Submerged liquid-cooled energy storage device circulation cooling control method and system

Abstract

The application provides an immersed liquid cooling energy storage equipment circulating cooling control method and system, relates to the technical field of new energy storage, and comprises the following steps: collecting temperature data to establish a thermodynamic diagram, identifying a temperature abnormal area, calculating a temperature transfer coefficient to determine a heat transfer path and dividing a control sub-area, activating a main refrigeration circuit in the temperature abnormal area and pre-activating an adjacent area auxiliary refrigeration circuit, cooperatively controlling heat transfer blockage and diffusion, and updating in real time until the temperature distribution meets a preset condition. The application can accurately identify hot spot areas, effectively prevent heat runaway spread, and improve the safety and stability of the energy storage system.

Description

Technical Field

[0001] This invention relates to new energy storage technology, and more particularly to a circulating cooling control method and system for an immersion liquid-cooled energy storage device. Background Technology

[0002] With the rapid development of the new energy industry, the safety and lifespan of energy storage battery modules have attracted much attention. Immersion liquid cooling systems, with their excellent heat dissipation performance and temperature uniformity, are widely used in the temperature management of large-scale energy storage equipment. However, existing immersion liquid cooling systems typically employ a fixed cooling strategy, which cannot dynamically adjust the control parameters of the cooling circuit according to the real-time temperature distribution characteristics of the energy storage battery module.

[0003] In existing technologies, immersion liquid cooling systems passively activate the cooling circuit to lower the temperature only when the local temperature exceeds the limit. This single temperature control method not only leads to low efficiency in the utilization of cooling resources, but also fails to effectively prevent the transfer and diffusion of heat within the energy storage battery module. Furthermore, due to the lack of real-time monitoring and prediction of heat transfer paths, control measures are often only taken after the temperature anomaly has spread over a large area, increasing the safety risks of temperature runaway.

[0004] Therefore, there is an urgent need for a circulating cooling control method for immersion liquid-cooled energy storage equipment that can achieve predictive control based on temperature distribution characteristics. By establishing a heat transfer path and realizing coordinated control of the main and auxiliary refrigeration circuits, the extension and diffusion of the heat transfer path can be effectively blocked while improving the refrigeration efficiency. Summary of the Invention

[0005] This invention provides a circulating cooling control method and system for an immersion liquid-cooled energy storage device, which can solve the problems in the prior art.

[0006] A first aspect of the present invention provides a method for controlling the circulating cooling of an immersion liquid-cooled energy storage device, comprising:

[0007] Temperature data of energy storage battery modules immersed in liquid cooling medium is collected. The temperature change rate and temperature fluctuation value of adjacent monitoring points are calculated based on the temperature data. A temperature distribution heat map is established. Monitoring points with temperature change rates exceeding preset change values ​​are marked in the temperature distribution heat map. The temperature fluctuation values ​​of the marked monitoring points are divided into zones according to their numerical values, and a temperature anomaly area distribution map is output.

[0008] Based on the temperature fluctuation values ​​of each region in the temperature anomaly distribution map, the temperature transfer coefficient between adjacent regions is calculated. The heat transfer path is determined according to the numerical order of the temperature transfer coefficient, and the heat transfer path is divided into multiple control sub-regions according to the preset interval of the temperature transfer coefficient.

[0009] Based on the spatial distribution characteristics of the heat transfer path, the main refrigeration circuit is activated in the control sub-region where the temperature transfer coefficient exceeds the preset temperature value. At the same time, the auxiliary refrigeration circuit of the adjacent control sub-region is pre-activated in the extension direction of the heat transfer path. The extension and diffusion of the heat transfer path are blocked by the coordinated control of the main refrigeration circuit and the auxiliary refrigeration circuit.

[0010] The system collects temperature data from each monitoring point in real time and updates the temperature distribution heat map. It then repeats the steps of identifying abnormal temperature areas and controlling the cooling circuit until the temperature distribution of the energy storage battery module meets the preset conditions.

[0011] In one alternative embodiment,

[0012] Temperature data of energy storage battery modules immersed in liquid cooling medium is collected. Based on the temperature data, the rate of temperature change and temperature fluctuation values ​​of adjacent monitoring points are calculated, and a temperature distribution heat map is established, including:

[0013] Temperature sensor arrays are set on the surface of the battery cells, the tab connection, the busbar and the edge of the energy storage battery module immersed in liquid cooling medium to acquire temperature data at each monitoring point. Wavelet transform is used to decompose the temperature data to obtain noise-reduced temperature data.

[0014] Based on the noise-reduced temperature data, the temperature difference between two adjacent monitoring points at two adjacent sampling times is calculated. The temperature difference is divided by the sampling time interval to obtain the temperature change rate. The sampling time interval is dynamically adjusted according to the temperature change trend. At the same time, the temperature fluctuation value of the monitoring point is calculated using an adaptive sliding variance algorithm to generate temperature feature data containing the temperature change rate and temperature fluctuation value.

[0015] The grid density is determined using temperature feature data, and the temperature interpolation of the grid nodes is calculated using radial basis functions. The temperature values ​​of the grid nodes are converted into color codes to generate an initial heat map reflecting the temperature distribution. For regions in the initial heat map where the temperature gradient is greater than a preset gradient threshold, the grid is subdivided, and the temperature values ​​of the subdivided grids are recalculated to generate a temperature distribution heat map with local densification characteristics.

[0016] In one alternative embodiment,

[0017] In the temperature distribution heatmap, monitoring points where the rate of temperature change exceeds a preset value are marked. The temperature fluctuation values ​​of the marked monitoring points are then divided into zones according to their numerical values, and a temperature anomaly area distribution map is output, including:

[0018] Based on the charging and discharging state of the energy storage battery module and the ambient temperature, a benchmark threshold for the rate of temperature change is set. Monitoring points that exceed the benchmark threshold in the temperature distribution heat map are marked as abnormal points. The temperature fluctuation values ​​of the abnormal points are extracted, and the temperature fluctuation values ​​are divided into multiple numerical intervals through adaptive interval division to establish a temperature feature distribution sequence of abnormal points.

[0019] Anomalies in the temperature feature distribution sequence whose temperature fluctuation feature similarity is greater than a preset similarity threshold are aggregated into the same group. Taking each group of anomalies as the center, the monitoring points whose temperature fluctuation values ​​are in the same range and whose positions are adjacent are included in the corresponding anomaly group. Adjacent anomaly groups with common boundaries and whose temperature fluctuation values ​​differ from the preset fluctuation value are merged to generate a temperature anomaly area distribution map labeled with anomaly level.

[0020] In one alternative embodiment,

[0021] Based on the temperature fluctuation values ​​of each region in the temperature anomaly distribution map, the temperature transfer coefficient between adjacent regions is calculated. The heat transfer path is determined according to the numerical order of the temperature transfer coefficient, and the heat transfer path is divided into multiple control sub-regions according to the preset interval of the temperature transfer coefficient, including:

[0022] Obtain temperature data from monitoring points within each abnormal region in the temperature anomaly distribution map, calculate the temperature change value at adjacent sampling times, determine the fluctuation amplitude, fluctuation period, and fluctuation direction of the temperature change, and combine them to form the temperature fluctuation value;

[0023] For adjacent abnormal regions, the correlation coefficient of temperature fluctuation values ​​within a specified time window is calculated to obtain the degree of matching. The time difference corresponding to the maximum correlation coefficient is determined to determine the matching delay. Combined with the weighting factor, the temperature transfer coefficient is determined.

[0024] The abnormal region with the largest temperature fluctuation value is selected as the starting region. The abnormal regions are connected along the direction of the maximum gradient of the temperature transfer coefficient to form a transfer path. The dividing point of the transfer path is determined according to the time change sequence of the temperature fluctuation value. The heat transfer paths are obtained by arranging them in descending order of the temperature transfer coefficient.

[0025] Based on the preset range of temperature transfer coefficients between adjacent abnormal areas along the heat transfer path, the heat transfer path is divided into multiple control sub-regions.

[0026] In one alternative embodiment,

[0027] The heat transfer paths are determined by identifying the dividing points of the heat transfer path based on the time variation sequence of temperature fluctuation values, and then arranged in descending order of the temperature transfer coefficient. The heat transfer paths include:

[0028] A sliding time window is used to analyze the time series of temperature fluctuation values. The statistical characteristics and the rate of change of temperature fluctuation values ​​within the window are calculated to obtain the time series feature vector.

[0029] Based on the time-series feature vector, the energy loss rate at adjacent time points is calculated to obtain the instantaneous transfer efficiency, the cumulative transfer efficiency within a fixed window is calculated, and the average transfer efficiency over the entire period is calculated. The instantaneous transfer efficiency, cumulative transfer efficiency, and average transfer efficiency are combined to obtain the transfer efficiency index.

[0030] Using the aforementioned transfer efficiency index, the degree of abrupt change in transfer efficiency, the inflection point amplitude, and the depth of local troughs are calculated respectively. The locations with the greatest degree of abrupt change, the locations with the greatest inflection point amplitude, and the locations with the deepest troughs are combined and scored to determine the dividing points of the heat transfer path.

[0031] The heat transfer path is segmented based on the dividing points, the weighted average of the temperature transfer coefficient of each segment is calculated, and the comprehensive score of the transfer path is calculated by combining the fluctuation characteristics of the temperature transfer coefficient and the distribution of the dividing points.

[0032] Based on the comprehensive score, the heat transfer paths are sorted from largest to smallest according to their temperature transfer coefficient values ​​to obtain the heat transfer paths.

[0033] In one alternative embodiment,

[0034] Based on the spatial distribution characteristics of the heat transfer path, the main refrigeration circuit is activated in the controlled sub-region where the temperature transfer coefficient exceeds the preset temperature value. Simultaneously, the auxiliary refrigeration circuit in the adjacent controlled sub-region is pre-activated in the direction of the heat transfer path's extension. The coordinated control of the main and auxiliary refrigeration circuits blocks the extension and diffusion of the heat transfer path, including:

[0035] Collect temperature field data within the controlled sub-region, calculate the spatial distribution and intensity variation of the temperature field data, and obtain the distribution characteristics of the heat transfer path;

[0036] Calculate the temperature transfer coefficient of each control sub-region based on the temperature field data. When the temperature transfer coefficient exceeds the preset temperature value, determine the over-limit region and set and activate the main refrigeration circuit according to the heat load.

[0037] Based on the distribution characteristics of the heat transfer path, the adjacent control sub-regions of the over-limit area in the direction of the heat transfer path extension are determined. When the heat transfer rate reaches the preset start-up condition, the auxiliary cooling circuit is set and started.

[0038] Collect temperature field data of the control sub-area, adjust the power of the main refrigeration circuit according to the temperature change rate of the control area of ​​the main refrigeration circuit, and adjust the power of the auxiliary refrigeration circuit according to the temperature change rate of the control area of ​​the auxiliary refrigeration circuit.

[0039] The system calculates and controls the heat transfer rate of the sub-area in real time, switches auxiliary cooling circuits with heat transfer rates exceeding a preset threshold to main cooling circuits, and switches main cooling circuits with heat transfer rates below a preset threshold to standby mode, thus blocking the extension of the heat transfer path.

[0040] A second aspect of the present invention provides a circulating cooling control system for an immersion liquid-cooled energy storage device, comprising:

[0041] The first unit is used to collect temperature data of energy storage battery modules immersed in liquid cooling medium, calculate the temperature change rate and temperature fluctuation value of adjacent monitoring points based on the temperature data, establish a temperature distribution heat map, mark the monitoring points whose temperature change rate exceeds the preset change value in the temperature distribution heat map, divide the temperature fluctuation value of the marked monitoring points into zones according to the value, and output a temperature anomaly area distribution map.

[0042] The second unit is used to calculate the temperature transfer coefficient between adjacent areas based on the temperature fluctuation values ​​of each area in the temperature anomaly distribution map, determine the heat transfer path according to the numerical order of the temperature transfer coefficient, and divide the heat transfer path into multiple control sub-areas according to the preset interval of the temperature transfer coefficient.

[0043] The third unit is used to activate the main refrigeration circuit in the control sub-region where the temperature transfer coefficient exceeds the preset temperature value according to the spatial distribution characteristics of the heat transfer path, and at the same time pre-activate the auxiliary refrigeration circuit in the adjacent control sub-region in the extension direction of the heat transfer path, so as to block the extension and diffusion of the heat transfer path through the coordinated control of the main refrigeration circuit and the auxiliary refrigeration circuit.

[0044] The fourth unit is used to collect temperature data from each monitoring point in real time and update the temperature distribution heat map. It repeatedly executes the steps of identifying abnormal temperature areas and controlling the cooling circuit until the temperature distribution of the energy storage battery module meets the preset conditions.

[0045] A third aspect of the present invention provides an electronic device, comprising:

[0046] processor;

[0047] Memory used to store processor-executable instructions;

[0048] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0049] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0050] In this embodiment, by collecting temperature data and constructing a temperature distribution heatmap, the temperature changes of the energy storage battery module can be monitored in real time, and abnormal temperature areas can be quickly identified. This provides data support for precise control of the cooling system, improving the system's response speed and accuracy to temperature anomalies. The method of calculating heat transfer paths based on temperature transfer coefficients and dividing them into multiple control sub-regions enables early prediction of heat diffusion trends, allowing the cooling system to perform differentiated control based on the thermal characteristics of different regions, thus optimizing the allocation efficiency of cooling resources. Through the collaborative working mechanism of the main cooling loop and the auxiliary cooling loop, while activating the main cooling system in the hot spot area, the auxiliary cooling system in the direction of heat propagation is pre-activated, forming a heat diffusion blocking barrier. This effectively prevents the occurrence and spread of thermal runaway, improving the safety and operational stability of the energy storage system. Attached Figure Description

[0051] Figure 1 This is a schematic flowchart of the circulating cooling control method for an immersion liquid-cooled energy storage device according to an embodiment of the present invention;

[0052] Figure 2 This is a comparison chart of the effects of differentiated cooling strategies in embodiments of the present invention;

[0053] Figure 3 This is a graph showing the relationship between temperature transfer coefficient and heat load in an embodiment of the present invention. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0056] Figure 1 This is a schematic flowchart of the circulating cooling control method for an immersion liquid-cooled energy storage device according to an embodiment of the present invention, as shown below. Figure 1 As shown, the method includes:

[0057] Temperature data of energy storage battery modules immersed in liquid cooling medium is collected. The temperature change rate and temperature fluctuation value of adjacent monitoring points are calculated based on the temperature data. A temperature distribution heat map is established. Monitoring points with temperature change rates exceeding preset change values ​​are marked in the temperature distribution heat map. The temperature fluctuation values ​​of the marked monitoring points are divided into zones according to their numerical values, and a temperature anomaly area distribution map is output.

[0058] Based on the temperature fluctuation values ​​of each region in the temperature anomaly distribution map, the temperature transfer coefficient between adjacent regions is calculated. The heat transfer path is determined according to the numerical order of the temperature transfer coefficient, and the heat transfer path is divided into multiple control sub-regions according to the preset interval of the temperature transfer coefficient.

[0059] Based on the spatial distribution characteristics of the heat transfer path, the main refrigeration circuit is activated in the control sub-region where the temperature transfer coefficient exceeds the preset temperature value. At the same time, the auxiliary refrigeration circuit of the adjacent control sub-region is pre-activated in the extension direction of the heat transfer path. The extension and diffusion of the heat transfer path are blocked by the coordinated control of the main refrigeration circuit and the auxiliary refrigeration circuit.

[0060] The system collects temperature data from each monitoring point in real time and updates the temperature distribution heat map. It then repeats the steps of identifying abnormal temperature areas and controlling the cooling circuit until the temperature distribution of the energy storage battery module meets the preset conditions.

[0061] In one optional implementation, temperature data of the energy storage battery module immersed in the liquid cooling medium is collected, and the temperature change rate and temperature fluctuation value of adjacent monitoring points are calculated based on the temperature data to establish a temperature distribution heat map, including:

[0062] Temperature sensor arrays are set on the surface of the battery cells, the tab connection, the busbar and the edge of the energy storage battery module immersed in liquid cooling medium to acquire temperature data at each monitoring point. Wavelet transform is used to decompose the temperature data to obtain noise-reduced temperature data.

[0063] Based on the noise-reduced temperature data, the temperature difference between two adjacent monitoring points at two adjacent sampling times is calculated. The temperature difference is divided by the sampling time interval to obtain the temperature change rate. The sampling time interval is dynamically adjusted according to the temperature change trend. At the same time, the temperature fluctuation value of the monitoring point is calculated using an adaptive sliding variance algorithm to generate temperature feature data containing the temperature change rate and temperature fluctuation value.

[0064] The grid density is determined using temperature feature data, and the temperature interpolation of the grid nodes is calculated using radial basis functions. The temperature values ​​of the grid nodes are converted into color codes to generate an initial heat map reflecting the temperature distribution. For regions in the initial heat map where the temperature gradient is greater than a preset gradient threshold, the grid is subdivided, and the temperature values ​​of the subdivided grids are recalculated to generate a temperature distribution heat map with local densification characteristics.

[0065] This embodiment provides a method for collecting temperature data of an energy storage battery module immersed in a liquid cooling medium and establishing a temperature distribution heatmap. The liquid cooling medium can be a non-conductive cooling liquid such as transformer oil or fluorinated liquid. The energy storage battery module is composed of multiple battery cells connected in series and parallel, and the module has busbars and tab connections inside. Temperature sensors are evenly distributed around the battery module to form a sensor array. The temperature sensors are PT100 models, with a measurement accuracy of ±0.1℃ and a response time of less than 1 second, suitable for long-term operation in a liquid-cooled environment. The sensor installation locations include: the center point of each battery cell surface, each tab connection, busbar nodes, and the perimeter of the module, forming a monitoring network covering the entire module.

[0066] The temperature data acquisition module employs a 16-bit ADC converter, with an initial sampling frequency set to 1Hz. The acquired raw temperature data undergoes noise reduction via wavelet transform. Specifically, the db4 wavelet function is selected as the basis function, and the temperature signal is decomposed into five levels. After decomposition, approximation coefficients and detail coefficients are obtained. The detail coefficients are processed using a soft thresholding method, with the threshold set to three times the noise standard deviation. The processed coefficients are then combined with the unprocessed approximation coefficients for wavelet reconstruction to obtain the denoised temperature data. For example, the original surface measurements of a battery cell are 39.7℃, 39.2℃, 40.1℃, 39.5℃, and 40.3℃. After wavelet transform noise reduction, the values ​​are 39.5℃, 39.6℃, 39.8℃, 39.9℃, and 40.0℃, effectively removing random fluctuations during the measurement process.

[0067] The temperature change rate of adjacent monitoring points is calculated using the noise-reduced temperature data. For any two adjacent monitoring points A and B, at sampling times t1 and t2, the temperature change rate is calculated as the temperature difference between points A and B at times t1 and t2 divided by the time interval (t2-t1). For example, if the temperatures of adjacent monitoring points A and B are 38.5℃ and 38.7℃ respectively at time t1, and the temperatures change to 39.1℃ and 40.2℃ at time t2 (10 seconds later), then the temperature change rate is (39.1-38.5) / (40.2-38.7) divided by 10 seconds, which is 0.06℃ / second and 0.15℃ / second.

[0068] The sampling interval is dynamically adjusted based on temperature change trends. When the detected temperature change rate exceeds 0.2℃ / second, the sampling frequency is automatically increased to 2Hz; when the change rate exceeds 0.5℃ / second, the sampling frequency is increased to 5Hz; when the change rate is below 0.1℃ / second for one minute, the sampling frequency drops back to 1Hz to balance data accuracy and system resource usage. Simultaneously, an adaptive sliding variance algorithm is used to calculate temperature fluctuation values. Twenty sampling points are selected as the initial window size, and the variance of the temperature data within the window is calculated as the fluctuation value. The window size is automatically adjusted according to the temperature change rate; when the temperature change rate increases, the window size decreases to 10 sampling points; when the temperature stabilizes, the window size expands to 30 sampling points. For example, the variance of temperature data from 20 consecutive samples at a certain monitoring point is 0.025℃. 2 The temperature fluctuation at that point is 0.158℃ (square root of variance).

[0069] Temperature characteristic data is generated, including the location coordinates of each monitoring point, real-time temperature value, temperature change rate, and temperature fluctuation value. The surface of the battery module is initially divided into a uniform grid of 50×30, and the grid density can be dynamically adjusted according to the temperature characteristic data. In areas where the temperature change rate is greater than 0.3℃ / second or the temperature fluctuation value is greater than 0.2℃, the grid density is doubled, that is, the grid size in that area is reduced to half of the original size.

[0070] Temperature interpolation of grid nodes is performed using radial basis functions (RBF). For each grid node, data from the eight nearest monitoring points are selected, and a quadratic RBF is applied to calculate the node's temperature. The radius of influence is set to 1.5 times the average distance between monitoring points. For example, if the temperatures of the eight monitoring points surrounding a grid node are 39.5℃, 40.2℃, 39.8℃, 40.5℃, 40.1℃, 39.7℃, 40.0℃, and 40.3℃, the interpolated temperature of that node calculated using RBF is 40.1℃.

[0071] The calculated grid node temperature values ​​are converted into color codes to generate an initial heatmap. The color mapping uses a gradient spectrum from dark blue (25℃) to dark red (60℃), with each 1℃ corresponding to a color level. For example, 38℃ is mapped to light blue, 45℃ to orange, and 55℃ to dark red.

[0072] The initial heatmap is optimized by identifying regions with temperature gradients greater than 2℃ / cm, and these regions are then subdivided into smaller grids. The subdivided grid size is one-quarter of the original, and the temperature interpolation for each subdivided grid is recalculated. For example, at the battery tab connection, the temperature change from 42℃ to 46℃ over a distance of only 1.8cm results in a temperature gradient of 2.22℃ / cm, exceeding a preset threshold. Therefore, this region is automatically subdivided into a 16×16 grid from the original 4×4 grid, resulting in a finer temperature distribution in the heatmap for that area.

[0073] The resulting temperature distribution heatmap features localized detailing, clearly displaying the temperature distribution on the battery module surface, especially providing higher-resolution temperature visualization in areas of significant temperature change. This heatmap is updated in real time to monitor the thermal behavior of the battery module, identify potential hotspots and abnormal temperature distributions early, and provide decision-making support for the battery thermal management system.

[0074] In this embodiment, comprehensive temperature data is acquired through a precisely arranged array of temperature sensors, and wavelet transform denoising improves data accuracy. Temperature change rate calculation and adaptive sampling time adjustment enable rapid response to temperature anomalies, while the adaptive sliding variance algorithm enhances sensitivity to temperature fluctuations. Adaptive grid partitioning based on temperature feature data and radial basis function interpolation significantly improve the accuracy of the heatmap, especially for the refined representation of hotspot regions. The heatmap with localized encryption visually presents the temperature distribution, enabling the system to promptly identify hotspot regions and implement targeted cooling, effectively preventing the risk of thermal runaway.

[0075] In one optional implementation, monitoring points where the rate of temperature change exceeds a preset value are marked on the temperature distribution heatmap. The temperature fluctuation values ​​of the marked monitoring points are then partitioned according to their numerical values, and a temperature anomaly area distribution map is output, including:

[0076] Based on the charging and discharging state of the energy storage battery module and the ambient temperature, a benchmark threshold for the rate of temperature change is set. Monitoring points that exceed the benchmark threshold in the temperature distribution heat map are marked as abnormal points. The temperature fluctuation values ​​of the abnormal points are extracted, and the temperature fluctuation values ​​are divided into multiple numerical intervals through adaptive interval division to establish a temperature feature distribution sequence of abnormal points.

[0077] Anomalies in the temperature feature distribution sequence whose temperature fluctuation feature similarity is greater than a preset similarity threshold are aggregated into the same group. Taking each group of anomalies as the center, the monitoring points whose temperature fluctuation values ​​are in the same range and whose positions are adjacent are included in the corresponding anomaly group. Adjacent anomaly groups with common boundaries and whose temperature fluctuation values ​​differ from the preset fluctuation value are merged to generate a temperature anomaly area distribution map labeled with anomaly level.

[0078] In this embodiment, it is first necessary to establish an adaptive temperature change rate benchmark threshold. The benchmark threshold setting depends on the operating state and environmental conditions of the energy storage battery module. For lithium-ion battery energy storage systems, the temperature change rate benchmark threshold can be set to 0.5℃ / min during charging; 0.7℃ / min during discharging; and 0.2℃ / min during resting. These benchmark thresholds can also be adjusted according to the ambient temperature. For example, the above standards can be maintained at an ambient temperature of 25℃, and each threshold can be increased by 0.1℃ / min for every 10℃ increase in ambient temperature; and each threshold can be decreased by 0.1℃ / min for every 10℃ decrease in ambient temperature.

[0079] After obtaining the temperature distribution heatmap, the temperature change rate at each monitoring point is calculated. The change rate is calculated using a time window method, with a 5-minute observation window. Temperature values ​​are recorded every 30 seconds, and the average temperature change rate within the window is calculated. Taking a 10×10 monitoring grid as an example, the temperature change rate at each monitoring point is compared with the baseline threshold under the corresponding operating conditions, and points with change rates exceeding the threshold are marked as outliers.

[0080] For marked outliers, their temperature fluctuation values ​​are extracted, which are the difference between the measured rate of temperature change and the baseline threshold. For example, if the rate of temperature change at a monitoring point during charging is 0.8℃ / min and the baseline threshold is 0.5℃ / min, then the temperature fluctuation value is 0.3℃ / min. After collecting the temperature fluctuation values ​​of all outliers, an adaptive interval partitioning method is applied to divide these values ​​into zones.

[0081] The adaptive interval division adopts a data distribution-based method. Assuming the collected temperature fluctuation values ​​range from 0.1 to 1.5℃ / minute, the quartiles of the data are first calculated: Q1 = 0.3℃ / minute, Q2 (median) = 0.6℃ / minute, and Q3 = 0.9℃ / minute. Based on the interquartile range (IQR) = Q3 - Q1 = 0.6℃ / minute, the temperature fluctuation values ​​are divided into four intervals: a slight anomaly interval [0.1, 0.3)℃ / minute, a moderate anomaly interval [0.3, 0.6)℃ / minute, a severe anomaly interval [0.6, 0.9)℃ / minute, and an extreme anomaly interval [0.9, 1.5]℃ / minute. The system assigns a different identifier and color to each interval: yellow for slight anomalies, orange for moderate anomalies, red for severe anomalies, and dark red for extreme anomalies.

[0082] Based on the defined intervals, a temperature feature distribution sequence is established, mapping each anomaly point to its corresponding interval to form a feature vector. For example, if a battery module has 25 anomalies, including 8 minor anomalies, 10 moderate anomalies, 5 severe anomalies, and 2 extreme anomalies, a feature sequence of the form {L1: 0.15, L1: 0.23, ..., L4: 1.2, L4: 1.4} can be established, where L1 to L4 represent the anomaly level, and the following values ​​are the specific temperature fluctuation values.

[0083] When aggregating similar outliers, a temperature feature similarity threshold of 0.85 is set. The similarity of temperature fluctuation features between two outliers is calculated. If the similarity is greater than 0.85 and the two points belong to the same outlier interval, they are aggregated into the same group. The similarity calculation is based on the proximity of temperature fluctuation values ​​and the consistency of their trends. For example, if two outliers have temperature fluctuation values ​​of 0.65℃ / min and 0.68℃ / min respectively, and their trends are consistent, the calculated similarity is 0.92, exceeding the threshold, and they are grouped together.

[0084] For each anomalous group formed by aggregation, a region expansion is performed. Taking the anomalous point in each group as the center, the monitoring points physically adjacent to it are checked. If the temperature fluctuation value of an adjacent point is within the same numerical range as the center point, then that point is included in the corresponding anomalous group. For example, if an anomalous point belongs to a severe anomalous range, and its adjacent monitoring point has a temperature fluctuation value of 0.7℃ / minute, which also belongs to a severe anomalous range, then the adjacent point is included in the same anomalous group.

[0085] After initial area expansion, the boundary relationships between different anomaly groups are checked. A preset fluctuation value difference threshold is set to 0.2℃ / minute. For two anomaly groups with a common boundary, if their average temperature fluctuation value difference is less than 0.2℃ / minute, the two groups are merged into a larger anomaly region. For example, if the average temperature fluctuation value of anomaly group A is 0.65℃ / minute and the average temperature fluctuation value of anomaly group B is 0.78℃ / minute, the difference is 0.13℃ / minute, which is less than the threshold, so the two groups are merged. Finally, a temperature anomaly region distribution map labeled with anomaly levels is generated. On this distribution map, different colors represent different anomaly level regions, and gradient colors are used to show the changing trend of temperature fluctuation values. For regions with blurred boundaries, a semi-transparent transition is used. The system also marks the core location, area size, and highest temperature fluctuation value of each anomaly region on the distribution map, facilitating maintenance personnel to quickly locate the problem area.

[0086] In this embodiment, by dynamically setting a baseline threshold based on the charging / discharging state and ambient temperature, anomaly monitoring becomes more targeted and adaptable. Adaptive interval division of temperature fluctuation values ​​improves the ability to identify temperature anomalies of different degrees. Anomaly point grouping methods based on similarity aggregation effectively identify regions with the same thermal characteristics, while outward expansion and merging mechanisms accurately define the actual boundaries of temperature anomaly regions. The generated temperature anomaly region distribution map visually displays the anomaly level, facilitating the system's rapid location of key hotspots, assessment of anomaly severity, and diffusion trends. This zoning and hierarchical anomaly monitoring mechanism allows cooling resources to be allocated preferentially according to anomaly severity, improving cooling efficiency and providing data support for predicting potential thermal runaway, significantly enhancing the safety management capabilities of the energy storage system.

[0087] In one optional implementation, based on the temperature fluctuation values ​​of each region in the temperature anomaly distribution map, the temperature transfer coefficient between adjacent regions is calculated. The heat transfer path is determined according to the numerical order of the temperature transfer coefficient, and the heat transfer path is divided into multiple control sub-regions according to a preset interval of the temperature transfer coefficient, including:

[0088] Obtain temperature data from monitoring points within each abnormal region in the temperature anomaly distribution map, calculate the temperature change value at adjacent sampling times, determine the fluctuation amplitude, fluctuation period, and fluctuation direction of the temperature change, and combine them to form the temperature fluctuation value;

[0089] For adjacent abnormal regions, the correlation coefficient of temperature fluctuation values ​​within a specified time window is calculated to obtain the degree of matching. The time difference corresponding to the maximum correlation coefficient is determined to determine the matching delay. Combined with the weighting factor, the temperature transfer coefficient is determined.

[0090] The abnormal region with the largest temperature fluctuation value is selected as the starting region. The abnormal regions are connected along the direction of the maximum gradient of the temperature transfer coefficient to form a transfer path. The dividing point of the transfer path is determined according to the time change sequence of the temperature fluctuation value. The heat transfer paths are obtained by arranging them in descending order of the temperature transfer coefficient.

[0091] Based on the preset range of temperature transfer coefficients between adjacent abnormal areas along the heat transfer path, the heat transfer path is divided into multiple control sub-regions.

[0092] During implementation, the temperature data of monitoring points within each abnormal area in the temperature anomaly distribution map is first acquired. Optionally, the sampling frequency is set to once every 0.5 seconds, and each abnormal area contains at least 3 monitoring points. The temperature change value is obtained by calculating the temperature difference between adjacent sampling times. Taking an abnormal area A as an example, the temperature data of one monitoring point within 10 consecutive seconds is [38.2℃, 38.5℃, 38.7℃, 38.9℃, 39.2℃, 39.4℃, 39.5℃, 39.3℃, 39.1℃, 38.8℃], and the temperature difference sequence between adjacent sampling times is [0.3℃, 0.2℃, 0.2℃, 0.3℃, 0.2℃, 0.1℃, -0.2℃, -0.2℃, -0.3℃].

[0093] The fluctuation amplitude is calculated based on the temperature change value, i.e., the absolute value of the temperature change. In the example above, the fluctuation amplitude sequence is [0.3℃, 0.2℃, 0.2℃, 0.3℃, 0.2℃, 0.1℃, 0.2℃, 0.2℃, 0.3℃]. The change in fluctuation amplitude per unit time is taken as the fluctuation period, which is obtained by dividing the difference between adjacent fluctuation amplitudes by the sampling interval. In this example, the fluctuation period sequence is [0.2℃ / s, 0℃ / s, 0.2℃ / s, -0.2℃ / s, -0.2℃ / s, 0.2℃ / s, 0℃ / s, 0.2℃ / s]. The difference in the direction of temperature change between adjacent moments is taken as the fluctuation direction, determined by comparing the signs of adjacent temperature change values: the same direction is denoted as 1, opposite directions as -1, and no change in temperature as 0. In this example, the fluctuation direction sequence is [1, 1, 1, 1, 1, -1, -1, -1]. The temperature fluctuation value is formed by combining the fluctuation amplitude, fluctuation period, and fluctuation direction. Specifically, the fluctuation amplitude is multiplied by the fluctuation direction, and then the fluctuation period is multiplied by 0.5.

[0094] For adjacent anomalous regions, the correlation coefficient of temperature fluctuation values ​​within a specified time window is calculated to determine the degree of matching. The time window length is set to 5 seconds, and the correlation between the temperature fluctuation value sequences of the two regions is calculated using the Pearson correlation coefficient method. For example, if the temperature fluctuation value sequences of region A and adjacent region B within the same time window are [0.4, 0.3, 0.3, 0.4, 0.1] and [0.2, 0.3, 0.4, 0.3, 0.2] respectively, the calculated correlation coefficient is 0.316, indicating that the temperature fluctuations of the two regions have a certain correlation. The time difference at which the correlation coefficient is at its maximum is determined by shifting the time window as the matching delay. Shifting the time window of region B forward by 1 second, the temperature fluctuation value sequence becomes [0.3, 0.4, 0.3, 0.2, 0.1], and the correlation coefficient rises to 0.685; shifting it forward another second, the correlation coefficient becomes 0.421. Therefore, the matching delay is determined to be 1 second, indicating that it takes approximately 1 second for heat to transfer from region A to region B.

[0095] The ratio of the contact boundary length to the shape feature value is used as a weighting factor. The contact boundary length is the boundary length shared by two adjacent regions, and the shape feature value is the ratio of the region's perimeter to its area. Assuming the contact boundary length between region A and region B is 25 mm, the perimeter of region B is 120 mm, and its area is 300 mm², then the shape feature value is 0.4, and the weighting factor is calculated as 25 / (120 / 300) = 62.5. The product of the matching degree, matching delay, and weighting factor is used as the temperature transfer coefficient. In this example, the temperature transfer coefficient is 0.685 × 1 × 62.5 = 42.8, representing the intensity of heat transfer from region A to region B.

[0096] The region with the largest temperature fluctuation was selected as the starting region. In actual monitoring, region C was found to have an average temperature fluctuation of 0.8℃, higher than other regions; therefore, region C was selected as the starting region. The temperature transfer coefficient gradient value, i.e., the rate of change of the temperature transfer coefficient between adjacent regions, was calculated. A transfer path was formed by connecting the abnormal regions along the direction of the largest gradient. Starting from region C, the temperature transfer coefficient of all its adjacent regions was calculated. Region D was found to have the largest temperature transfer coefficient at 56.2; therefore, regions C and D were connected. Then, starting from region D, the temperature transfer coefficient of region E was found to be 49.7, higher than other adjacent regions; region E was then connected, and so on.

[0097] The heat transfer efficiency is calculated based on the time-varying sequence of temperature fluctuations to determine the dividing point of the heat transfer path. The heat transfer efficiency is defined as the ratio of the temperature fluctuation value in the downstream region to that in the upstream region. When the heat transfer efficiency is below 0.6, it indicates a significant reduction in heat transfer, and this point is used as the dividing point of the heat transfer path. The heat transfer paths are then arranged in descending order of temperature transfer coefficient values. In practical applications, the obtained heat transfer path is C→D→E→G→H→J, with transfer coefficients of 56.2, 49.7, 42.3, 31.5, and 15.8, respectively. The heat transfer efficiency between G and H is 0.55, which is below the threshold of 0.6; therefore, a dividing point is set between G and H.

[0098] The temperature transfer coefficient is divided into preset intervals: a high transfer interval (40, 60], a medium transfer interval (20, 40], and a low transfer interval [0, 20]. Based on the interval to which the temperature transfer coefficient between adjacent abnormal regions on the heat transfer path belongs, the heat transfer path is divided into multiple control sub-regions. In this example, C→D→E belongs to the high transfer interval, forming control sub-region 1; G belongs to the medium transfer interval, forming control sub-region 2; and H→J belongs to the low transfer interval, forming control sub-region 3.

[0099] Based on the defined control sub-regions, differentiated cooling strategies were developed for the immersion liquid-cooled energy storage system. For control sub-region 1, due to its high heat transfer coefficient, active pressurized circulation cooling was adopted, increasing the liquid cooling medium flow rate by 30% while simultaneously reducing the medium temperature by 2°C. For control sub-region 2, standard flow rate circulation cooling was used. For control sub-region 3, intermittent circulation cooling was employed, starting for 10 seconds every 30 seconds. Through this differentiated cooling strategy, during a certain charging process, the highest temperature in control sub-region 1 decreased from 42°C to 38°C, control sub-region 2 decreased from 39°C to 37°C, and control sub-region 3 remained at around 36°C, effectively suppressing hotspot diffusion and optimizing the allocation of cooling resources.

[0100] In this embodiment, temperature fluctuation values ​​are constructed by introducing multi-dimensional features such as temperature fluctuation amplitude, fluctuation period, and direction. Furthermore, the temperature transfer coefficient is calculated based on the correlation between temperature fluctuation values ​​between adjacent abnormal regions. Compared to existing technologies that typically rely solely on temperature thresholds for simple judgment or use average temperature differences to assess heat conduction trends, this method more accurately captures the dynamic characteristics of temperature anomaly propagation. Furthermore, by analyzing the moment of maximum correlation through a moving time window and introducing contact boundary morphology as a weighting factor, the calculation of the temperature transfer coefficient takes into account both time delay and spatial coupling characteristics, unlike existing thermal path analysis methods that ignore spatial structural differences. In addition, this application starts from the region with the most severe temperature fluctuations, constructs a heat transfer path along the transfer coefficient gradient, and identifies key dividing points in the path based on the changing trends of fluctuation characteristics, achieving refined modeling of the heat transfer chain. Finally, the path is segmented and managed according to a preset interval of the transfer coefficient, effectively improving the accuracy and response efficiency of the thermal control strategy. The improvement aims to overcome the problems of existing technologies in identifying abnormal hot zones and grasping the evolution trend of thermal paths, thereby enabling real-time tracking and proactive intervention of the thermal diffusion process of energy storage systems, and improving the system's safety and the level of intelligent thermal management.

[0101] Figure 2 This is a comparison chart of the effects of differentiated cooling strategies in embodiments of the present invention, such as... Figure 2As shown in the figure, this diagram illustrates the performance of three different cooling methods across four key performance indicators. Regarding the reduction in maximum temperature, this technical solution achieves a reduction of 4.0℃, significantly outperforming the uniform cooling method (3.0℃) and the temperature threshold-based cooling method (3.4℃), demonstrating the effective control of hotspot temperatures by the differentiated cooling strategy. In terms of temperature uniformity, this technical solution achieves a uniformity of 1.6℃, superior to the other two methods (2.0℃ and 2.6℃), indicating smaller temperature differences across regions and a more uniform distribution of hotspots. Regarding cooling energy consumption, this technical solution consumes only 2.6 kW·h, demonstrating significant energy efficiency. In terms of system stabilization time, this technical solution achieves system temperature stabilization in just 1.4 minutes, indicating a faster response speed. Overall, considering all four indicators, this technical solution achieves significant optimization in cooling effect, uniformity, energy efficiency, and response speed through precise identification of heat transfer paths and a differentiated cooling strategy. While traditional uniform cooling methods are simple to implement, they result in high energy consumption and slow response because the same cooling parameters are applied to all areas. Although temperature threshold-based methods are an improvement over uniform cooling, they are difficult to achieve optimal cooling performance because they do not consider heat transfer paths.

[0102] In one optional implementation, the dividing points of the heat transfer path are determined based on the time variation sequence of temperature fluctuation values, and the heat transfer paths are obtained by arranging them in descending order of their temperature transfer coefficients, including:

[0103] A sliding time window is used to analyze the time series of temperature fluctuation values. The statistical characteristics and the rate of change of temperature fluctuation values ​​within the window are calculated to obtain the time series feature vector.

[0104] Based on the time-series feature vector, the energy loss rate at adjacent time points is calculated to obtain the instantaneous transfer efficiency, the cumulative transfer efficiency within a fixed window is calculated, and the average transfer efficiency over the entire period is calculated. The instantaneous transfer efficiency, cumulative transfer efficiency, and average transfer efficiency are combined to obtain the transfer efficiency index.

[0105] Using the aforementioned transfer efficiency index, the degree of abrupt change in transfer efficiency, the inflection point amplitude, and the depth of local troughs are calculated respectively. The locations with the greatest degree of abrupt change, the locations with the greatest inflection point amplitude, and the locations with the deepest troughs are combined and scored to determine the dividing points of the heat transfer path.

[0106] The heat transfer path is segmented based on the dividing points, the weighted average of the temperature transfer coefficient of each segment is calculated, and the comprehensive score of the transfer path is calculated by combining the fluctuation characteristics of the temperature transfer coefficient and the distribution of the dividing points.

[0107] Based on the comprehensive score, the heat transfer paths are sorted from largest to smallest according to their temperature transfer coefficient values ​​to obtain the heat transfer paths.

[0108] In this embodiment, a sliding time window is first used to analyze the time series of temperature fluctuation values. The window length is set to 15 seconds, and the sliding step is 3 seconds. For each sliding window, the statistical characteristics within the window are calculated, including the mean, standard deviation, maximum value, minimum value, and median. Taking the temperature fluctuation values ​​of an abnormal region A during the charging process as an example, the mean value of the first window is 0.632℃, the standard deviation is 0.142℃, the maximum value is 0.78℃, the minimum value is 0.35℃, and the median is 0.68℃.

[0109] Simultaneously, the rate of change characteristics of temperature fluctuation values ​​are calculated, including first-order and second-order differences. First-order differences calculate the difference between temperature fluctuation values ​​at adjacent time points, while second-order differences calculate the difference between the first-order difference sequences. Within the first window, the average value of the first-order differences is 0.02℃, and the average value of the second-order differences is -0.01℃. The statistical characteristics and rate of change characteristics are combined to form a time-series feature vector for subsequent analysis.

[0110] Based on the time-series feature vector, the instantaneous transfer efficiency is obtained by calculating the energy loss rate at adjacent time points. The energy loss rate is defined as the deviation between the ratio of temperature fluctuation values ​​in adjacent anomalous regions and the ideal transfer ratio. Assuming that the temperature fluctuation values ​​of region A and adjacent region B at a certain moment are 0.62℃ and 0.50℃ respectively, the instantaneous transfer efficiency is calculated to be 0.806. The cumulative transfer efficiency within a fixed window is calculated, defined as the weighted sum of the instantaneous transfer efficiencies at all time points within the window, with the weights proportional to the temperature fluctuation values ​​at those time points. The cumulative transfer efficiency for the first window is 0.827. The average transfer efficiency for the entire period is calculated, which is 0.817 in this example. The instantaneous transfer efficiency, cumulative transfer efficiency, and average transfer efficiency are combined to obtain the transfer efficiency index, with the following combination: instantaneous transfer efficiency accounts for 50%, cumulative transfer efficiency accounts for 30%, and average transfer efficiency accounts for 20%.

[0111] Using the transfer efficiency index, we calculate the degree of abrupt change, inflection point amplitude, and local trough depth of the transfer efficiency. The degree of abrupt change is defined as the change in the transfer efficiency index between two adjacent windows divided by the time interval. In the example, the transfer efficiency index for the first window is 0.818, and for the second window it is 0.815, with a degree of abrupt change of 0.001. The inflection point amplitude is defined as the local extremum of the second derivative of the transfer efficiency index curve. The local trough depth is defined as the difference between the local minimum of the transfer efficiency index and the average of the local maximum values ​​on its left and right sides.

[0112] During the analysis, it was found that the heat transfer efficiency index of one window was 0.736, a decrease of 0.048 compared to the previous window, with a sudden change of 0.016; the heat transfer efficiency index of another window showed an inflection point with an amplitude of 0.008; and another window formed a local trough with a depth of 0.063. The location with the greatest sudden change, the location with the largest inflection point amplitude, and the location with the deepest trough were combined and scored to determine the dividing point of the heat transfer path. The combined scoring method was: a weight of 0.5 for the sudden change, a weight of 0.3 for the inflection point amplitude, and a weight of 0.2 for the trough depth. The 27th second was calculated as the dividing point of the heat transfer path.

[0113] The heat transfer path is segmented based on dividing points, and the weighted average of the temperature transfer coefficient for each segment is calculated. The weighted average of the temperature transfer coefficient is calculated by summing the products of the temperature transfer coefficient and the corresponding temperature fluctuation value at each time point, and dividing by the sum of the temperature fluctuation values. In the example, the weighted average of the temperature transfer coefficient before the dividing point is 42.6, and the weighted average of the temperature transfer coefficient after the dividing point is 31.4. Combining the fluctuation characteristics of the temperature transfer coefficient and the distribution location of the dividing points, a comprehensive score for the heat transfer path is calculated. The fluctuation characteristics include the standard deviation and coefficient of variation of the temperature transfer coefficient. The standard deviation for the first segment is 5.2, and the coefficient of variation is 0.122; the standard deviation for the second segment is 3.8, and the coefficient of variation is 0.121. The distribution location characteristic of the dividing points is their relative position within the entire path, which is 0.6 in this case. The comprehensive score is calculated as follows: the weighted average of the temperature transfer coefficient accounts for 60%, the fluctuation characteristics account for 25%, and the distribution location of the dividing points accounts for 15%. The comprehensive score for the first segment is 38.9, and for the second segment it is 29.2.

[0114] The heat transfer paths are sorted from highest to lowest based on the comprehensive score. In a practical application, a certain submerged liquid-cooled energy storage system contains 10 heat transfer path segments. The comprehensive scores calculated according to the above method, from highest to lowest, are as follows: Path A (45.6), Path B (41.3), Path C (38.9), Path D (35.2), Path E (33.7), Path F (30.5), Path G (29.2), Path H (26.8), Path I (23.4), and Path J (19.7).

[0115] Based on the ranked heat transfer paths, the liquid cooling system prioritizes enhanced cooling for high-scoring paths. Path AC is cooled using high-pressure differential circulation; path DF is cooled using standard circulation; and path GJ is cooled using low-power intermittent circulation.

[0116] This application constructs a time-varying sequence of temperature fluctuation values ​​and uses a sliding window to extract statistical features, enabling dynamic reflection of the energy evolution of abnormal heat during the transfer process. Compared to existing technologies that construct heat transfer paths based solely on static temperature differences or fixed sampling points, this scheme introduces a comprehensive evaluation method using instantaneous, cumulative, and average transfer efficiency indicators, significantly improving the ability to perceive trends in heat transfer efficiency changes. By identifying abrupt changes, inflection points, and local troughs in the transfer efficiency indicators, it achieves automatic segmentation of key nodes in the heat transfer path, and then performs weighted modeling based on the temperature transfer characteristics of different segments to obtain a more representative comprehensive score. Existing technologies generally lack a detailed characterization of local characteristics and dynamic efficiency changes in the heat path, leading to delayed response or insufficient control precision in thermal control strategies. In contrast, this application starts from time-series behavior, identifies turning points in the path structure, and prioritizes the overall path, achieving accurate extraction and dynamic management of the heat propagation path, which helps improve the initiative, precision, and energy efficiency of the system's thermal control scheduling.

[0117] In one optional implementation, based on the spatial distribution characteristics of the heat transfer path, the main refrigeration circuit is activated in the controlled sub-region where the temperature transfer coefficient exceeds a preset temperature value, and the auxiliary refrigeration circuit in the adjacent controlled sub-region is pre-activated in the extension direction of the heat transfer path. The coordinated control of the main refrigeration circuit and the auxiliary refrigeration circuit blocks the extension and diffusion of the heat transfer path, including:

[0118] Collect temperature field data within the controlled sub-region, calculate the spatial distribution and intensity variation of the temperature field data, and obtain the distribution characteristics of the heat transfer path;

[0119] Calculate the temperature transfer coefficient of each control sub-region based on the temperature field data. When the temperature transfer coefficient exceeds the preset temperature value, determine the over-limit region and set and activate the main refrigeration circuit according to the heat load.

[0120] Based on the distribution characteristics of the heat transfer path, the adjacent control sub-regions of the over-limit area in the direction of the heat transfer path extension are determined. When the heat transfer rate reaches the preset start-up condition, the auxiliary cooling circuit is set and started.

[0121] Collect temperature field data of the control sub-area, adjust the power of the main refrigeration circuit according to the temperature change rate of the control area of ​​the main refrigeration circuit, and adjust the power of the auxiliary refrigeration circuit according to the temperature change rate of the control area of ​​the auxiliary refrigeration circuit.

[0122] The system calculates and controls the heat transfer rate of the sub-area in real time, switches auxiliary cooling circuits with heat transfer rates exceeding a preset threshold to main cooling circuits, and switches main cooling circuits with heat transfer rates below a preset threshold to standby mode, thus blocking the extension of the heat transfer path.

[0123] In submerged liquid-cooled energy storage devices, collecting temperature field data within the controlled sub-regions is fundamental to implementing this method. High-precision temperature sensor arrays are used for temperature field data acquisition, with at least nine temperature measurement points arranged in a 3×3 grid layout within each controlled sub-region. The temperature sensor sampling frequency is set to 2Hz, with an accuracy of ±0.1℃. When calculating the spatial distribution and intensity variation of the temperature field data, a bicubic spline interpolation algorithm is used to convert discrete temperature measurement point data into a continuous temperature field distribution. The specific process of bicubic spline interpolation is as follows: each temperature measurement point is considered a control point, and an interpolation control grid is established using its coordinates (x, y) and temperature value T; for any point (x', y') within the grid, the temperature value of that point is obtained by calculating its distance weight from the control point; the distance weight calculation uses a cubic polynomial kernel function to ensure that the interpolation surface passes through the control point and that the first derivative at the control point is continuous. In practical applications, nine temperature measurement points within the control sub-region A of an energy storage battery module are arranged in a 3×3 grid according to their spatial location, with coordinates in centimeters. The coordinates and temperature values ​​of the measurement points are as follows: (0, 0, 38.5℃), (5, 0, 39.2℃), (10, 0, 38.7℃), (0, 5, 39.6℃), (5, 5, 40.2℃), (10, 5, 39.8℃), (0, 10, 39.1℃), (5, 10, 39.5℃), and (10, 10, 38.9℃). The temperature field within the entire 10cm×10cm region is calculated using interpolation with a resolution of 0.5cm, resulting in a 21×21 temperature grid. Based on the interpolation results, the temperature gradient vector field is calculated to determine the direction and intensity of heat transfer at each point. The analysis results show that the heat mainly extends from the center (5, 5) point to the upper right, and the maximum temperature gradient is 0.42℃ / cm, which appears at coordinates (6.5, 5.5). The angle of the heat transfer direction is about 32°.

[0124] The temperature transfer coefficient of each controlled sub-region is calculated based on temperature field data. The temperature transfer coefficient is calculated based on the temperature difference and distance between adjacent temperature measurement points. The calculation method is to divide the temperature difference by the distance and then multiply by a correction factor for the material's thermal conductivity. Specifically, two adjacent points along the direction of the largest temperature gradient are selected, and the ratio of their temperature difference to distance is calculated, then multiplied by the correction factor for the thermal conductivity of the battery module material (this correction factor is related to the material's thermal conductivity and the characteristics of the liquid cooling medium; in this example, it is set to 8.4). For controlled sub-region A, the temperature transfer coefficient between points (5, 5) and (6.5, 5.5) is calculated as: (40.2 - 39.3)℃ ÷ 1.5cm × 8.4 = 46.8 W / m·K. The preset temperature value is set to 40 W / m·K. When the temperature transfer coefficient exceeds this value, the heat load of the current controlled sub-region needs to be calculated. The heat load calculation is based on the temperature field distribution, the area of ​​the region, and the rate of temperature change. Specifically, it is the difference between the average temperature of the region and the reference temperature multiplied by the area of ​​the region, the specific heat capacity, and the density. The reference temperature is set at 35℃, the area is 0.01m², the specific heat capacity of the battery module is 2100J / kg·K, and the effective density is 7.8kg / m³. 3 The average temperature in the area is 39.5℃. The heat load of the controlled sub-area A is calculated as (39.5-35)℃ × 0.01m³. 2 ×2100J / kg·K×7.8kg / m 3 =819J. The main refrigeration circuit is set and activated according to the heat load. The refrigeration power of the main refrigeration circuit is set to 1.2 times the heat load divided by the response time. Here, the response time is set to 30 seconds, and the refrigeration power is 819J×1.2÷30s=32.8W.

[0125] Based on the distribution characteristics of the heat transfer path, adjacent controlled sub-regions along the extension direction of the heat transfer path are identified for controlled sub-regions where the temperature transfer coefficient exceeds the preset temperature value. The extension direction of the heat transfer path is determined by the temperature gradient vector. In controlled sub-region A, heat is mainly transferred in the 32° direction, and the corresponding adjacent controlled sub-region is B. The heat transfer rate of controlled sub-region B is calculated based on the boundary temperature difference and thermal resistance, specifically by dividing the boundary temperature difference by the thermal resistance. The boundary temperature difference is the temperature difference at the boundary between region A and region B, which is 39.8℃ - 37.2℃ = 2.6℃ in this case. The thermal resistance depends on the material properties and the contact area, and here the thermal resistance value is 0.03m. 2 • K / W. The heat transfer rate in controlled sub-region B is 2.6℃ ÷ 0.03m. 2K / W = 86.7W. The preset start-up condition is that the heat transfer rate exceeds 80W, therefore the auxiliary cooling circuit needs to be activated. The cooling power of the auxiliary cooling circuit is set according to the heat transfer rate, and the calculation method is to multiply the heat transfer rate by the pre-cooling coefficient. Here, the pre-cooling coefficient is set to 0.8, and the cooling power is 86.7W × 0.8 = 69.4W.

[0126] After activating the main and auxiliary cooling circuits, temperature field data for the controlled sub-area is continuously collected. The first temperature change rate of the main cooling circuit's controlled area is calculated every 5 seconds, by dividing the change in the area's average temperature over 5 seconds by the time interval. During implementation, the average temperature of controlled sub-area A decreased from 39.5℃ to 38.7℃, with a first temperature change rate of (38.7-39.5)℃÷5s=-0.16℃ / s. The cooling power of the main cooling circuit is adjusted based on this first temperature change rate using the following formula: when the temperature change rate is greater than -0.1℃ / s, maintain the current power; when the temperature change rate is between -0.2℃ / s and -0.1℃ / s, reduce the power by 10%; when the temperature change rate is less than -0.2℃ / s, reduce the power by 20%. Therefore, the main cooling circuit power is adjusted to 32.8W×90%=29.5W. Simultaneously, the second temperature change rate of the auxiliary cooling circuit control area is calculated. The average temperature of the controlled sub-region B decreases from 37.2℃ to 36.9℃, and the second temperature change rate is (36.9-37.2)℃÷5s=-0.06℃ / s. The cooling power of the auxiliary cooling circuit is adjusted according to this second temperature change rate. The adjustment formula is as follows: when the temperature change rate is greater than -0.05℃ / s, the power is increased by 15%; when the temperature change rate is between -0.1℃ / s and -0.05℃ / s, the current power is maintained; when the temperature change rate is less than -0.1℃ / s, the power is reduced by 15%. Therefore, the auxiliary cooling circuit power is maintained at 69.4W.

[0127] The system calculates the heat transfer rate of the controlled sub-regions in real time and monitors changes in the heat transfer path. After 30 seconds of continuous operation, the heat transfer rate of controlled sub-region B rises to 103.2W, exceeding the preset rate threshold of 100W. Therefore, the auxiliary cooling circuit of sub-region B needs to be switched to the main cooling circuit. During the switchover, the cooling power of sub-region B is increased from 69.4W to 103.2W × 1.2 = 123.8W. Simultaneously, the auxiliary cooling circuit of sub-region C (the next sub-region in the heat transfer direction of sub-region B) is activated, with its cooling power set to the heat transfer rate of sub-region C (65.4W) multiplied by the pre-cooling coefficient of 1.03, resulting in 67.4W. At the same time, the heat transfer rate of controlled sub-region A drops to 62.3W, below the preset rate threshold of 70W. The main cooling circuit of sub-region A is switched to standby mode, maintaining minimum power operation, set to 20% of the initial power, i.e., 32.8W × 20% = 6.56W.

[0128] The aforementioned coordinated control strategy effectively blocked the extension of the heat transfer path. In a practical application, during the fast charging process of the energy storage battery module, the hot spot area would normally spread along the heat transfer path, causing the temperature of multiple areas to exceed the safety threshold. After applying this method, the highest temperature of the main hot spot area dropped from 40.2℃ to 38.3℃, the heat transfer path was confined to within the initial diffusion range, and the temperature of adjacent areas was controlled within the safe range, with the highest not exceeding 37.5℃.

[0129] In this embodiment, by analyzing the spatial distribution characteristics of heat transfer paths and dynamically calculating the temperature transfer coefficient and heat load of each sub-region using temperature field data, rapid identification of high-risk areas for thermal anomalies and precise activation of the main refrigeration loop are achieved. Analysis of heat transfer rate and direction is introduced to identify path extension trends in advance and pre-activate auxiliary refrigeration loops in adjacent regions before heat diffusion occurs, achieving forward intervention control of heat diffusion. Through coordinated adjustment of the main and auxiliary loops, not only is the flexibility of local cooling response enhanced, but the refrigeration power is also adjusted in real time based on the temperature change rate and heat transfer rate. The roles of the main and auxiliary loops are switched according to the dynamic heat transfer rate, ensuring both cooling efficiency and system stability. This effectively suppresses the diffusion and spread of heat within the energy storage device, improving the overall operational safety and energy efficiency ratio.

[0130] Figure 3 This is a graph showing the relationship between temperature transfer coefficient and heat load in an embodiment of the present invention, as shown below. Figure 3 As shown in the figure, this graph compares the relationship between the temperature transfer coefficient and the heat load, clearly demonstrating the significant difference between this technical solution and the traditional proportional threshold control method in heat load calculation. The horizontal axis represents the temperature transfer coefficient (W / m·K), ranging from 20 to 70; the vertical axis represents the heat load (J), ranging from 0 to 1000. The graph uses dashed lines to mark preset thresholds, dividing the heat load region into low heat load region (0-170), medium heat load region (170-330), and high heat load region (330-500). The traditional proportional threshold control method uses a simple linear relationship to calculate the heat load, represented by square data points and dashed lines in the graph, increasing linearly with the temperature transfer coefficient. In contrast, this technical solution (circular data points and solid lines) exhibits a non-linear relationship, particularly showing a more rapid upward trend in the medium-to-high heat load regions. This nonlinear relationship better reflects the actual physical characteristics of heat transfer. By considering the spatial distribution characteristics of the heat transfer path, this technical solution achieves more accurate heat load calculation. Especially after the temperature transfer coefficient exceeds a preset threshold, it can more accurately reflect the actual heat load changes, providing a reliable basis for precise control of the refrigeration circuit. Although the proportional threshold control method is simple and intuitive, it ignores the nonlinear characteristics and spatial distribution effects in the heat transfer process, resulting in insufficient heat load estimation in the high-temperature transfer coefficient region and inaccurate refrigeration control.

[0131] A second aspect of the present invention provides a circulating cooling control system for an immersion liquid-cooled energy storage device, the system comprising:

[0132] The first unit is used to collect temperature data of energy storage battery modules immersed in liquid cooling medium, calculate the temperature change rate and temperature fluctuation value of adjacent monitoring points based on the temperature data, establish a temperature distribution heat map, mark the monitoring points whose temperature change rate exceeds the preset change value in the temperature distribution heat map, divide the temperature fluctuation value of the marked monitoring points into zones according to the value, and output a temperature anomaly area distribution map.

[0133] The second unit is used to calculate the temperature transfer coefficient between adjacent areas based on the temperature fluctuation values ​​of each area in the temperature anomaly distribution map, determine the heat transfer path according to the numerical order of the temperature transfer coefficient, and divide the heat transfer path into multiple control sub-areas according to the preset interval of the temperature transfer coefficient.

[0134] The third unit is used to activate the main refrigeration circuit in the control sub-region where the temperature transfer coefficient exceeds the preset temperature value according to the spatial distribution characteristics of the heat transfer path, and at the same time pre-activate the auxiliary refrigeration circuit in the adjacent control sub-region in the extension direction of the heat transfer path, so as to block the extension and diffusion of the heat transfer path through the coordinated control of the main refrigeration circuit and the auxiliary refrigeration circuit.

[0135] The fourth unit is used to collect temperature data from each monitoring point in real time and update the temperature distribution heat map. It repeatedly executes the steps of identifying abnormal temperature areas and controlling the cooling circuit until the temperature distribution of the energy storage battery module meets the preset conditions.

[0136] A third aspect of the present invention provides an electronic device, comprising:

[0137] processor;

[0138] Memory used to store processor-executable instructions;

[0139] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0140] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0141] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0142] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for controlling the circulating cooling of an immersion liquid-cooled energy storage device, characterized in that, include: Collect temperature data of energy storage battery modules immersed in liquid cooling medium, calculate the temperature change rate and temperature fluctuation value of monitoring points at adjacent time points based on the temperature data, establish a temperature distribution heat map, mark the monitoring points whose temperature change rate exceeds the preset change value in the temperature distribution heat map, divide the temperature fluctuation value of the marked monitoring points into zones according to the value, and output a temperature anomaly area distribution map. Based on the temperature fluctuation values ​​of each region in the temperature anomaly distribution map, the temperature transfer coefficient between adjacent regions is calculated. The heat transfer path is determined according to the numerical order of the temperature transfer coefficient, and the heat transfer path is divided into multiple control sub-regions according to the preset interval of the temperature transfer coefficient. Based on the spatial distribution characteristics of the heat transfer path, the main refrigeration circuit is activated in the control sub-region where the temperature transfer coefficient exceeds the preset temperature value. At the same time, the auxiliary refrigeration circuit of the adjacent control sub-region is pre-activated in the extension direction of the heat transfer path. The extension and diffusion of the heat transfer path are blocked by the coordinated control of the main refrigeration circuit and the auxiliary refrigeration circuit. Real-time temperature data of each monitoring point is collected and the temperature distribution heat map is updated. The steps of temperature anomaly area identification and cooling circuit control are repeated until the temperature distribution of the energy storage battery module meets the preset conditions. Based on the temperature fluctuation values ​​of each region in the temperature anomaly distribution map, the temperature transfer coefficient between adjacent regions is calculated. The heat transfer path is determined according to the numerical order of the temperature transfer coefficient, and the heat transfer path is divided into multiple control sub-regions according to the preset interval of the temperature transfer coefficient, including: Obtain temperature data from monitoring points within each abnormal region in the temperature anomaly distribution map, calculate the temperature change value at adjacent sampling times, determine the fluctuation amplitude, fluctuation period, and fluctuation direction of the temperature change, and combine them to form the temperature fluctuation value; For adjacent abnormal regions, the correlation coefficient of temperature fluctuation values ​​within a specified time window is calculated to obtain the degree of matching. The time difference corresponding to the maximum correlation coefficient is determined to determine the matching delay. Combined with the weighting factor, the temperature transfer coefficient is determined. The abnormal region with the largest temperature fluctuation value is selected as the starting region. The abnormal regions are connected along the direction of the maximum gradient of the temperature transfer coefficient to form a transfer path. The dividing point of the transfer path is determined according to the time change sequence of the temperature fluctuation value. The heat transfer paths are obtained by arranging them in descending order of the temperature transfer coefficient. Based on the preset range of temperature transfer coefficients between adjacent abnormal areas along the heat transfer path, the heat transfer path is divided into multiple control sub-regions. Specifically, the ratio of the contact boundary length of adjacent abnormal regions to the shape feature value is used as a weighting factor, the time difference when the correlation coefficient is at its maximum is determined by the moving time window as the matching delay, and the product of the matching degree, the matching delay, and the weighting factor is used as the temperature transfer coefficient.

2. The method according to claim 1, characterized in that, Temperature data of energy storage battery modules immersed in liquid cooling medium is collected. Based on the temperature data, the rate of temperature change and temperature fluctuation values ​​of monitoring points at adjacent time points are calculated, and a temperature distribution heat map is established, including: Temperature sensor arrays are set on the surface of the battery cells, the tab connection, the busbar and the edge of the energy storage battery module immersed in liquid cooling medium to acquire temperature data at each monitoring point. Wavelet transform is used to decompose the temperature data to obtain noise-reduced temperature data. Based on the noise-reduced temperature data, the temperature difference between two monitoring points at adjacent sampling times is calculated. The temperature difference is divided by the sampling time interval to obtain the temperature change rate. The sampling time interval is dynamically adjusted according to the temperature change trend. At the same time, the temperature fluctuation value of the monitoring point is calculated using an adaptive sliding variance algorithm to generate temperature feature data containing the temperature change rate and temperature fluctuation value. The grid density is determined using temperature feature data, and the temperature interpolation of the grid nodes is calculated using radial basis functions. The temperature values ​​of the grid nodes are converted into color codes to generate an initial heat map reflecting the temperature distribution. For regions in the initial heat map where the temperature gradient is greater than a preset gradient threshold, the grid is subdivided, and the temperature values ​​of the subdivided grids are recalculated to generate a temperature distribution heat map with local densification characteristics.

3. The method according to claim 1, characterized in that, In the temperature distribution heatmap, monitoring points where the rate of temperature change exceeds a preset value are marked. The temperature fluctuation values ​​of the marked monitoring points are then divided into zones according to their numerical values, and a temperature anomaly area distribution map is output, including: Based on the charging and discharging state of the energy storage battery module and the ambient temperature, a benchmark threshold for the rate of temperature change is set. Monitoring points that exceed the benchmark threshold in the temperature distribution heat map are marked as abnormal points. The temperature fluctuation values ​​of the abnormal points are extracted, and the temperature fluctuation values ​​are divided into multiple numerical intervals through adaptive interval division to establish a temperature feature distribution sequence of abnormal points. Anomalies in the temperature feature distribution sequence whose temperature fluctuation feature similarity is greater than a preset similarity threshold are aggregated into the same group. Taking each group of anomalies as the center, the monitoring points whose temperature fluctuation values ​​are in the same range and whose positions are adjacent are included in the corresponding anomaly group. Adjacent anomaly groups with common boundaries and whose temperature fluctuation values ​​differ from the preset fluctuation value are merged to generate a temperature anomaly area distribution map labeled with anomaly level.

4. The method according to claim 1, characterized in that, The heat transfer paths are determined by identifying the dividing points of the heat transfer path based on the time variation sequence of temperature fluctuation values, and then arranged in descending order of the temperature transfer coefficient. The heat transfer paths include: A sliding time window is used to analyze the time series of temperature fluctuation values. The statistical characteristics and the rate of change of temperature fluctuation values ​​within the window are calculated to obtain the time series feature vector. Based on the time-series feature vector, the energy loss rate at adjacent time points is calculated to obtain the instantaneous transfer efficiency, the cumulative transfer efficiency within a fixed window is calculated, and the average transfer efficiency over the entire period is calculated. The instantaneous transfer efficiency, cumulative transfer efficiency, and average transfer efficiency are combined to obtain the transfer efficiency index. Using the aforementioned transfer efficiency index, the degree of abrupt change in transfer efficiency, the inflection point amplitude, and the depth of local troughs are calculated respectively. The locations with the greatest degree of abrupt change, the locations with the greatest inflection point amplitude, and the locations with the deepest troughs are combined and scored to determine the dividing points of the heat transfer path. The heat transfer path is segmented based on the dividing points, the weighted average of the temperature transfer coefficient of each segment is calculated, and the comprehensive score of the transfer path is calculated by combining the fluctuation characteristics of the temperature transfer coefficient and the distribution of the dividing points. Based on the comprehensive score, the heat transfer paths are sorted from largest to smallest according to their temperature transfer coefficient values ​​to obtain the heat transfer paths.

5. The method according to claim 1, characterized in that, Based on the spatial distribution characteristics of the heat transfer path, the main refrigeration circuit is activated in the controlled sub-region where the temperature transfer coefficient exceeds the preset temperature value. Simultaneously, the auxiliary refrigeration circuit in the adjacent controlled sub-region is pre-activated in the direction of the heat transfer path's extension. The coordinated control of the main and auxiliary refrigeration circuits blocks the extension and diffusion of the heat transfer path, including: Collect temperature field data within the controlled sub-region, calculate the spatial distribution and intensity variation of the temperature field data, and obtain the distribution characteristics of the heat transfer path; Calculate the temperature transfer coefficient of each control sub-region based on the temperature field data. When the temperature transfer coefficient exceeds the preset temperature value, determine the over-limit region and set and activate the main refrigeration circuit according to the heat load. Based on the distribution characteristics of the heat transfer path, the adjacent control sub-regions of the over-limit area in the direction of the heat transfer path extension are determined. When the heat transfer rate reaches the preset start-up condition, the auxiliary cooling circuit is set and started. Collect temperature field data of the controlled sub-area, adjust the power of the main refrigeration circuit according to the temperature change rate of the controlled area of ​​the main refrigeration circuit, and adjust the power of the auxiliary refrigeration circuit according to the temperature change rate of the controlled area of ​​the auxiliary refrigeration circuit. The system calculates and controls the heat transfer rate of the sub-area in real time, switches auxiliary cooling circuits with heat transfer rates exceeding a preset threshold to main cooling circuits, and switches main cooling circuits with heat transfer rates below a preset threshold to standby mode, thus blocking the extension of the heat transfer path.

6. A circulating cooling control system for an immersion liquid-cooled energy storage device, used to implement the method of any one of claims 1-5, characterized in that, include: The first unit is used to collect temperature data of the energy storage battery module immersed in the liquid cooling medium, calculate the temperature change rate and temperature fluctuation value of the monitoring points at adjacent time points based on the temperature data, establish a temperature distribution heat map, mark the monitoring points whose temperature change rate exceeds the preset change value in the temperature distribution heat map, divide the temperature fluctuation value of the marked monitoring points into zones according to the value, and output a temperature anomaly area distribution map. The second unit is used to calculate the temperature transfer coefficient between adjacent areas based on the temperature fluctuation values ​​of each area in the temperature anomaly distribution map, determine the heat transfer path according to the numerical order of the temperature transfer coefficient, and divide the heat transfer path into multiple control sub-areas according to the preset interval of the temperature transfer coefficient. The third unit is used to activate the main refrigeration circuit in the control sub-region where the temperature transfer coefficient exceeds the preset temperature value according to the spatial distribution characteristics of the heat transfer path, and at the same time pre-activate the auxiliary refrigeration circuit in the adjacent control sub-region in the extension direction of the heat transfer path, so as to block the extension and diffusion of the heat transfer path through the coordinated control of the main refrigeration circuit and the auxiliary refrigeration circuit. The fourth unit is used to collect temperature data from each monitoring point in real time and update the temperature distribution heat map. It repeatedly executes the steps of identifying abnormal temperature areas and controlling the cooling circuit until the temperature distribution of the energy storage battery module meets the preset conditions.

7. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 5.

Images