Battery module thermal safety cooperative control method and system

Abstract

The present application relates to the technical field of battery module heat control, and particularly relates to a battery module heat safety cooperative control method and system, the method comprising: acquiring a thermal characteristic parameter and constructing a time-series heat state benchmark set; constructing a module heat generation response quantity based on the time-series heat state benchmark set; establishing a heat coupling candidate connection between modules, calculating the heat coupling strength of the screened heat coupling candidate connection, and forming a heat coupling mapping relationship between modules; dividing a three-dimensional heat field calculation grid unit according to the heat coupling strength and the spatial position of the box body, and sequentially reconstructing a dynamic heat distribution map inside the box body frame by frame according to the time sequence; and generating a corresponding battery module heat dissipation control strategy based on the heat migration path and the heat accumulation center. The present application can realize partition linkage regulation and control for a heat diffusion chain, can both inhibit the continuous output of a heat source and weaken the heat propagation along a path, and can reduce the risk of heat accumulation in a downstream area in advance, thereby improving the pertinence and timeliness of control.

Description

Technical Field

[0001] This invention relates to the field of battery module thermal control technology, specifically to a method and system for collaborative control of battery module thermal safety. Background Technology

[0002] Submersible energy storage boxes are widely used in new energy storage, microgrids, and industrial energy management scenarios due to their high heat exchange efficiency and high integration. These boxes typically house a large number of battery modules densely, which can easily lead to abnormal localized temperature rises under high-rate charging / discharging or localized heat exchange constraints. Because of the compact module arrangement and complex medium flow direction, localized heat not only accumulates near the abnormal module but may also diffuse along the medium flow direction to adjacent modules, thus affecting the overall thermal safety of the storage box.

[0003] In existing technologies, the thermal status monitoring of battery modules in immersion energy storage boxes is mostly based on a small number of temperature sampling points combined with voltage and current parameters. This usually only reflects local discrete temperature information and is difficult to simultaneously characterize the temperature difference in different areas of the module surface, the heat transfer difference between the front and back sides of the medium, and the thermal field change process over a continuous period of time. Therefore, it is difficult to form a unified thermal status basic data corresponding to the module position, medium state, and time changes.

[0004] Furthermore, existing technologies for analyzing the thermal behavior inside the enclosure mostly rely on static temperature rise comparisons, threshold alarms, or simple empirical judgments. They lack quantitative characterization of the thermal coupling relationship between modules, making it difficult to identify the migration path of heat between adjacent modules and the evolution process of local heat re-accumulation. Therefore, it is also difficult to accurately reconstruct the dynamic thermal distribution map inside the enclosure.

[0005] In terms of control, existing methods mostly adopt overall enhanced heat dissipation, fixed flow circulation, or single module overheating and load reduction. They are usually passively adjusted after high temperature has already formed, making it difficult to distinguish between heat source modules, heat transfer modules, and downstream risk modules. They cannot implement source heat suppression, path heat interception, and downstream pre-cooling control for different areas, which can easily lead to control lag and unreasonable allocation of heat dissipation resources.

[0006] Therefore, it is necessary to propose a collaborative control method for battery module thermal safety to achieve precise monitoring, dynamic reconstruction, and zoned heat dissipation regulation of the internal thermal field of the immersion energy storage box. Summary of the Invention

[0007] To address the aforementioned issues, this invention provides a collaborative control method and system for battery module thermal safety. By directly converting thermal field analysis results into heat dissipation control criteria, it distinguishes between the core heat source module, channel module, and downstream risk module, and implements source heat suppression control, path heat interception control, and downstream pre-cooling control respectively. This achieves zoned linkage regulation oriented towards the heat diffusion chain, which can suppress continuous heat output from the heat source, weaken heat propagation along the path, and reduce the risk of heat accumulation in downstream areas in advance.

[0008] This invention employs a battery module thermal safety collaborative control method to construct a three-dimensional acquisition unit for the battery module in an immersion energy storage box, acquire thermal characteristic parameters, and construct a time-series thermal state reference set.

[0009] The module's heating response is constructed based on a time-series thermal state benchmark set;

[0010] Based on the spatial adjacency relationship, medium flow relationship, and module heating response relationship of battery modules, candidate thermal coupling connections between modules are established. The thermal coupling strength of the selected candidate thermal coupling connections is calculated to form a thermal coupling mapping relationship between modules.

[0011] Based on the thermal coupling strength and the spatial position of the box, a three-dimensional thermal field calculation grid is divided, and the dynamic thermal distribution map inside the box is reconstructed frame by frame in time sequence.

[0012] Heat migration paths and heat accumulation centers are obtained based on dynamic heat distribution map sequences;

[0013] Based on the heat migration path and heat accumulation center, a corresponding battery module heat dissipation control strategy is generated, and the heat dissipation actuator in the immersion energy storage box is controlled to implement zoned heat dissipation adjustment of the target battery module.

[0014] As a further description of the above technical solution: the method for constructing the stereoscopic acquisition unit includes:

[0015] A unified three-dimensional coordinate system is established within the immersion energy storage box, and all battery modules are uniquely numbered according to their layer, column, and row arrangement to obtain the module number and its corresponding spatial coordinates.

[0016] For each numbered battery module, its outer contour surface is taken as the internal boundary, and then extended outward along the normal direction of the six faces of the battery module to the perpendicular plane of the line connecting the centers of adjacent battery modules; when there is no adjacent battery module in a certain normal direction, the outward extension boundary in that direction is extended to the inner wall of the energy storage box, thereby forming an independent three-dimensional acquisition unit surrounding the local space of the battery module and its surrounding cooling medium.

[0017] As a further description of the above technical solution: the method for obtaining thermal characteristic parameters and constructing a time-series thermal state reference set includes:

[0018] A thermal characteristic parameter acquisition sensor group is arranged in each three-dimensional acquisition unit. The thermal characteristic parameter acquisition sensor group includes a module temperature sensor, a current sensor, a voltage sensor, a cooling medium temperature sensor, and a flow rate sensor.

[0019] A unified sampling period is set, and the collected results are structured to form a time-series thermal state reference set. The time-series thermal state reference set includes module number, spatial coordinates, temperature values ​​of five modules, module current values, module voltage values, upstream medium temperature values, downstream medium temperature values, upstream medium velocity values, and downstream medium velocity values.

[0020] As a further description of the above technical solution: the module temperature sensor is set up as follows: five module temperature sensors are fixed on the outer surface of each battery module. The five module temperature sensors are respectively arranged at: the center of the upper end face near the positive electrode junction area, the geometric center of the module's main heat dissipation surface, the center of the lower end face near the negative electrode junction area, the middle of the narrow side on the current-facing side, and the middle of the narrow side on the back-current side. Each module temperature sensor is attached to the outer surface of the module housing through an insulating and thermally conductive fixing layer. The temperature sensing end of the module temperature sensor is directly attached to the module housing, and the wires are led out through a medium-resistant corrosion-resistant sheath.

[0021] As a further description of the above technical solution: the method for constructing the module heating response based on the time-series thermal state reference set is as follows: the temperature values ​​of the five modules are weighted and summed to obtain the module temperature characterization value of the battery module at the current moment, which is recorded as the module temperature characterization value. The difference between the module temperature characterization value at the current moment and the module temperature characterization value at the previous moment is calculated and divided by the time interval between two adjacent timestamps to obtain the temperature rise response of the battery module.

[0022] The module current value is multiplied by the module voltage value to obtain the module power characterization value. The module power characterization value and the temperature rise response value are normalized and then weighted and summed to obtain the module heating response value.

[0023] As a further description of the above technical solution: the method for establishing inter-module thermal coupling candidate connections includes: for any target battery module, selecting the P nearest neighboring battery modules in the length direction, width direction, and height direction as candidate neighboring modules according to their spatial location;

[0024] Then, based on the flow velocity values ​​of the medium on the front and back sides, the mainstream direction of the cooling medium around the target battery module is determined, and only candidate adjacent modules located downstream of the mainstream, with a lateral offset less than a preset distance threshold and a vertical height difference less than a preset height threshold are retained as effective coupling objects.

[0025] After obtaining the temperature rise response of the target battery module, the delay time when the temperature rise response of the effectively coupled object changes in the same direction over time is determined. When the temperature rise response changes in the same direction and the delay time falls within the preset effective response window, it is determined that there is a thermal coupling relationship between the two.

[0026] As a further description of the above technical solution: the method for calculating thermal coupling strength includes: for thermal coupling candidate connections, extracting characteristic parameters between the target battery module and the effective coupling object, the characteristic parameters including center distance, forward projection matching degree along the mainstream direction of the medium, lateral offset distance, difference in module temperature characterization value, temperature difference of the medium on the front side, temperature difference of the medium on the back side, difference in medium flow velocity, and temperature rise response delay time.

[0027] Among them, the center distance, lateral offset distance, medium flow velocity difference, and temperature rise response delay time are used as attenuation factors, and the forward projection matching degree along the mainstream direction of the medium, the difference in module temperature characterization value, the temperature difference of the medium on the front side, the temperature difference of the medium on the back side, and the medium flow velocity difference are used as enhancement factors. The original thermal coupling strength is obtained by weighting the characteristic parameters after dimensionless processing. Then, all the original thermal coupling strengths corresponding to the same target battery module are normalized to obtain the thermal coupling strength between the target battery module and each effective coupling object.

[0028] As a further description of the above technical solution: the method for dividing the three-dimensional thermal field calculation mesh elements includes:

[0029] Using the internal space boundary of the submersible energy storage box as the range, a three-dimensional mesh is divided along the length, width, and height of the box. Each mesh unit has unique spatial coordinates and volume number.

[0030] The space occupied by the battery module is directly mapped to the heat-generating physical area;

[0031] Obtain the liquid level position inside the submersible energy storage tank, mark the area below the liquid level as the effective submersible area, and mark the area above the liquid level as the non-submersible area;

[0032] For each battery module location, its module temperature characterization value, module heat response value, front-flow side medium temperature value, back-flow side medium temperature value, medium flow rate, and thermal coupling strength are all assigned to the corresponding grid cell.

[0033] As a further description of the above technical solution: the method for reconstructing the dynamic thermal distribution map inside the box frame by frame in chronological order includes:

[0034] Based on the time-series thermal state reference set at the same timestamp, the module temperature characterization value, the corresponding heat response of the battery module, the thermal coupling strength characteristics, and the liquid level immersion state characteristics of the area where the grid unit is located are extracted to construct the input feature set of the grid temperature prediction model; the input feature set is input into the grid temperature prediction model to output the module temperature characterization value of each grid unit at the next time step; a three-dimensional thermal distribution map at the corresponding timestamp is generated based on the module temperature characterization values ​​of all grid units at the next time step.

[0035] Then, the three-dimensional heat distribution map generated at the current timestamp is used as the basis for the historical heat field prediction at the next timestamp. The temperature characterization values ​​of the grid cell modules are predicted continuously in the order of timestamps to generate a dynamic heat distribution map sequence.

[0036] As a further description of the above technical solution: the method for obtaining the heat migration path is as follows:

[0037] In the 3D thermal distribution map corresponding to each time stamp, the grid cell with the largest temperature gradient is obtained, and the gradient change direction from the high temperature region to the low temperature region is defined as the local thermal migration direction. Then, the local thermal migration directions with continuous spatial location, continuous direction, and continuous temperature change under multiple consecutive time stamps are connected in series to form a heat migration path.

[0038] As a further description of the above technical solution: the method for obtaining the heat accumulation center is as follows: in the three-dimensional thermal distribution map corresponding to each time stamp, select grid cells with a temperature higher than the average temperature of the box at the current time as high-temperature candidate cells, then aggregate spatially adjacent high-temperature candidate cells, and mark high-temperature candidate cells with an aggregation number greater than or equal to 3 as high-temperature accumulation areas to form multiple high-temperature accumulation areas; calculate the average temperature, maximum temperature, volume, geometric center coordinates, and duration of each high-temperature accumulation area; define the high-temperature accumulation area whose duration reaches a preset duration and whose maximum temperature continues to rise as the heat accumulation center.

[0039] As a further description of the above technical solution: the method for generating the corresponding battery module heat dissipation control strategy includes:

[0040] Based on the positional matching relationship between the geometric center coordinates of the heat accumulation center and the center coordinates of each battery module, the core heat source module corresponding to each heat accumulation center is determined.

[0041] Based on the heat flow direction of the heat migration path, the heat migration path is divided into the source section, the migration section and the accumulation section. The battery modules that fall within the coverage area of ​​each path section are marked as source modules, channel modules and downstream risk modules, respectively, to form corresponding heat dissipation control zones.

[0042] For source modules, implement source heat suppression control; for channel modules, implement path heat interception control; and for downstream risk modules, implement downstream pre-cooling control.

[0043] As a further description of the above technical solution: the method of dividing the heat migration path into a source segment, a migration segment, and an accumulation segment includes:

[0044] Extract all grid cells that constitute the heat migration path along the heat migration direction, and treat each grid cell as a path point to obtain the path point sequence P1, P2, ..., Pn, where P1 is the starting path point and Pn is the ending path point.

[0045] Extract the temperature value of each path point and the proportion of high-temperature candidate units in the preset neighborhood around each path point; determine the continuous path segment starting from the starting path point with a temperature higher than the preset source temperature threshold as the source segment; determine the continuous path segment immediately following the source segment with the temperature change of adjacent path points less than or equal to the preset temperature decay threshold and the change in the proportion of high-temperature candidate units in the preset neighborhood around the adjacent path points not exceeding the preset proportion change threshold as the migration segment; and divide the path immediately following the migration segment into the accumulation segment.

[0046] A battery module thermal safety collaborative control system, used to implement the aforementioned battery module thermal safety collaborative control method, includes:

[0047] The parameter acquisition module constructs a three-dimensional acquisition unit for the battery module inside the submersible energy storage box, acquires thermal characteristic parameters, and constructs a time-series thermal state reference set.

[0048] The response analysis module constructs the module's heating response quantity based on a time-series thermal state reference set;

[0049] The coupling mapping module establishes candidate thermal coupling connections between modules based on the spatial adjacency relationship, medium flow relationship, and module heating response relationship of the battery modules. It calculates the thermal coupling strength of the selected candidate thermal coupling connections to form a thermal coupling mapping relationship between modules.

[0050] The thermal reconstruction module divides the three-dimensional thermal field calculation grid units according to the thermal coupling strength and the spatial position of the box, and reconstructs the dynamic thermal distribution map inside the box frame by frame in time sequence.

[0051] The migration identification module obtains heat migration paths and heat accumulation centers based on dynamic heat distribution map sequences;

[0052] The zone control module generates corresponding battery module heat dissipation control strategies based on heat migration paths and heat accumulation centers, and controls the heat dissipation actuators in the immersion energy storage box to implement zoned heat dissipation adjustment for the target battery module.

[0053] The beneficial effects of this invention are as follows:

[0054] The battery module thermal safety collaborative control method provided by the present invention constructs the module heating response quantity based on the time-series thermal state reference set, and establishes the thermal coupling mapping relationship between modules by combining the spatial adjacency relationship of modules, the medium flow relationship and the temperature rise response relationship. This enables the thermal influence between adjacent modules inside the housing to be quantitatively characterized. Compared with the prior art, which only judges by the temperature rise of a single module or a fixed threshold, the present invention can identify the strength and coupling direction of heat transfer between modules, thereby improving the ability to identify the trend of local thermal anomaly propagation.

[0055] Then, by using three-dimensional thermal field calculation grid cells and a grid temperature prediction model, the dynamic thermal distribution map inside the box is reconstructed frame by frame in chronological order. This not only provides the temperature distribution results at a single moment but also continuously reflects the spatial and temporal evolution of the thermal field. Therefore, it is possible to further extract heat migration paths and heat accumulation centers, overcoming the shortcomings of existing technologies in identifying heat diffusion chains and local secondary heat accumulation processes, thus elevating hotspot identification from static discovery to dynamic tracking.

[0056] After obtaining the heat migration path and heat accumulation center, the thermal field analysis results are directly converted into the basis for heat dissipation control. The heat source core module, channel module and downstream risk module are distinguished, and source heat suppression control, path heat interception control and downstream pre-cooling control are implemented respectively. Compared with the existing technology of overall heat dissipation enhancement or passive load reduction after single module overheating, the present invention can realize zoned linkage control for the heat diffusion chain. It can suppress the continuous output of heat source, weaken the heat propagation along the path, and reduce the risk of heat accumulation in downstream areas in advance, thereby improving the pertinence and timeliness of control. Attached Figure Description

[0057] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0058] Figure 1 This is a flowchart of the battery module thermal safety collaborative control method provided in Embodiment 1 of the present invention;

[0059] Figure 2 This is a flowchart of the method for establishing inter-module thermal coupling candidate connections provided in Embodiment 1 of the present invention;

[0060] Figure 3 This is a schematic diagram of the heat migration path division provided in Embodiment 1 of the present invention;

[0061] Figure 4 This is a module connection diagram of the battery module thermal safety collaborative control system provided in Embodiment 2 of the present invention. Detailed Implementation

[0062] To make the technical means, creative features, objectives, and effects of this invention readily understandable, the invention is further described below with reference to specific illustrations. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0063] Example 1

[0064] Please see Figures 1-3 This invention provides a technical solution: a battery module thermal safety collaborative control method, comprising:

[0065] A three-dimensional acquisition unit is constructed for the battery module inside the immersion energy storage box to acquire thermal characteristic parameters and construct a time-series thermal state benchmark set;

[0066] Methods for constructing a stereoscopic acquisition unit include:

[0067] A unified three-dimensional coordinate system is established within the immersion energy storage box, and all battery modules are uniquely numbered according to their layer, column, and row arrangement to obtain the module number and its corresponding spatial coordinates.

[0068] Specifically, taking the bottom vertex of the inner cavity of the energy storage box near the cooling medium inlet as the origin, a first coordinate axis is established along the length of the box, a second coordinate axis is established along the width of the box, and a third coordinate axis is established along the height of the box. Then, according to the arrangement of the battery modules in the box, all battery modules are divided into several layers, columns, and rows, and each battery module is uniquely marked as the module in the y-th layer, m-th column, and p-th row. The geometric center of the corresponding battery module is used as the spatial reference point of the module, thereby completing the unified mapping between the box-level spatial coordinates and the module-level spatial coordinates.

[0069] A three-dimensional acquisition unit boundary is constructed with a single battery module as the core, thus forming a three-dimensional acquisition unit.

[0070] Specifically, the method for constructing the boundary of the acquisition unit includes: for each numbered battery module, taking its outer contour surface as the internal boundary, and then extending it outward along the normal direction of the six faces of the battery module to the perpendicular plane of the line connecting the centers of adjacent battery modules; when there is no adjacent battery module in a certain normal direction, the outward extension boundary in that direction is extended to the inner wall of the energy storage box, thereby forming an independent three-dimensional acquisition unit surrounding the local space of the battery module and its surrounding cooling medium.

[0071] Methods for obtaining thermal characteristic parameters and constructing time-series thermal state reference sets include:

[0072] A thermal characteristic parameter acquisition sensor group is arranged in each three-dimensional acquisition unit. The thermal characteristic parameter acquisition sensor group includes a module temperature sensor, a current sensor, a voltage sensor, a cooling medium temperature sensor, and a flow rate sensor.

[0073] In this embodiment, the module temperature sensors are configured as follows: five module temperature sensors are fixed on the outer surface of each battery module. The five module temperature sensors are respectively arranged at: the center of the upper end face near the positive electrode junction area, the geometric center of the module's main heat dissipation surface, the center of the lower end face near the negative electrode junction area, the middle of the narrow side on the current-facing side, and the middle of the narrow side on the back-current side. Each module temperature sensor is attached to the outer surface of the module housing through an insulating and thermally conductive fixing layer. The temperature sensing end of the module temperature sensor is directly attached to the module housing, and the wires are led out through a medium corrosion resistant sheath. Through the above five-point arrangement, the axial temperature difference, in-plane temperature difference, and flow-direction temperature difference are obtained respectively, thereby avoiding the problem that single-point temperature measurement cannot reflect the migration of local hot spots.

[0074] The current sensor and voltage sensor are configured as follows: a current sensor is fitted on the outside of the branch busbar of the corresponding battery module to collect the charging and discharging current of the battery module in real time; high-impedance voltage sampling lines are led out from the positive and negative output terminals of the battery module respectively, and the module terminal voltage is collected through an isolation voltage sensor.

[0075] The cooling medium temperature sensor and flow rate sensor are configured as follows: In each three-dimensional acquisition unit, along the main flow direction of the cooling medium, a first medium temperature sensor and a first medium flow rate sensor are arranged 10 mm in front of the outer surface of the battery module on the flow-facing side, and a second medium temperature sensor and a second medium flow rate sensor are arranged 10 mm behind the outer surface of the battery module on the back flow side. All four sensors are fixed on an insulating support frame, and the position of the support frame relative to the battery module remains constant, thereby characterizing the inflow and outflow states of the medium.

[0076] A unified sampling period is set, and the collected results are structured to form a time-series thermal state reference set. The time-series thermal state reference set includes module number, spatial coordinates, temperature values ​​of five modules, module current values, module voltage values, upstream medium temperature values, downstream medium temperature values, upstream medium velocity values, and downstream medium velocity values.

[0077] The main controller triggers synchronous sampling at each level of the acquisition nodes according to a unified timestamp. During normal operation, a 1-second sampling period is used, and when the temperature rise rate reaches a preset threshold, it switches to a 0.2-second sampling period.

[0078] Specifically, the preset threshold is determined based on the statistical boundary of the temperature rise rate of the battery module under normal operating conditions.

[0079] Specifically, the method for structuring the acquisition results to form a time-series thermal state reference set includes: using a single timestamp as an index, storing the module number, spatial coordinates, five module temperature values, module current values, module voltage values, upstream medium temperature values, downstream medium temperature values, upstream medium velocity values, and downstream medium velocity values ​​for each three-dimensional acquisition unit, thereby generating a time-series thermal state reference set.

[0080] Each record in the time-series thermal state benchmark set can be traced back to a unique battery module, a unique spatial coordinate, and a unique sampling time, thus providing a unified input for subsequent thermal field reconstruction, hotspot identification, and risk prediction.

[0081] The module's heating response is constructed based on a time-series thermal state benchmark set;

[0082] Specifically, the method for constructing the module's heating response based on a time-series thermal state reference set is as follows:

[0083] The temperature values ​​of the five modules are weighted and summed to obtain the module temperature characterization value of the battery module at the current moment, which is recorded as the module temperature characterization value. The difference between the module temperature characterization value at the current moment and the module temperature characterization value at the previous moment is divided by the time interval between two adjacent timestamps to obtain the temperature rise response of the battery module.

[0084] Among them, the center position of the main heat dissipation surface has the highest weight, followed by the narrow edges on the front and back sides, and then the upper and lower end faces, thus obtaining the module temperature characterization value of the battery module at the current moment. The specific weight values ​​are set by those skilled in the art.

[0085] The module current value is multiplied by the module voltage value to obtain the module power characterization value. The module power characterization value and the temperature rise response value are normalized and then weighted and summed to obtain the module heating response value.

[0086] Example: The expression for the module's heat response is: In the formula, For the first Each battery module at the timestamp The following heat response amount, No. The module power rating of each battery module For the first Temperature rise rate of each battery module Reference power; For reference temperature rise rate, and These are the weighting coefficients, and + =1. It should be noted that the weighting coefficients in the formula are set by those skilled in the art based on actual conditions or obtained through simulation with a large amount of data.

[0087] Candidate thermal coupling connections between modules are established based on the spatial adjacency relationship, medium flow relationship, and module heating response relationship of battery modules.

[0088] Specifically, the method for establishing candidate thermal coupling connections between modules includes: for any target battery module, selecting P (P≥1) nearest neighboring battery modules in the length, width, and height directions as candidate neighboring modules according to their spatial location;

[0089] Then, based on the flow velocity values ​​of the medium on the front and back sides, the mainstream direction of the cooling medium around the target battery module is determined, and only candidate adjacent modules located downstream of the mainstream, with a lateral offset less than a preset distance threshold and a vertical height difference less than a preset height threshold are retained as effective coupling objects.

[0090] After obtaining the temperature rise response of the target battery module, the delay time when the temperature rise response of the effectively coupled object changes in the same direction over time is obtained. When the temperature rise response changes in the same direction and the delay time falls within the preset effective response window, it is determined that there is a thermal coupling relationship between the two, thereby completing the screening of thermal coupling candidate connections.

[0091] It should be noted that the effective response window is obtained by combining the measured lag time of the candidate battery module after the target battery module has undergone a significant temperature rise during historical operation.

[0092] Calculate the thermal coupling strength of the selected thermal coupling candidate connections to form a thermal coupling mapping relationship between modules;

[0093] Specifically, the method for calculating thermal coupling strength includes: for thermal coupling candidate connections, extracting characteristic parameters between the target battery module and the effective coupling object, the characteristic parameters including center distance, forward projection matching degree along the mainstream direction of the medium, lateral offset distance, difference in module temperature characterization value, temperature difference of the medium on the front side, temperature difference of the medium on the back side, difference in medium flow velocity, and temperature rise response delay time;

[0094] The medium velocity difference is the absolute value of the difference between the average medium velocity in the local area where the target battery module is located and the average medium velocity in the local area where the effectively coupled object is located. It is used to characterize the degree to which the two modules are in the same flow channel or similar flow field state; the larger the medium velocity difference, the lower the degree to which the two modules are in the same flow channel or similar flow field state, and the lower the thermal coupling strength.

[0095] Among them, the center distance, lateral offset distance, medium flow velocity difference, and temperature rise response delay time are used as attenuation factors, and the forward projection matching degree along the mainstream direction of the medium, the difference in module temperature characterization value, the temperature difference of the medium on the front side, and the temperature difference of the medium on the back side are used as enhancement factors. The original thermal coupling strength is obtained by weighting the characteristic parameters after dimensionless processing. Then, all the original thermal coupling strengths corresponding to the same target battery module are normalized to obtain the thermal coupling strength between the target battery module and each effective coupling object, forming a thermal coupling strength map with the battery module as the node and the thermal coupling strength as the edge weight.

[0096] It should be noted that the forward projection matching degree along the mainstream direction of the medium is used to characterize whether the effective coupling object is located on the downstream thermal transport path of the target battery module. When the effective coupling object is located downstream of the target battery module and the proportion of the forward projection to the center distance is high, it is determined that it is more susceptible to the thermal wake of the target battery module. When the effective coupling object is located upstream or the forward projection is less than the preset projection threshold, the forward projection matching degree is zero or the weight is reduced.

[0097] The methods for obtaining the forward projection matching degree of the mainstream direction of the medium include:

[0098] In a three-dimensional coordinate system, obtain the center coordinates of the target battery module and the center coordinates of the effective coupling object, and construct a relative position vector pointing from the center of the target battery module to the center of the effective coupling object;

[0099] Obtain the unit vector of the mainstream direction of the medium in the region where the target battery module is located, and calculate the projection value of the relative position vector onto the unit vector of the mainstream direction of the medium;

[0100] When the projection value is greater than zero, it is determined that the effective coupling object is located in the downstream direction of the target battery module, and the basic forward projection matching degree is determined according to the ratio of the projection value to the center distance between modules.

[0101] When the projection value is less than or equal to zero, it is determined that the effective coupling object is not on the forward thermal transport path of the target battery module, and the forward projection matching degree is set to zero.

[0102] Based on the thermal coupling strength and the spatial position of the box, a three-dimensional thermal field calculation grid is divided, and the dynamic thermal distribution map inside the box is reconstructed frame by frame in time sequence.

[0103] Specifically, the method for dividing the three-dimensional thermal field calculation grid cells includes:

[0104] Using the internal space boundary of the submersible energy storage box as the range, a three-dimensional mesh is divided along the length, width, and height of the box. Each mesh unit has unique spatial coordinates and volume number.

[0105] The space occupied by the battery module is directly mapped to the heat-generating physical area;

[0106] Obtain the liquid level position inside the submersible energy storage tank, mark the area below the liquid level as the effective submersible area, and mark the area above the liquid level as the non-submersible area;

[0107] For each battery module location, its module temperature characterization value, module heat response value, front-flow side medium temperature value, back-flow side medium temperature value, medium flow velocity, and thermal coupling strength are all assigned to the corresponding grid cell.

[0108] It should be noted that the medium flow rate is the average of the medium flow rate values ​​on the front and back sides of the battery module.

[0109] The methods for reconstructing the dynamic thermal distribution map inside the enclosure frame by frame in chronological order include:

[0110] Based on the time-series thermal state reference set at the same timestamp, the module temperature characterization value, the corresponding heat response of the battery module, the thermal coupling strength characteristics, and the liquid level immersion state characteristics of the area where the grid unit is located are extracted to construct the input feature set of the grid temperature prediction model; the input feature set is input into the grid temperature prediction model to output the module temperature characterization value of each grid unit at the next time step; a three-dimensional thermal distribution map at the corresponding timestamp is generated based on the module temperature characterization values ​​of all grid units at the next time step.

[0111] Then, the three-dimensional heat distribution map generated at the current timestamp is used as the basis for the historical heat field prediction at the next timestamp. The temperature characterization values ​​of the grid cell modules are predicted continuously in the order of timestamps to generate a dynamic heat distribution map sequence.

[0112] The training method for the grid temperature prediction model includes:

[0113] Gradient boosting regression tree was selected as the grid temperature prediction model to accurately predict the module temperature characterization value of each grid unit at the next moment based on the multi-source thermal state characteristics of the three-dimensional grid unit of the energy storage box, thus supporting the reconstruction of the dynamic thermal distribution map inside the box.

[0114] Historical operational data of the submersible energy storage tank under all operating conditions were pre-collected. This data covered different charge / discharge rates, medium flow rates, liquid levels, and module operating states. It included module temperature characteristics, thermal response, thermal coupling strength characteristics, and liquid immersion status characteristics for each grid cell at each time stamp, as well as the actual temperature characteristics of the grid cell module at the next time point. The collected data was then cleaned, filtered, and normalized sequentially. Invalid samples due to equipment failure, sampling interruptions, or numerical anomalies were removed. Linear interpolation was used to supplement any missing data, ensuring sample continuity and completeness.

[0115] All continuous input features are dimensionless and normalized to eliminate dimensional differences; discrete features such as liquid level immersion state and spatial region type are numerically encoded to form a standardized input feature set. All processed samples are divided into training, validation, and test sets in a 7:2:1 ratio.

[0116] Initialization and configuration are completed before model training. The model training uses mean squared error (MSE) as the loss function to measure the deviation between the model's predicted grid cell module temperature values ​​and the actual values. The standardized input feature set is used as the model input, and the actual grid cell module temperature values ​​at the next time step are used as the training target. Model training is conducted based on the training set. Each new tree aims to fit the regression residual of the training set's loss function. The optimal splitting feature is selected using the MSE criterion to divide the samples into different child nodes until stopping conditions are met, including reaching the maximum depth, the number of samples in the child node being less than the minimum number of samples, or the splitting gain being lower than a threshold.

[0117] Gradient descent is used to optimize the weights of the leaf nodes of each new tree. The contribution of the new tree to the final prediction result is controlled by the learning rate, avoiding a single tree dominating the output and improving the stability and generalization ability of temperature prediction.

[0118] The hyperparameter optimization adopts the Bayesian optimization method to search for the optimal combination of hyperparameters within a preset range. The optimization objective is to minimize the root mean square error of the verification set.

[0119] An early stopping mechanism is implemented during training. The root mean square error (RMSE) of the validation set is calculated every 20 trees in each iteration. Training is stopped when the RMSE decreases by less than 0.001 for three consecutive iterations to prevent overfitting. After training, the model parameters with the lowest RMSE in the validation set are saved, including the splitting rules of all decision trees and the weights of the leaf nodes, to ensure that the model has stable and reliable temperature prediction capabilities for different operating conditions and grid cells at different locations.

[0120] The trained model was validated using an independent test set, and the root mean square error, mean absolute error, and coefficient of determination were calculated. If the root mean square error of the test set is ≤2% of the module temperature characterization value, the mean absolute error is ≤1% of the module temperature characterization value, and the coefficient of determination is ≥0.95, then the model performance evaluation met the standards and it can be deployed and applied to support frame-by-frame reconstruction of three-dimensional thermal fields and generation of dynamic thermal distribution map sequences.

[0121] Heat migration paths and heat accumulation centers are obtained based on dynamic heat distribution map sequences;

[0122] In the 3D thermal distribution map corresponding to each time stamp, the grid cell with the largest temperature gradient is obtained, and the gradient change direction from the high temperature region to the low temperature region is defined as the local thermal migration direction. Then, the local thermal migration directions with continuous spatial location, continuous direction, and continuous temperature change under multiple consecutive time stamps are connected in series to form a heat migration path.

[0123] The continuous temperature change means that, at adjacent time points, the high-temperature side and the low-temperature side corresponding to the local heat migration direction still maintain the same temperature difference direction.

[0124] It should be noted that consecutive timestamps refer to at least two or at least three timestamps in a three-dimensional heat map sequence that are arranged in chronological order and have no missing frames between adjacent timestamps.

[0125] In the 3D thermal distribution map corresponding to each timestamp, grid cells with temperatures higher than the average temperature of the box at the current time are selected as high-temperature candidate cells. Then, spatially adjacent high-temperature candidate cells are aggregated, and high-temperature candidate cells with an aggregation number greater than or equal to 3 are marked as high-temperature clusters, forming multiple high-temperature clusters. The average temperature, maximum temperature, volume, geometric center coordinates, and duration of each high-temperature cluster are calculated. The high-temperature cluster with a duration reaching a preset duration and a continuously rising maximum temperature is defined as the heat accumulation center.

[0126] Based on the heat migration path and heat accumulation center, a corresponding battery module heat dissipation control strategy is generated, and the heat dissipation actuator in the immersion energy storage box is controlled to implement zoned heat dissipation adjustment of the target battery module.

[0127] Specifically, the methods for generating corresponding battery module heat dissipation control strategies include:

[0128] Based on the positional matching relationship between the geometric center coordinates of the heat accumulation center and the center coordinates of each battery module, the core heat source module corresponding to each heat accumulation center is determined.

[0129] Based on the heat flow direction of the heat migration path, the heat migration path is divided into the source section, the migration section and the accumulation section. The battery modules that fall within the coverage area of ​​each path section are marked as source modules, channel modules and downstream risk modules, respectively, to form corresponding heat dissipation control zones.

[0130] For source modules, implement source heat suppression control; for channel modules, implement path heat interception control; and for downstream risk modules, implement downstream pre-cooling control.

[0131] Specific examples of implementing source heat suppression control for source modules

[0132] When the source module is identified as the main heat source module, the controller sends a load reduction command to the module power management unit, thereby directly reducing the instantaneous heat generation of the source module.

[0133] Specific example of path heat interception control for channel modules

[0134] The controller sends a heat-cutting command to the corresponding circulation branch of the channel area, which increases the frequency of medium renewal around the channel module, thereby accelerating the transfer of heat from the migration section to the main circulation loop and weakening the continued forward transmission of heat along the path.

[0135] Based on the direction of heat migration, the controller adjusts the guide plate of the channel module from the original parallel guide angle to the deflection guide angle, so that the cooling medium changes from flushing along the path to flushing laterally, thereby interrupting the continuous heat propagation chain of the channel module and reducing the possibility of heat being concentrated and transferred along a fixed path.

[0136] Specific examples of implementing downstream pre-cooling control for downstream risk modules

[0137] When a module is identified as a downstream risk module, the controller activates the pre-cooling branch in its corresponding area, creating a local low-temperature buffer zone around the downstream risk module.

[0138] Methods for dividing heat migration pathways into source segments, migration segments, and accumulation segments include:

[0139] Extract all grid cells that constitute the heat migration path along the heat migration direction, and treat each grid cell as a path point to obtain the path point sequence P1, P2, ..., Pn, where P1 is the starting path point of the heat migration path and Pn is the ending path point of the heat migration path.

[0140] Extract the temperature value of each path point and the proportion of high-temperature candidate units in the preset neighborhood around each path point; determine the continuous path segment starting from the starting path point with a temperature higher than the preset source temperature threshold as the source segment; determine the continuous path segment immediately following the source segment with the temperature change of adjacent path points less than or equal to the preset temperature decay threshold and the change in the proportion of high-temperature candidate units in the preset neighborhood around the adjacent path points not exceeding the preset proportion change threshold as the migration segment; and divide the path immediately following the migration segment into the accumulation segment.

[0141] It should be noted that the preset thresholds or ranges in the above analysis of heat migration path division were set by those skilled in the art based on actual conditions or obtained through large-scale data simulation.

[0142] Example 2

[0143] Please see Figure 4This invention provides a technical solution: a battery module thermal safety collaborative control system, which is used to implement the battery module thermal safety collaborative control method, including:

[0144] The parameter acquisition module constructs a three-dimensional acquisition unit for the battery module in the immersion energy storage box, acquires thermal characteristic parameters and constructs a time-series thermal state reference set.

[0145] The response analysis module constructs the module's heating response quantity based on a time-series thermal state reference set;

[0146] The coupling mapping module establishes candidate thermal coupling connections between modules based on the spatial adjacency relationship, medium flow relationship, and module heating response relationship of the battery modules. It calculates the thermal coupling strength of the selected candidate thermal coupling connections to form a thermal coupling mapping relationship between modules.

[0147] The thermal reconstruction module divides the three-dimensional thermal field calculation grid units according to the thermal coupling strength and the spatial position of the box, and reconstructs the dynamic thermal distribution map inside the box frame by frame in time sequence.

[0148] The migration identification module obtains heat migration paths and heat accumulation centers based on dynamic heat distribution map sequences;

[0149] The zone control module generates corresponding battery module heat dissipation control strategies based on heat migration paths and heat accumulation centers, and controls the heat dissipation actuators in the immersion energy storage box to implement zoned heat dissipation adjustment for the target battery module.

[0150] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for coordinated control of thermal safety in battery modules, characterized in that, include: A three-dimensional acquisition unit is constructed for the battery modules inside an immersion energy storage tank to acquire thermal characteristic parameters and construct a time-series thermal state reference set. The method for constructing the three-dimensional acquisition unit includes: establishing a unified three-dimensional coordinate system within the immersion energy storage tank, and uniquely numbering all battery modules according to their layer, column, and row arrangement to obtain module numbers and corresponding spatial coordinates; for each numbered battery module, its outer contour surface is taken as the internal boundary, and then extended outward along the normal directions of the six faces of the battery module to the perpendicular bisector of the line connecting the centers of adjacent battery modules; when there is no adjacent battery module in a certain normal direction, the outward extension boundary in that direction is extended to the inner wall of the energy storage tank, thereby forming an independent three-dimensional acquisition unit surrounding the battery module and the local space of the surrounding cooling medium; the time-series thermal state reference set includes the module number, spatial coordinates, five module temperature values, module current values, module voltage values, upstream medium temperature values, downstream medium temperature values, upstream medium flow velocity values, and downstream medium flow velocity values. The module's heating response is constructed based on a time-series thermal state reference set; The method for constructing the module's thermal response is as follows: The temperature values ​​of the five modules are weighted and summed to obtain the module temperature characterization value at the current moment, which is denoted as the module temperature characterization value. The difference between the current module temperature characterization value and the previous module temperature characterization value is calculated and divided by the time interval between two adjacent timestamps to obtain the temperature rise response of the battery module. The module current value and the module voltage value are multiplied to obtain the module power characterization value. The module power characterization value and the temperature rise response value are then normalized and weighted and summed to obtain the module's thermal response. Based on the spatial adjacency relationship, medium flow relationship, and module heating response relationship of battery modules, candidate thermal coupling connections between modules are established. The thermal coupling strength of the selected candidate thermal coupling connections is calculated to form a thermal coupling mapping relationship between modules. The method for establishing candidate thermal coupling connections between modules includes: for any target battery module, selecting the P nearest adjacent battery modules in the length, width, and height directions as candidate adjacent modules according to their spatial location; then determining the mainstream direction of the cooling medium around the target battery module based on the flow velocity values ​​of the upstream and downstream media, and retaining only the candidate adjacent modules located downstream of the mainstream, with a lateral offset less than a preset distance threshold and a vertical height difference less than a preset height threshold as effective coupling objects; obtaining the delay time when the temperature rise response of the target battery module changes and the temperature rise response of the effective coupling object changes in the same direction over time, and determining that there is a thermal coupling relationship between the two when the temperature rise response changes in the same direction and the delay time falls within the preset effective response window; Based on the thermal coupling strength and the spatial position of the box, a three-dimensional thermal field calculation grid is divided, and the dynamic thermal distribution map inside the box is reconstructed frame by frame in time sequence. The method for reconstructing the dynamic thermal distribution map inside the enclosure frame by frame according to time sequence includes: based on the time-series thermal state reference set at the same timestamp, extracting the module temperature characterization value of the grid cell where the battery module is located, the heat response amount of the corresponding battery module, the thermal coupling strength characteristics, and the liquid level immersion state characteristics of the area where the grid cell is located, and constructing the input feature set of the grid temperature prediction model; inputting the input feature set into the grid temperature prediction model, and outputting the module temperature characterization value of each grid cell at the next time moment; generating a three-dimensional thermal distribution map at the corresponding timestamp based on the module temperature characterization values ​​of all grid cells at the next time moment; then using the three-dimensional thermal distribution map generated at the current timestamp as the basis for the historical thermal field prediction at the next timestamp, and continuously executing the prediction of the module temperature characterization value of the grid cell according to the timestamp order to generate a dynamic thermal distribution map sequence; Heat migration paths and heat accumulation centers are obtained based on dynamic heat distribution map sequences; The method for obtaining the heat migration path is as follows: in the three-dimensional heat distribution map corresponding to each time point, the grid cell with the largest temperature gradient is obtained, and the gradient change direction from the high temperature area to the low temperature area is defined as the local heat migration direction; then, the local heat migration directions with continuous spatial location, continuous direction, and continuous temperature change under multiple consecutive time points are connected in series to form a heat migration path. The method for obtaining the heat accumulation center is as follows: In the three-dimensional heat distribution map corresponding to each time stamp, select grid cells with a temperature higher than the average temperature of the box at the current time as high-temperature candidate cells, then aggregate spatially adjacent high-temperature candidate cells, and mark high-temperature candidate cells with an aggregation number greater than or equal to 3 as high-temperature accumulation areas to form multiple high-temperature accumulation areas; calculate the average temperature, maximum temperature, volume, geometric center coordinates, and duration of each high-temperature accumulation area; define the high-temperature accumulation area whose duration reaches a preset duration and whose maximum temperature continues to rise as the heat accumulation center; Based on the heat migration path and heat accumulation center, a corresponding battery module heat dissipation control strategy is generated, and the heat dissipation actuator in the immersion energy storage box is controlled to implement zoned heat dissipation adjustment of the target battery module. The method for generating corresponding battery module heat dissipation control strategies includes: determining the core heat source module corresponding to each heat accumulation center based on the positional matching relationship between the geometric center coordinates of the heat accumulation center and the center coordinates of each battery module; dividing the heat migration path into source segment, migration segment, and accumulation segment according to the heat flow direction of the heat migration path, and marking the battery modules falling within the coverage area of ​​each path segment as source module, channel module, and downstream risk module, respectively, to form corresponding heat dissipation control zones; implementing source heat suppression control for source modules, implementing path heat interception control for channel modules, and implementing downstream pre-cooling control for downstream risk modules.

2. The battery module thermal safety collaborative control method according to claim 1, characterized in that, The method for acquiring thermal characteristic parameters and constructing a time-series thermal state reference set includes: A thermal characteristic parameter acquisition sensor group is arranged in each three-dimensional acquisition unit. The thermal characteristic parameter acquisition sensor group includes a module temperature sensor, a current sensor, a voltage sensor, a cooling medium temperature sensor, and a flow rate sensor. A uniform sampling period is set, and the collected results are organized in a structured manner to form a time-series thermal state reference set.

3. The battery module thermal safety collaborative control method according to claim 2, characterized in that, The module temperature sensors are configured as follows: five module temperature sensors are fixed on the outer surface of each battery module. The five module temperature sensors are respectively arranged at: the center of the upper end face near the positive electrode junction area, the geometric center of the module's main heat dissipation surface, the center of the lower end face near the negative electrode junction area, the middle of the narrow side on the current-facing side, and the middle of the narrow side on the back-current side. Each module temperature sensor is attached to the outer surface of the module housing through an insulating and thermally conductive fixing layer. The temperature sensing end of the module temperature sensor is directly attached to the module housing, and the wires are led out through a medium-resistant corrosion-resistant sheath.

4. The battery module thermal safety collaborative control method according to claim 1, characterized in that, Methods for calculating thermal coupling strength include: For thermal coupling candidate connections, feature parameters between the target battery module and the effective coupling object are extracted. These feature parameters include center distance, forward projection matching degree along the mainstream direction of the medium, lateral offset distance, difference in module temperature characterization value, temperature difference of the medium on the front side, temperature difference of the medium on the back side, difference in medium flow velocity, and temperature rise response delay time. Center distance, lateral offset distance, medium velocity difference, and temperature rise response delay time are used as attenuation factors, while forward projection matching degree along the mainstream direction of the medium, module temperature characterization value difference, front-flow side medium temperature difference, back-flow side medium temperature difference, and medium velocity difference are used as enhancement factors. The original thermal coupling strength is obtained by performing dimensionless processing on the feature parameters and then weighting the calculation. Then, the original thermal coupling strengths corresponding to the same target battery module are normalized to obtain the thermal coupling strength between the target battery module and each effective coupling object.

5. The battery module thermal safety collaborative control method according to claim 1, characterized in that, The method for dividing the three-dimensional thermal field calculation grid includes: Using the internal space boundary of the submersible energy storage box as the range, a three-dimensional mesh is divided along the length, width, and height of the box. Each mesh unit has unique spatial coordinates and volume number. The space occupied by the battery module is directly mapped to the heat-generating physical area; Obtain the liquid level position inside the submersible energy storage tank, mark the area below the liquid level as the effective submersible area, and mark the area above the liquid level as the non-submersible area; For each battery module location, its module temperature characterization value, module heat response value, front-flow side medium temperature value, back-flow side medium temperature value, medium flow rate, and thermal coupling strength are all assigned to the corresponding grid cell.

6. The battery module thermal safety collaborative control method according to claim 1, characterized in that, Methods for dividing heat migration pathways into source segments, migration segments, and accumulation segments include: Extract all grid cells that constitute the heat migration path along the heat migration direction, and treat the grid cells as path points to obtain the path point sequence P1, P2, ..., Pn, where P1 is the starting path point and Pn is the ending path point. Extract the temperature value of each path point and the proportion of high-temperature candidate units in the preset neighborhood around each path point; determine the continuous path segment starting from the starting path point with a temperature higher than the preset source temperature threshold as the source segment; determine the continuous path segment immediately following the source segment with the temperature change of adjacent path points less than or equal to the preset temperature decay threshold and the change in the proportion of high-temperature candidate units in the preset neighborhood around the adjacent path points not exceeding the preset proportion change threshold as the migration segment; and divide the path immediately following the migration segment into the accumulation segment.

7. A battery module thermal safety collaborative control system, used to implement the battery module thermal safety collaborative control method according to any one of claims 1-6, characterized in that, include: The parameter acquisition module constructs a three-dimensional acquisition unit for the battery module in the immersion energy storage box, acquires thermal characteristic parameters and constructs a time-series thermal state reference set. The response analysis module constructs the module's heating response quantity based on a time-series thermal state reference set; The coupling mapping module establishes candidate thermal coupling connections between modules based on the spatial adjacency relationship, medium flow relationship, and module heating response relationship of the battery modules. It calculates the thermal coupling strength of the selected candidate thermal coupling connections to form a thermal coupling mapping relationship between modules. The thermal reconstruction module divides the three-dimensional thermal field calculation grid units according to the thermal coupling strength and the spatial position of the box, and reconstructs the dynamic thermal distribution map inside the box frame by frame in time sequence. The migration identification module obtains heat migration paths and heat accumulation centers based on dynamic heat distribution map sequences; The zone control module generates corresponding battery module heat dissipation control strategies based on heat migration paths and heat accumulation centers, and controls the heat dissipation actuators in the immersion energy storage box to implement zoned heat dissipation adjustment for the target battery module.

Images