A heat exchange regulation system for battery pack

Abstract

The present application discloses a heat exchange regulation system for a battery pack, comprising a microchannel heat exchanger, a plurality of interdigital electrodes, a swinging assembly, and a local early warning module. The microchannel heat exchanger is provided with a flow channel, and the local early warning module induces electroosmosis through the interdigital electrodes, thereby removing bubbles near the interdigital electrodes. The swinging assembly comprises a first magnetic component and a plurality of swinging components. The local early warning module controls the intermittent power supply of the first magnetic component to drive the swinging component to swing, thereby shearing and peeling off bubbles attached to the wall of the flow channel. By superimposing the mainstream of the electroosmotic flow and the swinging eddy current, the bubbles in the flow channel are reduced, and the local thermal resistance is significantly improved. The local early warning module adopts the coupling control of the DRL adaptation layer and the LAMPC execution layer to achieve real-time defoaming and optimize the flow and heat exchange performance. Through the closed-loop control strategy of coordinated sensing of temperature and pressure, coordinated defoaming of electroosmotic flow and mechanical disturbance, intelligent early warning, and dynamic feedback optimization, efficient and stable operation of microchannel heat exchange under high heat flux density is achieved.

Description

Technical Field

[0001] The present application relates to the field of battery technology, and in particular to a heat exchange regulation system for a battery pack. Background Art

[0002] As the core of a new generation of clean energy technology, the thermal management system of batteries plays a vital role in the performance, lifespan, and safety of the entire device. In existing technologies, the battery pack is installed in a battery mounting frame, a cooling groove is provided at the bottom of the battery mounting frame, and the microchannel heat exchanger is installed in the inner cavity of the battery mounting frame and in the cooling groove. Due to its compact structure and high specific surface area, the microchannel heat exchanger has become an important means of dissipating high heat flux density of batteries and ensuring the temperature of the battery stack. The microchannel heat exchanger is a new type of high-efficiency heat exchange equipment based on microscale flow and heat transfer theory. Its core feature is that it significantly increases the heat exchange area and strengthens the fluid disturbance through a large number of parallel arranged tiny flow channels, thereby breaking through the performance limit of traditional heat exchangers.

[0003] In a microchannel heat exchanger, when the operating conditions become complex, such as under high heat flux density, although the microchannel heat exchanger has the potential for high heat transfer efficiency, the coupling of its microscale characteristics and extreme heat load will amplify the original problem. The fluid in the microchannel is dominated by surface tension, and tiny bubbles (diameter ≤ 100μm) are easy to nucleate on the wall and quickly merge to form a "steam plug" to block the flow channel. If the steam bubbles generated by boiling cannot be discharged in time, they will gather in the flow channel to form a "gas plug", blocking the liquid phase flow, resulting in local heat exchange interruption. The flow channel blockage rate increases exponentially with the operating time, and eventually causes local overheating failure. Summary of the Invention

[0004] The present application aims to solve at least one of the technical problems existing in the prior art. The present application provides a heat exchange regulation system for a battery pack, which can quickly and comprehensively remove bubbles in a microchannel heat exchanger, greatly improving the overall stability and safety of the system.

[0005] A heat exchange regulation system for a battery pack according to an embodiment of the present application includes:

[0006] A microchannel heat exchanger for dissipating heat for the battery pack; the microchannel heat exchanger includes an inlet end, an outlet end, and a plurality of baffles arranged along the length of the microchannel heat exchanger, with flow channels formed between adjacent baffles. The inlet end and the outlet end are respectively provided with a first pressure sensing component, each of the flow channels is provided with a temperature sensing component, and the inlet and outlet of each flow channel are respectively provided with a second pressure sensing component;

[0007] a plurality of interdigital electrodes, wherein the interdigital electrodes are disposed in the microchannel heat exchanger, and the interdigital electrodes are disposed in a one-to-one correspondence with the flow channels;

[0008] A swing assembly includes a first magnetic component and a plurality of swing components, wherein the swing components are arranged in a one-to-one correspondence with the flow channels;

[0009] A local early warning module applies an electric field to the flow channel through the interdigital electrodes to induce electroosmotic flow, and controls the first magnetic component to be intermittently energized to drive the swing component to swing up and down in the flow channel.

[0010] The heat exchange regulation system for a battery pack according to the embodiment of the present application has at least the following beneficial effects:

[0011] The heat exchange regulation system for battery packs of the present application includes a microchannel heat exchanger, a plurality of interdigitated electrodes, a swinging assembly and a local early warning module, and the interdigitated electrodes and the swinging assembly are arranged in the microchannel heat exchanger. The microchannel heat exchanger includes an inlet end, an outlet end and a plurality of partitions arranged along the length direction of the microchannel heat exchanger, and a flow channel is formed between two adjacent partitions. The interdigitated electrodes are arranged in a one-to-one correspondence with the flow channel. The local early warning module applies an electric field to the corresponding flow channel through each interdigitated electrode to induce electroosmosis, thereby removing bubbles near the interdigitated electrodes. The swinging assembly includes a first magnetic component and a plurality of swinging components. The swinging components are arranged in a one-to-one correspondence with the flow channel. The local early warning module controls the intermittent power supply of the first magnetic component to drive each swinging component to swing up and down in the corresponding flow channel, driving the fluid in the flow channel to form a local vortex, and the shear force of the vortex peels off the tiny bubbles attached to the wall of the flow channel. The superposition of the mainstream velocity generated by electroosmosis and the local velocity caused by the oscillating motion creates a composite flow of mainstream and eddy currents, accelerating the migration of bubbles toward the channel outlet. This significantly reduces bubbles in the channel, achieving efficient removal of all-size bubbles, with a total removal rate of ≥95%, and significantly improving local thermal resistance. A first pressure sensor is installed at the inlet and outlet of the microchannel heat exchanger to collect pressure measurements at each end, generating a pressure difference between the two ends. This pressure difference indirectly reflects the degree of blockage in the microchannel heat exchanger. Each channel is equipped with a temperature sensor to detect the temperature of the corresponding channel, indirectly reflecting the average temperature of the fluid in the channel. Second pressure sensors are installed at the inlet and outlet of each channel to calculate the pressure difference between the two sides. An abnormal increase in this pressure difference directly reflects the degree of bubble aggregation in that channel. The channel number, temperature, inlet pressure, and outlet pressure values ​​are simultaneously recorded at each timestamp, forming a four-dimensional time series dataset of time, location, temperature, and pressure, providing basic data for subsequent training of local early warning modules or other modules and for anomaly diagnosis.

[0012] According to some embodiments of the present application, the fixed end of the swing component is arranged at one end of the flow channel, the free end of the swing component is arranged at the other end of the flow channel, and the first magnetic component is arranged at the free end of the swing component.

[0013] According to some embodiments of the present application, the swing assembly further includes a second magnetic component, and the first magnetic component and the second magnetic component are respectively arranged on the upper and lower sides of the microchannel heat exchanger to drive the swing component to reciprocate in the flow channel by alternating power supply.

[0014] According to some embodiments of the present application, the inlet end and the outlet end are further provided with pressure pulsation sensing components respectively.

[0015] According to some embodiments of the present application, the system state variable of the microchannel heat exchanger is defined as the state vector of the control sequence at time k ,in are the flow channel temperature value, the inlet side pressure value, and the outlet side pressure value at time k respectively;

[0016] The system control input vector of the microchannel heat exchanger is: , = ,in are respectively the electric field intensity applied by the interdigital electrodes at time k, the swing frequency of the swing component, and the swing amplitude of the swing component, and T represents transposition.

[0017] According to some embodiments of the present application, the local early warning module includes a DRL adaptation layer and a LAMPC execution layer, the DRL adaptation layer formulates an optimal dynamic set point for the MPC controller in the LAMPC execution layer, and the LAPMC execution layer uses the weight vector and reference state provided by the DRL adaptation layer as targets to solve the optimization problem and output an optimal control sequence;

[0018] The input of the DRL adaptation layer is the current state ;

[0019]

[0020] in, is the pressure value at the inlet side, is the pressure value at the outlet side, and T is the temperature value of the flow channel;

[0021] The output of the DRL adaptation layer is the weight vector and reference state ;

[0022]

[0023]

[0024] in, is the pressure difference between the inlet side and the outlet side The error weight, is the error weight of the temperature value of the flow channel, is the reference state of the pressure value on the inlet side, is the reference state of the pressure value on the outlet side, is the reference state of the temperature value of the flow channel; wherein ;

[0025] The DRL adaptation layer is implemented by the reward function By penalizing the error, the system parameters are made to minimize the error of the control target;

[0026]

[0027] The DRL adaptation layer is based on the current state Adjust the weight vector , thereby optimizing long-term rewards, when the system feedback produces a new current state When the policy network of the DRL adaptation layer is updated and a new weight vector is calculated according to the current policy ;

[0028] The LAMPC execution layer receives the weight vector output by the DR1 adaptation layer and reference state , and according to the current state at time k Setting the optimization objective function :

[0029]

[0030] in, 、 、 are the inlet pressure value, outlet pressure value, and flow channel temperature value tracked at the future i-th step at k moments, N is the prediction time domain length of the LAMPC execution layer, To control the incremental penalty coefficient, Changes in control inputs to the fluid dynamics equations;

[0031] The LAMPC execution layer outputs the optimal control sequence ; Where N is the prediction time domain length of the MPC controller, , are the electric field intensity applied under the control timing at time k, the swing frequency of the swing component, and the swing amplitude of the swing component.

[0032] According to some embodiments of the present application, the inlet side pressure value , the outlet pressure value and the temperature value By frequency Smoothing estimation through Kalman filter;

[0033] The DRL adaptation layer is used to decide the frequency Output weight vector With reference state , and every interval The new weights and new reference states are sent to the MPC controller in seconds. The DRL adaptation layer captures long-term changes and adaptively corrects the control target of the LAMPC execution layer.

[0034] The LAMPC execution layer operates at a frequency The optimization problem is solved online to generate an optimal control sequence of the electric field E of the interdigital electrodes, the swing frequency f and the swing amplitude θ of the swing component, and the first control sequence in the optimal control sequence is executed.

[0035] According to some embodiments of the present application, the electric field strength of the interdigitated electrodes is ; Electroosmotic flow duration ; The swing frequency of the swing component ; The swing amplitude of the swing member ;

[0036] When multiple flow channels alarm at the same time, the system determines the priority based on the following formula:

[0037]

[0038] in, 、 are preset weights, is the temperature gradient, is the pressure difference deviation; The largest value means that the flow channel has the highest priority; the highest priority flow channel is applied with 100% electric field strength, the secondary priority flow channel is applied with 50% electric field strength, and the remaining flow channels are applied with the basic electric field;

[0039] The defoaming effect is evaluated by:

[0040] Temperature drop rate, if within 30s A drop of ≥10°C indicates that the bubble removal is effective;

[0041] If the pressure difference decreases, A decrease of ≥20% indicates that the flow resistance of the flow channel is reduced and the bubble volume fraction is reduced;

[0042] If the temperature gradient is uniform, The temperature dropped from ≥10℃ to ≤5℃, indicating that the local high temperature area has been eliminated.

[0043] According to some embodiments of the present application, a cloud-based verification module is also included; the inlet side pressure value, the outlet side pressure value and time form a pressure difference time series, and the flow channel temperature value and time form a temperature time series. The cloud-based verification module analyzes the pressure difference time series and the temperature time series, and combines the historical data of the local early warning module to determine the risk level of each of the flow channels, and instructs the local early warning module to perform heat exchange adjustment according to the risk level.

[0044] According to some embodiments of the present application, when the cloud-based verification module outputs a risk probability coefficient ≥ 0.8, the local early warning module marks the flow channel as high risk and sends an intervention request to the cloud-based verification module;

[0045] If the risk level is determined to be medium or high, the local early warning module is triggered to start the interdigital electrodes and the swing assembly corresponding to the flow channel;

[0046] During the electroosmotic flow and oscillation, the system continuously monitors ;

[0047] If the foam is removed within 30 seconds, Drop ≥ 10℃ and If it drops by ≥20%, it is determined that the defoaming is successful and the monitoring recovery state is entered;

[0048] If the target is not reached, the electroosmotic flow and the swinging action are triggered repeatedly;

[0049] When the system fails to remove bubbles three times in a row, the fault protection state is triggered; in the fault protection state, if the temperature continues to exceed 200°C, the electroosmotic flow and the first magnetic component are automatically shut down, the fluid inlet is closed and an alarm signal is issued; at the same time, the operating data of the past 30 minutes is uploaded to the cloud verification module. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The present application is further described below with reference to the accompanying drawings and embodiments, wherein:

[0051] Figure 1 This is a structural diagram of a heat exchange regulation system for a battery pack according to an embodiment of the present application;

[0052] Figure 2 for Figure 1 A structural diagram from another angle;

[0053] Figure 3 This is a schematic diagram of the internal structure of a heat exchange regulation system for a battery pack according to an embodiment of the present application;

[0054] Figure 4 for Figure 3 A structural diagram from another angle;

[0055] Figure 5 This is a flow chart of a heat exchange regulation system for a battery pack according to an embodiment of the present application.

[0056] Reference numerals:

[0057] Microchannel heat exchanger 1; partition 11; flow channel 12; interdigital electrode 13; inlet end 14; outlet end 15;

[0058] Swinging component 21; first magnetic component 22; second magnetic component 23. DETAILED DESCRIPTION

[0059] The following describes in detail embodiments of the present application. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application and are not to be construed as limiting the present application.

[0060] In the description of this application, it should be understood that if the terms "center", "middle", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "axial", "radial", "circumferential" and the like appear, the orientation or position relationship indicated is based on the orientation or position relationship shown in the accompanying drawings, which is only for the convenience of describing this application and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on this application. In addition, features defined as "first" and "second" may explicitly or implicitly include one or more of such features. In the description of this application, unless otherwise specified, "multiple" means two or more.

[0061] In the description of this application, it should be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they can refer to fixed connections, detachable connections, or integral connections; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; and internal connections between two components. Those skilled in the art will understand the specific meanings of the above terms in this application based on the specific circumstances.

[0062] Refer to the following Figures 1 to 5 A heat exchange regulation system for a battery pack in an embodiment of the present application is described.

[0063] The inventors of this case discovered that in a microchannel heat exchanger, when the operating conditions become complicated, such as under high heat flux density, although the microchannel heat exchanger has the potential for high heat exchange efficiency, the coupling of its microscale characteristics and extreme heat load will amplify the original problem. The fluid in the microchannel is dominated by surface tension, and tiny bubbles (diameter ≤ 100μm) are easy to nucleate on the wall and quickly merge to form a "steam plug" to block the flow channel. If the steam bubbles generated by boiling cannot be discharged in time, they will gather in the flow channel to form a "gas plug", blocking the flow of the liquid phase and causing local heat exchange interruption. The flow channel blockage rate increases exponentially with the operating time, and eventually causes local overheating failure. For this reason, according to Figures 1 to 4 As shown, a heat exchange regulation system for a battery pack in one embodiment of the present application includes a microchannel heat exchanger 1, a plurality of interdigitated electrodes 13, an oscillating assembly, and a local early warning module. The interdigitated electrodes 13 and the oscillating assembly are arranged in the microchannel heat exchanger 1 to eliminate bubbles in the microchannel heat exchanger 1. The microchannel heat exchanger 1 is used to dissipate heat for the battery pack. The microchannel heat exchanger 1 includes an inlet end 14, an outlet end 15, and a plurality of partitions 11 arranged along the length direction of the microchannel heat exchanger 1. The partitions 11 are parallel to each other and spaced apart. Narrow flow channels 12 are formed between any two adjacent partitions 11. The interdigitated electrodes 13 are arranged in a one-to-one correspondence with the flow channels 12. The local early warning module applies an electric field to the corresponding flow channels 12 through each interdigitated electrode 13 to induce electroosmosis. After the interdigitated electrodes 13 are energized, the electrode surface is negatively charged, and the liquid in the solution Positive ions will accumulate in the double layer and migrate toward the cathode under the action of the electric field force, driving the surrounding liquid to flow toward the cathode, thereby removing bubbles near the electrode. The swing assembly includes a first magnetic component 22 and multiple swing components 21. The swing components 21 are arranged in a one-to-one correspondence with the flow channel 12. The local early warning module controls the intermittent power supply of the first magnetic component 22. Through the bubble removal mechanism of mechanical disturbance, each swing component 21 is driven to swing up and down in the corresponding flow channel 12, thereby driving the fluid in the flow channel 12 to form a local vortex. The vortex intensity is proportional to the swing speed. The shear force of the vortex is used to peel off the tiny bubbles attached to the wall of the flow channel 12. When the swing component 21 swings, it squeezes the flow channel 12, causing the local pressure to rise instantly. The bubble shrinks in volume under high pressure and is more likely to break away from the wall of the flow channel 12 and migrate with the mainstream. By superimposing the mainstream velocity generated by the electroosmotic flow and the local flow velocity caused by the swinging action, a composite flow of mainstream and eddy current is formed, which accelerates the migration of bubbles to the outlet side of the flow channel 12, thereby greatly reducing the bubbles in the flow channel 12, achieving efficient removal of full-size bubbles, with a total removal rate of ≥95%, and significantly improving the local thermal resistance.

[0064] A first pressure sensing component is provided at the inlet end 14 and the outlet end 15 of the microchannel heat exchanger 1 to respectively collect the pressure values ​​at both ends and obtain the pressure difference between the two ends. The change in the pressure difference can indirectly reflect the degree of blockage in the microchannel heat exchanger 1.

[0065] Each flow channel 12 is provided with a temperature sensing component, which can indirectly reflect the average temperature of the fluid in the flow channel 12; each flow channel 12 is provided with a second pressure sensing component on the inlet and outlet sides to collect the pressure value on the inlet side. and the pressure value on the outlet side , calculate the pressure difference between the inlet and outlet sides When bubbles gather, the effective flow area of ​​the flow channel 12 decreases and the fluid flow rate increases, but the viscosity of the bubbles is much lower than that of the fluid, which will cause the local flow resistance to increase significantly; at the same time, the bubbles occupying the space of the flow channel 12 will hinder the flow of the fluid, resulting in pressure accumulation on the inlet side and pressure reduction on the outlet side, so the pressure difference An abnormal increase in pressure directly reflects the degree of bubble aggregation. The flow channel 12 sequence number, temperature, inlet pressure, and outlet pressure values ​​are synchronously recorded at each timestamp, forming a four-dimensional time series dataset of time, location, temperature, and pressure. This provides basic data for subsequent local early warning module or other module training and abnormality diagnosis.

[0066] The heat exchange regulation system for battery packs in this application achieves efficient and stable operation of microchannel heat exchange under high heat flux density through a full closed-loop control strategy that integrates temperature and pressure multi-parameter collaborative sensing, electroosmotic flow and mechanical disturbance collaborative defoaming, intelligent early warning, and dynamic feedback optimization. The electroosmotic flow and oscillating disturbance work together to achieve high bubble removal efficiency, with a total bubble removal rate of ≥95%, controlling the local maximum temperature below the safety threshold, extending the service life and extending the thermal fatigue life from the traditional 5000h to more than 10000h. In addition, through the low-power design of electroosmotic flow and reduced voltage drop, the system energy efficiency ratio is optimized and improved by more than 20%.

[0067] In addition, the inventors of this case also discovered that the size of the microchannel flow channel is comparable to the particle size (10-100μm) of particulate matter (such as metal debris and dust in the refrigerant fluid), which is easily captured and accumulated, resulting in a decrease in the effective cross-sectional area of ​​the flow channel (a decrease of 20%-50%) and a surge in flow resistance (pressure drop increases by more than 1 times). For example, when a fluid containing 10μm particulate impurities flows through a microchannel, the flow channel blockage rate increases exponentially with operating time (the blockage rate reaches 30% after 100 hours), and ultimately causes local overheating failure. The heat exchange regulation system for battery packs of this application can prevent impurities from accumulating in the flow channel 12 by driving the swing component 21 to swing up and down in the flow channel 12, greatly reducing the risk of local overheating failure.

[0068] In some embodiments, a temperature sensing component is provided in each flow channel 12, and the temperature sensing component coincides with the projection of the center of the flow channel 12 on the substrate surface of the microchannel heat exchanger 1, which can indirectly reflect the average temperature of the fluid in the flow channel 12; in addition, the inventors found that the bubble aggregation area will lead to low heat exchange efficiency, causing the temperature gradient in the flow channel 12 to abnormally increase, so a temperature sensing component is provided on the inlet and outlet sides of each flow channel 12, and the axial temperature gradient in each flow channel 12 is calculated by three temperature sensing components in each flow channel 12, which can assist in determining the specific aggregation position of the bubbles.

[0069] In some embodiments, the microchannel heat exchanger 1 is flat; it is made of aluminum alloy or copper alloy, so that it has the characteristics of light weight and high thermal conductivity; the thickness of its substrate is only 0.5 to 1 mm, ensuring that the thin-wall characteristics can achieve rapid heat transfer.

[0070] In some embodiments, the partitions 11 are made of aluminum alloy or copper alloy; the spacing between adjacent partitions 11 is 0.5 to 2 mm. In some embodiments, the partitions 11 and the substrate are integrally formed by 3D printing.

[0071] In some embodiments, the inner wall surface of each flow channel 12 is provided with a microscopic rough structure layer formed by plasma etching, thereby enhancing the surface hydrophilicity and inhibiting the secondary attachment of bubbles after they detach from the wall of the flow channel 12 .

[0072] In some embodiments, the inner surface of each flow channel 12 is provided with a silicon dioxide coating with a thickness of 50 to 100 mm, thereby improving corrosion resistance and extending service life.

[0073] In some embodiments, the interdigitated electrodes 13 are interdigitated structures and are integrated on the surface of the inlet substrate through a photolithography process.

[0074] In some embodiments, the surface of the interdigitated electrode 13 is provided with The composite coating has both high corrosion resistance and low hydrogen evolution overpotential.

[0075] In some embodiments, each interdigitated electrode 13 is connected to an external power supply module via a flexible circuit board, and the electric field strength of each flow channel 12 can be independently adjusted.

[0076] In some embodiments, the temperature sensing component is a temperature sensor, and the first pressure sensing component and the second pressure sensing component are both pressure sensors.

[0077] according to Figure 3 and Figure 4As shown, in one embodiment of the present application, the fixed end of the swinging member 21 is disposed at one end of the flow channel 12, and the free end of the swinging member 21 is disposed at the other end of the flow channel 12. The first magnetic member 22 is disposed near the free end of the swinging member 21 and is used to magnetically attract the free end of the swinging member 21, driving the free end of the swinging member 21 to swing relative to the fixed end, so that the swinging member 21 can swing up and down relative to the flow channel 12.

[0078] In some embodiments, the first magnetic component 22 is disposed outside the microchannel heat exchanger 1 .

[0079] In some embodiments, the fixed end of the swing component 21 is fixed to the side wall of the flow channel 12 through a micro hinge, and the free end of the swing component 21 extends to the middle and rear area of ​​the flow channel 12, and the free end can swing up and down around the hinge axis.

[0080] In some embodiments, the fixed end of the swing component 21 is fixed to the side wall of the flow channel 12 by welding, and the free end of the swing component 21 extends to the middle and rear area of ​​the flow channel 12, and the free end can swing up and down around the fixed point.

[0081] according to Figures 1 to 4 As shown, in one embodiment of the present application, the swing assembly further includes a second magnetic component 23, and the first magnetic component 22 and the second magnetic component 23 are respectively arranged on the upper and lower outer sides of the microchannel heat exchanger 1. The first magnetic component 22 and the second magnetic component 23 are alternately energized to generate an alternating magnetic field, driving the swing component 21 to periodically reciprocate in the flow channel 12. When the first magnetic component 22 is energized, the swing component 21 swings toward the side of the first magnetic component 22, and when the second magnetic component 23 is energized, the swing component 21 swings to the opposite side, thereby squeezing the bubbles in the flow channel 12.

[0082] In some embodiments, the first magnetic component 22 and the second magnetic component 23 are both electromagnets. In some embodiments, the first magnetic component 22 and the second magnetic component 23 are respectively disposed at the outlet end 15 of the microchannel heat exchanger 1. Both are elongated and disposed along the width direction of the microchannel heat exchanger 1, corresponding to the upper and lower sides of the free end of the swinging component 21, respectively, so that the free end of the swinging component 21 periodically reciprocates in the flow channel 12.

[0083] In some embodiments, the swinging member 21 is a swinging bar having a sheet-like elastic structure. In some embodiments, a pulsed DC power supply strategy is employed, with a current of 100 to 500 mA. When the first magnetic member 22 is energized, the free end of the swinging member 21 is subjected to a magnetic force, causing the free end of the swinging member 21 to swing toward the first magnetic member 22. When the power is removed, the swinging member 21 returns to its initial position under its own elastic force. When the second magnetic member 23 is energized, the free end of the swinging member 21 swings to the opposite side, and this cycle repeats, squeezing out bubbles.

[0084] In some embodiments, the first magnetic component 22 and the second magnetic component 23 are both fixed to the outside of the microchannel heat exchanger 1 by epoxy resin glue.

[0085] In one embodiment of the present application, the inlet end 14 and the outlet end 15 are also provided with pressure pulsation sensing components, respectively, to detect the pressure pulsation at the inlet end 14 and the pressure pulsation at the outlet end 15 of the microchannel heat exchanger 1, thereby identifying the pressure fluctuation caused by the bubbles leaving the flow channel 12, and providing a data basis for the subsequent local early warning module.

[0086] according to Figures 1 to 5 As shown, in one embodiment of the present application, according to the structural characteristics of the microchannel heat exchanger 1, the system state variable of the microchannel heat exchanger 1 is defined as the state vector of the control sequence of the system at time k: , select measurable macro variables, so choose ;

[0087] in is the temperature value of the flow channel 12 at time k, is the inlet pressure value at time k, is the outlet pressure value at time k.

[0088] It is an important derivative index that reflects the flow resistance of the fluid in the flow channel 12 and is closely related to the presence of bubbles.

[0089] According to the heat exchange regulation working mechanism of the microchannel heat exchanger 1, the system control input vector of the microchannel heat exchanger 1 can be defined as: , = ;

[0090] in, is the electric field strength applied by the interdigital electrode 13 at time k, is the oscillation frequency of the oscillating component 21 at time k, is the swing amplitude of the swing member 21 at time k, and T represents transposition.

[0091] The inventors of this case found that for the microchannel heat exchanger 1, its internal heat exchange regulation process is a multi-physical field coupled environment. The temperature, pressure difference, bubble aggregation area, electric field strength applied by the interdigital electrode 13, and the swing frequency and amplitude of the swing component affect each other. For example, local overheating aggravates phase change blockage, which in turn causes a surge in pressure drop. Therefore, the primary goal is to reduce the temperature Maintain the desired target temperature The second is to eliminate the bubbles inside, but the bubbles are difficult to measure directly, and can only be measured through the macroscopic phenomenon caused by the bubbles, that is, the pressure difference of the flow channel 12 caused by the accumulation of bubbles. Abnormal increase and violent pressure fluctuations are generated to indirectly measure and control; finally, under the premise of meeting performance requirements, it is necessary to reduce control energy consumption as much as possible and reduce The amplitude of .

[0092] Because the heat transfer regulation process inside the microchannel heat exchanger 1 is a multi-physics coupled environment, its accurate first-principles model is extremely complex and difficult to use for real-time control. The system can be abstractly described as a nonlinear discrete-time system:

[0093]

[0094] in, is an unknown nonlinear function, represents the complex dynamics of the system, represents the process noise and unmodeled dynamics in the system. The challenge for this nonlinear discrete-time system is The strong nonlinearity (such as the instability of the phase change process of the microchannel heat exchanger 1) and time-varying characteristics (such as the change in the degree of blockage of the flow channel 12 in the microchannel heat exchanger 1) of the microchannel heat exchanger are taken into consideration. To this end, a local warning module of a hierarchical coupling control framework is designed and constructed. The local warning model framework is a hybrid control framework that is tightly coupled with layered adaptive model predictive control (LAMPC) and deep reinforcement learning (DRL). Figures 1 to 5 As shown, in one embodiment of the present application, the local warning module includes an upper-layer DRL adaptation layer and a lower-layer LAMPC execution layer. The DRL adaptation layer is implemented based on deep reinforcement learning. It works on a slower time scale and is responsible for examining the system state from a macro perspective and formulating the optimal working instructions, i.e., dynamic set points, for the MPC controller in the LAMPC execution layer. The LAPMC execution layer uses the weight vector and reference state provided by the DRL adaptation layer as targets to solve the optimization problem and output the optimal control sequence. The DRL adaptation layer is responsible for long-term performance optimization, and the LAMPC is responsible for real-time trajectory tracking and constraint processing. The two are tightly coupled and work together.

[0095] The input of the DRL adaptation layer is the current state ;

[0096]

[0097] in, is the pressure value at the inlet side, is the pressure value at the outlet side, and T is the temperature value of the flow channel 12.

[0098] The output of the DRL adaptation layer is the weight vector and reference state ;

[0099]

[0100]

[0101] in, is the pressure difference between the inlet and outlet sides The error weight, is the error weight of the temperature value of flow channel 12, is the reference state of the pressure value on the inlet side, is the reference state of the pressure value on the outlet side, is the reference state of the temperature value of the flow channel 12; .

[0102] The DRL adaptation layer uses the reward function By penalizing the error, the system parameters are made to minimize the error of the control target; the reward function for:

[0103]

[0104] The DRL adaptation layer is trained using a deep reinforcement learning algorithm (such as DDPG, TD3, or SAC) to learn how to adapt to the current state. Adjust the weight vector , thereby optimizing long-term rewards, when the system feedback produces a new current state When , the policy network of the DRL adaptation layer is updated and a new weight vector is calculated according to the current policy .

[0105] The temperature sensing component and the second pressure sensing component periodically collect the current state of the flow channel 12 , transmit the sensor data to the MPC controller to construct the current state.

[0106] In each control cycle, the LAMPC execution layer receives the weight vector from the DRL adaptation layer. and reference state , and according to the current state at time k Set the optimization objective function of the LAMPC execution layer :

[0107]

[0108] in, is the pressure value of the inlet port 14 tracked at the next i-th step at k moments, is the pressure value of the outlet port 15 tracked at the next i-th step at k moments, is the temperature value of the flow channel 12 tracked at the future i-th step at k moments, N is the prediction time domain length of the LAMPC execution layer, To control the incremental penalty coefficient, Changes in control inputs to complex fluid dynamics equations;

[0109] The LAMPC execution layer outputs the optimal control sequence ;

[0110] Where N is the prediction time domain length of the MPC controller, , is the electric field strength applied under the control timing at time k, is the oscillation frequency of the oscillating component 21 under the control timing at time k, is the swing amplitude of the swing component 21 under the control timing at time k.

[0111] The LAMPC execution layer solves the problem through quadratic programming (QP) or interior point method. Under real-time requirements, a dedicated embedded QP solver can be used. The LAMPC execution layer solves the above optimization objective function The optimization problem is used to determine the next control input = ;in, is the electric field strength applied by the interdigital electrode 13 at time k, is the oscillation frequency of the oscillating component 21 at time k, is the swing amplitude of the swing member 21 at time k.

[0112] according to Figure 5 As shown, in one embodiment of the present application, the pressure value at the inlet end 14 is , outlet port 15 pressure value and temperature values Can be frequency The smoothing estimation is performed by Kalman filter.

[0113] DRL adaptation layer to make decisions based on frequency Through the strategic network Output the weight vector of the LAMPC execution layer With reference state ,in, Typically low, such as once a day or every few cycles of the LAMPC execution layer, and every Seconds will be the new weight With the new reference state The MPC controller is sent to the LAMPC execution layer, while the DRL adaptation layer is responsible for capturing long-term changes and adaptively correcting the control target of the LAMPC execution layer.

[0114] The LAMPC execution layer runs at a frequency The weight optimization problem is solved online to generate the optimal control sequence of the electric field E of the interdigital electrode 13, the swing frequency f and the swing amplitude θ of the swing component 21, that is, in each control cycle 1 / Internally, solve the optimization problem and output the action = , and sent to the interdigital electrodes 13 and the swing component to perform, wherein, is the electric field strength applied by the interdigital electrode 13 at time k, is the oscillation frequency of the oscillating component 21 at time k, is the swing amplitude of the swing component 21 at time k. The LAMPC execution layer only takes the first action of the optimal control sequence and sends it to the actuator as the actual control signal.

[0115] After the current control cycle ends, the LAMPC execution layer updates the state based on the feedback (pressure, temperature) from the sensing components (including the first sensing component, the second sensing component, and the temperature sensing component). The DRL adaptation layer accumulates experience through online interaction and simulation environment, and gradually adjusts the policy network. , optimize the reward function The LAMPC execution layer adjusts control inputs through real-time feedback to ensure stable system operation and minimize error. Therefore, during training, the DRL adaptation layer and the LAMPC execution layer have different update frequencies. The DRL adaptation layer's low-frequency updates avoid interfering with the LAMPC execution layer's frequent adjustments. The LAMPC execution layer quickly responds to the immediate state of flow channel 12, ensuring stable short-term control results.

[0116] Therefore, the local early warning module framework adopts the hierarchical coupling control structure of the DRL adaptation layer and the LAMPC execution layer. The lower LAMPC execution layer solves the optimal control variables by updating the prediction model online to adapt to the changes in working conditions and system parameter drifts. The upper DRL adaptation layer continuously optimizes its strategy through continuous interaction and learning with the environment, and configures execution parameters for the LAMPC execution layer. It can ensure that the LAMPC execution layer remains effective in the long term when the system working conditions and environment undergo fundamental changes, such as changes in heat transfer characteristics caused by channel aging, thereby realizing online iterative optimization and avoiding misjudgment caused by changes in the working conditions of a single flow channel.

[0117] In some embodiments, constraints can also be imposed on the local warning module framework, including constraints on the electric field strength. , swing amplitude Upper and lower limits and an integrated collision detection mechanism prevent the swing component from physically colliding with the flow channel 12.

[0118] according to Figures 1 to 5 As shown, in one embodiment of the present application, the electric field strength of the interdigitated electrodes 13 is Through many experiments, it was found that when the electric field strength of the interdigital electrode 13 When the electric field strength E is too high, it may cause water decomposition, produce gas byproducts, and increase the risk of clogging. Therefore, the voltage V needs to be limited to within the range.

[0119] Electroosmotic flow duration Through many experiments, it was found that when the ms, will lead to insufficient electroosmotic flow rate and inability to effectively loosen bubbles; if it lasts for a long time ms, which may cause the local temperature of the fluid to drop, or the surface of the interdigital electrode 13 to accelerate aging due to continuous discharge, thereby reducing the service life.

[0120] Oscillation frequency :when When the bubble removal efficiency ;when When it is too low, the risk of bubbles reattaching after peeling will be greatly increased; when When it is too high, the pressure drop caused by the swing increases, affecting the energy efficiency of the system.

[0121] Swing amplitude : ;when hour, Reaching peak value; when When the force is too large, the collision between the swing component 21 and the wall of the flow channel 12 is aggravated, resulting in wear and tear, and reducing the service life.

[0122] Swing period Electroosmotic flow time Synchronize to ensure that after the electroosmotic flow loosens the bubbles, the swinging immediately squeezes them out to prevent the bubbles from re-aggregating.

[0123] The local early warning module makes different controls according to different working conditions in the microchannel heat exchanger 1:

[0124] (1) Normal working condition. When bubbles are not gathered, the pressure difference in each flow channel 12 is Stable in a small range, temperature value Uniform distribution, temperature gradient ≤5℃;

[0125] (2) Bubble aggregation condition. Pressure difference in some flow channels 12 Significantly increased local temperature Abnormally elevated (eg , set the threshold ; and the temperature gradient Increase, such as ).

[0126] The inventors of this case also discovered that in a microchannel heat exchanger, multiple flow channels are usually in different heat flux densities and flow states. When multiple flow channels are simultaneously alarmed by bubble blockage, increased pressure difference, or abnormal temperature, if a unified regulation method is adopted (such as starting the electroosmotic flow and the swing component at full power at the same time), it will not only cause the overall power load of the system to be too large, but may also cause problems such as waste of resources, excessive intervention in low-risk flow channels, and insufficient response of high-risk flow channels. For this reason, when multiple flow channels 12 alarm at the same time, the system adopts a priority and power allocation strategy to avoid a surge in total power consumption, and determines the priority based on the following formula:

[0127]

[0128] in, 、 are preset weights, is the temperature gradient of flow channel 12, is the pressure difference deviation of the flow channel 12, i represents the sequence number of the flow channel 12; The highest value indicates that the flow channel 12 is at the highest priority. The highest priority flow channel 12 is applied with 100% electric field strength, the secondary priority flow channel 12 is applied with 50% electric field strength, and the remaining flow channels 12 are applied with the basic electric field to evenly distribute the flow.

[0129] After the interdigital electrodes 13 are activated, electroosmotic flow is induced and the defoaming effect is evaluated by the following indicators:

[0130] Temperature drop rate, if the local temperature value within 30s A drop of ≥10°C indicates that the bubble removal is effective;

[0131] Pressure difference reduction range, if the pressure difference deviation of flow channel 12 A decrease of ≥20% indicates that the flow resistance of flow channel 12 is reduced and the bubble volume fraction is reduced;

[0132] The temperature gradient is uniform. If the temperature gradient of the flow channel 12 The temperature dropped from ≥10℃ to ≤5℃, indicating that the local high temperature area has been eliminated.

[0133] according to Figure 5As shown, in one embodiment of the present application, the heat exchange regulation system for the battery pack further includes a cloud verification module. The inlet side pressure value, the outlet side pressure value and time form a pressure difference time series , the temperature value and time of flow channel 12 form a temperature time series , time series of operating parameters of multiple local warning modules ,

[0134] The cloud verification module is trained on a higher computing power platform and uses an LSTM neural network to capture the time correlation of temperature series and analyze the pressure difference time series. , temperature time series and operating parameter time series , and combined with the historical data of the local early warning module, determine and output the risk level of each flow channel 12 (such as high risk, medium risk, low risk), and instruct the local early warning module to adjust the heat exchange according to the risk level, as the basis for triggering the local early warning module to work.

[0135] In some embodiments, the cloud verification module is updated collaboratively with the local warning module through a federated learning framework, and the local warning module uploads the encrypted feature data to the cloud verification module.

[0136] according to Figure 5 As shown, in one embodiment of the present application, when the cloud verification module outputs a risk probability coefficient ≥ 0.8, the cloud verification module determines that the flow channel 12 is high risk, the local early warning module marks the flow channel 12 as high risk, and sends an intervention request to the cloud verification module;

[0137] If the cloud verification module determines that the risk level is medium or high, the local warning module is triggered to start the interdigital electrodes 13 and the swing assembly corresponding to the flow channel 12;

[0138] During the electroosmotic flow and oscillation, the system continuously monitors the inlet pressure value of the flow channel 12. , outlet pressure value , temperature value ;

[0139] Within 30 seconds of defoaming, if the temperature Drop ≥10℃ and pressure difference If it drops by ≥20%, it is determined that the defoaming is successful and the monitoring recovery state is entered;

[0140] If the target is not reached, the electroosmotic flow and the swinging action are triggered repeatedly;

[0141] When the system fails to remove bubbles three times in a row, the fault protection state is triggered. In the fault protection state, if the temperature continues to exceed 200°C, the electroosmotic flow and the first magnetic component 22 are automatically shut down, the fluid inlet is closed, and an audible and visual alarm signal is issued. At the same time, the operating data of the past 30 minutes is recorded and uploaded to the cloud verification module for fault diagnosis.

[0142] In some embodiments, normal operation data is uploaded to the cloud verification module through the communication module for model iteration; fault data is marked as abnormal samples for robustness training of the local early warning module.

[0143] The heat exchange regulation system for battery packs of the present application can construct a four-dimensional time series data set of time-position-temperature-pressure of each flow channel 12 through multiple temperature sensing components and a second pressure sensing component under high heat flux density conditions, and perform real-time perception of the internal temperature and pressure status of each flow channel 12, thereby realizing intelligent early warning at the early stage of bubble aggregation. By setting the interdigitated electrodes 13 and the swinging parts 21 in each flow channel 12, they can be activated independently as needed. The interdigitated electrodes 13 apply a DC electric field and use the double electric layer effect formed between the liquid and the wall to directly drive the fluid, realize DC electric field induced electroosmosis, and finely and quickly adjust the local flow velocity of the microchannel heat exchanger 1. By controlling the alternating current of the magnetic component, the swinging component is driven to generate mechanical vibrations of a specific frequency and amplitude. This vibration can effectively disturb the flow field, forcibly peel off the bubbles attached to the wall, enhance fluid mixing, promote heat transfer, and remove bubbles and restore the smoothness of the flow channel 12 through the synergistic effect of electroosmosis and mechanical disturbance, thereby achieving efficient removal of full-size bubbles, with a total removal rate of ≥95%, ensuring that the local thermal resistance is rapidly reduced and the temperature gradient is homogenized. To address the nonlinear, time-varying characteristics of the complex microscale fluid dynamics and extreme heat loads coupled within the microchannel heat exchanger (1), a hierarchical hybrid control strategy is implemented, tightly coupling the DRL adaptation layer with the LAMPC execution layer. This creates a multi-level, differentiated control mechanism. The DRL adaptively adjusts the control targets and weights of the interdigitated electrodes (13) and the oscillating assembly. The LAMPC execution layer solves the optimization problem online and generates high-quality optimal control sequences for electric field intensity, oscillation frequency, and amplitude. This system coordinates the three control inputs, electric field, oscillation frequency, and amplitude, each with distinct response characteristics and mechanisms, to ultimately stabilize the temperature. This strategy boasts excellent prediction accuracy and nonlinear adaptive capabilities, enabling real-time defoaming and optimizing flow and heat transfer performance. When multiple flow channels (12) experience simultaneous anomalies, the system automatically allocates the electric field in a hierarchical manner, focusing on the high-priority flow channels (12), ensuring overall power consumption is controlled and regulation effectiveness is maximized. A cloud-based verification module analyzes pressure differential time series, temperature time series, operating condition time series, and historical data to determine risk factors and levels, instructing the local early warning module to operate. Furthermore, it can trigger protective actions and upload data if defoaming fails, enabling self-diagnosis and self-evolution throughout the system's lifecycle.

[0144] In the description of this specification, if the reference terms "one embodiment," "some examples," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples" appear, it means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples.

[0145] The above describes the implementation methods of the present application in detail in conjunction with the accompanying drawings, but the present application is not limited to the above implementation methods. Various changes can be made within the knowledge scope of ordinary technicians in the technical field without departing from the purpose of the present application.

Claims

1. A heat exchange regulation system for a battery pack, characterized in that: include A microchannel heat exchanger for dissipating heat for the battery pack; the microchannel heat exchanger includes an inlet end, an outlet end, and a plurality of baffles arranged along the length of the microchannel heat exchanger, with flow channels formed between adjacent baffles. The inlet end and the outlet end are respectively provided with a first pressure sensing component, each of the flow channels is provided with a temperature sensing component, and each of the flow channels is provided with a second pressure sensing component on the inlet side and the outlet side. a plurality of interdigital electrodes, wherein the interdigital electrodes are disposed in the microchannel heat exchanger, and the interdigital electrodes are disposed in a one-to-one correspondence with the flow channels; a swing assembly comprising a first magnetic component, a second magnetic component, and a plurality of swing components, wherein the swing components are arranged in a one-to-one correspondence with the flow channels, the fixed end of the swing component is arranged at one end of the flow channel, and the free end of the swing component is arranged at the other end of the flow channel, the first magnetic component is arranged at the free end of the swing component, and the first magnetic component and the second magnetic component are respectively arranged at the upper and lower sides of the microchannel heat exchanger, so that the swing components are alternately energized to drive reciprocating motion in the flow channel; A local early warning module applies an electric field to the flow channel through the interdigital electrodes to induce electroosmotic flow, and controls the first magnetic component to be intermittently energized to drive the swing component to swing up and down in the flow channel.

2. The heat exchange and regulation system for a battery pack according to claim 1, characterized in that: The inlet end and the outlet end are also provided with pressure pulsation sensing components respectively.

3. The heat exchange regulation system for a battery pack according to claim 1, characterized in that: The system state variable of the microchannel heat exchanger is defined as the state vector of the control sequence at time k ,in are the temperature value of the flow channel, the pressure value of the inlet side, and the pressure value of the outlet side at time k, respectively; The system control input vector of the microchannel heat exchanger is: , = ,in are respectively the electric field intensity applied by the interdigital electrodes at time k, the swing frequency of the swing component, and the swing amplitude of the swing component, and T represents transposition.

4. The heat exchange and regulation system for a battery pack according to claim 1, characterized in that: The local warning module includes a DRL adaptation layer and a LAMPC execution layer. The DRL adaptation layer formulates an optimal dynamic set point for the MPC controller in the LAMPC execution layer. The LAMPC execution layer uses the weight vector and reference state provided by the DRL adaptation layer as targets to solve the optimization problem and output an optimal control sequence. The input of the DRL adaptation layer is the current state ; in, is the pressure value at the inlet side, is the pressure value at the outlet side, and T is the temperature value of the flow channel; The output of the DRL adaptation layer is the weight vector and reference state ; in, is the pressure difference between the inlet side and the outlet side The error weight, is the error weight of the temperature value of the flow channel, is the reference state of the pressure value on the inlet side, is the reference state of the pressure value on the outlet side, is the reference state of the temperature value of the flow channel; ; The DRL adaptation layer is implemented by the reward function By penalizing the error, the system parameters are made to minimize the error of the control target; The DRL adaptation layer is based on the current state Adjust the weight vector , thereby optimizing long-term rewards, when the system feedback produces a new current state When the policy network of the DRL adaptation layer is updated and a new weight vector is calculated according to the current policy ; The LAMPC execution layer receives the weight vector output by the DRL adaptation layer and reference state , and according to the current state at time k Setting the optimization objective function : in, 、 、 are the pressure value of the inlet side, the pressure value of the outlet side, and the temperature value of the flow channel tracked at the future i-th step at k moments, N is the prediction time domain length of the LAMPC execution layer, To control the incremental penalty coefficient, Changes in control inputs to the fluid dynamics equations; The LAMPC execution layer outputs the optimal control sequence ; Where N is the prediction time domain length of the MPC controller, , are the electric field intensity applied under the control timing at time k, the swing frequency of the swing component, and the swing amplitude of the swing component.

5. The heat exchange and regulation system for a battery pack according to claim 4, characterized in that: The pressure value at the inlet side , the pressure value on the outlet side and the temperature value of the flow channel By frequency Smoothing estimation through Kalman filter; The DRL adaptation layer is used to decide the frequency Output weight vector With reference state , and every interval The new weights and new reference states are sent to the MPC controller in seconds. The DRL adaptation layer captures long-term changes and adaptively corrects the control target of the LAMPC execution layer. The LAMPC execution layer operates at a frequency The optimization problem is solved online to generate an optimal control sequence of the electric field E of the interdigital electrodes, the swing frequency f and the swing amplitude θ of the swing component, and the first control sequence in the optimal control sequence is executed.

6. The heat exchange and regulation system for a battery pack according to claim 1, characterized in that: The electric field strength of the interdigitated electrodes ; Electroosmotic flow duration ; The swing frequency of the swing component ; The swing amplitude of the swing member ; When multiple flow channels alarm at the same time, the system determines the priority based on the following formula: in, 、 are preset weights, is the temperature gradient, is the pressure difference deviation; The largest value means that the flow channel has the highest priority; the highest priority flow channel is applied with 100% electric field strength, the secondary priority flow channel is applied with 50% electric field strength, and the remaining flow channels are applied with the basic electric field; The defoaming effect is evaluated in the following way: Temperature drop rate, if within 30s A drop of ≥10°C indicates that the bubble removal is effective; If the pressure difference decreases A decrease of ≥20% indicates that the flow resistance of the flow channel is reduced and the bubble volume fraction is reduced; If the temperature gradient is uniform, The temperature dropped from ≥10℃ to ≤5℃, indicating that the local high temperature area has been eliminated.

7. The heat exchange and regulation system for a battery pack according to claim 4, characterized in that: It also includes a cloud-based verification module; the pressure value on the inlet side, the pressure value on the outlet side and the time form a pressure difference time series, and the temperature value and time of the flow channel form a temperature time series. The cloud-based verification module analyzes the pressure difference time series and the temperature time series, and combines the historical data of the local early warning module to determine the risk level of each of the flow channels, and instructs the local early warning module to perform heat exchange adjustment according to the risk level.

8. The heat exchange and regulation system for a battery pack according to claim 7, characterized in that: When the cloud verification module outputs a risk probability coefficient ≥ 0.8, the local warning module marks the flow channel as high risk and sends an intervention request to the cloud verification module; If the risk level is determined to be medium or high, the local early warning module is triggered to activate the interdigital electrodes and the swing assembly corresponding to the flow channel; During the electroosmotic flow and oscillation, the system continuously monitors ; If the foam is removed within 30 seconds, Drop ≥ 10℃ and If it drops by ≥20%, it is determined that the defoaming is successful and the monitoring recovery state is entered; If the target is not reached, the electroosmotic flow and the swinging action are triggered repeatedly; When the system fails to remove bubbles three times in a row, the fault protection state is triggered; In the fault protection state, if the temperature continues to exceed 200°C, the electroosmotic flow and the first magnetic component are automatically shut down, the fluid inlet is closed and an alarm signal is issued; at the same time, the operating data of nearly 30 minutes is uploaded to the cloud verification module.

Images