Battery thermal runaway early warning method based on multi-point temperature field

Abstract

The invention discloses a battery thermal runaway early warning method based on a multi-point temperature field. The method comprises the following steps: arranging a plurality of temperature sensors in a battery pack, and collecting temperature vectors; temperature field characteristic quantities are extracted based on the temperature vector, and the temperature field characteristic quantities comprise the maximum temperature, the space temperature difference, the maximum temperature rise rate and the temperature field distortion index; calculating a corresponding sliding statistic for each temperature field characteristic quantity; and a comprehensive risk index is calculated through the sliding statistics, and then graded early warning is carried out through a graded threshold value. According to the invention, the temperature field distortion index and the comprehensive risk index are constructed to realize battery thermal runaway early warning.

Description

Technical Field

[0001] This invention relates to the field of lithium-ion power battery safety management technology, and more specifically, to a battery thermal runaway early warning method based on a multi-point temperature field. Background Technology

[0002] Lithium-ion batteries may experience thermal runaway under conditions such as overcharging, internal short circuits, aging, high-temperature environments, and cooling failures. Thermal runaway typically includes the following stages: 1. Localized areas of the battery cell or electrode post first experience temperature rise; 2. Temperature diffuses outwards, creating a temperature gradient. 3. The rate of temperature rise gradually increases and a clear inflection point appears; 4. If no measures are taken, an irreversible thermal runaway process will occur.

[0003] Existing technologies mostly rely on single-point temperature, fixed threshold alarms, or changes in electrical parameters (voltage / internal resistance), but these methods cannot effectively identify spatial anomalies and dynamic trends in the early stages of thermal runaway, resulting in delayed warnings and high false alarm rates.

[0004] Therefore, it is necessary to develop a battery thermal runaway early warning method based on a multi-point temperature field.

[0005] The information disclosed in the background section of this invention is intended only to enhance the understanding of the general background of this invention, and should not be construed as an admission or in any way implying that such information constitutes prior art known to those skilled in the art. Summary of the Invention

[0006] This invention proposes a battery thermal runaway early warning method based on a multi-point temperature field, which can realize battery thermal runaway early warning by constructing a temperature field distortion index and a comprehensive risk index.

[0007] This disclosure provides a battery thermal runaway early warning method based on a multi-point temperature field, including: Multiple temperature sensors are arranged in the battery pack to collect temperature vectors; Temperature field features are extracted based on temperature vectors, including maximum temperature, spatial temperature difference, maximum temperature rise rate, and temperature field distortion index. Calculate the corresponding sliding statistics for each temperature field characteristic quantity; A comprehensive risk index is calculated using sliding statistics, and then graded early warnings are issued based on tiered thresholds.

[0008] Preferably, the maximum temperature is:

[0009] in, This is the maximum temperature.

[0010] Preferably, the temperature difference in the space is:

[0011] in, This refers to the temperature difference in space.

[0012] Preferably, the maximum temperature rise rate is:

[0013] in, This represents the maximum rate of temperature rise.

[0014] Preferably, the extraction of the temperature field distortion index includes: The first temperature field distortion index is constructed based on the Euclidean deviation of the reference temperature field; A second temperature field distortion index is constructed based on neighborhood gradient energy; A third temperature field distortion index is constructed based on the residual bias from principal component analysis; The temperature field distortion index is determined based on at least one of the first temperature field distortion index, the second temperature field distortion index, and the third temperature field distortion index.

[0015] Preferably, the first temperature field distortion index is:

[0016] in, This is the reference temperature field under normal operating conditions; The distortion index of the second temperature field is:

[0017] E is the set of adjacency relationships between sensors; The distortion index of the first temperature field is:

[0018] in Reconstruction results of the normal temperature field for PCA modeling.

[0019] Preferably, the corresponding sliding statistics are calculated for each temperature field characteristic quantity:

[0020] in, Let be the i-th temperature field characteristic at time t. , Let be the sliding statistic corresponding to the i-th temperature field characteristic at time t. , Let α be the sliding statistic corresponding to the i-th temperature field feature at time t-1, where α∈(0.01,0.1).

[0021] Preferably, the calculation of the comprehensive risk index using sliding statistics includes: Calculate the normalized characteristic quantities based on the temperature field characteristic quantities and the corresponding sliding statistics:

[0022] Constructing a risk fusion model:

[0023] Among them, w i As weight, ; The output of the risk fusion model is averaged to obtain the comprehensive risk index:

[0024] in, This is a comprehensive risk index.

[0025] Preferably, graded early warning based on graded thresholds includes: Set grading threshold ; when If so, a Level 1 warning will be issued, and the charging rate will be limited; when If so, a level two warning will be issued, limiting power and reducing load; when If the situation is detected, a Level 3 warning will be issued, and emergency protection measures will be implemented immediately, including cutting off high-voltage power, downgrading operation, and activating the fire extinguishing system.

[0026] The methods and apparatus of the present invention have other features and advantages that will be apparent from or will be set forth in detail in the accompanying drawings and following detailed description, which together serve to explain the particular principles of the invention. Attached Figure Description

[0027] The above and other objects, features and advantages of the present invention will become more apparent from the accompanying drawings, in which like reference numerals generally denote like parts.

[0028] Figure 1 A flowchart illustrating the steps of a battery thermal runaway early warning method based on a multi-point temperature field according to an embodiment of the present invention is shown.

[0029] Figure 2A diagram showing the arrangement of a multi-point temperature sensor according to an embodiment of the present invention is illustrated. Detailed Implementation

[0030] Preferred embodiments of the invention will now be described in more detail. While preferred embodiments of the invention are described below, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0031] To facilitate understanding of the solutions and effects of the embodiments of the present invention, a specific application example is given below. Those skilled in the art should understand that this example is merely for the purpose of understanding the present invention, and any specific details therein are not intended to limit the present invention in any way.

[0032] Example 1

[0033] Figure 1 A flowchart illustrating the steps of a battery thermal runaway early warning method based on a multi-point temperature field according to an embodiment of the present invention is shown.

[0034] like Figure 1 As shown, this battery thermal runaway early warning method based on multi-point temperature fields includes: Step 101: Arrange multiple temperature sensors in the battery pack to collect temperature vectors; Step 102: Extract temperature field features based on temperature vectors, including maximum temperature, spatial temperature difference, maximum temperature rise rate, and temperature field distortion index. Step 103: Calculate the corresponding sliding statistics for each temperature field characteristic quantity; Step 104: Calculate the comprehensive risk index using sliding statistics, and then use the tiered thresholds to issue tiered early warnings.

[0035] In one example, the maximum temperature is:

[0036] in, This is the maximum temperature.

[0037] In one example, the temperature difference in the space is:

[0038] in, This refers to the temperature difference in space.

[0039] In one example, the maximum rate of temperature rise is:

[0040] in, This represents the maximum rate of temperature rise.

[0041] In one example, extracting the temperature field distortion index includes: The first temperature field distortion index is constructed based on the Euclidean deviation of the reference temperature field; A second temperature field distortion index is constructed based on neighborhood gradient energy; A third temperature field distortion index is constructed based on the residual bias from principal component analysis; The temperature field distortion index is determined based on at least one of the first temperature field distortion index, the second temperature field distortion index, and the third temperature field distortion index.

[0042] In one example, the distortion index of the first temperature field is:

[0043] in, This is the reference temperature field under normal operating conditions; The distortion index of the second temperature field is:

[0044] E is the set of adjacency relationships between sensors; The distortion index of the first temperature field is:

[0045] in Reconstruction results of the normal temperature field for PCA modeling.

[0046] In one example, the corresponding sliding statistics are calculated for each temperature field characteristic:

[0047] in, Let be the i-th temperature field characteristic at time t. , Let be the sliding statistic corresponding to the i-th temperature field characteristic at time t. , Let α be the sliding statistic corresponding to the i-th temperature field feature at time t-1, where α∈(0.01,0.1).

[0048] In one example, calculating the composite risk index using the sliding statistic includes: Calculate the normalized characteristic quantities based on the temperature field characteristic quantities and the corresponding sliding statistics:

[0049] Constructing a risk fusion model:

[0050] Among them, wi As weight, ; The output of the risk fusion model is averaged to obtain the comprehensive risk index:

[0051] in, This is a comprehensive risk index.

[0052] In one example, tiered alerts based on tiered thresholds include: Set grading threshold ; when If so, a Level 1 warning will be issued, and the charging rate will be limited; when If so, a level two warning will be issued, limiting power and reducing load; when If the situation is detected, a Level 3 warning will be issued, and emergency protection measures will be implemented immediately, including cutting off high-voltage power, downgrading operation, and activating the fire extinguishing system.

[0053] Figure 2 A diagram showing the arrangement of a multi-point temperature sensor according to an embodiment of the present invention is illustrated.

[0054] Specifically, such as Figure 2 As shown, N temperature sensors (recommended N≥8, typically 8–32) should be placed in the critical areas of the battery pack. The number should be determined based on the following principles: 1. The larger the battery pack volume, the denser the battery pack layout. 2. Must cover the main surface of the battery cell, terminals, end plates, upper and lower layers of the enclosure, and air duct inlet / outlet; 3. The density of sampling points is higher in areas with drastic temperature changes; 4. All temperature points must be collected synchronously within the same sampling period.

[0055] The sampling period is 0.1–1 second, which can capture the trend of dT / dt changes. The acquisition results form a temperature vector:

[0056] The following four key indicators are extracted from the temperature field: 1) Maximum temperature

[0057] 2) Spatial temperature difference

[0058] It is used to reflect whether local hotspots have formed.

[0059] 3) Maximum temperature rise rate

[0060] Where Δt is the sampling period.

[0061] It is used to detect the significant increase in the rate of temperature rise during the early stages of thermal runaway.

[0062] 4) Temperature field distortion index

[0063] The distortion index is used to measure whether the temperature field deviates from the normal spatial pattern.

[0064] At least three feasible calculation methods should be provided: Used to identify anomalies in the spatial distribution of temperature fields.

[0065] (1) Euclidean deviation based on the reference temperature field:

[0066] in: : Reference temperature field under normal operating conditions (updated using a sliding window).

[0067] (2) Based on neighborhood gradient energy:

[0068] E is the set of adjacency relationships between sensors.

[0069] This method reflects the "roughness" of the temperature field.

[0070] (3) Residual bias based on principal component analysis:

[0071] in Reconstruction results of the normal temperature field for PCA modeling.

[0072] In practical applications, one or more combinations can be selected.

[0073] During the normal operation phase of the system, the sliding statistics are calculated for each characteristic quantity f(t):

[0074] Where α∈(0.01,0.1).

[0075] Adaptive threshold:

[0076] The typical range of λ is 1.5–3, which can be calibrated according to the safety level.

[0077] This mechanism can avoid false alarms caused by factors such as ambient temperature, season, and aging.

[0078] Construct a comprehensive risk index: (1) Feature normalization

[0079] (2) Risk fusion model

[0080] Where: w_i is the weight, satisfying... It can be calibrated through experiments or automatically adjusted according to working conditions.

[0081] To suppress transient noise, R(t) can be calculated using a moving average:

[0082] Multi-level early warning judgment based on risk index: Set the grading threshold:

[0083] when If so, a Level 1 warning will be issued: 1. An initial trend of rising temperatures has emerged. 2. Can prompt you to check the cooling system. 3. Limit charging rate when If so, a Level 2 warning will be issued: 1. Significant distortion in the temperature field. 2. The rate of temperature rise is accelerated. 3. The system enters controlled operation. 4. Limit power and reduce load. when Then a level three warning will be issued: 1. High risk of thermal runaway 2. Immediately implement emergency protection: ①. Disconnect the high-voltage power. ②. Degraded operation ③. Activate the fire suppression system (ship / energy storage) The coordinated protection and control measures following the early warning include: Different control signals are triggered based on different warning levels: PWM regulation of cooling pump / fan (Level 1); EMS / VCU control reduces power consumption of the entire ship / vehicle (level 2); The ship's electrical system triggered a "section-level isolation" (level 3). The energy storage station triggered the fire suppression system and disconnected the fault cluster (Level 3); The internal power failure module (fuse / solid-state switch) of the battery pack immediately activates (level three).

[0084] The coordinated execution strategy follows: Time-delay protection (such as delayed shutdown of the fan); Redundancy check (triggered only when R(t) continuously exceeds the threshold). Scenario adaptation (strategy for differentiating between sailing, docking, and charging status).

[0085] Those skilled in the art should understand that the above description of the embodiments of the present invention is only intended to illustrate the beneficial effects of the embodiments of the present invention, and is not intended to limit the embodiments of the present invention to any of the examples given.

[0086] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments.

Claims

1. A battery thermal runaway early warning method based on multi-point temperature field, characterized in that, include: Multiple temperature sensors are arranged in the battery pack to collect temperature vectors; Temperature field features are extracted based on temperature vectors, including maximum temperature, spatial temperature difference, maximum temperature rise rate, and temperature field distortion index. Calculate the corresponding sliding statistics for each temperature field characteristic quantity; A comprehensive risk index is calculated using sliding statistics, and then graded early warnings are issued based on tiered thresholds.

2. The battery thermal runaway early warning method based on multi-point temperature field according to claim 1, wherein, The maximum temperature is: in, This is the maximum temperature.

3. The battery thermal runaway early warning method based on multi-point temperature field according to claim 1, wherein, The temperature difference in the space is: in, This refers to the temperature difference in space.

4. The battery thermal runaway early warning method based on multi-point temperature field according to claim 1, wherein, The maximum rate of temperature rise is: in, This represents the maximum rate of temperature rise.

5. The battery thermal runaway early warning method based on multi-point temperature field according to claim 1, wherein, The temperature field distortion index is extracted as follows: The first temperature field distortion index is constructed based on the Euclidean deviation of the reference temperature field; A second temperature field distortion index is constructed based on neighborhood gradient energy; A third temperature field distortion index is constructed based on the residual bias from principal component analysis; The temperature field distortion index is determined based on at least one of the first temperature field distortion index, the second temperature field distortion index, and the third temperature field distortion index.

6. The battery thermal runaway early warning method based on multi-point temperature field according to claim 5, wherein, The distortion index of the first temperature field is: in, This is the reference temperature field under normal operating conditions; The distortion index of the second temperature field is: E is the set of adjacency relationships between sensors; The distortion index of the first temperature field is: in Reconstruction results of the normal temperature field for PCA modeling.

7. The battery thermal runaway early warning method based on multi-point temperature field according to claim 1, wherein, Calculate the corresponding sliding statistics for each temperature field characteristic: in, Let be the i-th temperature field characteristic at time t. , Let be the sliding statistic corresponding to the i-th temperature field characteristic at time t. , Let α be the sliding statistic corresponding to the i-th temperature field feature at time t-1, where α∈(0.01,0.1).

8. The battery thermal runaway early warning method based on multi-point temperature field according to claim 1, wherein, The comprehensive risk index is calculated using sliding statistics, including: Calculate the normalized characteristic quantities based on the temperature field characteristic quantities and the corresponding sliding statistics: Constructing a risk fusion model: Among them, w i As weight, ; The output of the risk fusion model is averaged to obtain the comprehensive risk index: in, This is a comprehensive risk index.

9. The battery thermal runaway early warning method based on multi-point temperature field according to claim 1, wherein, Graded early warning based on tiered thresholds includes: Set grading threshold ; when If so, a Level 1 warning will be issued, and the charging rate will be limited; when If so, a level two warning will be issued, limiting power and reducing load; when If the situation is detected, a Level 3 warning will be issued, and emergency protection measures will be implemented immediately, including cutting off high-voltage power, downgrading operation, and activating the fire extinguishing system.

Images