Intelligent liquid-cooling self-adaptive heat dissipation control method for sodium-ion battery energy storage cabinet

Abstract

The invention discloses an intelligent liquid cooling self-adaptive heat dissipation control method for a sodium ion battery energy storage cabinet, and relates to the technical field of battery energy storage heat management, and the method comprises the steps: periodically collecting thermal and electrochemical detection signal data of multiple positions in a battery module through a sensor array arranged in the energy storage cabinet; an instantaneous three-dimensional temperature field distribution model is constructed according to thermal data, and the optimal heat dissipation demand weight corresponding to each battery monomer is calculated through a self-adaptive coupling algorithm in combination with electrochemical data. And dynamically generating a distribution instruction of cooling medium flow of each branch pipeline of the liquid cooling system according to the weight, adjusting the opening degree of a valve and the rotating speed of a main circulating pump in real time, and synchronously collecting a feedback signal. According to the method, accurate heat dissipation demand judgment from overall three-dimensional thermal field sensing to the single electrochemical state is achieved, cooling resources can be actively distributed in a differentiated mode, and therefore the heat dissipation efficiency and the temperature uniformity of the battery module are improved.

Description

Technical Field

[0001] This invention belongs to the field of battery energy storage thermal management technology, specifically a smart liquid-cooled adaptive heat dissipation control method for sodium-ion battery energy storage cabinets. Background Technology

[0002] Currently, heat dissipation control in sodium-ion battery energy storage cabinets largely relies on temperature sensors placed on the outside of the battery module or at a few key points. This conventional technology monitors the temperature at a limited number of points; when the temperature at a certain point exceeds a set threshold, it triggers the liquid cooling system to cool the entire module or a specific area with uniform intensity. This control strategy based on point or localized temperature feedback is the mainstream thermal management approach.

[0003] This conventional method has its limitations. Heat generation within the battery module is uneven, with complex spatial temperature gradients. Limited monitoring points cannot accurately reflect the three-dimensional thermal state inside the battery, especially in the core area, and may miss localized hotspots. Furthermore, the electrochemical state has a decisive influence on heat generation and distribution; relying solely on temperature feedback cannot predict the risk of thermal runaway in advance, nor can it differentiate the varying heat dissipation needs of different cells due to differences in their electrochemical states. This results in lagging and inefficient heat dissipation control, and an unreasonable allocation of cooling resources.

[0004] A technology is needed that can accurately sense the three-dimensional thermal distribution inside the battery module and integrate electrochemical state information for forward-looking and refined heat dissipation control. It needs to address the problems of vague understanding of the internal thermal state and the disconnect between heat dissipation response and actual battery requirements in existing technologies, from both the overall thermal field perception and the judgment of differentiated needs, in order to achieve precise, efficient, and adaptive thermal management. Summary of the Invention

[0005] This invention aims to solve at least one of the technical problems existing in the prior art; Therefore, this invention proposes an intelligent liquid-cooled adaptive heat dissipation control method for sodium-ion battery energy storage cabinets, comprising: The thermal and electrochemical detection signal data are periodically collected from multiple locations within the battery module by a sensor array arranged inside the energy storage cabinet. A three-dimensional instantaneous temperature field distribution model of the battery module is constructed based on the thermal detection signal data; Based on the electrochemical detection signal data and the instantaneous three-dimensional temperature field distribution model, the optimal heat dissipation requirement weight for each battery cell is calculated using an adaptive coupling algorithm. Based on the optimal heat dissipation requirement weight, a dynamic allocation command for the cooling medium flow rate in the branch pipeline of the liquid cooling system is generated; The valve opening degree of each branch of the liquid cooling system and the speed of the main circulation pump are adjusted in real time according to the dynamic allocation command, and the adjusted thermal feedback signal data are collected simultaneously.

[0006] Furthermore, the step of constructing an instantaneous three-dimensional temperature field distribution model of the battery module based on the thermal detection signal data specifically includes: Read the real-time measurements of all temperature sensors and their three-dimensional spatial coordinates inside the battery module from the sensor array; For each sampling moment, the real-time measured value and the corresponding three-dimensional spatial coordinates are used to form a spatial temperature point set; The Kriging space interpolation method is used to perform interpolation calculations on the spatial temperature point set in the entire three-dimensional geometric space of the battery module. Based on the interpolation results, a temperature distribution matrix with a regular voxel grid structure covering the entire volume of the battery module is generated. The temperature distribution matrix is ​​spatially mapped and aligned with the structural CAD model of the battery module, and the voxel regions with temperatures higher than a preset threshold are marked to form the instantaneous three-dimensional temperature field distribution model.

[0007] Furthermore, based on the electrochemical detection signal data and the instantaneous three-dimensional temperature field distribution model, the optimal heat dissipation requirement weight for each battery cell is calculated using an adaptive coupling algorithm, specifically as follows: The real-time operating current, operating voltage, and surface temperature change rate of each battery cell are extracted from the electrochemical detection signal data. Based on the real-time operating current and operating voltage, calculate the heat generation power of each battery cell within the current sampling interval; The heat generation power is coupled with the average temperature value of the corresponding battery cell location extracted from the instantaneous three-dimensional temperature field distribution model to obtain the heat load index of the battery cell. Historical health status assessment data of individual battery cells are introduced, which are calculated based on the long-term cycle capacity decay rate and internal resistance growth trend of individual battery cells. A weight calculation rule base based on fuzzy reasoning is established, with the heat load index, the surface temperature change rate, and the historical health status assessment data as input variables; The adaptive coupling algorithm executes the fuzzy inference process in the weight calculation rule base and outputs a continuous value between zero and one, which is the optimal heat dissipation requirement weight.

[0008] Furthermore, the step of generating dynamic allocation instructions for the cooling medium flow rate in the branch pipes of the liquid cooling system based on the optimal heat dissipation demand weight specifically includes: Obtain the topology diagram of the liquid cooling system, which defines the correspondence between the main circulation pipeline, each branch pipeline and the battery cell cluster; Read the total flow rate and inlet temperature of the cooling medium in the main circulation pipeline of the current liquid cooling system; Based on the topology diagram, the optimal heat dissipation requirements of all battery cells served by the same branch pipeline are weighted and summed to obtain the comprehensive heat dissipation requirement value of the branch pipeline. Based on the sum of the total heat dissipation requirements of all branch pipes, calculate the theoretical flow distribution ratio of each branch pipe; Calculate the target flow rate value for each branch pipe based on the theoretical flow rate distribution ratio and the total flow rate of the cooling medium; By combining the inlet temperature and the historical heat dissipation efficiency coefficient of the branch pipeline, the target flow rate value is finely adjusted and compensated to generate the dynamic allocation instruction containing the final target flow rate value of each branch pipeline.

[0009] Furthermore, the step of adjusting the valve openings of each branch of the liquid cooling system and the rotational speed of the main circulation pump in real time according to the dynamic allocation command specifically includes: The dynamic allocation command is sent to the local controller of the liquid cooling system, the local controller including a flow controller and a pump speed controller; The flow controller queries the flow characteristic curve of the regulating valve in the branch pipeline based on the final target flow value of the branch pipeline, and calculates the valve opening setting value required to achieve the target flow. The flow controller sends position control signals to the electric regulating valves of each branch, driving the valve core to move towards the valve opening set value; Meanwhile, the pump speed controller calculates the main circulation pump speed setpoint required to meet the total flow demand based on the sum of the final target flow values ​​of all branch pipelines and the system pipeline resistance characteristic curve. The pump speed controller sends a speed control signal to the frequency converter driver of the main circulation pump to adjust the motor speed of the main circulation pump to the set speed value.

[0010] Furthermore, the synchronous collection of adjusted thermal feedback signal data, and the closed-loop evaluation and parameter correction of the control effect based on this data, specifically includes: After adjusting the valve opening and pump speed, wait for a preset thermal equilibrium time window; After the thermal equilibrium time window ends, a new round of thermal detection signal data is collected again through the sensor array; Based on the new round of thermal detection signal data, an adjusted instantaneous three-dimensional temperature field distribution model of the battery module is constructed. The two instantaneous three-dimensional temperature field distribution models before and after adjustment are compared differentially to calculate the temperature change in each cell region. The number and spatial distribution of battery cells whose temperature drop has not reached the expected target are counted. If the number of battery cells exceeds a preset threshold, the current control effect is determined to be poor. The electrochemical detection signal data, optimal heat dissipation demand weight, and actual allocated cooling flow rate of the battery cells corresponding to the poor control effect are extracted to form a correction analysis dataset. Based on the corrected analysis dataset, the key coupling parameters in the adaptive coupling algorithm are adjusted using a nonlinear regression method for weight calculation in the next cycle.

[0011] Furthermore, the waiting period is a preset thermal equilibrium time window, the duration of which is dynamically determined in the following manner: Based on the flow rate change of each branch caused by the dynamic allocation command, calculate the theoretical hydraulic balance time required for the cooling medium distribution of the entire liquid cooling system to reach stability. Based on the expected change in the overall average temperature of the battery module before and after adjustment, and combined with the specific heat capacity and mass of the battery materials, the theoretical thermal inertia time required for the overall temperature response of the battery module is estimated. Compare the theoretical hydraulic equilibrium time with the theoretical thermal inertia time. If the theoretical hydraulic equilibrium time is longer than the theoretical thermal inertia time, then the theoretical hydraulic equilibrium time is used as the basic waiting time; otherwise, the theoretical thermal inertia time is used as the basic waiting time. Considering the current average temperature level of the battery module, a temperature-related correction coefficient is obtained from a pre-set empirical lookup table; Multiply the base waiting time by the correction factor to obtain the final preset thermal equilibrium time window.

[0012] Furthermore, the process of coupling the heat generation power with the average temperature value of the corresponding battery cell location extracted from the instantaneous three-dimensional temperature field distribution model to obtain the heat load index of the battery cell is specifically as follows: The theoretical thermal conductivity is calculated based on the material thermal conductivity and geometric dimensions of the battery cell. Based on the theoretical thermal conductivity and the heat generation power, the theoretical temperature difference between the core and surface of the battery cell under the current heat generation state is calculated using a one-dimensional steady-state heat conduction formula. The average temperature value corresponding to the location of the battery cell is extracted from the instantaneous three-dimensional temperature field distribution model and used as the measured average surface temperature of the battery cell. Calculate the measured temperature difference between the measured average surface temperature and the preset ambient reference temperature; The theoretical temperature difference is compared with the measured temperature difference, and the ratio is calculated as a matching factor characterizing the degree of matching between heat dissipation capacity and heat generation intensity. A nonlinear weighting function is established with the heat generation power as the main variable and the matching degree factor as the correction coefficient. The heat generation power and the matching factor are input into the nonlinear weighting function for calculation, and a standardized dimensionless value is output, which is the heat load index of the battery cell.

[0013] Furthermore, the introduction of historical health status assessment data for individual battery cells, which is calculated based on the long-term cycle capacity decay rate and internal resistance growth trend of the individual battery cells, specifically includes: The nominal capacity, current measured capacity, nominal internal resistance, and current measured internal resistance of each battery cell are periodically obtained from the historical database of the battery management system. Calculate the cumulative capacity decay rate and internal resistance growth rate of each battery cell since it has been put into operation; A time series smoothing method is used to denoise the cumulative capacity decay rate and the internal resistance growth rate to obtain their long-term trend lines. The slope values ​​of the long-term trend lines are mapped to the health scores for capacity decay and internal resistance growth, respectively. According to the predefined weighted fusion rules, the capacity decay health score and the internal resistance growth health score are merged into a comprehensive health status score, which is the historical health status assessment data used for weight calculation.

[0014] Furthermore, the process of periodically collecting thermal and electrochemical detection signal data from multiple locations within the battery module using a sensor array arranged within the energy storage cabinet also includes a self-test and calibration step for the sensor array. Send a self-test command to the sensor array and read the raw output signals of each temperature sensor and voltage / current sensor in sequence; The temperature sensor is placed in a constant temperature reference source, and its reading is compared with the standard value of the reference source to calculate the zero drift and gain error of each temperature sensor. When the battery is in a static, no-load state, the outputs of all voltage and current sensors are read and compared with the readings of a high-precision standard meter to calculate the measurement deviation. If the zero drift, gain error, or measurement deviation exceeds the allowable range, a calibration coefficient is generated for the corresponding sensor, and the calibration coefficient is applied to compensate the original signal in real time during subsequent data acquisition. The self-test and calibration results are recorded in the system log. If a sensor malfunction is detected, an early warning is issued and the redundant sensor switching process is initiated.

[0015] Compared with the prior art, the beneficial effects of the present invention are: By periodically collecting dense thermal signals from a sensor array deployed at multiple locations within the energy storage cabinet, an instantaneous three-dimensional temperature field distribution model of the battery module is constructed. This technology represents a leap from discrete-point monitoring of the battery's internal thermal state to continuous spatial field sensing. This allows for precise and intuitive location of the actual temperature at any point within the module, particularly identifying hidden hotspots and spatial temperature gradients that are difficult to detect using conventional methods. It provides a comprehensive, three-dimensional, and realistic thermal state map for heat dissipation control, avoiding localized overheating or overcooling caused by temperature measurement blind spots.

[0016] Based on electrochemical detection signal data and the aforementioned instantaneous three-dimensional temperature field distribution model, an adaptive coupling algorithm dynamically calculates the optimal heat dissipation demand weight for each battery cell. This technology performs real-time correlation analysis between electrochemical parameters characterizing the battery's operating state and health and the physical thermal field. This transforms the heat dissipation decision-making basis from a single temperature parameter to a multi-dimensional coupled variable of temperature and electrochemical state. The system can predict the intensity of heat dissipation demand for each cell based on its real-time combined "thermal-electric" state, thereby generating differentiated cooling commands. This ensures that the flow distribution of the cooling medium precisely matches the actual, dynamic heat generation demand of each cell, improving heat dissipation efficiency and uniformity. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the steps of the intelligent liquid-cooled adaptive heat dissipation control method for the sodium-ion battery energy storage cabinet described in this invention. Figure 2 A flowchart of the adaptive coupling algorithm for calculating the optimal heat dissipation requirement weights; Figure 3 A professional analysis diagram of a sodium-ion battery liquid cooling system; Figure 4 A comparative analysis chart of multiple indicators of the thermal properties of sodium-ion battery modules; Figure 5 This is an analysis diagram of the sensor redundancy switching effect in a sodium-ion battery energy storage cabinet. Detailed Implementation

[0018] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. 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.

[0019] See Figure 1A sensor array arranged inside the energy storage cabinet periodically collects thermal and electrochemical detection signal data from multiple locations within the battery module. Based on the thermal detection signal data, an instantaneous three-dimensional temperature field distribution model of the battery module is constructed. Based on the electrochemical detection signal data and the instantaneous three-dimensional temperature field distribution model, an adaptive coupling algorithm is used to calculate the optimal heat dissipation requirement weight for each battery cell. According to the optimal heat dissipation requirement weight, a dynamic allocation command for the cooling medium flow rate in the branch pipes of the liquid cooling system is generated. According to the dynamic allocation command, the valve opening degree of each branch of the liquid cooling system and the speed of the main circulation pump are adjusted in real time, and the adjusted thermal feedback signal data are collected simultaneously.

[0020] In one embodiment of the present invention, real-time measurements of all temperature sensors and their three-dimensional spatial coordinates within the battery module are read from a sensor array arranged within the energy storage cabinet. For each sampling moment, the real-time measurements and corresponding three-dimensional spatial coordinates are used to construct a spatial temperature point set. The spatial temperature point set is stored in array form, containing the location information and temperature reading of each temperature sensor. In some embodiments, the data structure of the spatial temperature point set is designed as a list of tuples containing three-dimensional coordinates and temperature values, facilitating direct access for subsequent interpolation calculations. The update frequency of the spatial temperature point set is consistent with the sampling period of the sensor array. The Kriging spatial interpolation method is used to interpolate the spatial temperature point set within the entire three-dimensional geometric space of the battery module. The Kriging spatial interpolation method estimates the temperature values ​​of unsampled points and generates a continuous temperature field based on spatial statistical theory. It can be understood that the Kriging spatial interpolation method requires the construction of a variogram model to characterize the spatial autocorrelation of the temperature data. The interpolation calculation process is implemented through the following formula:

[0021] in: Indicates the point to be estimated. Predicted temperature value at the location, Represents a known point Temperature measurement at the location, Indicates assigning a given point The weighting coefficients, The weighting factor represents the number of known temperature measurements used in the interpolation. The Kriging equations are solved to ensure unbiased predictions with minimal variance. A temperature distribution matrix with a regular voxel grid structure covering the entire volume of the battery module is generated based on the interpolation results. The dimension of the temperature distribution matrix is ​​consistent with the number of voxels in the three-dimensional geometric space of the battery module, and each matrix element stores the temperature value of the corresponding voxel. Optionally, the size of the voxel grid is configured according to the actual physical size of the battery module and the distribution density of temperature sensors; a smaller voxel size can improve the temperature field resolution but increases the computational load. The temperature distribution matrix is ​​spatially mapped and aligned with the structural CAD model of the battery module, marking voxel regions with temperatures higher than a preset threshold to form an instantaneous three-dimensional temperature field distribution model. The spatial mapping and alignment process converts the voxel index coordinates of the temperature distribution matrix into the world coordinate system coordinates of the structural CAD model through coordinate transformation. In some embodiments, the preset threshold is set to a fixed value based on the thermal safety boundary of the sodium-ion battery material or dynamically adjusted based on the battery's operating state. When marking high-temperature voxel regions, their position indices are stored in an independent data layer in the three-dimensional model.

[0022] It is understandable that the instantaneous three-dimensional temperature field distribution model exists in the control system as a multi-dimensional array, supporting visualization rendering for real-time monitoring. Model updates are synchronized with sensor array sampling to ensure the timeliness of temperature field data. In specific implementation, after generating the temperature distribution matrix, a data validity check is performed to remove abnormal interpolation results caused by sensor malfunctions. Invalid data points are replaced with the temperature values ​​of adjacent valid voxels. Optionally, parameters of the Kriging space interpolation method, such as the variogram model type and search radius, are pre-calibrated based on the thermal conductivity characteristics of the battery module to improve the interpolation accuracy under complex thermal boundary conditions.

[0023] In one embodiment of the present invention, see [reference] Figure 2The system extracts the real-time operating current, operating voltage, and surface temperature change rate of each battery cell from electrochemical detection signal data. Based on the real-time operating current and operating voltage, it calculates the heat generation power of each battery cell within the current sampling interval. In some embodiments, the calculation of heat generation power considers the Joule heat generated by the ohmic internal resistance and polarization internal resistance of the battery cell; the value of heat generation power is obtained by multiplying the real-time operating current by the operating voltage drop. The theoretical thermal conductivity of the battery cell is calculated based on its material thermal conductivity and geometric dimensions. The theoretical thermal conductivity reflects the thermal conductivity of the battery cell material itself and the influence of the battery cell's shape and size on the heat conduction path. Based on the theoretical thermal conductivity and heat generation power, the theoretical temperature difference between the theoretical core and surface of the battery cell under the current heat generation state is calculated using a one-dimensional steady-state heat conduction formula. This one-dimensional steady-state heat conduction formula simplifies the internal heat conduction model of the battery cell. The average temperature value corresponding to the location of the battery cell is extracted from the instantaneous three-dimensional temperature field distribution model as the measured average surface temperature of the battery cell. The extraction process involves locating the voxel region occupied by the battery cell in the temperature distribution matrix and calculating the arithmetic mean of the temperature values ​​of all voxels within that region. The measured temperature difference between the actual average surface temperature and the preset ambient reference temperature is calculated. The ambient reference temperature is set as the inlet temperature of the cooling medium in the liquid cooling system or the average ambient temperature inside the energy storage cabinet. The theoretical temperature difference is compared with the measured temperature difference, and their ratio is calculated as a matching factor characterizing the degree of matching between heat dissipation capacity and heat generation intensity. A matching factor greater than one indicates that the heat dissipation capacity is better than the current heat generation demand, and a matching factor less than one indicates that the heat dissipation capacity is insufficient. A nonlinear weighting function is established with heat generation power as the main variable and the matching factor as the correction coefficient. The design of the nonlinear weighting function allows battery cells with the same heat generation power but different matching factors to obtain different heat load indices. The heat generation power and the matching factor are input into the nonlinear weighting function for calculation, and a standardized dimensionless value is output, which is the heat load index of the battery cell. The form of the nonlinear weighting function is as follows:

[0024] in: This indicates the heat load index of a single battery cell. This represents the normalized heat generation power of a single battery cell. Represents the matching degree factor. This represents a natural constant. The nominal capacity, current measured capacity, nominal internal resistance, and current measured internal resistance of each battery cell are periodically retrieved from the historical database of the battery management system. The historical database stores the operating parameter data of each battery cell in time-series format. The cumulative capacity decay rate and internal resistance growth rate of each battery cell since it was put into operation are calculated. The cumulative capacity decay rate is obtained by the ratio of the difference between the nominal capacity and the current measured capacity to the nominal capacity, and the internal resistance growth rate is obtained by the ratio of the difference between the current measured internal resistance and the nominal internal resistance to the nominal internal resistance. A time-series smoothing method is used to denoise the cumulative capacity decay rate and internal resistance growth rate to obtain their long-term trend lines. The time-series smoothing method uses an exponentially weighted moving average to reduce the impact of short-term fluctuations. The slope values ​​of the long-term trend lines are mapped to capacity decay health scores and internal resistance growth health scores, respectively. The mapping process uses a linear or piecewise linear function to transform the slope of the trend line to a scoring range of zero to one. According to the predefined weighted fusion rules, the capacity decay health score and the internal resistance growth health score are merged into a comprehensive health status score. The comprehensive health status score is the historical health status assessment data used for weight calculation. The weighted fusion rules assign different weight coefficients to the capacity decay health score and the internal resistance growth health score.

[0025] It is understandable that a lower overall health status score indicates a worse historical health status of the battery cell, which should be given priority in heat dissipation resource allocation. A weight calculation rule base based on fuzzy inference is established. This rule base contains multiple fuzzy rules expressed in if-law form to define the logical relationship between input variables and output weights. Heat load index, surface temperature change rate, and historical health status assessment data are used as input variables. Before entering the fuzzy inference process, the input variables need to be fuzzified and converted into membership degrees of the corresponding fuzzy sets. The fuzzy inference process in the weight calculation rule base is executed through an adaptive coupling algorithm. The fuzzy inference process adopts the Mamdani inference method and uses the centroid method for defuzzification. A continuous value between zero and one is output; this continuous value is the optimal heat dissipation demand weight. A higher optimal heat dissipation demand weight indicates that the corresponding battery cell should receive more cooling medium flow allocation in the next control cycle. Optionally, the fuzzy rules in the weight calculation rule base can be adjusted and expanded according to the thermal characteristics of different sodium-ion battery models. In some embodiments, the surface temperature change rate is obtained by calculating the ratio of the difference between the surface temperature of the battery cell at the current sampling time and the previous sampling time to the sampling time interval. A positive surface temperature change rate indicates that the temperature is rising.

[0026] In one embodiment of the present invention, a topology diagram of the liquid cooling system is obtained. The topology diagram defines the correspondence between the main circulation pipeline, each branch pipeline, and the battery cell clusters. The topology diagram is stored in the control system in the form of network node connections, clearly indicating the subordinate relationship between each branch pipeline and a specific battery cell cluster. The total flow rate and inlet temperature of the cooling medium in the current main circulation pipeline of the liquid cooling system are read. The total flow rate of the cooling medium is measured by an electromagnetic flowmeter on the main circulation pipeline, and the inlet temperature is measured by a platinum resistance temperature sensor installed at the inlet of the main circulation pipeline. In some embodiments, the construction of the topology diagram is based on a digital mapping of the actual physical layout of the liquid cooling system. Each branch pipeline node is associated with a list of battery cell cluster identifiers. The reading of the total flow rate and inlet temperature of the cooling medium is performed at a fixed sampling frequency and filtered to eliminate instantaneous fluctuations.

[0027] Based on the topology diagram, the optimal heat dissipation demand weights of all battery cells served by the same branch pipeline are weighted and summed to obtain the comprehensive heat dissipation demand value of the branch pipeline. The weighted summation operation iterates through the list of battery cell cluster identifiers associated with the branch pipeline and accumulates the optimal heat dissipation demand weight value for each battery cell in the list. It can be understood that the optimal heat dissipation demand weight, derived from the adaptive coupling algorithm, is a continuous value between zero and one. The comprehensive heat dissipation demand value reflects the urgency of the overall heat dissipation demand of the battery cell cluster served by the branch pipeline. The theoretical flow allocation ratio for each branch pipeline is calculated based on the sum of the comprehensive heat dissipation demand values ​​of all branch pipelines. The theoretical flow allocation ratio is calculated using the following formula:

[0028] in: Indicates branch pipeline The theoretical flow allocation ratio, Indicates branch pipeline The overall heat dissipation requirements, This represents the sum of the total heat dissipation requirements of all branch pipes. This indicates the total number of branch pipes in the liquid cooling system. The target flow rate for each branch pipe is calculated based on the theoretical flow distribution ratio and the total flow rate of the cooling medium. The target flow rate is then determined using the theoretical flow distribution ratio. Total flow rate of cooling medium The product is obtained as follows The target flow rate is fine-tuned by combining the inlet temperature and the historical heat dissipation efficiency coefficient of the branch pipeline. The fine-tuning process obtains the temperature correction coefficient by querying the preset temperature-flow correction curve based on the inlet temperature value, and obtains the efficiency correction coefficient based on the historical heat dissipation efficiency coefficient of the branch pipeline.

[0029] Optionally, the historical heat dissipation efficiency coefficient is calculated based on the ratio of the actual cooling effect to the theoretical cooling effect of the branch pipeline in past operating cycles and is updated over time. The efficiency correction coefficient is obtained by looking up a table. A dynamic allocation instruction containing the final target flow value of each branch pipeline is generated. The data structure of the dynamic allocation instruction includes the branch pipeline number and the corresponding final target flow value field. The final target flow value is obtained by multiplying the target flow value by the temperature correction coefficient and then by the efficiency correction coefficient.

[0030] In some embodiments, when calculating the comprehensive heat dissipation demand value using a weighted summation method, the optimal heat dissipation demand weights for individual battery cells are not additionally weighted before summing. The calculation of the theoretical flow allocation ratio is performed in real time within each control cycle to ensure that the allocation ratio reflects the latest heat dissipation demand status. It can be understood that the total flow rate of the cooling medium... This is a real-time measured value that may vary with the main circulation pump speed; therefore, the latest total flow rate of the cooling medium is used when calculating the target flow rate. During fine-tuning compensation, a temperature correction factor is used to compensate for changes in heat exchange capacity caused by variations in the cooling medium inlet temperature. When the inlet temperature is high, the temperature correction factor is greater than one to increase flow distribution. Historical heat dissipation efficiency coefficients are used to compensate for differences in fluid resistance in different branch pipes due to length, number of bends, or minor blockages. Optionally, after the dynamic allocation command is generated, it is encapsulated into a data packet with a specific communication protocol and sent to the local controller of the liquid cooling system. The local controller parses the data packet and extracts the final target flow rate value for each branch pipe.

[0031] See Figure 3 This is a professional analysis chart of a sodium-ion battery liquid cooling system, showing a comparison between the standardized comprehensive heat dissipation requirements and the final target flow rates of eight branch pipes. The trends of the standardized comprehensive heat dissipation requirements and the final target flow rates for all branch pipes are highly consistent, indicating that the flow rate allocation strictly matches the urgency of the heat dissipation needs, consistent with the core logic of "dynamic flow rate allocation" in your project. Both data points for branch pipe 3 are the highest, indicating that the battery cluster served by this pipe has the most urgent heat dissipation needs, hence the allocation of the largest cooling flow rate. Both data points for branch pipe 7 are the lowest, corresponding to the battery cluster with the most moderate heat dissipation needs. The final flow rate of some pipes is slightly higher than the standardized requirements (such as branches 2 and 7), because fine-tuning compensation is applied using temperature correction coefficients and efficiency correction coefficients, reflecting the dynamic adaptability of the algorithm.

[0032] In one embodiment of the present invention, a dynamic allocation command is sent to the local controller of the liquid cooling system. The local controller includes a flow controller and a pump speed controller. The flow controller queries the flow characteristic curve of the regulating valve of the branch pipeline according to the final target flow value of the branch pipeline. The flow characteristic curve stores the correspondence between valve opening and flow in the form of a data table and calculates the valve opening set value required to achieve the target flow. The flow controller sends a position control signal to the electric regulating valve of each branch to drive the valve core to move towards the valve opening set value. The position control signal is an analog voltage signal or a digital pulse signal, depending on the interface type of the electric regulating valve. At the same time, the pump speed controller calculates the main circulation pump speed set value required to meet the total flow requirement based on the sum of the final target flow values ​​of all branch pipelines and the system pipeline resistance characteristic curve. The system pipeline resistance characteristic curve describes the relationship between the total system flow and the total pipeline resistance loss. The pump speed controller sends a speed control signal to the frequency converter driver of the main circulation pump to adjust the motor speed of the main circulation pump to the speed set value. In some embodiments, the flow characteristic curve is obtained from the valve factory calibration data and verified and calibrated on-site during system commissioning, while the system pipeline resistance characteristic curve is obtained from the hydraulic calculation model simulation of the liquid cooling system and corrected by measured data in the early stage of operation.

[0033] After adjusting the valve opening and pump speed, a preset thermal equilibrium time window is waited for. Based on the flow rate changes in each branch caused by the dynamic allocation command, the theoretical hydraulic equilibrium time required for the cooling medium distribution of the entire liquid cooling system to reach stability is calculated. This theoretical hydraulic equilibrium time is calculated based on the transient flow theory of fluid mechanics, considering the hydraulic inertia of the pipe network. The theoretical thermal inertia time required for the overall temperature response of the battery module is estimated based on the expected change in the overall average temperature of the battery module before and after adjustment, combined with the specific heat capacity and mass of the battery materials. The estimation of the theoretical thermal inertia time treats the battery module as a homogeneous body using a heat capacity and thermal resistance model. The theoretical hydraulic equilibrium time is compared with the theoretical thermal inertia time. If the theoretical hydraulic equilibrium time is longer than the theoretical thermal inertia time, the theoretical hydraulic equilibrium time is used as the base waiting time; otherwise, the theoretical thermal inertia time is used as the base waiting time. Considering the current average temperature level of the battery module, a temperature-related correction coefficient is obtained from a preset experience lookup table. This preset experience lookup table is based on historical operating data statistics and establishes a scaling ratio between different temperature ranges and the system thermal time constant. The final preset thermal equilibrium time window is obtained by multiplying the base waiting time by the correction factor. The thermal equilibrium time window is calculated using the following formula:

[0034] in: This indicates the duration of the final thermal equilibrium time window. Indicates the base waiting time. This represents the temperature-related correction factor obtained from the empirical lookup table. See Table 1, Empirical Lookup Table.

[0035] Table 1: Lookup Table for Temperature-Related Correction Factors

[0036] After the thermal equilibrium time window ends, a new round of thermal detection signal data is collected again using the sensor array. Based on this new round of thermal detection signal data, an adjusted instantaneous three-dimensional temperature field distribution model of the battery module is constructed. The two instantaneous three-dimensional temperature field distribution models before and after adjustment are compared differentially to calculate the temperature change of each battery cell region. The differential comparison generates a temperature change distribution matrix by subtracting the values ​​of the corresponding voxel grids. The number of battery cells whose temperature reduction does not reach the expected target and their spatial distribution are counted. The expected temperature reduction target is set as a fixed temperature reduction threshold or a percentage based on the initial temperature according to the heat dissipation requirements of the dynamic allocation command. If the number of battery cells exceeds the preset threshold, the current control effect is considered poor. The preset threshold is set according to a certain proportion of the total number of battery cells in the battery module. The electrochemical detection signal data, optimal heat dissipation requirement weight, and actual allocated cooling flow of the battery cells corresponding to the poor control effect are extracted to form a correction analysis dataset. The correction analysis dataset is stored in a structured table format, containing battery cell identifiers and related parameters. Based on the corrected analysis dataset, key coupling parameters in the adaptive coupling algorithm are adjusted using a nonlinear regression method for weight calculation in the next cycle. The nonlinear regression method employs least squares to fit the relationship between the actual heat dissipation effect and the weights in the corrected analysis dataset to update the coupling parameters. Optionally, the theoretical hydraulic equilibrium time is calculated using the standard time constant output by the fluid network transient simulation software, and the theoretical thermal inertia time is estimated as the product of the total heat capacity and thermal resistance of the battery module. It can be understood that the dynamic determination of the thermal equilibrium time window ensures that the system is evaluated only after the temperature field has stabilized, avoiding misjudgments due to transient processes. The temperature-related correction coefficient compensates for the influence of changes in material thermophysical properties at different temperatures on the thermal response rate.

[0037] In some embodiments, the preset threshold is set to 5% of the total number of battery cells. If the control effect is unsatisfactory, a parameter correction process is triggered; otherwise, the system maintains the current parameters. Optionally, after updating key coupling parameters using the nonlinear regression method, boundary checks are required to ensure the parameters remain within a reasonable range and avoid divergence. The corrected parameters take effect the next time the adaptive coupling algorithm is executed. It can be understood that when calculating temperature changes using differential comparison, regions with small temperature increases or changes are ignored, focusing only on regions with insufficient temperature decreases. The actual allocated cooling flow rate is obtained from the execution records of the flow controller. In specific implementations, the calculation of the theoretical hydraulic equilibrium time ignores the mutual hydraulic interference between branch pipes and simplifies using an independent pipe model. The estimation of the theoretical thermal inertia time ignores the temperature difference between battery cells and assumes uniform temperature change in the module.

[0038] See Figure 4 This is a multi-indicator comparative analysis chart of the thermal characteristics of sodium-ion battery modules, visually displaying core parameters such as temperature, heat generation power, heat dissipation demand weight, and temperature change for eight battery modules. Modules with higher heat generation power (such as modules 4, 3, and 8) have higher corresponding heat dissipation demand weights, indicating that the flow allocation algorithm of the liquid cooling system can accurately match the heat generation intensity. The operating temperature of all modules is below the warning threshold, and the temperature change is negative, indicating that the current heat dissipation control strategy effectively ensures that the battery operates within a safe temperature range. Module 4 has a significantly higher heat generation power than other modules, and its thermal stability during long-term operation needs to be closely monitored. More redundant heat dissipation capacity could be allocated to it in subsequent optimizations. By comparing heat generation power and temperature change, the efficiency of the heat dissipation control strategy can be evaluated, providing data support for algorithm iteration.

[0039] In one embodiment of the invention, a self-test command is sent to the sensor array to sequentially read the raw output signals of each temperature sensor and voltage / current sensor. The self-test command is sent in digital instruction form via the system bus and triggers the sensors to perform a complete measurement and data reporting process. In some embodiments, the raw output signals include the resistance or voltage value of the temperature sensor and the analog voltage value of the voltage / current sensor. After reading, the raw output signals are converted from analog to digital and stored in a temporary buffer. The temperature sensors are placed in a constant-temperature reference source, and their readings are compared with the standard value of the reference source. The constant-temperature reference source is a calibration device with high precision and stable temperature control capabilities, and its standard value is transmitted by a higher-level standard. The zero drift and gain error of each temperature sensor are calculated. The zero drift is calculated by dividing the difference between the actual output value and the theoretical output value of the temperature sensor in the constant-temperature reference source by the sensor sensitivity. The gain error is calculated by comparing the slope of the fitted line of the output value of the temperature sensor at multiple known temperature points with the standard slope. When the battery is in a quiescent, no-load state, the outputs of all voltage and current sensors are compared with the readings of a high-precision standard meter. This quiescent, no-load state is controlled by the battery management system, which disconnects all loads and maintains this state for a period to ensure stable battery terminal voltage and internal resistance. The high-precision standard meter is connected to the measurement point via a calibration link. The measurement deviations of the voltage and current sensors are calculated. The voltage measurement deviation is calculated as the difference between the sensor output value and the high-precision standard meter reading. The current measurement deviation is calculated as the difference between the sensor output value and the high-precision standard meter reading when a small, constant test current is applied.

[0040] If zero drift, gain error, or measurement deviation exceeds the allowable range, a calibration coefficient is generated for the corresponding sensor. The allowable range is preset according to the accuracy specifications in the sensor's technical data sheet. The calibration coefficient is generated using the following formula:

[0041] in: This represents the compensated sensor output value. Indicates the original output signal. This indicates that the generated gain calibration coefficients are used to correct for gain errors. The generated offset calibration coefficient is used to correct zero drift or measurement deviation. In subsequent data acquisition, the calibration coefficient is applied to compensate the original signal in real time. This real-time compensation process is completed synchronously in the signal processing stage of the data acquisition module, performing compensation calculations on the original output signal of each sampling point. Self-test and calibration results are recorded in the system log, which is stored as a timestamp file containing the sensor identifier, pre-calibration data, calibration coefficient, post-calibration data, and a flag indicating whether the error is out of tolerance. If a sensor fault is detected, an early warning is issued and a redundant sensor switching process is initiated. Sensor faults are determined based on zero drift, gain error, or measurement deviation exceeding the allowable range and reaching a preset fault threshold. The early warning information is displayed through the human-machine interface and uploaded to the remote monitoring center.

[0042] Optionally, the redundant sensor switching process involves removing the faulty sensor from the active sensor list and activating a backup sensor configured at the same measurement point for inclusion in the data acquisition sequence. In some embodiments, the temperature setpoint of the isothermal reference source is selected as the midpoint of the normal operating temperature range of the battery module, such as 35 degrees Celsius, and the accuracy level of the high-precision standard is two orders of magnitude higher than that of the calibrated sensor. It is understood that the self-test command can be sent automatically at a preset cycle or upon system power-on initialization or upon receiving a manual command; the duration of the no-load rest state must ensure sufficient relaxation of the battery polarization voltage. The calibration coefficients are calculated using the least squares method to fit the relationship between the sensor output and the standard value under multiple sets of calibration data. The gain calibration coefficient and offset calibration coefficient are stored in the sensor's independent configuration storage area. While recording the self-test and calibration results in the system log, the sensor's health status identifier is updated. This health status identifier is used in subsequent data acquisition and processing to determine whether to adopt the sensor's data. It is understood that after the redundant sensor switching process is executed, the system needs to re-map the sensor array topology to ensure the continuity of temperature field construction and electrochemical data acquisition. The warning information for the faulty sensor includes its physical location identifier for maintenance and positioning.

[0043] See Figure 5 This is a professional analysis chart showing the effect of redundant switching of sensors in a sodium-ion battery energy storage cabinet. The core feature is the trend of data integrity changes before and after the redundant sensor switching. Before the switching, the problem was extremely fluctuating data integrity, dropping from approximately 98% initially to approximately 82% in cycle 3, and further to approximately 78% in cycle 6, far below the target value of 98%, indicating that faulty sensors had severely impacted data acquisition stability. After the switching, the data integrity remained consistently above 98% (maximum 100%), fully meeting the system objectives and verifying the effectiveness of the redundant sensor switching process. Redundant sensor switching is a key means of ensuring continuous data acquisition, enabling rapid restoration of system reliability in the event of a fault. This chart can serve as a verification basis for the redundant switching function, demonstrating its emergency support capability in sensor failure scenarios.

[0044] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A method for intelligent liquid-cooled adaptive heat dissipation control of a sodium-ion battery energy storage cabinet, characterized in that, Perform the following steps: The thermal and electrochemical detection signal data are periodically collected from multiple locations within the battery module by a sensor array arranged inside the energy storage cabinet. A three-dimensional instantaneous temperature field distribution model of the battery module is constructed based on the thermal detection signal data; Based on the electrochemical detection signal data and the instantaneous three-dimensional temperature field distribution model, the optimal heat dissipation requirement weight for each battery cell is calculated using an adaptive coupling algorithm. Based on the optimal heat dissipation requirement weight, a dynamic allocation command for the cooling medium flow rate in the branch pipeline of the liquid cooling system is generated; The valve opening degree of each branch of the liquid cooling system and the speed of the main circulation pump are adjusted in real time according to the dynamic allocation command, and the adjusted thermal feedback signal data are collected simultaneously.

2. The intelligent liquid-cooled adaptive heat dissipation control method for a sodium-ion battery energy storage cabinet according to claim 1, characterized in that, The construction of the instantaneous three-dimensional temperature field distribution model of the battery module based on the thermal detection signal data specifically includes: Read the real-time measurements of all temperature sensors and their three-dimensional spatial coordinates inside the battery module from the sensor array; For each sampling moment, the real-time measured value and the corresponding three-dimensional spatial coordinates are used to form a spatial temperature point set; The Kriging space interpolation method is used to perform interpolation calculations on the spatial temperature point set in the entire three-dimensional geometric space of the battery module. Based on the interpolation results, a temperature distribution matrix with a regular voxel grid structure covering the entire volume of the battery module is generated. The temperature distribution matrix is ​​spatially mapped and aligned with the structural CAD model of the battery module, and the voxel regions with temperatures higher than a preset threshold are marked to form the instantaneous three-dimensional temperature field distribution model.

3. The intelligent liquid-cooled adaptive heat dissipation control method for a sodium-ion battery energy storage cabinet according to claim 2, characterized in that, Based on the electrochemical detection signal data and the instantaneous three-dimensional temperature field distribution model, the optimal heat dissipation requirement weight for each battery cell is calculated using an adaptive coupling algorithm, specifically as follows: The real-time operating current, operating voltage, and surface temperature change rate of each battery cell are extracted from the electrochemical detection signal data. Based on the real-time operating current and operating voltage, calculate the heat generation power of each battery cell within the current sampling interval; The heat generation power is coupled with the average temperature value of the corresponding battery cell location extracted from the instantaneous three-dimensional temperature field distribution model to obtain the heat load index of the battery cell. Historical health status assessment data of individual battery cells are introduced, which are calculated based on the long-term cycle capacity decay rate and internal resistance growth trend of individual battery cells. A weight calculation rule base based on fuzzy reasoning is established, with the heat load index, the surface temperature change rate, and the historical health status assessment data as input variables; The adaptive coupling algorithm executes the fuzzy inference process in the weight calculation rule base and outputs a continuous value between zero and one, which is the optimal heat dissipation requirement weight.

4. The intelligent liquid-cooled adaptive heat dissipation control method for a sodium-ion battery energy storage cabinet according to claim 3, characterized in that, The step of generating dynamic allocation instructions for the cooling medium flow rate in the branch pipes of the liquid cooling system based on the optimal heat dissipation demand weight specifically includes: Obtain the topology diagram of the liquid cooling system, which defines the correspondence between the main circulation pipeline, each branch pipeline and the battery cell cluster; Read the total flow rate and inlet temperature of the cooling medium in the main circulation pipeline of the current liquid cooling system; Based on the topology diagram, the optimal heat dissipation requirements of all battery cells served by the same branch pipeline are weighted and summed to obtain the comprehensive heat dissipation requirement value of the branch pipeline. Based on the sum of the total heat dissipation requirements of all branch pipes, calculate the theoretical flow distribution ratio of each branch pipe; Calculate the target flow rate value for each branch pipe based on the theoretical flow rate distribution ratio and the total flow rate of the cooling medium; By combining the inlet temperature and the historical heat dissipation efficiency coefficient of the branch pipeline, the target flow rate value is finely adjusted and compensated to generate the dynamic allocation instruction containing the final target flow rate value of each branch pipeline.

5. The intelligent liquid-cooled adaptive heat dissipation control method for a sodium-ion battery energy storage cabinet according to claim 4, characterized in that, The real-time adjustment of valve openings and main circulation pump speed in each branch of the liquid cooling system according to the dynamic allocation command is specifically as follows: The dynamic allocation command is sent to the local controller of the liquid cooling system, the local controller including a flow controller and a pump speed controller; The flow controller queries the flow characteristic curve of the regulating valve in the branch pipeline based on the final target flow value of the branch pipeline, and calculates the valve opening setting value required to achieve the target flow. The flow controller sends position control signals to the electric regulating valves of each branch, driving the valve core to move towards the valve opening set value; Meanwhile, the pump speed controller calculates the main circulation pump speed setpoint required to meet the total flow demand based on the sum of the final target flow values ​​of all branch pipelines and the system pipeline resistance characteristic curve. The pump speed controller sends a speed control signal to the frequency converter driver of the main circulation pump to adjust the motor speed of the main circulation pump to the set speed value.

6. The intelligent liquid-cooled adaptive heat dissipation control method for a sodium-ion battery energy storage cabinet according to claim 5, characterized in that, The synchronous collection of adjusted thermal feedback signal data, and the subsequent closed-loop evaluation and parameter correction of the control effect, specifically include: After adjusting the valve opening and pump speed, wait for a preset thermal equilibrium time window; After the thermal equilibrium time window ends, a new round of thermal detection signal data is collected again through the sensor array; Based on the new round of thermal detection signal data, an adjusted instantaneous three-dimensional temperature field distribution model of the battery module is constructed. The two instantaneous three-dimensional temperature field distribution models before and after adjustment are compared differentially to calculate the temperature change in each cell region. The number and spatial distribution of battery cells whose temperature drop has not reached the expected target are counted. If the number of battery cells exceeds a preset threshold, the current control effect is determined to be poor. The electrochemical detection signal data, optimal heat dissipation demand weight, and actual allocated cooling flow rate of the battery cells corresponding to the poor control effect are extracted to form a correction analysis dataset. Based on the corrected analysis dataset, the key coupling parameters in the adaptive coupling algorithm are adjusted using a nonlinear regression method for weight calculation in the next cycle.

7. The intelligent liquid-cooled adaptive heat dissipation control method for a sodium-ion battery energy storage cabinet according to claim 6, characterized in that, The process involves waiting for a preset thermal equilibrium time window, the duration of which is dynamically determined in the following manner: Based on the flow rate change of each branch caused by the dynamic allocation command, calculate the theoretical hydraulic balance time required for the cooling medium distribution of the entire liquid cooling system to reach stability. Based on the expected change in the overall average temperature of the battery module before and after adjustment, and combined with the specific heat capacity and mass of the battery materials, the theoretical thermal inertia time required for the overall temperature response of the battery module is estimated. Compare the theoretical hydraulic equilibrium time with the theoretical thermal inertia time. If the theoretical hydraulic equilibrium time is longer than the theoretical thermal inertia time, then the theoretical hydraulic equilibrium time is used as the base waiting time. Otherwise, the theoretical thermal inertia time is used as the base waiting time; Considering the current average temperature level of the battery module, a temperature-related correction coefficient is obtained from a pre-set empirical lookup table; Multiply the base waiting time by the correction factor to obtain the final preset thermal equilibrium time window.

8. The intelligent liquid-cooled adaptive heat dissipation control method for sodium-ion battery energy storage cabinet according to claim 3, characterized in that, The process of coupling the heat generation power with the average temperature value of the corresponding battery cell location extracted from the instantaneous three-dimensional temperature field distribution model to obtain the heat load index of the battery cell is as follows: The theoretical thermal conductivity is calculated based on the material thermal conductivity and geometric dimensions of the battery cell. Based on the theoretical thermal conductivity and the heat generation power, the theoretical temperature difference between the core and surface of the battery cell under the current heat generation state is calculated using a one-dimensional steady-state heat conduction formula. The average temperature value corresponding to the location of the battery cell is extracted from the instantaneous three-dimensional temperature field distribution model and used as the measured average surface temperature of the battery cell. Calculate the measured temperature difference between the measured average surface temperature and the preset ambient reference temperature; The theoretical temperature difference is compared with the measured temperature difference, and the ratio is calculated as a matching factor characterizing the degree of matching between heat dissipation capacity and heat generation intensity. A nonlinear weighting function is established with the heat generation power as the main variable and the matching degree factor as the correction coefficient. The heat generation power and the matching factor are input into the nonlinear weighting function for calculation, and a standardized dimensionless value is output, which is the heat load index of the battery cell.

9. The intelligent liquid-cooled adaptive heat dissipation control method for a sodium-ion battery energy storage cabinet according to claim 3, characterized in that, The introduction of historical health status assessment data for individual battery cells, which is calculated based on the long-term cycle capacity decay rate and internal resistance growth trend of the individual battery cells, specifically includes: The nominal capacity, current measured capacity, nominal internal resistance, and current measured internal resistance of each battery cell are periodically obtained from the historical database of the battery management system. Calculate the cumulative capacity decay rate and internal resistance growth rate of each battery cell since it has been put into operation; A time series smoothing method is used to denoise the cumulative capacity decay rate and the internal resistance growth rate to obtain their long-term trend lines. The slope values ​​of the long-term trend lines are mapped to the health scores for capacity decay and internal resistance growth, respectively. According to the predefined weighted fusion rules, the capacity decay health score and the internal resistance growth health score are merged into a comprehensive health status score, which is the historical health status assessment data used for weight calculation.

10. The intelligent liquid-cooled adaptive heat dissipation control method for a sodium-ion battery energy storage cabinet according to claim 1, characterized in that, The process involves periodically collecting thermal and electrochemical detection signal data from multiple locations within the battery module using a sensor array arranged within the energy storage cabinet. Prior to this, a self-test and calibration step for the sensor array is also included. Send a self-test command to the sensor array and read the raw output signals of each temperature sensor and voltage / current sensor in sequence; The temperature sensor is placed in a constant temperature reference source, and its reading is compared with the standard value of the reference source to calculate the zero drift and gain error of each temperature sensor. When the battery is in a static, no-load state, the outputs of all voltage and current sensors are read and compared with the readings of a high-precision standard meter to calculate the measurement deviation. If the zero drift, gain error, or measurement deviation exceeds the allowable range, a calibration coefficient is generated for the corresponding sensor, and the calibration coefficient is applied to compensate the original signal in real time during subsequent data acquisition. The self-test and calibration results are recorded in the system log. If a sensor malfunction is detected, an early warning is issued and the redundant sensor switching process is initiated.

Images