Immersed battery pack heat management system assisted by heat pipe

Abstract

The invention discloses a heat pipe-assisted immersed battery pack heat management system, which belongs to the technical field of battery control, and comprises a sealed shell, a battery pack is installed in the sealed shell, a gap is arranged between two adjacent battery packs in the battery pack, a heat dissipation pipe is inserted in the gap, and the heat dissipation pipe is connected with the sealed shell. The lower half portion of the heat dissipation pipe is located in the sealing shell, the upper half portion of the heat dissipation pipe penetrates through the sealing shell and extends to the outside, and a fan is installed at the position, close to the heat dissipation pipe, of the outside of the sealing shell. The heat dissipation efficiency of the battery pack can be remarkably improved by combining the heat pipe, immersed cooling and an intelligent control strategy, the temperature gradient problem caused by uneven flowing of cooling liquid is avoided by combining the advantages of dynamic and static immersed cooling, the heat dissipation effect is enhanced through the heat pipe and the wet core structure, and the service life of the battery pack is prolonged. The uniformity of temperature distribution in the battery pack is ensured, and the service life of the battery is effectively prolonged.

Description

Technical Field

[0001] The present invention relates to the field of battery control technology, and more particularly to a heat pipe-assisted immersed battery pack thermal management system. Background Art

[0002] Power batteries generate significant amounts of heat during the charging and discharging process. If this heat cannot be dissipated promptly and effectively, the battery temperature will rise. When the battery temperature exceeds a certain range, it not only accelerates battery aging and shortens its service life, but also triggers a series of safety issues such as thermal runaway, fire, and even explosion. Furthermore, temperature differences between individual cells within a battery pack can affect battery consistency and overall performance, thereby reducing the reliability and range of electric vehicles.

[0003] In order to effectively dissipate heat from batteries, immersion cooling is generally used in existing technologies. Immersion cooling technology is divided into two categories: dynamic and static, each with different characteristics and existing problems: Dynamic immersion cooling achieves efficient heat dissipation through circulating coolant. The coolant circulates around and within the battery pack, quickly dissipating heat generated by the batteries and achieving high heat dissipation efficiency. However, due to the flow characteristics of the coolant, uneven flow can occur during actual operation, leading to significant temperature differences between different parts of the battery pack and the formation of temperature gradients. This temperature gradient can affect battery consistency and overall performance, reducing the reliability and stability of the thermal management system.

[0004] The static immersion cooling solution completely immerses the battery pack in coolant, relying on the coolant's natural convection and heat conduction to maintain uniform battery temperature. This approach effectively avoids the temperature gradients caused by uneven coolant flow in dynamic solutions, ensuring relatively consistent temperatures across the battery pack. However, the static solution has relatively low heat dissipation efficiency, especially when the battery is operating at high loads. Natural convection in the coolant can't quickly and effectively dissipate the large amounts of heat, which can easily cause the battery temperature to rise, affecting its performance and lifespan.

[0005] In order to solve the above contradictions and problems existing in the current power battery thermal management technology, the present invention proposes a heat pipe-assisted immersion battery pack thermal management system. Summary of the Invention

[0006] In response to the problems existing in the prior art, the purpose of the present invention is to provide a heat pipe-assisted immersion battery pack thermal management system, which can significantly improve the heat dissipation efficiency of the battery pack by combining heat pipes, immersion cooling and intelligent control strategies, and combine the advantages of dynamic and static immersion cooling to avoid the temperature gradient problem caused by uneven flow of coolant, and enhance the heat dissipation effect through the heat pipe and wet core structure, thereby ensuring the uniformity of temperature distribution in the battery pack and effectively extending the battery life.

[0007] To solve the above problems, the present invention adopts the following technical solutions.

[0008] A heat pipe-assisted submerged battery pack thermal management system includes a sealed housing, a battery pack mounted therein, a gap provided between two adjacent battery groups in the battery pack, a heat dissipation pipe inserted in the gap, a lower half of the heat dissipation pipe located within the sealed housing, an upper half of the heat dissipation pipe extending through the sealed housing to the exterior, a fan mounted on the exterior of the sealed housing near the heat dissipation pipe, a wet wick mounted on the upper surface of the sealed housing at an end of the heat dissipation pipe remote from the fan, a coolant contained within the sealed housing, and a water pump mounted therein for pumping coolant from the sealed housing into the wet wick, with the coolant then being discharged from the lower end of the wet wick and flowing back into the sealed housing. Also includes, The perception and assessment module collects battery temperature, coolant status, and environmental parameters in real time and generates a multidimensional feature vector. Based on the collected data, it calculates the battery thermal risk index, which is a weighted combination of the battery's maximum temperature, temperature rise rate, and state of charge. The predictive modeling module is used to input multidimensional feature vectors into the time series prediction model; predict the spatial distribution of battery temperature field in future time periods based on historical temperature data; and calibrate the prediction accuracy through the loss function embedded in the thermodynamic equation; The decision optimization module implements a hierarchical control strategy based on the predicted temperature field and thermal risk index; it uses an evolutionary algorithm to solve the energy-efficient optimal control instructions that meet the temperature constraints; The execution monitoring module is used to convert control instructions into actuator drive signals; monitor the execution status and actual battery temperature data in real time through a closed-loop controller; monitor execution deviations in real time and activate the fault tolerance mechanism when flow anomalies are detected; The learning evolution module updates system parameters based on the feedback execution status and actual temperature data through a reinforcement learning mechanism, with the objective function taking into account both temperature stability and energy efficiency; and regularly calibrates the time series prediction model and control parameters.

[0009] Furthermore, the perception assessment module further includes: Determine the data acquisition sensor configuration for the battery and cooling system, including installing corresponding sensor devices on the battery cells, coolant pipes, and the outside of the sealed housing to monitor battery cell temperature, coolant status parameters, environmental parameters, and obtain battery state of charge; Extract key data features based on the sensor configuration, including the maximum temperature of the battery cell, temperature rise rate, coolant temperature difference, flow fluctuation, temperature and humidity outside the sealed housing, and battery state of charge, and integrate them into a feature information group; According to the characteristic information group, different weights are assigned to the maximum temperature of the battery cell, the temperature rise rate and the battery state of charge, and a comprehensive calculation is made to obtain a thermal risk index that reflects the battery thermal risk.

[0010] Furthermore, the predictive modeling module also includes: Retrieve the historical temperature data set associated with the current feature information group, including the temperature change records of the battery cells and the temperature field distribution pattern, construct a continuous time series in chronological order and mark the time nodes; A time series prediction model based on time dependence is constructed. This model integrates the temperature correlation characteristics of battery cells and uses coolant status and environmental parameters as auxiliary variable inputs to learn the temperature field evolution pattern through historical temperature change trends. Input the time series into the prediction model, predict the temperature change trend of each battery cell and the spatial distribution of the battery pack temperature field in the future period based on the current feature information, and mark the time node of the prediction result; Based on the principles of thermodynamics, a verification standard for prediction results is established. By comparing the consistency of the prediction results with the laws of heat conduction, the model parameters are adjusted until the prediction results meet the thermodynamic constraints.

[0011] Furthermore, the decision optimization module also includes: Establish a thermal risk classification threshold system, including determining the thermal risk index threshold and temperature field prediction index threshold based on the battery safety temperature range and historical failure data; Compare the real-time thermal risk index and temperature field prediction results with the threshold system, and output low, medium, and high control level signals based on the degree of risk index exceeding the standard and the scope of the temperature field violation area; In response to the control level signal, the preset basic control rules are called to generate the corresponding level of cooling actuator parameter adjustment plan, including the linkage control strategy of the fan and water pump; Under the constraints of the parameter adjustment scheme, the optimal control instruction set that meets the temperature safety constraints and has the lowest total power consumption is screened through simulation and energy consumption evaluation of multiple groups of control instructions.

[0012] Furthermore, the monitoring module is executed, including: Convert the control instructions of the heat dissipation actuator into a driving electrical signal that matches its signal type, including converting the fan speed instruction into a pulse modulation signal and the water pump flow instruction into a voltage signal; Send a drive signal to the heat dissipation actuator to start operation. At the same time, the actual operating status of the actuator, battery temperature, and coolant parameters are collected in real time through sensors to form a closed-loop monitoring data set synchronized with the target parameters and time stamps. Calculate the deviation between the actual operating parameters of the actuator and the target parameters, and judge the operating abnormality based on the preset deviation threshold, including judging the flow deviation exceeding the limit as flow abnormality; In response to flow anomalies, the fault-tolerance mechanism is activated, increasing the water pump power to the preset safety upper limit and collaboratively enhancing the fan's heat dissipation capacity; if the anomaly persists, an alarm is triggered and the anomaly time and deviation data are recorded to ensure that the battery temperature meets thermal management safety constraints.

[0013] Furthermore, the learning evolution module also includes; Compare the deviation parameters between the actual operating state of the actuator and the optimal control instructions, and analyze the spatial distribution differences between the actual battery temperature data and the predicted temperature field; synchronously collect the system's real-time energy consumption data and temperature fluctuation characteristics; Based on the deviation parameters and fluctuation characteristics, a reinforcement learning reward rule is set with temperature stability and energy efficiency as the core optimization goals, and the reward value is dynamically allocated according to the degree of goal achievement.

[0014] Furthermore, the learning evolution module also includes: Deviation data and reward rules are input into the reinforcement learning mechanism. Through iterative optimization, the thermal risk judgment threshold, time series prediction model feature weight, and actuator response coefficient in the control strategy are adjusted to make the system converge to the reward maximization state. Periodically retrain the temperature field prediction model and calibrate the control parameters, dynamically adjust the model structure and threshold range based on the actual operating results, and deploy the optimization results to the next round of control processes.

[0015] Furthermore, the wet core is made of multiple groups of metal tubes that are connected horizontally and vertically and staggered with each other, and the interiors of the multiple groups of metal tubes are interconnected, and spaces for liquid flow are left between the staggered metal tubes.

[0016] Furthermore, the upper end of the wet core is fixedly connected to a diversion box, and the lower end thereof is fixedly connected to a junction box. Both the diversion box and the junction box are hollow in design, and the interiors of the two are interconnected through the metal tube on the wet core. A guide hole is provided on the upper surface of the sealed shell below the junction box, and the interior of the junction box is interconnected with the interior of the sealed shell through the guide hole. A guide pipe is fixedly connected to the liquid outlet of the water pump, and the end of the guide pipe away from the water pump passes through the sealed shell and is connected to the diversion box, and the interior of the guide pipe is interconnected with the interior of the diversion box.

[0017] Furthermore, the upper and lower ends of the heat dissipation pipe are sealed, and the interior of the heat dissipation pipe is filled with evaporative liquid. Metal fins are fixedly installed on the upper half of the heat dissipation pipe, and a protective cover is fixedly installed on the upper end of the sealed shell. The wet core, fan and the upper half of the heat dissipation pipe are all arranged in the protective cover.

[0018] Compared with the prior art, the present invention has the following beneficial effects: (1) The heat pipe-assisted immersed battery pack thermal management system proposed in this scheme significantly improves the heat dissipation efficiency of the battery pack by combining heat pipes, wet cores, immersion cooling and intelligent control strategies. It also combines the advantages of dynamic and static immersion cooling, avoiding the temperature gradient problem caused by uneven coolant flow, and enhancing the heat dissipation effect through the heat pipe and wet core structure, ensuring the uniformity of temperature distribution within the battery pack and effectively extending the battery life.

[0019] (2) This solution forms a closed-loop intelligent control system through built-in perception evaluation, predictive modeling, decision optimization and execution monitoring modules. By collecting battery temperature, coolant status and environmental parameters in real time, it can accurately calculate the battery thermal risk index and predict the battery temperature field distribution in the future period. The decision optimization module executes a hierarchical control strategy based on the prediction results to ensure that the most appropriate heat dissipation measures are taken under different thermal risk levels, which not only ensures battery safety but also improves energy efficiency.

[0020] (3) This solution introduces learning evolution to enable the system to have the ability of self-learning and continuous optimization. Through the reinforcement learning mechanism, the system can dynamically adjust the control strategy and model parameters according to the actual operating results, and continuously optimize the heat dissipation performance and energy efficiency. This adaptive capability enables the system to cope with various complex working conditions, ensuring that the battery pack always maintains the best thermal management state during long-term operation, greatly improving the reliability and stability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort.

[0022] Figure 1 This is an external view of the overall structure of the present invention; Figure 2 This is a diagram showing the structure of the upper surface components of the sealed housing of the present invention; Figure 3 It is a structural schematic diagram of the heat dissipation pipe of the present invention; Figure 4 This is a schematic diagram of the structure inside the sealed housing of the present invention; Figure 5 This is a structural diagram of the sealed housing of the present invention; Figure 6 Schematic diagram of the structure of the battery pack of the present invention; Figure 7 This is a schematic flow chart of a heat pipe-assisted immersed battery pack thermal management system of the present invention.

[0023] Description of the numbers in the figure: 1. Sealed shell; 2. Positioning slot; 3. Battery pack; 4. Heat dissipation pipe; 5. Metal fins; 6. Fan; 7. Water pump; 8. Flow guide tube; 9. Diverter box; 10. Wet core; 11. Junction box; 12. Liquid filling port; 13. Protective cover; 14. Flow guide hole; 15. Pressure plate. DETAILED DESCRIPTION

[0024] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention; it is obvious that the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0025] See also Figures 1 to 7 , a heat pipe assisted immersion battery pack thermal management system, comprising a sealed shell 1, wherein a battery pack 3 is installed inside the sealed shell 1, a gap is provided between two adjacent groups of batteries in the battery pack 3, a heat dissipation pipe 4 is inserted in the gap, the lower half of the heat dissipation pipe 4 is located inside the sealed shell 1, and the upper half thereof passes through the sealed shell 1 and extends to the outside, a fan 6 is installed on the outside of the sealed shell 1 near the heat dissipation pipe 4, a wet core 10 is installed on the upper surface of the sealed shell 1 at the end of the heat dissipation pipe 4 away from the fan 6, the wet core 10 is made of multiple groups of metal tubes connected horizontally and vertically and staggered with each other, and the multiple groups of metal tubes are interconnected, and space for liquid flow is left between the staggered metal tubes, the interior of the sealed shell 1 is filled with coolant, and a water pump 7 is also installed inside the sealed shell 1, the water pump 7 is used to draw coolant from the sealed shell 1 and inject it into the wet core 10, and the coolant is discharged from the lower end of the wet core 10 and flows back into the sealed shell 1; The upper end of the wet core 10 is fixedly connected to a diverter box 9, and the lower end thereof is fixedly connected to a junction box 11. The diverter box 9 and the junction box 11 are both hollow in design, and the interiors of the two are communicated with each other through the metal tube on the wet core 10. The upper surface of the sealed housing 1 is provided with a guide hole 14 below the junction box 11. The interior of the junction box 11 is communicated with the interior of the sealed housing 1 through the guide hole 14. The liquid outlet of the water pump 7 is fixedly connected to a guide pipe 8. The end of the guide pipe 8 away from the water pump 7 passes through the sealed housing 1 and is connected to the diverter box 9, and the interior of the guide pipe 8 is communicated with the interior of the diverter box 9; the upper and lower ends of the heat dissipation pipe 4 are sealed, and the interior of the heat dissipation pipe 4 is filled with evaporative liquid. The upper half of the heat dissipation pipe 4 is fixedly mounted with a metal fin 5. The upper end of the sealed housing 1 is fixedly mounted with a protective cover 13. The wet core 10, the fan 6 and the upper half of the heat dissipation pipe 4 are all arranged in the protective cover 13.

[0026] By adopting the above technical solution, the battery pack 3 generates heat when outputting current and charging. As the heat accumulates, the temperature of the battery pack 3 will become too high, affecting the operation of the battery pack 3. The heat of the battery pack 3 will be transferred to the coolant in the sealed housing 1. After the coolant heats up, it will heat the heat pipe 4. The heat pipe 4 can be made of copper or aluminum materials, which can perform heat conduction well. After the lower half of the heat pipe 4 is heated, the evaporative liquid therein is heated and vaporized. The vapor rises and transfers heat to the upper half of the heat pipe 4. The upper part of the heat pipe 4 is cooled, and the evaporative liquid therein is cooled and returned to liquid state and returns to the lower part. The upper half of the heat pipe 4 is installed with metal fins 5. The metal fins 5 increase the contact area with the air, thereby improving the heat dissipation of the heat pipe 4. The faster heat dissipation of the heat pipe 4 also accelerates the heat dissipation of the coolant, thereby accelerating the heat dissipation of the battery pack 3.

[0027] A temperature sensor, such as a thermistor, is installed within the sealed housing 1 to detect the coolant temperature. When the coolant temperature exceeds a first preset temperature, fan 6 is activated, blowing air to accelerate air flow around the metal fins 5, thereby accelerating the cooling of the metal fins 5. By accelerating the cooling of the metal fins 5, the coolant is cooled faster, ensuring that the battery pack 3 remains within a stable temperature range. If the coolant temperature exceeds a second preset temperature, water pump 7 is activated to pump coolant from the sealed housing 1. The coolant then flows through flow guide tube 8 and ultimately into manifold box 9. From manifold box 9, the coolant flows into the metal tubes of wet wick 10. Because the multiple sets of metal tubes are staggered and interconnected, the coolant flows downward, flowing into each of the metal tubes, ultimately discharging from the lower end of the wet wick 10 and flowing into manifold box 11. The coolant in manifold box 11 then flows into the sealed housing 1 through flow guide holes 14, thus achieving a circulating flow of the coolant. When the coolant circulates, combined with the cooling of the fan 6, the air passes through the gaps in the wet core 10. The flowing air combined with the fast-flowing coolant in the metal tube can achieve rapid cooling of the coolant, thereby effectively cooling the battery pack 3 and achieving thermal control of the battery pack 3.

[0028] The sealed housing 1 has a sidewall with a liquid injection port 12 through which coolant can be injected into the sealed housing 1. The liquid injection port 12 is sealed with a sealing cover to prevent coolant from spilling out of the liquid injection port 12. The inner bottom wall of the sealed housing 1 has a positioning groove 2. The battery pack 3 is composed of multiple individual battery cells, each of which is installed in a corresponding positioning groove 2. A pressure plate 15 is also provided at the upper end of the battery cell, which provides a stable force at the upper and lower ends of the battery cell while preventing coolant from splashing, thereby ensuring that the battery pack 3 is stably installed in the sealed housing 1.

[0029] In some embodiments, the system further includes a perception and evaluation module configured to collect battery temperature parameters, coolant status parameters, and environmental parameters in real time and generate a multi-dimensional feature vector; calculate a battery thermal risk index based on the collected data, the index being generated by a weighted combination of the battery's maximum temperature, temperature rise rate, and state of charge; The predictive modeling module is used to input multidimensional feature vectors into the time series prediction model; predict the spatial distribution of battery temperature field in future time periods based on historical temperature data; and calibrate the prediction accuracy through the loss function embedded in the thermodynamic equation; The decision optimization module implements a hierarchical control strategy based on the predicted temperature field and thermal risk index; it uses an evolutionary algorithm to solve the energy-efficient optimal control instructions that meet the temperature constraints; The execution monitoring module is used to convert control instructions into actuator drive signals; monitor the execution status and actual battery temperature data in real time through a closed-loop controller; monitor execution deviations in real time and activate the fault tolerance mechanism when flow anomalies are detected; The learning evolution module updates system parameters based on the feedback execution status and actual temperature data through a reinforcement learning mechanism, with the objective function taking into account both temperature stability and energy efficiency; and regularly calibrates the time series prediction model and control parameters.

[0030] The Perceptual Assessment Module also includes: Determine the data acquisition sensor configuration for the battery and cooling system, including setting corresponding sensor devices on the battery cells, coolant pipelines, and the outside of the sealed housing 1 to monitor the battery cell temperature, coolant status parameters, environmental parameters, and obtain the battery state of charge; Extract key data features based on the sensor configuration, including the maximum temperature of the battery cell, temperature rise rate, coolant temperature difference, flow fluctuation, temperature and humidity outside the sealed housing 1, and battery state of charge, and integrate them into a feature information group; According to the characteristic information group, different weights are assigned to the maximum temperature of the battery cell, the temperature rise rate and the battery state of charge, and a comprehensive calculation is made to obtain a thermal risk index that reflects the battery thermal risk.

[0031] By adopting this technical solution, thermocouples or NTC thermistors are used as temperature sensors at the battery cell level. Each battery cell is tightly bonded to at least one sensing unit. Thermally conductive silicone grease fills the gap between the sensing unit and the cell housing to reduce thermal resistance, ensuring a temperature monitoring error within ±0.5°C. The signal output of the sensing unit is connected to the data acquisition module via a shielded cable, enabling independent, real-time monitoring of each cell's temperature. The sensing configuration of the coolant pipeline adopts a distributed layout: temperature sensors are installed at the inlet section where the coolant flows into the sealed shell 1 and the outlet section where the coolant flows out of the sealed shell 1, with the distance between the two being no less than 5 times the diameter of the pipeline to avoid local flow field interference, which is used to capture the heat exchange temperature difference between the coolant and the battery pack 3; an electromagnetic flowmeter is installed in the straight pipe section in the middle of the guide pipe 8, with a measurement accuracy of no less than ±2%, and the flow field distribution is optimized through the rectification structure in the pipeline to reduce the impact of flow fluctuations on measurement stability; temperature sensing points are set at the upper and lower ends of the heat dissipation pipe 4, with the distance between the two points being no less than 1 / 3 of the length of the effective heat exchange section of the heat dissipation pipe, which is used to monitor the phase change heat transfer effect of the evaporating liquid. The external sensing configuration of sealed enclosure 1 includes a temperature and humidity sensor mounted on the windward side of the top of the enclosure. Its measurement range covers -20°C to 60°C and 0 to 100% RH (humidity), and its sampling frequency is synchronized with the internal sensing device. The battery state of charge (SOC) is calculated via the communication interface of the battery management system (BMS). Data exchange is achieved via the CAN bus or RS485 protocol, ensuring that the SOC data update delay does not exceed 1 second. The power supply modules of all sensing devices are isolated to prevent data drift caused by electromagnetic interference.

[0032] Based on the above sensor configuration, the original monitoring data is analyzed in real time, and the temperature data of all battery cells are compared hour by hour. The instantaneous maximum value is screened out and the corresponding cell number and occurrence time are marked to form the highest temperature sequence; Taking 10 seconds as the time window, the ratio of the maximum temperature change in each window to time is calculated. The instantaneous fluctuations are filtered out by the sliding average algorithm to obtain a smooth temperature rise rate curve. The difference between the temperature sensor readings of the coolant flowing into the sealed housing 1 and the coolant flowing out of the sealed housing 1 is calculated in real time. When the difference is less than a preset threshold, an abnormal flag is triggered, indicating that the heat exchange efficiency has decreased; Perform spectrum analysis on the instantaneous flow data output by the flow meter, extract the flow standard deviation within 1 minute, and quantify the flow stability in the pipeline; According to the integrated temperature and humidity sensor installed on the outside of the sealed housing 1, the real-time data of the external sensor is directly read, and the hourly average value is calculated as the environmental reference parameter; After obtaining the original SOC data from the BMS, it is corrected based on the current charge and discharge rate to eliminate the SOC jump caused by instantaneous current impact. After the above characteristic parameters are aligned by timestamps, they are packaged into characteristic information groups at a fixed period (e.g., every 2 seconds). Each group of data contains 7 types of key features and corresponding collection time stamps, providing input for subsequent thermal risk assessment.

[0033] The thermal risk index based on the characteristic information group is calculated using a weighted fusion mechanism. For example, the maximum temperature of the battery cell is used as the core parameter and is given the highest weight because it directly reflects the critical state of thermal runaway. As the battery temperature approaches the upper limit of safety, the weight of this parameter is dynamically increased. The temperature rise rate is given the second highest weight to evaluate the heat accumulation rate. When the rate exceeds the preset safety threshold, its contribution to the thermal risk is amplified through weight adjustment. The battery state of charge is assigned a basic weight, and combined with its correlation with thermal stability (e.g., increased thermal sensitivity at high SOC), the weight is appropriately increased when the SOC is above 80% or below 20%. Through this weighting, the normalized eigenvalues ​​of the three are combined to generate a thermal risk index ranging from 0 to 100. An index below 30 is considered low risk, 30-70 is medium risk, and above 70 is high risk. The index value is dynamically updated as the characteristic parameters change in real time, providing a quantitative basis for the cooling system's control strategy.

[0034] In some embodiments, the predictive modeling module further includes: Retrieve the historical temperature data set associated with the current feature information group, including the temperature change records of the battery cells and the temperature field distribution pattern, construct a continuous time series in chronological order and mark the time nodes; A time series prediction model based on time dependence is constructed. This model integrates the temperature correlation characteristics of battery cells and uses coolant status and environmental parameters as auxiliary variable inputs to learn the temperature field evolution pattern through historical temperature change trends. Input the time series into the time series prediction model, predict the temperature change trend of each battery cell and the spatial distribution of the temperature field of the battery pack 3 in the future period based on the current feature information, and mark the time node of the prediction result; Based on the principles of thermodynamics, a verification standard for prediction results is established. By comparing the consistency of the prediction results with the laws of heat conduction, the model parameters are adjusted until the prediction results meet the thermodynamic constraints.

[0035] By employing this technical solution, data retrieval can be used to match related datasets within a historical database based on the current set of characteristic information (including parameters such as the maximum battery cell temperature, coolant temperature difference, and ambient temperature and humidity). The retrieval logic employs multi-dimensional feature matching: using the current ambient temperature and humidity and the battery state of charge (SOC) as primary search criteria, it selects historical operating periods with a similarity exceeding 85%. This is followed by secondary filtering based on the coolant flow rate fluctuation range and initial temperature distribution pattern, ultimately identifying historical samples that are highly consistent with the current operating conditions. Linear interpolation is performed on the selected historical data along the time axis to fill in gaps created by data collection intervals and ensure that the time resolution is consistent with the current monitoring frequency (e.g., 2 seconds / time). A sliding time window (with a window length of 1 hour) is used to segment the historical temperature data, extracting the temperature change records of the battery cells within each time period (including the temperature time series curve of each cell) and the temperature field distribution patterns (e.g., the migration path of high-temperature areas and the characteristics of temperature gradient changes). This data is then concatenated with the current feature information group in chronological order to form a continuous time series. Each data point is annotated with the time node accurate to the millisecond to ensure time series continuity and timestamp alignment. The time series prediction model adopts a spatiotemporal fusion architecture. The core layer is a sequence modeling module that processes time dependencies. It captures the short-term time series characteristics of temperature data through a sliding time window (the window size is set to 30 sampling points), and embeds a long-term memory mechanism to learn the temperature evolution law on a scale of several hours.

[0036] The model input layer adopts a multi-channel design. The main channel is connected to the battery cell temperature time series sequence, and the spatial correlation between battery cells is encoded through the adjacency matrix (such as the thermal conductivity coefficient of adjacent cells and the position distance weight), realizing the quantitative modeling of the cell temperature correlation characteristics; the auxiliary channel synchronously inputs the coolant state parameters (flow fluctuation, inlet and outlet temperature difference) and environmental parameters (temperature and humidity, atmospheric pressure), and converts non-temperature parameters into influencing factors related to the temperature field (such as the correction coefficient of coolant flow change to temperature gradient) through the feature mapping layer. During the model training phase, historical time series sequences are used for iterative optimization. The temperature field evolution data of the past 24 hours is used as training samples. The internal weight parameters are adjusted through the back-propagation mechanism, so that the model can gradually learn the temperature field evolution patterns under different working conditions (such as the temperature aggregation effect under high SOC state and the temperature response lag characteristics after a sudden change in coolant flow). Finally, a time series prediction model is formed that can simultaneously capture temporal dynamics and spatial correlations. The constructed continuous time series is fed into the time series prediction model to initiate the multi-step prediction process. The model first encodes the current feature information group and extracts key parameters (such as the current position of the high-temperature unit and the coolant heat exchange efficiency) as initial prediction conditions. Then, using a time series extrapolation algorithm and historically learned temperature evolution patterns, it generates point-by-point predictions for the next 10-30 minutes. The forecast output includes two dimensions: first, the temperature trend of each battery cell. Temperature curves are formed from temperature predictions at consecutive time nodes, marking possible temperature peaks and their arrival times. Second, the spatial distribution of the battery pack's overall temperature field. A gridding method (with grid accuracy matching the density of battery cell arrangement) is used to characterize the spatial temperature gradient. Color mapping is used to encode the distribution ranges of high-temperature regions (>40°C), medium-temperature regions (25-40°C), and low-temperature regions (<25°C), and the predicted expansion rate of each region is annotated. All forecast results are annotated with corresponding time nodes in chronological order, consistent with the timestamp format of historical data. The prediction result verification standard is established based on the basic principles of thermodynamics. The core verification items include: heat conduction directionality (heat transfer trend from high-temperature cells to low-temperature cells), heat conservation (the balance between the heat removed by the coolant and the heat generated by the battery), and phase change heat transfer law (the matching of the temperature difference between the upper and lower ends of the heat pipe 4 and the phase change rate of the evaporating liquid). The predicted results are compared item by item with the above criteria. If a temperature inversion occurs, with heat transfer from a low-temperature cell to a high-temperature cell, or if the predicted temperature does not decrease despite an increase in coolant flow, this is considered a violation of thermodynamic laws. At this point, the model parameter calibration mechanism is activated: the heat transfer coefficient in the spatial correlation matrix is ​​corrected, increasing the weight consistent with the actual heat transfer direction. The influencing factors of auxiliary variables (such as coolant flow rate) are adjusted to strengthen their constraints on the temperature field prediction. Multiple rounds of iterative verification are performed until the predicted results achieve a compliance level of at least 95% across all verification items, ensuring the physical plausibility of the predictions.

[0037] In some embodiments, the decision optimization module further includes: Establish a thermal risk classification threshold system, including determining the thermal risk index threshold and temperature field prediction index threshold based on the battery safety temperature range and historical failure data; Compare the real-time thermal risk index and temperature field prediction results with the threshold system, and output low, medium, and high control level signals based on the degree of risk index exceeding the standard and the scope of the temperature field violation area; In response to the control level signal, the preset basic control rules are called to generate the corresponding level of heat dissipation actuator parameter adjustment plan, including the linkage control strategy of the fan 6 and the water pump 7; Under the constraints of the parameter adjustment scheme, the optimal control instruction set that meets the temperature safety constraints and has the lowest total power consumption is screened through simulation and energy consumption evaluation of multiple groups of control instructions.

[0038] By adopting the above technical solution, the construction of a thermal risk classification threshold system is based on the fusion analysis of battery safe operating boundaries and historical failure data. First, based on the material properties and safety standards of the battery cells, the core temperature parameter threshold is determined: 80% (40°C) of the battery's maximum temperature tolerance (e.g., 50°C) is used as the critical value for temperature field prediction. If the area of ​​high temperature areas (>40°C) accounts for more than 10%, it is determined to be a temperature field violation. If the temperature rise rate exceeds 2°C / min, it is marked as a rapid temperature rise risk. At the same time, statistical modeling was performed on historical failure data to extract the thermal risk index characteristics before the failure occurred: by analyzing 50 thermal risk warning events in the past three months, it was determined that the failure probability was less than 0.1% when the thermal risk index was 30, and the failure probability suddenly rose to 15% when the thermal risk index was 70. Based on this, 30 was set as the first threshold (low-medium risk critical point), and 70 was set as the second threshold (medium-high risk critical point). The final threshold system includes two judgment dimensions: the thermal risk index threshold (30, 70) and the temperature field index threshold (10% of high temperature area, temperature rise rate 2°C / min). The two are jointly judged through logical AND operations to ensure that the threshold setting covers both sudden risks (such as a sudden increase in the index) and cumulative risks (such as the expansion of high temperature areas). Control level matching uses a dynamic comparison mechanism of real-time data and threshold system, real-time collected thermal risk index and predicted temperature field results, and then makes a judgment: Low risk level judgment: When the thermal risk index is continuously lower than 30, and the high temperature area ratio in the temperature field prediction result is less than 5%, and the temperature rise rate is less than 0.5℃ / min, it is judged as low risk and a low-level control signal is output; Medium risk level determination: When the thermal risk index fluctuates between 30-70, or the high temperature area accounts for 5%-10%, and the temperature rise rate is 0.5-2°C / min (any of these conditions is met), it is determined to be a medium risk and a medium-level control signal is output; High-risk level determination: When the thermal risk index exceeds 70, or the high-temperature area accounts for >10%, or the temperature rise rate is >2°C / min (any of these conditions is met), it is determined to be high risk and a high-level control signal is output. A hysteresis compensation mechanism is introduced into the judgment process. There is no delay when switching from medium risk to high risk. The switch from high risk to medium risk requires the indicator to fall back to the threshold for 30 seconds to avoid actuator loss caused by frequent switching. The working principle of the heat dissipation actuator parameter adjustment scheme generation is that the parameter adjustment scheme calls the preset rule base based on the control level signal. The rule base is pre-generated and solidified through the orthogonal test method: Low-risk solution: Maintain current operating parameters, with fan 6 running at 30% of its rated speed and pump 7 controlled at 2 L / min. A pulse width modulation (PWM) signal is used to lock the actuator state. Medium-risk solution: Activate the collaborative enhancement strategy, increase the speed of fan 6 to 60% of the rated value, and increase the flow rate of water pump 7 to 3 L / min at a ratio of 1:1.5, using a combination of air cooling and liquid cooling. High-risk solution: Execute full-load cooling, with fan 6 running at 100% of the rated speed and water pump 7 flow increased to a maximum of 5 L / min. Constraints are embedded during solution generation: the flow rate of water pump 7 does not exceed the upper pressure limit of the pipeline (8 L / min), and the speed of fan 6 does not fall below the minimum stable speed (10% of the rated value), ensuring that the actuator operates within a safe range.

[0039] The selection of optimal instructions is carried out by performing multi-objective optimization within the range of parameter adjustment solutions: Instruction generation is based on the scenario parameter range (e.g., fan 6 speed 50%-70% and water pump 7 flow 2.5-3.5 L / min at medium risk). Uniform sampling is used to generate 20 groups of candidate instructions, each group containing specific parameters such as fan 6 PWM duty cycle and water pump 7 voltage. Simulation verification: Input candidate instructions into the thermal balance model to simulate the temperature field changes in the next 10 minutes to verify whether the safety constraints of "maximum temperature ≤ 40°C and high temperature area cleared" are met; The thermal balance model can be constructed using a neural network. Inputs include battery status (current temperature, charge and discharge power), cooling parameters (fan speed, water pump flow), and ambient temperature. Outputs are temperature indicators for the next 10 minutes (average temperature every two minutes, final maximum cell temperature). Historical operating data covering multiple operating conditions is selected, normalized, and divided into training and validation sets. Time-sliced ​​data is then sliced ​​into input-output pairs. A shallow feedforward architecture is employed, consisting of an input layer (5 neurons), one hidden layer (32 neurons), and an output layer (6 neurons). The parameter size is controlled to ensure fast computation. An adaptive loss function (focusing on penalizing maximum temperature deviation) and an optimizer are used for training, incorporating an early stopping mechanism to avoid overfitting. Testing and calibration with new operating condition data ensure that the maximum temperature prediction error is ≤2°C for more than 95% of samples. The model is then solidified into a lightweight file and integrated into the simulation. Upon inputting candidate control commands, temperature predictions are rapidly output, allowing direct verification of compliance with safety constraints. Energy consumption evaluation: Through power monitoring, calculate the real-time power consumption of each set of instructions (the power consumption of fan 6 is proportional to the square of the rotation speed, and the power consumption of water pump 7 is linearly related to the flow rate), and accumulate the total energy consumption for 10 minutes; Optimal screening: Using the Pareto optimality principle, the set with the lowest total energy consumption is selected from candidate instructions that meet safety constraints. This is then output as the optimal control instruction set. The instruction format complies with the frame structure requirements of the actuator communication protocol (such as Modbus-RTU). The screening process is iterated every two minutes to ensure that the control instructions can dynamically adapt to changes in battery status.

[0040] The specific principle for using the Pareto optimality principle to select the group with the lowest total energy consumption is as follows: first, candidate instructions that meet safety constraints (temperature requirements) are screened to form a set of feasible solutions. Each feasible solution is then evaluated from two dimensions: "safety redundancy" (the difference between the temperature and the safety threshold) and "total energy consumption." Compare feasible solutions pairwise and eliminate "dominated solutions" (i.e., solutions with lower energy consumption and high safety redundancy). The remaining solutions are the "Pareto optimal solution set"; select the instruction with the lowest total energy consumption from the optimal solution set. If the energy consumption is the same, select the instruction with the highest safety redundancy as the final optimal instruction.

[0041] In some embodiments, executing the monitoring module includes: Convert the control instructions of the heat dissipation actuator into a driving electrical signal that matches its signal type, including converting the fan 6 speed instruction into a pulse modulation signal and the water pump 7 flow instruction into a voltage signal; Send a drive signal to the heat dissipation actuator to start operation. At the same time, the actual operating status of the actuator, battery temperature, and coolant parameters are collected in real time through sensors to form a closed-loop monitoring data set synchronized with the target parameters and time stamps. Calculate the deviation between the actual operating parameters of the actuator and the target parameters, and judge the operating abnormality based on the preset deviation threshold, including judging the flow deviation exceeding the limit as flow abnormality; In response to flow anomalies, the fault-tolerance mechanism is activated, increasing the power of the water pump 7 to a preset safety upper limit and collaboratively enhancing the heat dissipation capacity of the fan 6; if the anomaly persists, an alarm is triggered and the anomaly time and deviation data are recorded to ensure that the battery temperature meets the thermal management safety constraints.

[0042] By adopting the above technical solution, the speed command for Fan 6 is converted using pulse width modulation (PWM) technology. The control module maps the target speed percentage (such as 30% or 60%) to the corresponding duty cycle signal. For example, the rated speed corresponds to a 50Hz PWM signal, and 30% speed corresponds to a 15% duty cycle. A dedicated PWM generator chip outputs a 0-5V square wave signal, with the steepness of the rising and falling edges of the signal controlled within 100ns, ensuring the response delay of Fan 6's motor is ≤50ms. Pump 7's flow command is converted via a digital-to-analog converter (DAC) module, linearly converting the flow rate percentage (e.g., 2 L / min corresponds to 40% of the rated flow rate) into a 0-12V DC voltage signal. The voltage-flow mapping is calibrated using Pump 7's factory calibration curve. For example, 12V corresponds to a maximum flow rate of 5 L / min, and 6V corresponds to 2.5 L / min. Conversion accuracy is controlled within ±0.1V, ensuring flow rate deviation ≤5%. During the conversion process, a signal isolation circuit is embedded, achieving electrical isolation between the control module and the actuator via an optocoupler. This suppresses signal distortion caused by common-mode interference and ensures the stability of the drive signal. The actuator of the present invention is a hardware device that directly implements heat dissipation, including a control fan 6 and a water pump 7. Specifically, the control module sends a drive signal to regulate the operation of the actuator, and the actual state of the actuator is fed back to the control module via a sensor, achieving a dynamic coordination of "command-execution-monitoring-adjustment." After the driving signal is sent, the multi-dimensional real-time monitoring link is started: Regarding the acquisition of actuator status, a Hall effect speed sensor is installed at the fan 6 end, which outputs 2 pulse signals per revolution, and the actual speed is converted through the counter module (sampling frequency 1kHz); a flow meter is installed at the outlet of the water pump 7, which outputs a 4-20mA current signal, which is converted into a 0-5V signal through the current-voltage conversion circuit to reflect the actual flow in real time (sampling frequency 500Hz). Regarding the collection of battery and coolant parameters, the signals of the battery cell temperature sensor (sampling frequency 10Hz), the coolant inlet and outlet temperature sensors (sampling frequency 5Hz), and the pipeline pressure sensor (sampling frequency 1Hz) are first aggregated through a multiplexer and then synchronously converted into digital signals through a 16-bit ADC module.

[0043] All collected signals are time-stamped by a local high-precision clock module (accuracy ±10ppm) and packaged into data sets at 20ms intervals. Each data packet contains the actual speed of fan 6, the actual flow rate of water pump 7, the temperature of each battery cell, the coolant temperature difference and pressure value, ensuring that the time deviation between the target parameter (command value) and the actual parameter is ≤1ms, forming a complete closed-loop feedback link. Deviation calculation uses the instantaneous difference analysis method between the real-time value and the target value. For example, the deviation rate between the actual speed of fan 6 and the command speed = (actual value - target value) / target value × 100%. The flow deviation rate of water pump 7 uses the same calculation logic. The preset deviation thresholds are dynamically adjusted based on actuator characteristics: the normal deviation threshold for fan 6 is ±10% (relaxed to ±15% at low speeds), and the normal deviation threshold for pump 7 flow is ±8% (relaxed to ±12% for low flow rates ≤ 2 L / min). If the deviation rate exceeds the threshold for three consecutive sampling periods (60 ms) and no sensor fault is detected, an abnormal operation is detected. When the flow deviation rate of water pump 7 is greater than 20%, or the absolute value of the flow is ≤0.5L / min (lasting 1s), and the pipeline pressure is ≥1.2 times the rated value, it is determined to be a pipeline blockage; if the flow deviation rate is greater than 20% and the pressure is ≤0.8 times the rated value, it is determined to be a power failure of water pump 7. Both situations are classified as flow abnormalities. After a traffic anomaly is triggered, the hierarchical fault tolerance process is started: Primary compensation: Immediately increase the water pump drive voltage to the safe upper limit (12V) for 2 seconds in an attempt to break through the blockage or overcome insufficient power. At the same time, increase the PWM duty cycle of fan 6 by 20% (not exceeding the rated value) to compensate for the decrease in liquid cooling efficiency by enhancing air cooling. Condition monitoring: If the flow rate returns to the deviation threshold (≤±12%) within 2 seconds, it will be gradually adjusted back to the original command value at a rate of 5% / s to avoid system fluctuations caused by sudden parameter changes. Level 2 treatment: If the abnormality lasts for more than 5 seconds, current limiting protection is activated, and fan 6 runs at 100% of the rated speed to forcibly maintain the heat dissipation capacity. Alarm and Recording: If an anomaly persists for 10 seconds, a fault code (such as 0x00A3 for flow anomaly) is transmitted via the CAN bus, activating the alarm. Data such as the anomaly occurrence time, maximum deviation, and maximum battery temperature are written to non-volatile memory to ensure fault traceability. Throughout the fault-tolerance process, real-time temperature monitoring ensures that the maximum battery temperature does not exceed 45°C. If the temperature approaches the threshold, an emergency shutdown is triggered. This alarm device can be a buzzer installed in the sealed housing 1. When the alarm device is activated, the buzzer can be charged, causing it to sound and trigger an alarm.

[0044] In some embodiments, the learning evolution module further includes: Compare the deviation parameters between the actual operating state of the actuator and the optimal control instructions, and analyze the spatial distribution differences between the actual battery temperature data and the predicted temperature field; synchronously collect the system's real-time energy consumption data and temperature fluctuation characteristics; Based on the deviation parameters and fluctuation characteristics, a reinforcement learning reward rule is set with temperature stability and energy efficiency as the core optimization goals, and the reward value is dynamically allocated according to the degree of goal achievement; Deviation data and reward rules are input into the reinforcement learning mechanism. Through iterative optimization, the thermal risk judgment threshold, time series prediction model feature weight, and actuator response coefficient in the control strategy are adjusted to make the system converge to the reward maximization state. Periodically retrain the time series prediction model and calibrate the control parameters, dynamically adjust the model structure and threshold range based on the actual operation results, and deploy the optimization results to the next round of control process.

[0045] By employing this technical solution, a sliding average of the deviation rate between fan 6's actual speed and the optimal command (calculated every 100ms) is first performed to generate a speed deviation curve. The instantaneous difference between pump 7's actual flow rate and the commanded flow rate is then integrated to obtain the cumulative flow deviation. Both deviation parameters are stored by actuator type, with the duration of the deviation and the time of peak occurrence noted. Using a grid method, battery pack 3 was divided into 10×10 spatial cells. The actual temperature of each cell (collected by distributed sensors) was compared with the corresponding value of the predicted temperature field, and the absolute value and mean square error of the cell temperature difference were calculated. The overlap between the actual high temperature area (>40°C) and the predicted high temperature area was highlighted. If the overlap rate was less than 60%, it was determined to be a significant deviation in the spatial distribution. The battery cell number and duration of the deviation area were also recorded. System energy consumption data is sampled at 500Hz by power sensors, and the minute-by-minute average power consumption is calculated and compared with historical values ​​for the same period. Temperature fluctuations are quantified by calculating the standard deviation of the difference between the maximum and minimum temperatures of battery pack 3 (with a 1-minute sliding window), capturing transient spikes and dips. All data is aligned using timestamps to form a four-dimensional dataset containing deviation parameters, temperature differences, energy consumption, and fluctuation characteristics. The working principle of setting the reinforcement learning reward rule is dynamically generated based on the dual-objective optimization system: A temperature stability reward is defined, with the core indicators set as "temperature fluctuation ≤ 3°C" and "actual temperature deviation from target temperature (35±2°C) ≤ 2°C." When both conditions are met, a base reward is distributed linearly based on the fluctuation (smaller fluctuations result in higher rewards). If a single cell temperature exceeds 40°C, a penalty mechanism is triggered (a 30% reduction in base rewards is applied). An additional stability reward is awarded for five consecutive minutes of no temperature overshoot.

[0046] Energy efficiency rewards are defined, with the average energy consumption of the historical optimal operating conditions as the baseline. For every 10% below the baseline for actual energy consumption, the reward value increases by 15%; if energy consumption is higher than the baseline but temperature stability is significantly better than the historical level (fluctuation is reduced by 50%), a reward compensation mechanism can be activated (maximum compensation is 20%).

[0047] Under high-risk conditions (thermal risk index > 50), the reward weight for temperature stability is automatically increased to 70%, while the weight for energy consumption is reduced to 30%. The opposite is true for low-risk conditions. Rewards are calculated on a minute-by-minute basis, with a reward decay coefficient (0.95 / minute) to discourage short-term speculation and encourage long-term optimization.

[0048] The specific principle of reinforcement learning parameter optimization is to achieve system optimization through iterative adjustment: The state space is defined, with a four-dimensional data set (deviation parameter, temperature difference, energy consumption, and fluctuation characteristics) as the input state, covering actuator response deviation (such as fan 6 speed hysteresis), time series prediction model error (such as high temperature area deviation rate) and system dynamic characteristics (such as the rate of change of energy consumption with temperature).

[0049] The action space is designed to include three types of adjustable parameters: thermal risk judgment threshold (the first threshold is fine-tuned between 25 and 35, and the second threshold is fine-tuned between 65 and 75), time series prediction model feature weights (such as the weight coefficient of the coolant flow feature is adjusted by ±10%), and actuator response coefficients (such as the delay compensation coefficient of the PWM signal of fan 6). Iterative Optimization Process: Each 10-minute learning cycle feeds the current state and reward into the reinforcement learning mechanism. The policy gradient method adjusts the action parameters. If a parameter adjustment increases the reward, the adjustment is retained and the exploration step size is increased. If the reward decreases, the step size is reduced and fine-tuned in the opposite direction. Through 100-200 cycles of iteration, the system converges to an optimal state where the reward value remains stable above 80%. A retraining cycle is set to retrain the temperature field time series prediction model every 24 hours based on the 100,000+ operating data (including extreme operating condition samples) accumulated that day. New samples are prioritized for high-error operating conditions with a temperature deviation rate >5%. The model's ability to fit complex operating conditions is improved by increasing the number of hidden layer neurons (dynamically adjusted from 32 to 64). Control parameters are calibrated. The thermal risk determination threshold is calibrated weekly. A sliding window (data from the past seven days) is used to calculate the average reward value under different thresholds, and the threshold combination that maximizes the reward is selected. The actuator response coefficient is regularly updated through pulse response testing. For example, a step command is sent to fan 6 every three days, and the response coefficient is corrected based on the actual speed rise time. The optimized model and parameters are deployed through a grayscale release mechanism, and are first trial-run for 2 hours in 5% of the battery modules to verify that the temperature stability and energy consumption indicators meet the standards. Then, the full model and parameters are updated to the next round of control processes to ensure the security and continuity of the iterative process.

[0050] The above description is merely a preferred embodiment of the present invention; however, the scope of protection of the present invention is not limited thereto. Any person skilled in the art who, within the technical scope disclosed by the present invention, makes equivalent substitutions or modifications based on the technical solutions and improved concepts of the present invention shall be covered by the scope of protection of the present invention.

Claims

1. A heat pipe-assisted submerged battery pack thermal management system, comprising a sealed housing (1), characterized in that: A battery pack (3) is installed inside the sealed shell (1), a gap is provided between two adjacent battery packs in the battery pack (3), a heat dissipation pipe (4) is inserted in the gap, the lower half of the heat dissipation pipe (4) is located inside the sealed shell (1), and the upper half thereof passes through the sealed shell (1) and extends to the outside, a fan (6) is installed on the outside of the sealed shell (1) near the heat dissipation pipe (4), a wet core (10) is installed on the upper surface of the sealed shell (1) at the end of the heat dissipation pipe (4) away from the fan (6), the inside of the sealed shell (1) is filled with coolant, and a water pump (7) is also installed inside the sealed shell (1), the water pump (7) is used to extract coolant from the sealed shell (1) and inject it into the wet core (10), and the coolant is discharged from the lower end of the wet core (10) and flows back into the sealed shell (1); return include, A perception and evaluation module, which collects battery temperature parameters, coolant status parameters, and environmental parameters in real time and generates a multi-dimensional feature vector; Calculate the battery thermal risk index based on the collected data. The index is generated by a weighted combination of the battery's maximum temperature, temperature rise rate, and state of charge; The predictive modeling module is used to input multidimensional feature vectors into the time series prediction model; predict the spatial distribution of battery temperature field in future time periods based on historical temperature data; and calibrate the prediction accuracy through the loss function embedded in the thermodynamic equation; The decision optimization module implements a hierarchical control strategy based on the predicted temperature field and thermal risk index; it uses an evolutionary algorithm to solve the energy-efficient optimal control instructions that meet the temperature constraints; The execution monitoring module is used to convert control instructions into actuator drive signals; monitor the execution status and actual battery temperature data in real time through a closed-loop controller; monitor execution deviations in real time and activate the fault tolerance mechanism when flow anomalies are detected; The learning evolution module updates system parameters based on the feedback execution status and actual temperature data through a reinforcement learning mechanism, with the objective function taking into account both temperature stability and energy efficiency; and regularly calibrates the time series prediction model and control parameters.

2. The heat pipe-assisted submerged battery pack thermal management system according to claim 1, characterized in that: The Perceptual Assessment Module also includes: Determine the data acquisition sensor configuration of the battery and cooling system, including setting corresponding sensor devices on the battery cells, coolant pipes and the outside of the sealed housing (1) to monitor the battery cell temperature, coolant state parameters, environmental parameters and obtain the battery charge state; Extract key data features based on the sensor configuration, including the maximum temperature of the battery cell, the temperature rise rate, the coolant temperature difference, the flow fluctuation, the external temperature and humidity of the sealed housing (1) and the battery charge state, and integrate them into a feature information group; According to the characteristic information group, different weights are assigned to the maximum temperature of the battery cell, the temperature rise rate and the battery state of charge, and a comprehensive calculation is made to obtain a thermal risk index that reflects the battery thermal risk.

3. The heat pipe assisted submerged battery pack thermal management system according to claim 2, characterized in that: The predictive modeling module also includes: Retrieve the historical temperature data set associated with the current feature information group, including the temperature change records of the battery cells and the temperature field distribution pattern, construct a continuous time series in chronological order and mark the time nodes; A time series prediction model based on time dependence is constructed. This model integrates the temperature correlation characteristics of battery cells and uses coolant status and environmental parameters as auxiliary variable inputs to learn the temperature field evolution pattern through historical temperature change trends. Input the time series into the time series prediction model, predict the temperature change trend of each battery cell and the temperature field spatial distribution of the battery pack (3) in the future period based on the current feature information, and mark the time node of the prediction result; Based on the principles of thermodynamics, a verification standard for prediction results is established. By comparing the consistency of the prediction results with the laws of heat conduction, the model parameters are adjusted until the prediction results meet the thermodynamic constraints.

4. The heat pipe-assisted submerged battery pack thermal management system according to claim 3, characterized in that: The decision optimization module also includes: Establish a thermal risk classification threshold system, including determining the thermal risk index threshold and temperature field prediction index threshold based on the battery safety temperature range and historical failure data; Compare the real-time thermal risk index and temperature field prediction results with the threshold system, and output low, medium, and high control level signals based on the degree of risk index exceeding the standard and the scope of the temperature field violation area; In response to the control level signal, a preset basic control rule is called to generate a corresponding level of heat dissipation actuator parameter adjustment scheme, including a linkage control strategy for the fan (6) and the water pump (7); Under the constraints of the parameter adjustment scheme, the optimal control instruction set that meets the temperature safety constraints and has the lowest total power consumption is screened through simulation and energy consumption evaluation of multiple groups of control instructions.

5. The heat pipe assisted submerged battery pack thermal management system according to claim 4, characterized in that: Execution monitoring module, including; Converting the control instruction of the heat dissipation actuator into a driving electrical signal that matches its signal type, including converting the fan (6) speed instruction into a pulse modulation signal and the water pump (7) flow instruction into a voltage signal; Send a drive signal to the heat dissipation actuator to start operation. At the same time, the actual operating status of the actuator, battery temperature, and coolant parameters are collected in real time through sensors to form a closed-loop monitoring data set synchronized with the target parameters and time stamps. Calculate the deviation between the actual operating parameters of the actuator and the target parameters, and judge the operating abnormality based on the preset deviation threshold, including judging the flow deviation exceeding the limit as flow abnormality; In response to flow anomalies, the fault tolerance mechanism is activated to increase the power of the water pump (7) to a preset safety upper limit and to coordinately enhance the heat dissipation capacity of the fan (6); if the anomaly persists, an alarm is triggered and the anomaly time and deviation data are recorded to ensure that the battery temperature meets the thermal management safety constraints.

6. The heat pipe-assisted submerged battery pack thermal management system according to claim 5, characterized in that: The Learning Evolution module also includes; Compare the deviation parameters between the actual operating state of the actuator and the optimal control instructions, and analyze the spatial distribution differences between the actual battery temperature data and the predicted temperature field; synchronously collect the system's real-time energy consumption data and temperature fluctuation characteristics; Based on the deviation parameters and fluctuation characteristics, a reinforcement learning reward rule is set with temperature stability and energy efficiency as the core optimization goals, and the reward value is dynamically allocated according to the degree of goal achievement.

7. The heat pipe-assisted submerged battery pack thermal management system according to claim 6, characterized in that: The Learning Evolution module also includes: Deviation data and reward rules are input into the reinforcement learning mechanism. Through iterative optimization, the thermal risk judgment threshold, time series prediction model feature weight, and actuator response coefficient in the control strategy are adjusted to make the system converge to the reward maximization state. Periodically retrain the time series prediction model and calibrate the control parameters, dynamically adjust the model structure and threshold range based on the actual operation results, and deploy the optimization results to the next round of control process.

8. The heat pipe-assisted submerged battery pack thermal management system according to claim 1, characterized in that: The wet core (10) is made of a plurality of groups of metal tubes that are connected horizontally and vertically and are staggered with each other. The interiors of the plurality of groups of metal tubes are interconnected, and spaces for liquid flow are left between the staggered metal tubes.

9. The heat pipe-assisted submerged battery pack thermal management system according to claim 8, characterized in that: The upper end of the wet core (10) is fixedly connected to a diversion box (9), and the lower end thereof is fixedly connected to a junction box (11). The diversion box (9) and the junction box (11) are both hollow in design, and the interiors of the two are interconnected through the metal tube on the wet core (10). The upper surface of the sealed shell (1) is provided with a guide hole (14) below the junction box (11). The interior of the junction box (11) is interconnected with the interior of the sealed shell (1) through the guide hole (14). The liquid outlet of the water pump (7) is fixedly connected to a guide pipe (8). The end of the guide pipe (8) away from the water pump (7) passes through the sealed shell (1) and is connected to the diversion box (9), and the interior of the guide pipe (8) is interconnected with the interior of the diversion box (9).

10. The heat pipe-assisted submerged battery pack thermal management system according to claim 9, characterized in that: The upper and lower ends of the heat dissipation tube (4) are sealed, and the interior of the heat dissipation tube (4) is filled with evaporative liquid. Metal fins (5) are fixedly mounted on the upper half of the heat dissipation tube (4). A protective cover (13) is fixedly mounted on the upper end of the sealed housing (1). The wet core (10), the fan (6) and the upper half of the heat dissipation tube (4) are all arranged in the protective cover (13).

Images