffer zone within the post-vacuum analysis time interval. Then, a normalization operation is performed: the cumulative net energy throughout the heating process is multiplied by the external heating power, and the product is divided by the cumulative net energy in the buffer zone to generate a single numerical power scale parameter. This achieves dimensional unification through the energy-power coupling relationship. The parameter's value is clearly negatively correlated with the length of the post-vacuum buffer time window—a smaller parameter value indicates a higher proportion of accumulated energy in the buffer zone under the same heating power, resulting in more stable buffering behavior and a longer window; a larger parameter value indicates that the buffer window is more susceptible to shortening due to power disturbances, and thus exhibits greater sensitivity.

[0034] Understandably, by normalizing the correlation between total energy, heating power, and energy in the buffer zone, non-power factors such as individual cell differences and environmental fluctuations are eliminated, and the essential influence mechanism of heating power variables on buffer behavior is accurately focused. The negative correlation between parameter values ​​and buffer window length makes the sensitivity representation intuitive and clear, avoiding the risk of misjudgment caused by relying solely on the absolute value comparison of buffer duration in traditional methods.

[0035] In some embodiments of this application, when obtaining cell test data of several power batteries under different heating power conditions, the following are included: Standardized heating of the same type of power battery cells is performed under several preset heating power conditions; Collect temperature time-series data; determine the timing of the valve release based on voltage signal abrupt change characteristics and gas release detection signals; determine the timing of thermal runaway based on temperature step rise characteristics and voltage zeroing characteristics.

[0036] Specifically, standardized heating experiments were conducted on battery cells of the same model and batch in a constant temperature environment chamber. Multiple preset heating power levels (such as 100W, 200W, 300W, and 400W) were set, and no fewer than five sets of samples were tested at each power level to ensure statistical validity. Throughout the experiment, high-precision K-type thermocouples were used to continuously collect the core temperature time-series data of the battery cells at a sampling frequency of 1 Hz, and the voltage signal and the gas release pressure signal of the safety valve were recorded simultaneously. The determination of the valve release time adopted a dual verification mechanism: when the voltage signal showed a sudden change (the drop exceeded 5% of the rated voltage) and the gas release pressure signal simultaneously showed a step increase (the pressure increment exceeded 10 kPa from the baseline), the time point was marked as the valve release time. The determination of the thermal runaway time also adopted a dual feature recognition: when the temperature data showed a continuous step increase (the temperature change rate was higher than 5 degrees Celsius per second for 3 consecutive seconds) and the voltage signal returned to zero (below 0.1 volts), the time point was marked as the thermal runaway time.

[0037] Understandably, the statistical representativeness and operational coverage of the data were improved through multi-level power gradient design and multiple sets of sample tests. The timing of the valve release was calibrated by dual verification of voltage surge and gas release signals, and the timing of thermal runaway was identified by dual features of temperature step and voltage zeroing. This avoided the risk of misjudgment based on a single signal and ensured the objectivity and repeatability of the calibration at key time points. The standardized experimental procedure and high-precision synchronous acquisition mechanism eliminated interference factors such as environmental fluctuations and sensor drift, providing a high-quality and highly consistent data foundation for subsequent steps such as net energy accumulation calculation and power scale parameter construction.

[0038] In some embodiments of this application, determining the global statistical dispersion includes: Based on the effective power scale parameter set, the standard deviation of all power scale parameters within the effective power scale parameter set is determined, and the standard deviation is used as the global statistical dispersion.

[0039] Specifically, the effective power scale parameter set generated after outlier detection and removal is invoked. This set completely contains all effective power scale parameter values ​​measured under different heating power conditions. Subsequently, the arithmetic mean of this set is calculated strictly according to statistical standards. Then, the sum of squares of the deviations of each parameter value from the mean is calculated, divided by the total number of parameters (using the population standard deviation calculation method, since the set represents a complete test sample), and the square root is taken. The resulting value is the standard deviation. This standard deviation is explicitly defined as the global statistical dispersion, and its magnitude directly quantifies the overall fluctuation and distribution dispersion characteristics of the power scale parameters within the test power range.

[0040] In some embodiments of this application, determining the sensitivity level of the post-vacuum behavior of the power battery to changes in heating power includes: Compare the statistical dispersion with the dispersion threshold; When the statistical dispersion is less than or equal to the dispersion threshold, it is determined that the post-spray valve buffer behavior is not sensitive to changes in heating power. When the statistical dispersion is greater than the dispersion threshold, it is determined that the post-spray valve buffer behavior is sensitive to changes in heating power.

[0041] Specifically, the system retrieves the global statistical dispersion value calculated after outlier removal and obtains a preset dispersion threshold. This threshold is based on the upper limit of the dispersion statistics of samples with stable buffering behavior verified in the historical experimental database of the same battery model (such as the upper bound of the 95% confidence interval) and combined with safety margin calibration. It also supports dynamic adjustment and version management according to industry standards or enterprise specifications. A numerical comparison is then performed: if the global statistical dispersion is less than or equal to the dispersion threshold, it is determined that the post-vacuum buffering behavior of this type of power battery is not sensitive to changes in heating power, indicating that the buffering time window remains relatively stable within the test power range and is only slightly affected by power fluctuations. If the global statistical dispersion is greater than the dispersion threshold, it is determined that the post-vacuum buffering behavior is sensitive to changes in heating power, indicating that the buffering window length fluctuates significantly with changes in heating power, and small power disturbances may cause a sudden shortening of the buffering time, posing an accelerated risk of thermal runaway.

[0042] Understandably, by using scientific thresholds to achieve objective quantitative classification of buffer behavior power sensitivity, the traditional subjective assessment method that relies on manual experience to compare buffer durations is abandoned, thus improving the rigor, repeatability, and cross-experiment comparability of the judgment results. The threshold setting is closely based on historical data statistical patterns and safety margins, ensuring that the judgment criteria are highly consistent with the actual thermal safety characteristics of the battery and distinguishing material-level differences. The judgment conclusions directly drive the differentiated implementation of thermal management strategies—automatically triggering power fluctuation suppression mechanisms and dynamically tightening thermal runaway early warning thresholds for sensitive systems, while optimizing monitoring logic and improving operating efficiency for insensitive systems, thus achieving a precise balance between safety and performance.

[0043] In some embodiments of this application, the heating power range is divided into multiple continuous power sub-intervals, the local statistical dispersion of the effective power scale parameter in each power sub-interval is determined, and a dispersion distribution sequence is generated.

[0044] In some embodiments of this application, the difference value of the local statistical dispersion corresponding to adjacent power sub-intervals is determined according to the dispersion distribution sequence. When the difference value is greater than the difference change threshold, it is determined that the post-vacuum behavior of the corresponding power boundary region is sensitive to the change of heating power.

[0045] Specifically, the heating power range covered by the experiment (e.g., 100 watts to 400 watts) is divided into multiple continuous power sub-intervals at equal intervals (e.g., every 50 watts is an interval, forming [100, 150), [150, 200), etc.). For each sub-interval, all effective power scale parameters falling within that power range are selected, and their standard deviation is calculated as the local statistical dispersion of that interval. The sub-intervals are arranged in ascending order of power, and their local statistical dispersions are combined sequentially to generate a dispersion distribution sequence. Subsequently, the absolute difference of the local statistical dispersion of adjacent sub-intervals in the sequence is calculated as the difference value. The threshold for this difference change is preset to 0.05 (determined based on the 95% confidence interval of the statistical distribution of historical data of the same battery model, and also supports dynamic calibration according to the battery system characteristics). When a certain difference value is greater than this threshold, it is determined that the post-vacuuming behavior of the valve in the boundary region (e.g., the 150 watt critical point) of the corresponding two power sub-intervals is sensitive to changes in heating power, indicating that the buffering mechanism has changed significantly near the power critical point, and the region is automatically marked as a high-risk power boundary region.

[0046] Understandably, by accurately capturing the nonlinear evolution of buffer behavior in response to heating power and the critical point of sensitive abrupt changes through the discrete distribution sequence, local high-risk areas that may be masked by global judgment are identified; the generated power boundary sensitive area markers provide a basis for refined segmented optimization of the thermal management system—automatically triggering power fluctuation suppression strategies, dynamically tightening thermal runaway early warning thresholds, or adding buffer time redundancy in sensitive boundary areas, thereby improving the pertinence and effectiveness of risk prevention and control.

[0047] Example 1: Twenty samples each of high-nickel ternary lithium batteries (NCM811, nominal capacity 50Ah) and lithium iron phosphate batteries (LFP, nominal capacity 50Ah) were selected and a standardized heating experiment was conducted in a constant temperature 25℃ environmental chamber. Four heating power levels were set: 100W, 200W, 300W, and 400W, with five samples tested at each power level. Temperature time-series data (sampling frequency 1Hz), voltage signals, and safety valve gas pressure signals were collected simultaneously. Taking one sample of NCM811 at 400W power as an example: the voltage suddenly dropped by 5.3% (exceeding the threshold by 5%) and the gas pressure suddenly increased by 12.5kPa (exceeding the baseline by 10kPa), and the calibration time of the valve ejection was 286 seconds; the temperature change rate reached 6.2℃ / s for 3 consecutive seconds (exceeding 5℃ / s) and the voltage returned to zero (0.08V), and the calibration time of thermal runaway was 318 seconds. The LFP sample with the same power had a valve ejection time of 392 seconds and a thermal runaway time of 415 seconds. The raw temperature data were filtered using a 5-point moving average to effectively suppress high-frequency noise.

[0048] Step 1: Define a continuous time window, starting from the moment the valve opens and ending at the moment thermal runaway occurs: for NCM811 samples, the buffer zone is [286 seconds, 318 seconds], with a duration of 32 seconds; for LFP samples, the buffer zone is [392 seconds, 415 seconds], with a duration of 23 seconds. This interval strictly corresponds to the complete thermal buffering phase from the opening of the safety valve to the violent thermal runaway reaction.

[0049] Step 2: Obtain thermal properties using standard aluminum blocks: NCM811 equivalent heat capacity 1850 J / ℃, equivalent heat dissipation coefficient 8.2 W / ℃; LFP equivalent heat capacity 1920 J / ℃, equivalent heat dissipation coefficient 7.9 W / ℃. Integrate the instantaneous net power (heating power - heat dissipation power) in the time domain according to the heat balance equation: NCM811 (400W sample): Net energy accumulation over the entire range (0–318 seconds) E0 = 118.6 kJ, and net energy accumulation during the buffer period (286–318 seconds) E1 = 18.3 kJ; LFP (400W sample): E0 = 152.4kJ throughout (0–415 seconds), E1 = 38.7kJ between buffer (392–415 seconds).

[0050] Step 3: Perform ratio calculation: P =(E0×P) / E1.

[0051] NCM811 (400W): (118.6×400) / 18.3=2592.35; LFP (400W): (152.4×400) / 38.7=1577.78.

[0052] The valid data from all 20 samples were summarized (two outliers were removed using the 3σ principle): the mean of the NCM811 power scale parameter set was 2150.4, and the mean of the LFP set was 1285.6.

[0053] Step 4: Calculate the standard deviation of the effective power scale parameter set: NCM811: Standard deviation σ = 648.2; LFP: Standard deviation σ = 62.1.

[0054] This standard deviation is the global statistical dispersion, which quantitatively characterizes the degree of fluctuation of the parameter within the test power range.

[0055] Step 5: Set the dispersion threshold σ_th = 150 (calibrated based on the 95% confidence interval of the historical stable system): NCM811: 648.2>150, indicating that the post-spray valve buffer behavior is sensitive to changes in heating power; LFP: 62.1 < 150, indicating that the post-spray valve buffer behavior is insensitive to changes in heating power.

[0056] Step 6: Divide the heating power range into three continuous sub-intervals: [100W, 200W), [200W, 300W), and [300W, 400W]. Calculate the local standard deviation of the power scale parameter for the five groups of samples within each interval: [100W, 200W): σ1 = 128.5; [200W, 300W): σ² = 315.3; [300W, 400W]: σ3=582.7.

[0057] Generate a discrete distribution sequence: [128.5, 315.3, 582.7]. Calculate the difference between adjacent intervals: 200W boundary: |315.3-128.5|=186.8; 300W boundary: |582.7-315.3|=267.4.

[0058] The difference change threshold Δ_th is set to 50. Since 186.8>50 and 267.4>50, the two power boundary areas of 200W and 300W are determined to be abrupt change areas where the post-spray valve buffer behavior is sensitive to changes in heating power.

[0059] In summary, by constructing power-scale parameters and quantifying their statistical dispersion, an objective and quantifiable determination of the degree to which heating power affects the buffering behavior after the nozzle is achieved. Based on the precise definition of the buffering analysis interval at the nozzle and thermal runaway moments, and by integrating ambient temperature and equivalent heat dissipation coefficients to physically correct the temperature time-series data, the accuracy of net energy accumulation calculation is improved. By normalizing and correlating the heating power, total net energy accumulation, and net energy accumulation between buffer zones, the generated power-scale parameters remove the absolute energy differences from a single experiment, highlighting the relative impact mechanism of power changes on buffering behavior. Through intelligent outlier removal and global statistical dispersion analysis of the parameter set, sensitivity levels are differentiated; high dispersion indicates that buffering behavior is highly sensitive to power fluctuations, requiring strengthened power fluctuation suppression and dynamic adjustment of early warning thresholds in thermal management; low dispersion reflects stable buffering behavior, simplifying thermal runaway prevention strategies.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for determining the power sensitivity of the post-injection valve buffer behavior of a power battery, characterized in that, include: Cell detection data of several power batteries under different heating power conditions are obtained. The cell detection data of each power battery includes temperature time series data, valve injection time and thermal runaway time. The temperature time series data is then processed by moving average filtering. Based on the timing of the valve ejection and thermal runaway, the post-valve buffer analysis time interval is determined; based on the ambient temperature, equivalent heat dissipation coefficient, and the filtered temperature time series data, the total net energy accumulation from the heating start time to the thermal runaway time is determined, as well as the net energy accumulation between buffer zones within the post-valve buffer analysis time interval; based on the corresponding heating power, the total net energy accumulation, and the net energy accumulation between buffer zones, power scale parameters are generated. Outlier detection and removal are performed on the power scale parameters to generate an effective power scale parameter set; the global statistical dispersion is determined based on the effective power scale parameter set. Based on the global statistical dispersion, the sensitivity level of the power battery's post-valve buffer behavior to changes in heating power is determined.

2. The power sensitivity determination method for the post-injection valve buffer behavior of a power battery according to claim 1, characterized in that, When determining the post-spray valve buffer analysis time interval, it includes: The continuous time window between the start point and the end point, taking the time when the valve occurs as the starting point and the time when thermal runaway occurs as the ending point, is the post-valve buffer analysis time interval.

3. The power sensitivity determination method for the post-injection valve buffer behavior of a power battery according to claim 2, characterized in that, When determining the total net energy accumulation from the start of heating to the occurrence of thermal runaway, the following should be included: Based on the temperature time series data, the integral value of the instantaneous difference between heating power and heat dissipation power in the time domain from the start of heating to the occurrence of thermal runaway is determined as the total net energy accumulation.

4. The power sensitivity determination method for the post-injection valve buffer behavior of a power battery according to claim 3, characterized in that, Determining the net energy accumulation between buffer zones within the post-spray valve buffer analysis time interval includes: Based on the temperature time series data, the integral value of the instantaneous difference between heating power and heat dissipation power in the time domain within the buffer analysis time interval after the spray valve is determined as the net energy accumulation in the buffer zone.

5. The power sensitivity determination method for the post-injection valve buffer behavior of a power battery according to claim 4, characterized in that, When generating power scale parameters, the following are included: The power scale parameter is generated by calculating the ratio of the product of the total net energy accumulation and the corresponding heating power to the net energy accumulation between the buffer zones.

6. The power sensitivity determination method for the post-injection valve buffer behavior of a power battery according to claim 1, characterized in that, When acquiring cell test data for several power batteries under different heating power conditions, the following are included: Standardized heating of the same type of power battery cells is performed under several preset heating power conditions; Collect temperature time-series data; determine the timing of the valve release based on voltage signal abrupt change characteristics and gas release detection signals; determine the timing of thermal runaway based on temperature step rise characteristics and voltage zeroing characteristics.

7. The power sensitivity determination method for the post-injection valve buffer behavior of a power battery according to claim 1, characterized in that, Determining the global statistical dispersion includes: Based on the effective power scale parameter set, the standard deviation of all the power scale parameters within the effective power scale parameter set is determined, and the standard deviation is used as the global statistical dispersion.

8. The power sensitivity determination method for the post-injection valve buffer behavior of a power battery according to claim 1, characterized in that, When determining the sensitivity level of the buffer behavior of the power battery injection valve to changes in heating power, the following should be considered: The statistical dispersion is compared with the dispersion threshold; When the statistical dispersion is less than or equal to the dispersion threshold, it is determined that the post-spray valve buffer behavior is insensitive to changes in heating power. When the statistical dispersion is greater than the dispersion threshold, it is determined that the post-spray valve buffer behavior is sensitive to changes in heating power.

9. The power sensitivity determination method for the post-injection valve buffer behavior of a power battery according to claim 1, characterized in that, Also includes: The heating power range is divided into multiple continuous power sub-intervals, and the local statistical dispersion of the effective power scale parameter in each power sub-interval is determined to generate a dispersion distribution sequence.

10. The power sensitivity determination method for the post-injection valve buffer behavior of a power battery according to claim 9, characterized in that, Also includes: Based on the discrete distribution sequence, the difference value of the local statistical discreteness corresponding to adjacent power sub-intervals is determined. When the difference value is greater than the difference change threshold, it is determined that the post-vacuum behavior of the corresponding power boundary region is sensitive to the change of heating power.

Images