Board-level thermoelectric co-control system

Abstract

This invention discloses a board-level thermoelectric collaborative control system, belonging to the field of board-level thermoelectric control technology. It includes a heat flux dynamic mapping model construction module, an adaptive fluid disturbance injection module, a phase conjugate velocity compensation control module, a Hamiltonian variational heat diffusion regulation module, and a multi-layer coupled cooling path reconstruction module. The heat flux dynamic mapping model construction module constructs a heat flux density dynamic mapping model based on spatiotemporal multi-field coupling, using real-time acquired temperature gradients, power distributions, and cooling medium velocity vectors within the board as input parameters for multi-physics joint solution. This invention achieves synchronization of cooling flow and heat conduction directions through multi-field coupled heat flux dynamic mapping and adaptive fluid regulation, reducing local thermal resistance and improving thermal stability. Through Hamiltonian variational control and multi-layer cooling path reconstruction, it achieves unidirectional heat flux diffusion and energy self-balancing, significantly improving board heat dissipation efficiency and operational reliability.

Description

Technical Field

[0001] This invention relates to the field of board-level thermoelectric control technology, and more specifically to a board-level thermoelectric coordinated control system. Background Technology

[0002] A board-level thermoelectric co-control system is an intelligent system that comprehensively monitors and dynamically regulates the thermal and electrical interaction effects generated by various types of circuit boards (such as power boards, control boards, and computing boards) within electronic devices under high loads and long-term operation. The system uses the board as the smallest control unit, embedding multi-dimensional sensors for temperature, current, voltage, and power density to collect electrical parameters and thermal field distribution information in real time. It constructs a thermoelectric coupling model to analyze the causal relationships of energy flow, thermal drift, and current surges. Through multi-physics joint solution and feedback algorithms, the system can automatically adjust power distribution strategies, dynamically optimize cooling paths, and adaptively correct the operating thresholds of key components when it detects local hot spots or voltage fluctuations. It not only upgrades traditional passive heat dissipation to active thermal regulation but also achieves integrated electrical control and thermal management, ultimately forming a high-precision, low-power, self-healing thermoelectric dynamic balance system at the board level, significantly improving the stability, energy efficiency, and reliability of equipment in high-density integrated environments.

[0003] The existing technology has the following shortcomings:

[0004] In the process of board-level thermoelectric synergistic control, abrupt changes in heat flux density distribution over time and space are prone to occur during the dynamic cooling regulation stage. When a nonlinear jump occurs in local power load, the flow stability of the cooling medium is disrupted, and a backflow trap structure may form in the flow field, causing some heat flux to reverse and become trapped. This reverse accumulation effect leads to a significant increase in local thermal resistance of the cooling path, and heat that should have been dissipated in time is trapped in microscale channels, thus forming hidden hot spots. As heat continues to accumulate at local nodes, the temperature stress of solder joints, package interfaces, and critical heat conduction channels increases continuously, which can easily lead to failure phenomena such as solder fatigue peeling, microcrack propagation at metal interfaces, and nonlinear drift of conduction impedance, ultimately causing thermal instability or even ablation damage to the device.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a board-level thermoelectric coordinated control system to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a board-level thermoelectric synergistic control system, including a heat flow dynamic mapping model construction module, an adaptive fluid disturbance injection module, a phase conjugate flow velocity compensation control module, a Hamiltonian variational heat diffusion regulation module, and a multi-layer coupled cooling path reconstruction module;

[0008] The heat flux dynamic mapping model construction module constructs a heat flux density dynamic mapping model based on spatiotemporal multi-field coupling. It uses the temperature gradient, power distribution and flow velocity vector of the cooling medium collected in real time inside the board as input parameters to perform multi-physics joint solution and generate a dynamic field distribution matrix that characterizes the microscale flow trajectory and energy return characteristics inside the board.

[0009] The adaptive fluid disturbance injection module establishes an adaptive fluid disturbance injection chain based on the energy return feature region identified in the dynamic field distribution matrix. It applies a pulse disturbance signal to the local flow field of the cooling medium to change the microscale momentum distribution of the cooling medium, thereby eliminating the closed loop of heat flow stagnation identified by the dynamic field distribution matrix and providing dynamic input data for subsequent flow velocity compensation.

[0010] The phase conjugate velocity compensation control module generates a phase conjugate velocity compensation sequence based on the disturbance signal output by the adaptive fluid disturbance injection chain. Using the velocity change rate in the energy return characteristic region as a feedback variable, it performs vectorized phase rotation control on the main channel of the cooling medium, so that the flow direction and the heat conduction direction form a synchronous adjustment baseline.

[0011] The Hamiltonian variational heat diffusion control module, based on the synchronous adjustment baseline formed by the phase conjugate flow velocity compensation sequence, constructs the Hamiltonian variational heat diffusion control function. Using the synchronous characteristics of the flow direction and the heat conduction direction as constraints, it redistributes the heat flux density of the cooling medium inside the board to achieve unidirectional diffusion distribution of energy along the cooling path.

[0012] The multi-layer coupled cooling path reconstruction module performs a multi-layer coupled cooling path reconstruction process based on the output of the Hamiltonian variational heat diffusion control function. It combines the heat flux parameters of unidirectional diffusion distribution to coordinately regulate the driving rate of the cooling medium and the impedance of the board's heat conduction channel, forming a heat flow field structure with reversible equilibrium characteristics, thus suppressing energy backflow and local heat flow accumulation from the source.

[0013] Preferably, the steps for constructing a dynamic mapping model of heat flux density based on spatiotemporal multi-field coupling include:

[0014] In the initial stage of building a spatiotemporal multi-field coupling model, the temperature, power and cooling flow characteristics inside the board under different working conditions are acquired in real time and continuously. Multiple regions are arranged on the board according to the energy density distribution characteristics. Sensing points are set in each region to detect the temperature gradient, power density and the flow velocity vector of the cooling medium, forming multi-dimensional synchronous sampling data.

[0015] Based on the collected temperature gradient, power distribution, and flow velocity vector of the cooling medium, an interactive mapping relationship between the three is established on both spatial and temporal scales. By comparing the consistency between the power distribution and the flow velocity direction, energy transfer paths and potential backflow regions are identified, and a heat flux density distribution matrix that includes energy transfer direction and flow stability is constructed.

[0016] After completing the initial construction of the heat flux density distribution matrix, the microscale flow trajectory and energy return characteristics inside the board are analyzed in detail by combining the dynamic response in the time dimension. By comparing the temperature gradient and flow velocity change trend over a continuous time period, a dynamic field distribution matrix reflecting the energy retention, return and dissipation process is generated.

[0017] Preferably, during the generation of the dynamic field distribution matrix, the velocity vector change trajectory of the energy return characteristic region in the heat flux density distribution matrix is ​​jointly compared with the heat flux direction of the adjacent region. When a locally closed or nearly closed flow loop is detected, the flow loop is identified as an energy return characteristic region, and its spatial expansion range and energy density change trend are recorded in the dynamic field distribution matrix in a time series manner.

[0018] Preferably, the steps for establishing an adaptive fluid disturbance injection chain include:

[0019] After determining the energy return characteristic region, the local flow characteristics of the cooling medium within the energy return characteristic region are identified and the disturbance parameters are initialized. Based on the spatial coordinates and temporal evolution characteristics of the energy return characteristic region in the dynamic field distribution matrix, the flow rewind or stagnation location is located, and the amplitude, duration and frequency of the disturbance injection are determined.

[0020] After completing the location of the energy return characteristic region and the initialization of the disturbance parameters, the local flow field of the cooling medium is acted on by the pulse disturbance signal to redistribute the velocity vector of the stagnant fluid, break the closed rollback path, and dynamically correct the disturbance amplitude and interval according to the time evolution characteristics of the dynamic field distribution matrix to maintain the phase consistency between the disturbance process and the mainstream direction.

[0021] After the pulse disturbance signal ends, the disturbance area is continuously monitored and adjusted. When the flow velocity direction returns to the same direction as the heat flux and the flow velocity amplitude is stable, the disturbance is gradually weakened until it is terminated, thus forming a stable closed-loop disturbance response process.

[0022] Preferably, during the action of the pulsed disturbance signal, the disturbance is preferentially applied at the boundary position of the energy return characteristic region, so that the momentum of the cooling medium is transferred along the mainstream direction, and compensatory disturbances are performed at the key turning points or velocity drop zones of the flow channel to ensure that the local velocity vector and the heat flux direction remain consistent, thereby further enhancing the momentum redistribution effect in the heat flow retention region.

[0023] Preferably, the step of performing vectorized phase rotation control on the main channel of the cooling medium includes:

[0024] After the adaptive fluid disturbance injection chain is completed, the flow state in the disturbance area and its surrounding cooling channels is identified and dynamically quantified. By monitoring the rate of change of the flow velocity of the cooling medium in the energy return characteristic area, the relative phase difference distribution between the flow velocity of the cooling medium and the direction of heat flow is obtained.

[0025] After obtaining the velocity change rate and relative phase difference distribution, a phase conjugate velocity compensation sequence is generated based on the time characteristics of the disturbance signal and the dynamic response relationship of the flow field. This enables the flow direction vector of the cooling medium and the heat conduction direction vector to achieve phase reversal cancellation in time, thereby forming a dynamically synchronized velocity compensation trajectory.

[0026] The phase conjugate velocity compensation sequence is applied to the main channel of the cooling medium, and vectorized phase rotation control is performed on the overall flow direction to make the flow direction consistent with the heat conduction direction in space, thus forming a stable flow field.

[0027] After completing phase rotation control, a synchronous adjustment baseline is established. By continuously monitoring the changes in heat flux density and flow velocity vector, dynamic synchronization between the flow direction and the heat conduction direction is achieved, maintaining the long-term stability of energy transfer.

[0028] Preferably, during the process of establishing the synchronous adjustment baseline, the changes in heat flux density and flow velocity vector at different flow levels in the main channel of the cooling medium are continuously monitored over time. When the rate of change of the two remains linearly corresponding and the directional deviation is less than a preset threshold, it is determined that the flow direction and the heat conduction direction are in a synchronous state. This state is used as the maintenance condition for the synchronous adjustment baseline to ensure that the cooling medium flow field achieves continuous phase locking and stable energy transfer during operation.

[0029] The preset threshold is determined by comprehensively considering the average angle deviation between the cooling medium flow direction and the heat conduction direction under steady-state conditions and the energy transfer stability requirements, and is generally obtained through experimental or simulation calibration.

[0030] Specifically, in typical liquid-cooled or air-cooled circuit board structures, when the cooling medium flow rate is at Within the range, heat flux density is Within the specified range, controlling the angle deviation between the flow direction and the heat conduction direction to within 5° can ensure that the energy transfer efficiency remains above 95%; when the deviation exceeds 10°, the local thermal resistance increases significantly and the heat diffusion delay is obvious.

[0031] Therefore, the present invention preferably sets the preset threshold as directional deviation angle ≤ 5° and linear correlation coefficient between heat flux density and flow rate change rate ≥ 0.95, which are used as the standard for judging the synchronization state of flow direction and heat conduction direction, thereby ensuring that the cooling medium flow field achieves efficient steady-state heat transfer and phase locking in dynamic operation.

[0032] Preferably, the steps for constructing the Hamiltonian variational thermal diffusion control function include:

[0033] Based on the synchronous adjustment baseline formed by the phase conjugate velocity compensation sequence, the correspondence between the flow direction distribution of the cooling medium and the heat conduction direction distribution inside the board is obtained, as well as the angle difference, phase difference characteristics and velocity gradient change trend of each region inside the board.

[0034] After obtaining the synchronous relationship between the flow direction and the heat conduction direction, the basic state of the energy distribution boundary and heat flux density inside the board is determined, a thermal energy distribution map inside the cooling medium is established, and a distribution model with a variable energy gradient is formed.

[0035] After clarifying the synchronization characteristics and energy boundary distribution, the core structure of the Hamiltonian variational heat diffusion control function is established. By dynamically adjusting the energy diffusion path, the heat flux density is redistributed in space.

[0036] The output of the Hamiltonian variational heat diffusion control function is applied to the cooling medium flow field. By adjusting the local flow pressure, channel thermal impedance and cooling velocity distribution, energy is diffused unidirectionally along the cooling path and maintained in a stable manner.

[0037] A dynamic feedback mechanism is established based on the Hamiltonian variational heat diffusion control function. By continuously monitoring the changes in heat flux density and energy diffusion direction, the heat diffusion path can be self-corrected and adaptively adjusted.

[0038] Preferably, in the dynamic feedback mechanism, by real-time monitoring of the change in heat flux density and flow direction deviation of the cooling medium under the constraint of synchronous adjustment baseline, when a new unbalanced region of energy distribution is detected, the heat diffusion direction of the unbalanced region is automatically corrected according to the output result of the Hamilton variational heat diffusion control function, so that the energy transfer of the cooling medium inside the board can be restored to a unidirectional diffusion and stable equilibrium state.

[0039] Preferably, the steps of performing the multi-layer coupled cooling path reconfiguration process include:

[0040] After the Hamiltonian variational heat diffusion control function is output, its heat flux unidirectional diffusion distribution result is spatially mapped to determine the energy transfer state and thermal conductivity of each layer in the cooling medium flow path, and the cooling path is divided into the main heat channel, the auxiliary diffusion channel and the boundary equilibrium channel.

[0041] After obtaining the spatial layered structure of the cooling channels, the driving rate of the cooling medium is adjusted in layers based on the unidirectional diffusion distribution of heat flux, so that the momentum distribution of the cooling medium matches the thermal conductivity of each channel, forming a continuous and balanced energy transfer process.

[0042] After the cooling medium driving rate is adjusted in layers, the impedance of the heat conduction channel of the board is optimized in a coordinated manner according to the heat flux distribution parameters, so that the heat conduction path and the flow characteristics of the cooling medium are coupled and balanced in space.

[0043] After completing the coordinated regulation of driving rate and thermal conductivity impedance, the cooling path is dynamically stabilized and reversible balance is verified, so that the heat flux density change rate in the multi-layer channel tends to be stable, thereby forming a heat flow field structure with adaptive and reversible balance characteristics.

[0044] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0045] This invention constructs a spatiotemporally multi-field coupled dynamic heat flux density mapping model within the circuit board, and based on this, achieves adaptive fluid disturbance and phase conjugate velocity compensation. This ensures that the flow direction of the cooling medium aligns with the heat conduction direction in both time and space, significantly improving the problem of sudden increases in thermal resistance caused by localized heat flow stagnation and energy backflow. By identifying energy backflow characteristic regions in real time and dynamically correcting fluid momentum at the microscale, proactive control can be achieved before heat accumulation, ensuring that the cooling path is always in a state of efficient heat transfer. As a result, the heat flux distribution of the circuit board is more uniform under high power density operation, the local temperature rise is significantly reduced, the risk of thermal drift and device failure is effectively suppressed, and the thermal stability and continuous operation reliability of the system are significantly improved.

[0046] This invention establishes a Hamiltonian variational heat diffusion control function and a multi-layer coupled cooling path reconstruction mechanism to dynamically match the cooling medium driving rate with the impedance of the board's thermal conductive channels, fundamentally achieving unidirectional diffusion of heat flux density and self-balancing regulation of energy flow. Through this collaborative regulation process, heat is dynamically distributed among different thermal conductive layers of the board, and the cooling path structure can be automatically optimized according to changes in heat load, forming a heat flow field with reversible equilibrium characteristics. This heat flow field can maintain consistent energy flow direction and constant conduction efficiency during long-term operation, effectively avoiding material fatigue, solder joint peeling, and interface instability caused by thermal stress accumulation, thus improving the overall energy efficiency and long-term service life of the equipment in complex thermal environments. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0048] Figure 1 This is a schematic diagram of the module of the board-level thermoelectric coordinated control system of the present invention.

[0049] Figure 2 This is a schematic diagram illustrating the working principle of the board-level thermoelectric coordinated control system of the present invention.

[0050] Figure 3 A flowchart for establishing the adaptive fluid disturbance injection chain for this invention.

[0051] Figure 4 This is a flowchart illustrating the vectorized phase rotation control of the main cooling medium channel according to the present invention. Detailed Implementation

[0052] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0053] This invention provides, for example Figures 1 to 4 The board-level thermoelectric collaborative control system shown includes a heat flow dynamic mapping model construction module, an adaptive fluid disturbance injection module, a phase conjugate velocity compensation control module, a Hamiltonian variational heat diffusion regulation module, and a multi-layer coupled cooling path reconstruction module.

[0054] The heat flux dynamic mapping model construction module constructs a heat flux density dynamic mapping model based on spatiotemporal multi-field coupling. It uses the temperature gradient, power distribution and flow velocity vector of the cooling medium collected in real time inside the board as input parameters to perform multi-physics joint solution, generate a dynamic field distribution matrix that characterizes the microscale flow trajectory and energy return characteristics inside the board, and use the dynamic field distribution matrix as the spatiotemporal boundary basis for subsequent fluid disturbance control.

[0055] A dynamic mapping model of heat flux density based on spatiotemporal multi-field coupling is constructed to generate a dynamic field distribution matrix characterizing the microscale flow trajectory and energy reflection characteristics inside the board. The specific steps are as follows:

[0056] In the initial stage of constructing the spatiotemporal multi-field coupling model, the temperature, power, and cooling flow characteristics inside the board under different operating states are acquired in real time and continuously. To this end, several key regions are selected on the board according to the energy density distribution characteristics. Each region covers a typical high-power-density device area, an energy conduction transition zone, and a cooling medium flow channel. Multiple sensing points are set in each region to detect the temperature gradient, power density, and flow velocity vector within that region. The temperature gradient is obtained by deploying miniature thermistors or thermal sensors to measure instantaneous temperature changes at different heights and horizontal distances, thus reflecting the direction and intensity of the temperature gradient within the board's heat conduction path. Power distribution is measured by high-frequency sampling of the current and voltage parameters of each functional unit and obtaining transient power changes through time synchronization, thereby determining the energy input density per unit area in each region. The cooling medium flow velocity vector is determined by setting velocity detection points within the cooling channels and recording instantaneous changes in flow velocity along the flow direction and perpendicular to the flow direction, to obtain the velocity direction and magnitude of the cooling medium as it flows through each hot zone. All sensing points are sampled synchronously using time as a unified benchmark, ensuring strict correspondence between temperature, power, and flow velocity data in the time dimension and avoiding errors caused by time delay. Spatially, the sensing points are arranged in layers along the horizontal, vertical, and thickness directions of the board to achieve three-dimensional sampling coverage. Through long-term continuous acquisition, temperature gradient sequences, power distribution sequences, and flow velocity vector sequences under different operating cycles can be generated, providing high-precision raw data for subsequent multi-field joint solutions.

[0057] After obtaining the complete time series of temperature gradient, power distribution, and cooling medium velocity vector, the interaction mapping relationship between these physical quantities is further established on both spatial and temporal scales, thus forming a continuous distribution field of heat flux density inside the board. In this stage, regions with relatively concentrated energy input on the board are first identified based on the power distribution data, and these regions are used as the starting points for heat transfer. Subsequently, the temperature gradient change trend at the corresponding locations is matched with the cooling medium velocity direction. By comparing the consistency between the temperature rise rate and the cooling flow direction, the continuity of the energy conduction path is determined. When the spatial distribution of power input is relatively uniform and the cooling medium flow direction is consistent with the temperature gradient direction, the energy transfer path is in a stable diffusion state. When the power input density in a certain region suddenly increases, while the cooling medium velocity direction deviates from the temperature gradient direction, it indicates that there is a tendency for energy accumulation in that region. Further time comparison of the temperature gradient change rate and the power input change rate in this region reveals that if the temperature rise lags behind the power surge, it can be inferred that heat has not diffused in time, potentially leading to local energy backflow. To further confirm this backflow trend, the velocity vector of the cooling medium in this region is compared point by point to analyze the changes in the direction of the flow velocity. If the velocity direction reverses or vortexes in time, it indicates that local flow reversal has formed in this region. Through this continuous comparison and matching process, a spatial distribution map with energy transfer direction, heat accumulation intensity, and flow stability as core parameters can be gradually formed. This spatial distribution map clearly depicts the energy transfer path and potential backflow locations within the board. At this point, the preliminary structure of the heat flux density distribution matrix is ​​established. Each spatial cell in the matrix contains a combined data cell consisting of temperature gradient intensity, power input density, and cooling medium velocity vector, used to describe the transient heat flux transfer state at that location.

[0058] After initially constructing the heat flux density distribution matrix, the microscale flow trajectory and energy backflow characteristics within the board are further refined and dynamically characterized by incorporating the dynamic response over time. This stage uses the previously formed distribution matrix as a foundation to analyze the trends of temperature, power, and flow velocity changes over continuous time periods. Specifically, the rate of change of temperature gradient within adjacent time intervals is compared with the rate of change of cooling medium velocity vector to determine the response speed of the cooling flow field. When a region exhibits repeated changes in flow velocity direction or periodic fluctuations in velocity magnitude across multiple time intervals, while the rate of change of temperature gradient in that region continuously increases and fails to keep pace with changes in power input, it indicates the presence of spatial reverse energy accumulation within that region. To further confirm this energy backflow phenomenon, the velocity vector change trajectory of the aforementioned region is jointly analyzed with the heat flux direction of adjacent regions. If a closed or nearly closed flow loop is formed between adjacent regions, that region is identified as an energy backflow characteristic region. The spatiotemporal distribution information of these characteristic regions is marked with high-precision time markers in the dynamic field distribution matrix, thus forming a dynamic energy distribution field containing time evolution characteristics. Based on this, the spatial expansion range, duration, and energy density variation trend of the energy return characteristic region within each time period are recorded one by one to generate a dynamic field distribution matrix reflecting the energy flow behavior inside the board. This dynamic field distribution matrix not only reflects the intensity and direction of heat transfer between different regions, but also reflects the continuous evolution of heat flow retention, return, and dissipation processes within the microchannels. By dynamically updating this dynamic field distribution matrix, the changing trend of energy distribution inside the board can be tracked in real time, potential abnormal heat flow regions can be identified in a timely manner, and spatiotemporal coordinates can be provided for subsequent injection of cooling medium disturbances and adjustment of flow field direction.

[0059] It should be noted that:

[0060] An approximately closed flow loop refers to a situation where, during the local flow of the cooling medium inside a circuit board, although the velocity vector does not form a completely closed loop path in terms of spatial distribution and temporal evolution, the beginning and end regions of the flow direction are highly correlated in terms of geometric position and momentum direction, exhibiting a tendency for energy to roll back and stagnate in the local region.

[0061] Specifically, when the velocity vector of the cooling medium deviates from the mainstream direction in a certain region, it undergoes reverse or lateral flow within a short period of time, and then converges again near the initial flow path in adjacent spatial units. This causes the fluid momentum trajectory to form a nearly closed continuous curve structure in space, while the flow kinetic energy exhibits periodic fluctuations within this loop, preventing heat from being smoothly transferred to the external region. At this point, the inlet and outlet vector angle of the fluid momentum is less than 30°, and the rate of change of velocity exhibits periodic reversal characteristics in the time dimension, indicating that although the local flow has not completely returned to the origin, it has already met the conditions for forming a stable stagnant vortex region. Therefore, this nearly closed state essentially represents a local flow pattern where the energy cycle is not fully open. In a thermodynamic sense, it has similar energy accumulation and delayed heat diffusion effects as a completely closed loop, and is considered a potential energy return characteristic region.

[0062] Through the above steps, the constructed spatiotemporal multi-field coupled dynamic heat flux density mapping model can reflect the energy input and heat diffusion processes of the board at different operating stages in time, and reveal the synergistic relationship between the cooling medium flow path and the heat conduction direction in space. The resulting dynamic field distribution matrix can serve as the boundary input for subsequent cooling control processes, guiding the injection location and direction of cooling medium disturbances, ensuring continuous energy transfer in space and stable energy transfer in time, ultimately providing a control reference for solving energy backflow and local heat accumulation.

[0063] The adaptive fluid disturbance injection module establishes an adaptive fluid disturbance injection chain based on the energy return feature region identified in the dynamic field distribution matrix. It applies a pulse disturbance signal to the local flow field of the cooling medium to change the microscale momentum distribution of the cooling medium, thereby eliminating the closed loop of heat flow stagnation identified by the dynamic field distribution matrix and providing dynamic input data for subsequent flow velocity compensation.

[0064] Based on the energy return characteristic region identified in the dynamic field distribution matrix, an adaptive fluid disturbance injection chain is established. The specific steps are as follows:

[0065] After identifying the energy return characteristic region, the local flow characteristics of the cooling medium within this region are refined and the disturbance parameters are initialized. This step, based on the previously constructed dynamic field distribution matrix, extracts the spatial coordinates and temporal evolution characteristics of the identified energy return characteristic regions in the matrix to pinpoint the specific locations where backflow, stagnation, or momentum decay occurs in the cooling medium flow. To ensure the targetedness and effectiveness of the disturbance injection, a microscale momentum distribution description of the cooling medium flow is established within these regions. This involves continuously monitoring the gradient changes in velocity in all directions to clarify the shear stress distribution and flow energy density state of the local fluid layer. During this process, the spatial boundary of the energy return characteristic region is compared with the temperature gradient change trend of adjacent regions to determine the specific morphology of the heat flow stagnation boundary. If the flow velocity of the cooling medium within this region experiences a sudden drop, a deflection of the flow direction, or the formation of a closed flow loop, this region can be identified as the priority point for the pulsed disturbance signal. Subsequently, based on the combined characteristics of the cooling medium flow rate, temperature gradient intensity, and energy accumulation time within the region, the amplitude, duration, and frequency of the disturbance injection are determined, establishing initial control parameters for the precise injection of subsequent pulse signals. This method allows for the full quantification of the flow field characteristics before disturbance injection, providing an operational basis for the spatial coupling of subsequent disturbance energy.

[0066] After locating the energy return characteristic region and initializing the disturbance parameters, the momentum distribution characteristics of the cooling medium within the target region are directly altered by the application of pulsed disturbance signals, allowing the locally stagnant fluid to regain directional flow driving force. The key to this stage is applying external momentum disturbance to the cooling medium through intermittent pulsed disturbance signals, causing an ordered change in the local fluid velocity field. The pulsed disturbance signal can be understood as a periodic injection of kinetic energy into the flow state of the cooling medium, manifesting as instantaneous pressure changes, momentary flow direction shifts, or local velocity amplitude increases. By applying this disturbance at the boundary of the energy return characteristic region, the velocity vector of the stagnant fluid layer is redistributed, breaking the original closed loop path. Specifically, disturbance signals are applied at key turning points, local expansion zones, or velocity drop zones in the cooling medium flow channel, allowing the disturbance energy to be transmitted along the mainstream direction of the cooling medium. When the pulse signal acts on the boundary of the stagnant zone, the local velocity of the cooling medium is instantaneously increased, and the fluid layer, which was originally in a near-static state, undergoes momentum migration, forming a new flow gradient. At this point, the fluid in the stagnant zone is gradually drawn directionally by the high-speed fluid in the mainstream zone, thus weakening the closed loop of heat flow recirculation. To avoid excessive disturbance leading to instability in the overall flow field, the interval and amplitude of the pulse disturbance are dynamically modified based on the time evolution characteristics in the dynamic field distribution matrix, ensuring that the disturbance process remains phase-consistent with the main flow direction of the cooling medium. After the pulse ends, the local velocity distribution of the cooling medium gradually changes from its original uneven state to a continuous distribution, restoring the spatial continuity of the heat conduction path. Energy diffuses outward from the original recirculation zone, thereby achieving local elimination of heat flow stagnation.

[0067] After the pulsed disturbance signal ends, to ensure that the local flow structure of the cooling medium maintains a stable momentum transfer state, an adaptive feedback adjustment process is established for the post-disturbance flow field to achieve a dynamic balance between the disturbance energy and the cooling medium flow field. In this stage, the recovery effect of the flow field after the pulse is evaluated by continuously monitoring the velocity change trend of the cooling medium in the disturbance area. When the local velocity vector direction recovers to be consistent with the heat flux direction, and the velocity amplitude tends to stabilize over time, it indicates that the disturbance injection has achieved the expected momentum redistribution effect. If the monitoring results show that there are still large velocity fluctuations or energy transfer delays in some areas, small-amplitude compensatory disturbances need to be performed at the same or adjacent locations to further correct the local flow state. During this process, the adaptive fluid disturbance injection chain uses the real-time update results of the dynamic field distribution matrix as feedback. When the matrix shows that the energy reflection characteristics gradually disappear and the heat flux density distribution tends to smooth out, the disturbance injection process gradually weakens until it terminates. Meanwhile, to prevent the generation of new flow instabilities, the boundary layer of the cooling medium's flow channel is continuously observed to ensure that the momentum transfer direction of the fluid remains consistent after pulsed perturbation. Through continuous tracking and feedback correction of the perturbation region, a closed-loop perturbation response process is formed, enabling stable energy transfer of the cooling medium in the microscale channel. In this stage, the momentum distribution of the cooling medium not only regains spatial continuity but also achieves dynamic stability in time. Heat re-exhibits unidirectional diffusion characteristics in the conduction path, and energy backflow is effectively suppressed. Thus, the adaptive fluid perturbation injection chain completes the entire process from energy backflow identification, perturbation signal application to feedback correction, providing continuous and reliable dynamic input data for subsequent velocity compensation and further optimization of flow direction.

[0068] The adaptive fluid disturbance injection chain established through the above implementation steps achieves full-process control from energy backflow feature identification to fluid momentum redistribution. During this process, the flow state of the cooling medium gradually transforms from unsteady stagnation to steady-state transport, and the heat flow channel changes from partially closed to continuously open, allowing heat within the board to regain an orderly transfer path in both time and space. This implementation breaks the closed flow structure formed by localized heat flow backflow, thereby effectively reducing local thermal resistance and achieving synchronous correction of the momentum regeneration and energy flow direction of the cooling medium, providing a dynamic input basis for subsequent phase conjugate velocity compensation.

[0069] The phase conjugate velocity compensation control module generates a phase conjugate velocity compensation sequence based on the disturbance signal output by the adaptive fluid disturbance injection chain. Using the velocity change rate in the energy return characteristic region as a feedback variable, it performs vectorized phase rotation control on the main channel of the cooling medium, so that the flow direction and the heat conduction direction form a synchronous adjustment baseline.

[0070] Vectorized phase rotation control is applied to the main channel of the cooling medium to synchronize the flow direction with the heat conduction direction. The specific steps are as follows:

[0071] After the adaptive fluid disturbance injection chain is completed, the flow state in the disturbance region and its surrounding cooling channels is refined and dynamically quantified to provide a physical basis for generating the phase conjugate velocity compensation sequence. In this stage, the flow response characteristics after the disturbance signal are captured by monitoring the velocity change rate of the cooling medium within the energy return characteristic region. Because the disturbance signal induces transient momentum redistribution in the local flow field, the velocity of the cooling medium exhibits multidirectional changes in a short period, including velocity direction deflection, velocity amplitude oscillation, and shear stress fluctuations between flow layers. Therefore, the evolution trajectory of the velocity change rate over time is continuously recorded in this region to clarify the phase change trend of the cooling medium velocity vector. By comparing these velocity change rate data with the application time series of the disturbance signal, the response delay and phase shift direction of the velocity change can be identified. At this point, the flow state within the energy return characteristic region is transformed into a set of time-dependent vector change characteristics, providing a reference for subsequently establishing the phase conjugate relationship. The key to this stage is to obtain the relative phase difference distribution between the cooling medium flow rate and the heat flow direction, thereby clarifying the dynamic response mode of the cooling medium in the energy return characteristic region.

[0072] After obtaining the velocity change rate and relative phase difference distribution within the energy return characteristic region, a corresponding phase-conjugate velocity compensation sequence is generated based on the temporal characteristics of the disturbance signal and the dynamic response relationship of the flow field. The generation of this compensation sequence is essentially a phase-matching reconstruction process. Its core lies in achieving phase reversal cancellation between the flow direction vector of the cooling medium and the heat conduction direction vector in time, thereby forming a dynamically synchronized velocity compensation trajectory. In specific implementation, the local areas with larger velocity change rates within the energy return characteristic region are selected as priority response points for the compensation sequence to ensure concentrated energy compensation. The velocity change trend at each response point is mapped to a phase reference signal, which is then back-matched with the phase relationship of the disturbance signal to generate a compensation velocity change sequence with opposite phase characteristics. At this point, the mainstream direction vector of the cooling medium gradually begins to form a mirror-symmetric relationship with the phase of the heat diffusion direction, forming a phase-conjugate characteristic. Through continuous time stepping, the velocity vectors of each local region sequentially complete phase alignment, thereby forming an overall flow trend consistent with the heat flow direction in space. This stage not only achieves phase inversion in local areas, but also establishes multi-point synchronous response within the global flow channel, gradually transforming the velocity distribution of the cooling medium from a disordered disturbance state to a stable flow field with unified direction.

[0073] After generating the phase conjugate velocity compensation sequence, this sequence is applied to the main channel of the cooling medium to perform vectorized phase rotation control on the overall flow direction. This stage is the key to transforming the compensation sequence into the actual effect of the flow field, and its goal is to achieve complete synchronization between the cooling medium flow direction and the heat conduction direction through phase rotation. Specifically, based on the phase matching results of each local region in the aforementioned phase conjugate compensation sequence, the velocity vector in the main channel is corrected in direction. This correction is not a simple adjustment of the velocity magnitude, but rather a gradual alignment of the cooling medium flow direction through continuous phase rotation. To ensure a smooth transition in the flow direction, phase rotation operations are performed sequentially on different cross-sections of the main channel, so that the change in flow direction presents a gradual transition from upstream to downstream. As the cooling medium flows along the channel, its local velocity vector will gradually approach the heat conduction direction as the phase rotation angle changes continuously. At the same time, at the boundary of the energy return characteristic region, the momentum direction change generated by phase rotation can effectively counteract the reverse velocity component caused by local backflow, allowing the stagnant fluid to reintegrate into the mainstream direction. In this way, the flow direction of the cooling medium and the heat diffusion path are aligned in space. The kinetic energy and thermal energy transfer in the flow no longer cancel each other out, but instead form a superposition effect in the same direction, thereby improving the heat transfer efficiency and reducing local thermal resistance.

[0074] After completing the phase rotation control of the main channel, a synchronization adjustment baseline is established to ensure that the flow direction and the heat transfer direction remain consistent over time. The establishment of this synchronization adjustment baseline is the final step in the entire phase conjugate velocity compensation sequence, and its function is to achieve dynamic synchronization between the velocity direction and the heat diffusion direction through continuous phase comparison and energy flow monitoring. Specifically, using the velocity vector field after phase rotation control as the initial reference, continuous monitoring is performed at different flow levels of the cooling medium, recording the trends of heat flux density change and velocity vector change. By comparing the rates of change and relative angles of these two parameters, the degree of synchronization between the flow direction and the heat transfer direction can be determined. When the changes in heat flux density and velocity show a linear correspondence and the directional deviation remains within a very small range, it indicates that the flow and transfer have reached a stable synchronous state, and the flow direction at this point constitutes the synchronization adjustment baseline. If phase shifts or energy transfer lags are found in certain local areas during long-term monitoring, the phase can be readjusted through small-amplitude velocity corrections to keep the overall flow field in a synchronized state. This synchronized baseline not only maintains the long-term consistency between the cooling medium flow direction and the heat conduction direction, but also provides a continuous dynamic boundary reference for the subsequent heat diffusion control stage. By establishing the synchronized baseline, the cooling medium flow field achieves phase locking in the time dimension and directional consistency in the spatial dimension, enabling efficient diffusion of heat energy in a single direction within the board, thereby fundamentally improving the dynamic stability and heat transfer efficiency of board-level thermoelectric synergistic control.

[0075] Through the above steps, the phase conjugate velocity compensation process achieves a complete physical closed loop from disturbance signal response to velocity direction synchronization. This implementation uses the dynamic changes of the energy return characteristic region as the feedback core, enabling the cooling medium to quickly restore directional transmission after disturbance, ultimately forming a stable structure where the flow direction is consistent with the heat conduction direction. Through the coordinated adjustment of phase rotation and synchronization baseline, the diffusion path of heat energy inside the board is optimized, and local heat accumulation and energy retention are effectively eliminated, thereby achieving a continuous balance of heat flux density in the spatiotemporal dimensions.

[0076] The Hamiltonian variational heat diffusion control module, based on the synchronous adjustment baseline formed by the phase conjugate flow velocity compensation sequence, constructs the Hamiltonian variational heat diffusion control function. Using the synchronous characteristics of the flow direction and the heat conduction direction as constraints, it redistributes the heat flux density of the cooling medium inside the board to achieve unidirectional diffusion distribution of energy along the cooling path.

[0077] The Hamiltonian variational heat diffusion control function is constructed to redistribute the heat flux density of the cooling medium within the board. The specific steps are as follows:

[0078] In the initial stage of constructing the Hamiltonian variational heat diffusion control function, a synchronous adjustment baseline formed by a phase conjugate velocity compensation sequence is used as the basis to obtain the correspondence between the flow direction distribution of the cooling medium and the heat conduction direction distribution within the plate. The key point of this step is to clarify the angular difference, phase difference characteristics, and velocity gradient change trends between the flow direction and the heat conduction direction in different regions using the direction matching data provided by the synchronous adjustment baseline. By setting spatial sampling points in the cooling channel, the changes in heat flux density and flow direction at each location are compared in real time, forming a dynamic mapping relationship between the flow direction and the heat conduction direction. When the flow direction of the cooling medium and the heat conduction direction are completely synchronized, their energy transfer efficiency is at its maximum; when a deviation occurs, energy accumulation or conduction delay will form locally. Therefore, in this stage, it is necessary to identify which regions are in a synchronized state and which regions have deviations, and use these spatial differences as constraint inputs for subsequent variational control. Through this step, an initial characteristic field reflecting the coupling degree between the flow direction and the heat conduction direction can be constructed, providing spatially distributed boundary conditions for the establishment of the Hamiltonian variational heat diffusion control function.

[0079] After obtaining the synchronization relationship between the flow direction and the heat conduction direction, the basic state of the energy distribution boundary and heat flux density inside the board is determined as the energy input condition for constructing the Hamiltonian variational heat diffusion control function. This step establishes a thermal energy distribution map inside the cooling medium by continuously measuring the temperature distribution, heat capacity change, and local impedance characteristics of the heat conduction channel along the flow path of the cooling medium, based on the synchronous adjustment baseline. In specific implementation, different regions of the board are divided into multiple heat conduction units, each containing a local heat flux input end and an output end, representing the energy flow path in space. By statistically analyzing the energy input and output characteristics of these units, the initial distribution of heat flux density across the entire board can be obtained. Subsequently, using the synchronous region between the flow direction and the heat conduction direction as the constraint boundary, and the asynchronous region with a large rate of change in heat flux density as the key adjustment area, a distribution model with a variable energy gradient is formed. This distribution model describes the flow trend and imbalance characteristics of energy inside the board, providing the initial conditions for constructing the energy potential field in the Hamiltonian variational heat diffusion control function.

[0080] After clarifying the synchronization characteristics and energy boundary distribution, the core structure of the Hamiltonian variational heat diffusion control function is established, enabling dynamic adjustment of the heat flux path of the cooling medium. This step aims at maintaining the continuity of heat flow by normalizing the coupling relationship between the flow direction, heat conduction direction, and energy potential field, forming a constraint system reflecting the energy transfer trend in the path. At this point, the core significance of the Hamiltonian variational heat diffusion control function lies in achieving spatial redistribution of heat flux density through dynamic solution of the energy diffusion path. When the heat transfer rate in a certain region is lower than the heat transfer efficiency in the flow direction of the cooling medium, the Hamiltonian variational heat diffusion control function redefines the heat flux transfer ratio in that region based on the vector deviation of the flow direction and the difference in energy gradient, thereby redistributing heat energy in the path. Simultaneously, within the synchronization region, the Hamiltonian variational heat diffusion control function maintains the consistency of the ratio of heat flux density to velocity gradient, ensuring that the cooling medium maintains a kinetic energy transfer structure consistent with the heat diffusion direction throughout the flow process. Through this process, the energy flow of the cooling medium is no longer driven solely by local temperature differences, but rather forms a controlled thermal diffusion state under constraints on a global scale, thereby coordinating the direction and speed of energy transfer within the board as a whole.

[0081] After constructing the Hamiltonian variational heat diffusion control function, its output is applied to the cooling medium flow field to spatially redistribute and correct the path of heat flux density. This step, based on the energy diffusion trend output by the Hamiltonian variational heat diffusion control function, adjusts the local flow pressure, channel thermal resistance, and cooling velocity distribution along the cooling medium's flow path to unidirectionally distribute the heat flux diffusion. Specifically, for regions with high energy density and conduction direction deviating from the mainstream direction, the heat transfer capacity is increased by raising the local cooling velocity or lowering the thermal resistance; for regions with low energy density but high flow velocity, the cooling velocity is appropriately reduced to balance energy diffusion over time. Through this dynamic adjustment, heat diffuses in a single direction throughout the cooling path, forming a continuous, smooth, and non-reversible energy flow trajectory. Simultaneously, in regions where the flow direction is synchronized with the heat conduction direction, energy transfer efficiency is maintained at its optimal state, allowing the cooling medium to quickly remove excess heat as it passes through high-temperature regions, effectively preventing localized heat accumulation. After this stage of heat flux density redistribution, the energy flow direction in the cooling path gradually stabilizes, and the spatial non-uniformity of the heat diffusion process is significantly reduced.

[0082] Finally, to ensure the long-term stability and reversible adjustability of the thermal diffusion control, a dynamic feedback mechanism is established based on the output of the Hamiltonian variational thermal diffusion control function, ensuring that the thermal diffusion process of the cooling medium maintains a unidirectional orientation throughout continuous operation. This stage involves continuously comparing the control function output with the flow state of the cooling medium to monitor the changing trend of heat flux density and the direction of energy diffusion in real time. When new heat accumulation or flow velocity deviation is detected in a local area, the Hamiltonian variational thermal diffusion control function automatically corrects the thermal diffusion direction based on changes in the synchronous adjustment baseline, restoring the energy diffusion to a unidirectional distribution state. Simultaneously, by continuously monitoring the relative changes in temperature gradient and flow velocity in the cooling channels of the board, the adjustment effect of the Hamiltonian variational thermal diffusion control function can be dynamically verified. If the system maintains a stable balance of heat flux density during continuous operation, it indicates that the Hamiltonian variational thermal diffusion control function has achieved effective thermal diffusion optimization. When external operating conditions change or the power load distribution is adjusted, the control function can recalculate the energy distribution trend under new boundary conditions, automatically adapting the thermal diffusion path of the cooling medium to the new operating state, thereby maintaining the internal thermal balance structure of the board. Through this closed-loop feedback mechanism, the heat diffusion control process has self-correction and self-adaptation capabilities, enabling energy to diffuse in a unidirectional and stable manner throughout the cooling path, forming an efficient, stable, and dynamically reversible heat conduction structure.

[0083] It should be noted that:

[0084] The construction of the Hamiltonian variational heat diffusion control function is based on the dynamic coupling relationship between the flow direction of the cooling medium, the heat conduction direction, and the energy potential field. The optimal control of heat flux density is achieved through the balance redistribution of energy in space and time.

[0085] Specifically, firstly, using the synchronous adjustment baseline formed by the phase conjugate velocity compensation sequence as a reference, the velocity distribution, heat flux density distribution, and heat conduction channel impedance characteristics of the cooling medium inside the board are stratified and quantified. For example, in a typical high-density power board, the inlet velocity of the cooling channel is set to 1.2 m / s, and the average heat flux density inside the channel is approximately... The thermal resistance of the thermal interface is approximately At this point, if an approximately 8° deviation between the flow direction and the heat diffusion direction is detected in a localized high-power region (such as below the core chip), and the heat flux density increases to [a certain value], [the situation will be similarly affected]. If this deviation is not observed, the region is identified as an energy asynchrony zone. The Hamiltonian variational heat diffusion control function uses this deviation as a constraint input during construction and dynamically adjusts the energy distribution ratio between channels based on the changing trends of the cooling medium flow rate and thermal resistance. For example, it increases the flow rate of the dominant channel to 1.6 m / s and reduces the local thermal resistance by 10%, thus redistributing the heat flux to a high-efficiency region. Low Zone This allows for the realignment of the heat diffusion direction. Through the iterative action of this function, the system can maintain a heat flux density change rate within ±5% under different load cycles (such as power fluctuations within 80-120W), and the deviation between the flow direction and the heat conduction direction is less than 3°, ensuring that energy continues to diffuse in a single direction in the cooling path, forming a stable and controllable thermal equilibrium structure.

[0086] Through the above implementation steps, the Hamiltonian variational thermal diffusion control function realizes the complete process from flow direction constraint to energy diffusion optimization. It not only achieves the unification of cooling medium momentum and thermal energy transfer direction at the physical level, but also establishes a dynamically coupled thermal diffusion path in the spatiotemporal dimension. This allows energy transfer inside the board to no longer rely on passive heat conduction, but to achieve active diffusion through orderly heat flux control, thereby significantly improving the accuracy, response speed and long-term stability of thermoelectric synergistic control.

[0087] The multi-layer coupled cooling path reconstruction module executes the multi-layer coupled cooling path reconstruction process based on the output of the Hamiltonian variational heat diffusion control function. Combined with the heat flux parameters of unidirectional diffusion distribution, it coordinates the driving rate of the cooling medium and the impedance of the heat conduction channel of the board to form a heat flow field structure with reversible equilibrium characteristics, thereby suppressing energy backflow and local heat flow accumulation from the source.

[0088] Based on the output of the Hamiltonian variational heat diffusion control function, a multi-layer coupled cooling path reconstruction process is performed, with the following specific steps:

[0089] After the Hamiltonian variational heat diffusion control function is output, its given unidirectional heat flux diffusion distribution result is spatially mapped to determine the energy transfer state and thermal conductivity characteristics of different layers in the cooling medium flow path. This stage maps the heat flux density distribution parameters output by the Hamiltonian variational heat diffusion control function to the cooling channel structure inside the board. By comparing the thickness of the flow layer, fluid velocity gradient, thermal resistance of the thermally conductive material interface, and temperature distribution at the fluid-solid interface in the cooling channel, the differences in energy transfer efficiency in each region under the current heat diffusion state can be identified. Based on this, according to the energy density and the consistency of the heat flux direction, the cooling path is divided into three structural layers: the dominant heat channel, the auxiliary diffusion channel, and the boundary equilibrium channel. The dominant heat channel is responsible for high-speed energy transport, the auxiliary diffusion channel is used to regulate the uniformity of heat flux distribution, and the boundary equilibrium channel is used to prevent local energy stagnation. This spatial hierarchical division allows subsequent cooling path reconstruction to be performed within a clear hierarchical structure, thus laying a directional foundation for coupled control.

[0090] After obtaining the spatial layered structure of the cooling channels, the driving rate of the cooling medium is adjusted layer by layer based on the unidirectional diffusion distribution of heat flux, so that the fluid momentum distribution matches the thermal conductivity of each channel. This step is based on the description of the energy potential field in the Hamiltonian variational control function. By identifying the relationship between heat flux density and velocity gradient in different channels, the driving rate of the cooling medium should be distributed among different layers. When the region with high heat flux density corresponds to the dominant heat channel, the flow rate of the cooling medium in that channel is increased to enhance the instantaneous heat removal capability of the high-heat region; while for the auxiliary diffusion channels with low heat flux density, the flow rate should be appropriately reduced to allow for sufficient heat conduction and equalization over time. Through this layered driving method, the kinetic energy distribution of the cooling medium forms a positive spatial correspondence with the heat flux distribution, keeping the energy transfer process of the overall cooling path continuous and balanced. In practice, precise control of the cooling flow rate at different layers can be achieved by finely adjusting the inlet pressure of the cooling medium, the cross-sectional geometry of the flow channel, and the fluid viscosity parameters. When the cooling medium flows through the multi-layer channels, the high flow velocity in the main channels carries away the heat from the high heat density areas, the auxiliary channels achieve lateral heat diffusion through slow flow, and the boundary channels maintain the overall balanced heat diffusion state, thus forming a dynamically stable heat flow structure on a time scale.

[0091] After the cooling medium driving rate is adjusted in layers, the local impedance of the heat conduction channels on the board is optimized in a coordinated manner based on the heat flux distribution parameters, so that the heat conduction path and the flow characteristics of the cooling medium are spatially coupled and balanced. This stage is based on the previously established multi-layer channel structure, and further improves the continuity of energy transfer in the cooling path by adjusting the thermal resistance distribution of the heat conduction channels. Specifically, for the dominant channels with concentrated heat flux and high cooling flow rate, the contact thermal resistance of the heat conduction interface is reduced to reduce heat loss during the conduction process from the solid material to the cooling medium; while for the auxiliary diffusion channels with lower heat flux, the thermal resistance is appropriately increased to slow down the heat flow diffusion rate and maintain the heat transfer balance between different channels. In this way, the thermal resistance distribution of the heat conduction channels is no longer fixed, but is adjusted in real time according to the dynamic flow state of the cooling medium, so that the heat flow distribution in the path remains continuous and stable. At the same time, in the boundary region of the channel, the surface roughness and contact pressure of the heat conduction interface material can be finely adjusted to form a local thermal resistance adjustment zone to compensate for transient thermal resistance changes caused by fluid disturbance or sudden temperature gradient changes. Through this collaborative optimization process, the heat transfer structure inside the board gradually approaches a state of energy balance in each layer, so that the flow direction of the cooling medium and the direction of heat diffusion are perfectly matched in space.

[0092] After the coordinated regulation of the cooling medium driving rate and the heat conduction channel impedance is completed, the entire cooling path undergoes dynamic stability correction and reversible equilibrium verification to ensure that the thermal flow field structure can maintain long-term self-equilibrium characteristics under different operating conditions. This stage involves continuous monitoring of heat flux density changes, cooling medium velocity distribution, and thermal interface temperature gradients in the multi-layer cooling channels to determine whether the cooling path has reached thermodynamic stability. When the rate of change of heat flux density in each channel approaches zero and the velocity gradient remains stable over a continuous time interval, it indicates that the cooling path has formed an adaptive equilibrium structure. If heat accumulation or velocity fluctuations in a certain channel are detected during monitoring, the driving rate or thermal impedance of the corresponding channel needs to be readjusted according to the constraints of the Hamiltonian variational control function to quickly restore the system to equilibrium. To ensure the reversibility of this thermal flow field structure, when load conditions change or power surges, the cooling medium driving layer and the heat conduction channels can coordinate their responses to redistribute the energy transfer path in a short time, automatically restoring the unidirectional nature of heat diffusion. This self-balancing characteristic ensures that the circuit board maintains a stable thermal management state even under high heat density operation, avoiding localized heat accumulation or thermal drift caused by energy retention. Continuous operation has verified that when the flow velocity of the cooling medium and the impedance distribution of the heat conduction path remain dynamically symmetrical in the multi-layer structure, a reversible equilibrium structure is formed within the heat flow field. This means that if the energy input changes at any time, the system can redistribute heat using the path of minimum energy consumption, fundamentally suppressing the recurrence of energy backflow effects.

[0093] If, during monitoring, heat accumulation or flow rate fluctuations exceeding a threshold are detected in a certain channel, the driving rate or thermal impedance of the corresponding channel must be readjusted according to the constraints of the Hamiltonian variational control function to quickly restore the system to equilibrium. The method for setting this threshold is explained in detail below:

[0094] The threshold value is determined based on the coupling relationship between the rate of change of cooling medium flow rate, the rate of change of heat flux density, and the temperature difference at the heat conduction interface within the cooling channel. It is generally obtained through statistical analysis of steady-state operating data and calibration using thermal flow field simulation results.

[0095] For example, in a typical liquid-cooled board structure, when the cooling medium flow rate is... Within the range, heat flux density is If, within the specified range, the rate of change of flow velocity in the channel exceeds 8% or the rate of change of heat flux density exceeds 10% during a continuous monitoring period, and the temperature difference fluctuation at the heat conduction interface exceeds 2K, it can be determined that the region has exceeded the self-balancing threshold, and the control function needs to be triggered to redistribute the heat flux.

[0096] Conversely, when the rate of change of flow velocity is less than 5%, the rate of change of heat flux density is less than 7%, and the temperature difference fluctuation does not exceed 1K, the cooling path can be considered to be in a stable equilibrium state and no adjustment is required.

[0097] This threshold range can maintain heat transfer efficiency while avoiding excessive disturbance, ensuring that the system has long-term adaptive and reversible adjustment capabilities.

[0098] Through the above implementation steps, the multi-layer coupled cooling path reconstruction process achieves complete closed-loop control from heat flux spatial stratification, flow rate distribution, thermal impedance adjustment to dynamic balance correction. This not only macroscopically optimizes the heat conduction path structure within the board but also dynamically coordinates the energy transfer process at the microscopic level, ensuring that the flow of the cooling medium and the direction of heat diffusion remain consistent. By forming a heat flow field structure with reversible equilibrium characteristics, energy diffuses efficiently, continuously, and unidirectionally within the board, completely eliminating the heat flow recirculation and local energy accumulation problems present in traditional cooling structures. This significantly improves the overall stability and operational efficiency of thermoelectric synergistic control.

[0099] This invention constructs a spatiotemporally multi-field coupled dynamic heat flux density mapping model within the circuit board, and based on this, achieves adaptive fluid disturbance and phase conjugate velocity compensation. This ensures that the flow direction of the cooling medium aligns with the heat conduction direction in both time and space, significantly improving the problem of sudden increases in thermal resistance caused by localized heat flow stagnation and energy backflow. By identifying energy backflow characteristic regions in real time and dynamically correcting fluid momentum at the microscale, proactive control can be achieved before heat accumulation, ensuring that the cooling path is always in a state of efficient heat transfer. As a result, the heat flux distribution of the circuit board is more uniform under high power density operation, the local temperature rise is significantly reduced, the risk of thermal drift and device failure is effectively suppressed, and the thermal stability and continuous operation reliability of the system are significantly improved.

[0100] This invention establishes a Hamiltonian variational heat diffusion control function and a multi-layer coupled cooling path reconstruction mechanism to dynamically match the cooling medium driving rate with the impedance of the board's thermal conductive channels, fundamentally achieving unidirectional diffusion of heat flux density and self-balancing regulation of energy flow. Through this collaborative regulation process, heat is dynamically distributed among different thermal conductive layers of the board, and the cooling path structure can be automatically optimized according to changes in heat load, forming a heat flow field with reversible equilibrium characteristics. This heat flow field can maintain consistent energy flow direction and constant conduction efficiency during long-term operation, effectively avoiding material fatigue, solder joint peeling, and interface instability caused by thermal stress accumulation, thus improving the overall energy efficiency and long-term service life of the equipment in complex thermal environments.

[0101] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A board card level thermoelectric collaborative management system, characterized in that, The heat flow dynamic mapping model construction module, the adaptive fluid disturbance injection module, the phase conjugate flow velocity compensation control module, the Hamiltonian variation heat diffusion regulation module, and the multi-layer coupled cooling path reconstruction module; The heat flow dynamic mapping model construction module constructs a heat flow density dynamic mapping model based on space-time multi-field coupling, takes the temperature gradient, power distribution, and flow velocity vector of the cooling medium collected in real time inside the board card as input parameters, jointly solves multiple physical fields, and generates a dynamic field distribution matrix representing the micro-scale flow trajectory and energy return characteristics inside the board card; The adaptive fluid disturbance injection module establishes an adaptive fluid disturbance injection chain based on the energy return characteristic area identified in the dynamic field distribution matrix, changes the momentum distribution of the cooling medium by acting on the local flow field of the cooling medium through a pulse disturbance signal; The phase conjugate flow velocity compensation control module generates a phase conjugate flow velocity compensation sequence according to the disturbance signal output by the adaptive fluid disturbance injection chain, takes the flow velocity change rate of the energy return characteristic area as a feedback variable, performs vectorized phase rotation control on the cooling medium main channel, and makes the flow direction and heat conduction direction form a synchronous adjustment baseline; The Hamiltonian variation heat diffusion regulation module constructs a Hamiltonian variation heat diffusion control function based on the synchronous adjustment baseline formed by the phase conjugate flow velocity compensation sequence, takes the synchronization characteristics of the flow direction and the heat conduction direction as a constraint condition, and redistributes the heat flux density of the cooling medium inside the board card; The multi-layer coupled cooling path reconstruction module performs a multi-layer coupled cooling path reconstruction process according to the output result of the Hamiltonian variation heat diffusion control function, cooperatively regulates the cooling medium driving rate and the board card heat conduction channel impedance by combining the heat flux parameters of one-way diffusion distribution, and forms a heat flow field structure.

2. The board-level thermoelectric co-management system of claim 1, wherein, The steps of constructing a heat flow density dynamic mapping model based on space-time multi-field coupling include: In the initial stage of constructing the space-time multi-field coupling model, the temperature, power, and cooling flow characteristics of the board card under different working conditions are acquired in real time and continuously, multiple regions are arranged on the board card according to the energy density distribution characteristics, sensing points for detecting the temperature gradient, power density, and flow velocity vector of the cooling medium are arranged in each region, and multi-dimensional synchronous sampling data is formed; According to the collected temperature gradient, power distribution, and flow velocity vector of the cooling medium, the interactive mapping relationship among the three is established in the space and time dimensions, the consistency of the power distribution and the flow direction is compared, the energy transfer path and the potential return area are identified, and a heat flow density distribution matrix containing the energy transfer direction and the flow stability is constructed; After the preliminary construction of the heat flow density distribution matrix, the micro-scale flow trajectory and energy return characteristics inside the board card are analyzed in detail in combination with the dynamic response of the time dimension, and the dynamic field distribution matrix reflecting the energy retention, return, and dissipation process is generated by comparing the temperature gradient and flow velocity change trend in the continuous time period.

3. The board-level thermoelectric co-management system of claim 2, wherein, In the generation process of the dynamic field distribution matrix, the flow velocity vector variation trajectory of the energy return characteristic region in the heat flux density distribution matrix is compared with the heat flow direction of the adjacent region. When a closed or approximately closed flow loop is detected, the flow loop is identified as an energy return characteristic region, and its spatial expansion range and energy density variation trend are recorded in the dynamic field distribution matrix in time sequence.

4. The board-level thermoelectric co-management system of claim 2, wherein, The steps of establishing the adaptive fluid disturbance injection chain include: After determining the energy return characteristic region, the local flow characteristics of the cooling medium in the energy return characteristic region are identified and the disturbance parameters are initialized. According to the spatial coordinates and time evolution characteristics of the energy return characteristic region in the dynamic field distribution matrix, the flow backflow or stagnation position is located, and the amplitude, duration and action frequency of the disturbance injection are determined; After completing the positioning of the energy return characteristic region and the initialization of the disturbance parameters, the local flow field of the cooling medium is acted on by the pulse disturbance signal, so that the velocity vector of the stagnant fluid is redistributed, the closed backflow path is broken, and the disturbance amplitude and interval are dynamically corrected according to the time evolution characteristics of the dynamic field distribution matrix, so that the phase consistency of the disturbance process and the main flow direction is maintained; After the action of the pulse disturbance signal ends, the disturbed region is continuously monitored and feedback adjusted. When the flow direction and the heat flux direction are consistent and the flow amplitude is stable, the disturbance is gradually weakened until it is terminated.

5. The board-level thermoelectric co-management system of claim 4, wherein, During the action of the pulse disturbance signal, the disturbance is preferentially applied at the boundary position of the energy return characteristic region, so that the momentum of the cooling medium is transmitted along the main flow direction, and compensatory disturbance is performed at the key turning points or flow velocity sudden drop zones of the flow channel.

6. The board-level thermoelectric co-management system of claim 4, wherein, The steps of performing vectorized phase rotation control on the main channel of the cooling medium include: After the adaptive fluid disturbance injection chain is completed, the flow state in the disturbed region and its surrounding cooling channel is identified and dynamically quantified. By monitoring the flow velocity variation rate of the cooling medium in the energy return characteristic region, the relative phase difference distribution between the cooling medium flow velocity and the heat flow direction is obtained; After obtaining the flow velocity variation rate and the relative phase difference distribution, a phase conjugate flow velocity compensation sequence is generated according to the time characteristics of the disturbance signal and the dynamic response relationship of the flow field, so that the flow direction vector of the cooling medium and the heat conduction direction vector are time-reversing offset, forming a dynamically synchronized flow velocity compensation trajectory; The phase conjugate flow velocity compensation sequence is applied to the main channel of the cooling medium to perform vectorized phase rotation control on the overall flow direction, so that the flow direction and the heat conduction direction are consistent in space, forming a stable flow field; After completing the phase rotation control, a synchronous adjustment baseline is established, and the dynamic synchronization of the flow direction and the heat conduction direction is realized by continuously monitoring the heat flux density variation and the flow velocity vector variation trend.

7. The board-level thermoelectric co-management system of claim 6, wherein, In the process of establishing the synchronous adjustment baseline, the heat flux density variation and the flow velocity vector variation of the cooling medium in the main channel are time-continuously monitored. When the variation rates of the two are linearly corresponding and the direction deviation is less than the preset threshold, it is determined that the flow direction and the heat conduction direction are in a synchronous state, and this state is used as the maintenance condition of the synchronous adjustment baseline.

8. The board-level thermoelectric co-management system of claim 1, wherein, The steps of constructing the Hamiltonian variational heat diffusion control function include: On the basis of the synchronous adjustment baseline formed by the phase conjugate flow velocity compensation sequence, the correspondence between the flow direction distribution and the heat conduction direction distribution of the cooling medium inside the board card is obtained, the angle difference, phase difference characteristics and flow velocity gradient change trend of each region inside the board card are obtained; After obtaining the synchronous relationship between the flow direction and the heat conduction direction, the energy distribution boundary inside the board card and the basic state of the heat flux density are determined, the heat energy distribution diagram inside the cooling medium is established, and the distribution model with variable energy gradient is formed; After the synchronous characteristics and energy boundary distribution are determined, the core structure of the Hamiltonian variational heat diffusion control function is established, and the redistribution of heat flux density in space is realized through dynamic adjustment of the energy diffusion path; The output result of the Hamiltonian variational heat diffusion control function is applied to the cooling medium flow field, and by adjusting the local flow pressure, channel heat conduction impedance and cooling flow velocity distribution, the energy is diffused unidirectionally along the cooling path and the stable conduction is maintained; On the basis of the Hamiltonian variational heat diffusion control function, a dynamic feedback mechanism is established, and by continuously monitoring the heat flux density change and energy diffusion direction, the self-correction and adaptive adjustment of the heat diffusion path are realized.

9. The board-level thermoelectric co-management system of claim 8, wherein, In the dynamic feedback mechanism, by monitoring the heat flux density change and flow direction deviation of the cooling medium under the constraint of the synchronous adjustment baseline, when a new unbalanced area of energy distribution is detected, the heat diffusion direction of the unbalanced area is automatically corrected according to the output result of the Hamiltonian variational heat diffusion control function, so that the energy transfer of the cooling medium inside the board card re-maintains unidirectional diffusion and stable equilibrium state.

10. The board-level thermoelectric co-management system of claim 8, wherein, The steps of performing the multi-layer coupled cooling path reconstruction process include: After the output of the Hamiltonian variational heat diffusion control function, the spatial mapping of the heat flux unidirectional diffusion distribution result is performed, the energy transfer state and heat conduction characteristics of each layer in the cooling medium flow path are determined, and the cooling path is divided into main heat channels, auxiliary diffusion channels and boundary balance channels; After obtaining the spatial layered structure of the cooling channel, based on the heat flux unidirectional diffusion distribution result, the cooling medium driving rate is adjusted layer by layer, so that the momentum distribution of the cooling medium matches the heat conduction capacity of each channel, forming a continuous balanced energy transfer process; After the cooling medium driving rate completes the layered adjustment, according to the heat flux distribution parameters, the board card heat conduction channel impedance is optimized, so that the heat conduction path and the cooling medium flow characteristics form a coupled balance in space; After the collaborative regulation of the driving rate and the heat conduction impedance is completed, the dynamic stability correction and reversible balance verification of the cooling path are performed, so that the heat flux density change rate in the multi-layer channel tends to be stable, forming a heat flow field structure with adaptive and reversible balance characteristics.

Images