A phase change enhanced thermal management system based on an artificial diamond superconducting thermal layer

Abstract

This invention relates to a phase change enhanced thermal management system based on an artificial diamond superconducting thermal layer, belonging to the field of thermal management efficiency control technology. The method includes: constructing a phase change thermal management surface evolution network; combining phase change heat load sensitivity constraints with local hotspots to obtain graded phase change thermal management parameters; using local hotspots as the driving basis, defining enhanced thermal management trigger thresholds through a bidirectional phase change thermal management model, and outputting multi-region phase change heat transfer efficiency improvement schemes; modeling inter-regional heat flow guidance according to heat flux density distribution, constructing forced convection paths, and combining multi-domain performance states at different load stages to obtain heat flux feedback values ​​that satisfy enhanced response constraints; and retrieving the phase change thermal evolution stage and response mode through a phase change thermal management surface state reconstruction mechanism to adjust the configuration path of the bidirectional phase change thermal management model.

Description

Technical Field

[0001] This invention belongs to the field of thermal management efficiency control technology, specifically relating to a phase change enhanced thermal management system based on an artificial diamond superconducting thermal layer. Background Technology

[0002] With the rapid development of high-power electronic devices, highly integrated chips, and new energy vehicle power systems, the heat flux density generated during system operation continues to increase. Local areas are prone to forming high heat flux and strong non-uniform temperature field distribution, which in turn leads to problems such as device performance degradation, shortened lifespan, or even failure.

[0003] Traditional heat dissipation methods primarily rely on thermally conductive metallic materials (such as copper and aluminum) combined with air-cooled or liquid-cooled structures for heat conduction and convection. However, these methods often suffer from drawbacks such as slow response, limited heat conduction paths, and restricted heat diffusion when facing transient high heat flux impacts and localized hotspots, making it difficult to meet the refined thermal management requirements under high dynamic heat load scenarios. Therefore, introducing the latent heat of phase change materials (PCMs) to achieve temperature control through the absorption or release of latent heat has gradually become an important direction for enhancing thermal management. However, existing phase change thermal management technologies are mostly focused on single-region or passive control, lacking systematic modeling of complex multi-regional heat flux coupling relationships, making it difficult to achieve precise suppression of localized hotspots and dynamic equilibrium of the overall temperature field.

[0004] On the other hand, with the development of artificial diamond material preparation technology, its extremely high thermal conductivity and excellent anisotropic thermal conductivity provide a new technical path for constructing efficient heat conduction channels. However, the application of artificial diamond superconducting layers in existing technologies is mostly limited to static thermal conductivity enhancement, and has not yet fully incorporated the dynamic control mechanism of the phase transition process, nor does it lack a systematic method for adaptive reconstruction of heat flow paths and multi-regional synergistic control.

[0005] In addition, current thermal management systems generally lack closed-loop optimization capabilities, meaning they do not make sufficient use of heat flow feedback information and cannot adjust thermal management strategies in real time according to different load stages, resulting in limited adaptability and stability of the system under complex operating conditions.

[0006] Therefore, it is necessary to propose a thermal management system that integrates artificial diamond superconducting thermal layer and phase change enhancement mechanism. By constructing a multi-region heat flow evolution model and a two-way coupled control mechanism, it can achieve precise constraint of local hot spots, dynamic optimization of heat flow path and coordinated control of global temperature field, thereby significantly improving the heat dissipation efficiency and response capability of the system in complex high heat flow scenarios. Summary of the Invention

[0007] To address the aforementioned problems in the prior art, this invention provides a phase change enhanced thermal management system based on an artificial diamond superconducting thermal layer. The objective of this invention can be achieved through the following technical solutions: Phase change heat management parameter acquisition module: acquires heat flux data of phase change heat transfer surface, constructs phase change heat management surface evolution network, and detects local hot spots by combining phase change heat load sensitivity, and acquires graded phase change heat management parameters; Regional heat transfer enhancement scheme output module: Based on the graded phase change heat management parameters, it responds to the heat exchange scenario requirements and takes local hot spots as the driving basis. It defines the enhanced heat management trigger threshold through a two-way phase change heat management model and outputs a multi-region phase change heat transfer efficiency enhancement scheme. Heat flow feedback calculation module: Executes the multi-region phase change heat transfer efficiency improvement scheme, models the heat flow guidance between regions according to the heat flow density distribution, constructs a forced convection path, and obtains the heat flow feedback value that satisfies the enhanced response constraint relationship by combining the multi-domain performance status at different load stages. Closed-loop optimization and control module: The heat flow feedback value is used as the empowerment sample of the phase change thermal management surface to establish closed-loop management optimization logic. The phase change thermal evolution stage and response mode are retrieved through the phase change thermal management surface state reconstruction mechanism to adjust the configuration path of the bidirectional phase change thermal management model.

[0008] Specifically, the method for constructing the phase change thermal management surface evolution network is as follows: Based on the heat flux data of the phase change heat surface, the phase change heat surface is divided into heat flux density clusters to obtain thermal region association units; The phase change heat flow data includes: instantaneous heat flow density value, average heat flow density value, and heat flow change rate of each region of the phase change heat flow surface; The thermal region association unit is abstracted as a network node, and the basic thermal state data is used as the node attribute. The heat flow transfer relationship between regions is used as the node edge connection. A phase change state transition mechanism is introduced to expand the connection relationship between the network nodes, and a phase change thermal management surface evolution network is constructed.

[0009] Specifically, the process of local hotspot detection includes the following steps: The heat source contact surface is divided into multiple thermal response coupling sub-regions. An independent thermal response state vector is extracted in each coupling sub-region, and the heat flow increase change and phase change accumulation change are mapped into local hot spot constraint functions. The intersection time between the hot spot critical curve and the actual heat flow boundary is calculated using the local hot spot constraint function. The reciprocal of the intersection time is then weighted and coupled with the hot spot constraint growth gradient. Combined with the phase change heat load sensitivity, the hot spot area of ​​the heat source contact surface is defined to obtain the graded phase change heat management parameters. The method for obtaining the graded phase change thermal management parameters is as follows: extract the basic thermal response data of each region according to a preset sampling period; Based on the basic thermal response data, local hotspot areas and hotspot diffusion trends are identified, and each area is classified into different levels. The corresponding phase change trigger temperature thresholds are extracted, and graded phase change thermal management parameters are generated. The graded phase change thermal management parameters include: enhanced thermal management parameters, regulated thermal management parameters, and basic thermal management parameters.

[0010] Specifically, the process by which the staged phase change heat management parameters respond to the needs of the heat exchange scenario is as follows: The heat flux density distribution of the phase change heat exchange surface is analyzed based on the graded phase change heat management parameters to extract the scene features of the current heat exchange scenario; The scene features include: the overall heat flux density level of the current phase transformation heat surface and the heat flux density distribution characteristics of each region; Based on the characteristics of the scenario, the parameters are matched with a pre-established hierarchical phase change heat management parameter library, and with the core objectives of local hot spot suppression and overall temperature field equilibrium, parameter combinations suitable for the current heat exchange scenario are selected.

[0011] Specifically, the construction process of the bidirectional phase change thermal management model includes the following steps: Based on local hotspots, the diffusion path of heat flow within the interface is reconstructed and described. Using the hierarchical phase change thermal management parameters as the configuration path, the thermal response characteristics of different spatial regions are mapped to construct a bidirectional coupled feedback architecture. Based on the bidirectional coupled feedback architecture, the phase change heat in both directions is cross-boundary empowered and linked in the time and space domains, and the heat conduction path weights and regional heat flow distribution coefficients are archived periodically to construct a bidirectional phase change heat management model.

[0012] Specifically, the output process of the multi-region phase change heat transfer efficiency improvement scheme includes the following steps: The phase change heat surface is discretized, the heat flux density of each region is extracted, and the local hot spot regions are classified and calibrated according to the heat flux density to determine the thermal management priority of different regions. Based on the aforementioned thermal management priority, the graded phase change thermal management parameters are mapped to each region's phase change unit to construct the corresponding phase change control path; Based on the phase change control path, the enhanced thermal management trigger threshold is defined, and combined with the heat flow distribution relationship of each region, the phase change execution strategy is adaptively set to output a multi-region phase change heat transfer efficiency improvement scheme. The enhanced thermal management trigger threshold is used to determine whether the phase change heat transfer surface has entered the enhanced thermal management state and to trigger the phase change control strategy and heat conduction path reconstruction operation in the corresponding region.

[0013] Specifically, the bidirectional coupled feedback architecture includes: a forward phase change driving structure and a reverse heat flow feedback control structure; The execution process of the forward phase change driving structure is as follows: based on the graded phase change thermal management parameters and the enhanced thermal management trigger threshold, the phase change heat transfer surface is divided into regions and prioritized. Combined with the anisotropic thermal conductivity of the artificial diamond superconducting thermal layer, the heat flow conduction path is configured in a directional manner to construct an enhanced thermal conduction channel for hot spot areas, driving the rapid diffusion and redistribution of heat between multiple regions. The execution process of the reverse heat flow feedback control structure is as follows: receiving the heat flow feedback value and regional heat flow distribution coefficient output by the heat flow feedback calculation module, performing inverse analysis on the thermal response state of the current phase change heat surface, and adjusting the heat conduction path weight update and regional heat flow distribution ratio in the bidirectional phase change heat management model in reverse.

[0014] Specifically, the process of modeling the inter-regional heat flow guidance and constructing forced convection paths includes the following steps: Each discrete region is abstracted as a network node, and the heat flow transfer intensity and thermal conductivity efficiency are used as edge weights to weight and modulate the heat flow transmission priority of different paths to construct a multi-region heat flow guiding topology network. Based on the heat flow guiding topology network, the heat flow transfer relationship and path weight between each node are globally analyzed to identify the optimal heat transfer path between the high heat flow density region and the low temperature sink region. By adjusting the thermal conductivity directionality, interfacial thermal resistance, and local phase transition triggering state of the artificial diamond superconducting layer, a forced convection path is constructed by enhancing the configuration of the selected path.

[0015] Specifically, the process of obtaining the heat flow feedback value by combining the multi-domain performance states at different load stages includes the following steps: The operation process of the forced convection path is divided into load stages, and the multi-domain performance state parameters corresponding to each stage are extracted. Sensible heat transfer, latent heat exchange during phase change, and interfacial thermal resistance loss are used as multi-domain coupled heat flow constraints. The heat flow deviations in different phase change domains are weighted and the heat flow corrections for each region are obtained. The correction results are then filtered in conjunction with the phase lag response relationship to generate heat flow feedback values.

[0016] Specifically, the execution process of the empowered sample includes the following steps: By introducing phase change heat load sensitivity and heat conduction path information, feature enhancement is performed on the empowered sample to establish a mapping relationship between the empowered sample and the node state in the phase change heat management surface evolution network. Based on the phase change thermal management surface state reconstruction mechanism, the thermal management strategy parameters corresponding to the matching results are extracted and fed back to the phase change thermal management optimization sample library to iterate the bidirectional phase change thermal management model.

[0017] Specifically, the closed-loop management optimization logic includes: a strategy decision-making layer and an execution feedback layer; The strategy decision layer is used to: analyze and determine the thermal management requirements of the current heat exchange scenario based on the empowered sample and the graded phase change thermal management parameters, and generate corresponding phase change control strategies; The execution feedback layer is used to: perform actual control execution on the phase change heat exchange surface based on the phase change control strategy, calculate the heat flow feedback value in combination with the multi-domain energy heat balance function, and transmit the heat flow feedback value back to the strategy decision layer.

[0018] Specifically, the process by which the phase change thermal management surface state reconstruction mechanism retrieves the phase change thermal evolution stage and response mode includes the following steps: Based on the operational data in the phase change thermal management optimization sample library, the different phase change evolution curves are normalized in intervals, and the strengthening demand level and phase change thermal management surface state are used as auxiliary alignment factors for working condition labeling, resulting in a multi-dimensional strengthening feature vector. The multidimensional enhancement feature vector is flexibly registered on the time axis, and the phase change risk changes of different enhancement operations under the same phase change thermal management surface state conditions are matched in time sequence to retrieve the corresponding phase change thermal evolution stage and response mode.

[0019] The beneficial effects of this invention are as follows: The present invention proposes a phase change enhanced thermal management system based on an artificial diamond superconducting thermal layer. By introducing hierarchical phase change thermal management parameters and a multi-region synergistic control mechanism, it has significant performance improvements and application advantages compared with existing technologies.

[0020] By constructing a phase change thermal management surface evolution network and combining it with a local hotspot constraint mechanism, accurate identification and dynamic constraint of high heat flux density regions are achieved, effectively avoiding device performance degradation caused by hotspot overheating. Secondly, by utilizing the high thermal conductivity and anisotropic properties of the artificial diamond superconducting layer, the heat flow conduction path is optimized in a directional manner, enabling heat to spread and redistribute rapidly among multiple regions, significantly improving overall heat dissipation efficiency.

[0021] Furthermore, a bidirectional phase change thermal management model was established, organically combining forward phase change drive with reverse heat flow feedback, achieving dynamic adjustment and adaptive optimization of the thermal management process. Under different load stages, the system can adjust the heat conduction path weights and regional heat flow distribution ratios in real time based on heat flow feedback values, thereby ensuring that the thermal management strategy is always in an optimal state. In addition, the introduction of a multi-region phase change heat transfer efficiency improvement scheme enables the system to flexibly configure phase change execution strategies according to actual heat exchange scenarios, taking into account both local enhancement and global equilibrium, thereby improving the system's stability and robustness.

[0022] Finally, by using a closed-loop optimization control mechanism and a phase change thermal management surface state reconstruction method, continuous learning and iterative optimization of the thermal management strategy are achieved, enabling the system to have good self-adaptability and long-term operational reliability, and making it suitable for thermal management needs under high power density and complex operating conditions. Attached Figure Description

[0023] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0024] Figure 1 This is a schematic diagram of the framework of a phase change enhanced thermal management system based on an artificial diamond superconducting thermal layer according to the present invention.

[0025] Figure 2 This is a schematic diagram illustrating the execution of the bidirectional coupling feedback architecture in a phase change enhanced thermal management system based on an artificial diamond superconducting thermal layer according to the present invention. Detailed Implementation

[0026] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0027] Please see Figure 1 This invention also provides a phase change enhanced thermal management system based on an artificial diamond superconducting thermal layer, specifically including: Phase change heat management parameter acquisition module: acquires heat flux data of phase change heat transfer surface, constructs phase change heat management surface evolution network, and detects local hot spots by combining phase change heat load sensitivity, and acquires graded phase change heat management parameters; Regional heat transfer enhancement scheme output module: Based on the graded phase change heat management parameters, it responds to the heat exchange scenario requirements and takes local hot spots as the driving basis. It defines the enhanced heat management trigger threshold through a two-way phase change heat management model and outputs a multi-region phase change heat transfer efficiency enhancement scheme. Heat flow feedback calculation module: Executes the multi-region phase change heat transfer efficiency improvement scheme, models the heat flow guidance between regions according to the heat flow density distribution, constructs a forced convection path, and obtains the heat flow feedback value that satisfies the enhanced response constraint relationship by combining the multi-domain performance status at different load stages. Closed-loop optimization and control module: The heat flow feedback value is used as the empowerment sample of the phase change thermal management surface to establish closed-loop management optimization logic. The phase change thermal evolution stage and response mode are retrieved through the phase change thermal management surface state reconstruction mechanism to adjust the configuration path of the bidirectional phase change thermal management model.

[0028] In this embodiment, taking the thermal management system of a high-power amplifier for a base station manufactured by a communication equipment manufacturer as an example, the single-chip heat flux density of the base station PA module can reach 150W / cm² when running at full load.2 In summer, when the outdoor temperature reaches 45℃, traditional aluminum heat sinks are unable to control the junction temperature below 85℃, which can easily lead to device performance degradation, frequency drift, or even failure.

[0029] The system employs an artificial diamond superconducting thermal layer (e.g., CVD diamond film, 0.3 mm thick, with a thermal conductivity greater than 1800 W / m·K) combined with phase change materials (e.g., paraffin-based composite phase change materials, with a phase change temperature of 38-42℃ and a latent heat of 180 kJ / kg) to achieve multi-stage thermal management of "rapid hotspot diffusion + phase change latent heat storage + forced convection enhancement".

[0030] The hardware includes: diamond heat sink + phase change cavity + microchannel liquid cooling plate, temperature / heat flow sensor array (32 points), PLC + edge computing controller (Siemens S7-1500 + Jetson Orin).

[0031] The software platform is based on the Python main control framework and integrates COMSOL Multiphysics as a digital twin simulation engine.

[0032] The specific technical solution is as follows: Phase change thermal management parameter acquisition module; Heat flux data at each point on the phase transition heat exchange surface are collected in real time. This data is input into a clustering algorithm and automatically divided into high heat flux region (heat flux density greater than 250 W / cm²), medium heat flux region (150 to 250 W / cm²), and low heat flux region (less than 150 W / cm²) according to the heat flux density, resulting in a total of 12 thermal region association units.

[0033] Each associated unit is abstracted as a node in the network, and the node attributes include the current temperature value, the state of the phase change material (solid, liquid or partially molten) and the remaining heat capacity.

[0034] Nodes are connected by the actual heat flow intensity between regions. For example, the weight of an edge from a high heat flow region to a low heat flow region is set as a strong connection.

[0035] When the node temperature reaches above 68 degrees Celsius, the system automatically introduces a phase change state transition mechanism, expanding the original edge connections into new paths that include latent heat absorption, thereby dynamically constructing a phase change heat management surface evolution network. This network is updated once per second to ensure that the heat diffusion path can be tracked in real time.

[0036] During the constraint process of local hot spots, the heat source contact surface is divided into 9 equal-area thermal response coupling sub-regions. Each sub-region independently extracts a thermal response state vector, including the heat flux increase rate and the phase change accumulation rate.

[0037] The increase in heat flux and the cumulative change in phase transition are mapped to a local hot spot constraint function.

[0038] In high-load peak cases (e.g., GPUs performing large-scale AI training, where heat flux surges from 150 watts per square centimeter to 320 watts per square centimeter in 10 seconds), the heat flux in a sub-region increases while the cumulative change in phase transition decreases.

[0039] Combined with the phase change heat load sensitivity (e.g., the sensitivity threshold is set to 0.85, and a strong constraint is triggered when this value is exceeded), this sub-region is strictly limited to a hot spot region, and the boundary temperature must not exceed a preset degree Celsius. At the same time, the system pre-allocates additional heat conduction paths to adjacent sub-regions to avoid local heat accumulation.

[0040] In actual testing, this constraint mechanism reduces the peak temperature of hot spots, effectively preventing chip performance throttling.

[0041] Based on the above process, the module finally outputs graded phase change thermal management parameters (high, medium, and low levels) and passes them to the next module. In normal operation, when the GPU is under light rendering load, the network shows that only 2 nodes enter the liquid phase change, improving the heat transfer efficiency between the high and low heat flux regions and enhancing the overall temperature field uniformity.

[0042] Output module for regional heat transfer enhancement scheme; This module responds to the needs of heat transfer scenarios based on the hierarchical phase change heat management parameters. It takes local hot spots as the driving basis and hierarchical phase change heat management parameters as the configuration path. It defines the enhanced heat management trigger threshold through a bidirectional phase change heat management model and outputs a multi-region phase change heat transfer efficiency improvement scheme.

[0043] The heat flux density distribution of the phase change heat exchange surface is analyzed to extract the scene characteristics of the current heat exchange scenario, such as the current total heat load of 420 watts, the number of hot spots of 3, and the average temperature gradient of 15 degrees Celsius per centimeter. This is then matched with a pre-established graded phase change heat management parameter library, which contains preset combinations of low, medium, and high levels.

[0044] With the core objectives of suppressing local hotspots and balancing the overall temperature field, parameter combinations suitable for the current scenario were selected: The high heat flux region employs a first-stage phase change trigger (melting point 68 degrees Celsius, with latent heat absorption being activated first). The intermediate heat flow region employs a two-stage phase change (melting point 72 degrees Celsius, to assist sensible heat transfer); The low heat flux region employs a three-stage phase transition (melting point 76 degrees Celsius, focusing on heat removal).

[0045] In the fault recovery case (simulating a temporary failure of a sensor leading to a loss of heat flow data), the system can still match the correct parameter combination by interpolating features from adjacent nodes, ensuring the continuity of the solution.

[0046] Construction of a two-way phase change thermal management model: Based on local hot spots, the diffusion path of heat flow within the interface is reconstructed and described. Using the hierarchical phase change thermal management parameters as the configuration path, the thermal response characteristics of different spatial regions are mapped to construct a two-way coupled feedback architecture.

[0047] like Figure 2 The bidirectional coupled feedback architecture shown includes a forward phase change driving structure and a reverse heat flow feedback control structure.

[0048] The forward structure is based on hierarchical parameters and enhanced thermal management trigger thresholds (e.g., the trigger threshold is set to a heat flux density exceeding 220 W / cm² or a temperature exceeding 75°C). It divides the phase change heat transfer surface into regions and prioritizes them. Combined with the anisotropic thermal conductivity of the artificial diamond superconducting thermal layer, it constructs enhanced thermal conduction channels for hot spots by directional configuration of heat flow conduction paths through an external 0.5V electric field. This allows heat to diffuse and redistribute rapidly between multiple regions, increasing the diffusion speed to several times that of ordinary copper layers.

[0049] The reverse structure receives subsequent heat flow feedback values ​​and regional heat flow distribution coefficients, performs inverse analysis on the thermal response state of the current phase change heat surface, and adjusts the heat conduction path weights in the model (e.g., increasing the weight of hotspots pointing to low heat flow regions) and the regional heat flow distribution ratio in reverse.

[0050] The training process of the bidirectional phase change thermal management model is as follows: input the training part of the training sample set into the initial bidirectional phase change thermal management model; In each iteration, the model input layer receives preprocessed input features, which include heat flux density distribution, regional temperature gradient, local hot spot location, phase change triggering state, heat conduction path weight, and regional heat flux distribution coefficient. Feature extraction, coupled calculation and state mapping are performed by a forward phase change driving network and a reverse heat flow feedback network. The output layer outputs heat conduction path adjustment value, regional heat flow redistribution value and heat flow feedback prediction value. Based on the feedback prediction values, the model parameters are solidified to obtain the final bidirectional phase change thermal management model, which is stored in the model library for subsequent steps. If it does not meet the requirements, the model iteration parameters are adjusted, including increasing the number of iterations, adjusting the learning rate, or expanding the training sample size. The model training process is then re-executed until it passes the verification.

[0051] The output process of the multi-region phase change heat transfer efficiency improvement scheme is as follows: the phase change heat transfer surface is discretized into 256 tiny grid units, the heat flux density of each region is extracted, and the local hot spot regions are classified and calibrated according to the heat flux density (for example: first-level hot spot: heat flux greater than 280 W / cm², highest priority; second-level: 220 to 280 W / cm²; third-level: less than 220 W / cm²).

[0052] Driven by thermal management priority, the graded phase change thermal management parameters are mapped to phase change units in each region to construct corresponding phase change control paths (for example, the primary hot spot path is set as "immediately trigger latent heat + diamond directional heat conduction").

[0053] Based on this path definition, the thermal management trigger threshold is enhanced, and combined with the heat flow distribution relationship of each region, the phase change execution strategy is adaptively set, and finally a multi-region phase change heat transfer efficiency improvement scheme is output.

[0054] In normal operating cases, the solution improves overall heat transfer efficiency; in high load peak cases, the solution accelerates the heat flow removal rate in hot spots and reduces the standard deviation of the overall temperature field.

[0055] Heat flow feedback calculation module; This module executes the multi-region phase change heat transfer efficiency improvement scheme. Based on the heat flux density distribution, it models the heat flux guidance between regions, constructs a forced convection path, and combines the multi-domain performance status at different load stages to obtain the heat flux feedback value that satisfies the enhanced response constraint relationship.

[0056] The specific process is as follows: each discrete region is abstracted into a network node, and the heat flow transfer intensity and thermal conductivity efficiency are used as edge weights to weight and modulate the heat flow transmission priority of different paths, thereby constructing a multi-region heat flow guiding topology network.

[0057] Based on the heat flow-oriented topology network, the heat flow transfer relationship and path weight between each node are globally analyzed to identify the optimal heat transfer path between high heat flow density areas and low temperature sink areas (for example, the path from the hot spot at the center of the chip to the edge heat sink fins is selected first).

[0058] By adjusting the thermal conductivity directionality (electric field direction pointing to the optimal path), interface thermal resistance, and local phase transition triggering state of the artificial diamond superconducting layer, the selected path is enhanced and a forced convection path is ultimately constructed.

[0059] The forced convection path can directionally transport heat to the low-temperature zone during actual operation.

[0060] The process of obtaining heat flow feedback values ​​by combining the multi-domain performance status of different load stages is as follows: the operation process of the forced convection path is divided into load stages (low load stage: 0-30% of rated power; medium load: 30-70%; high load: above 70%), and the multi-domain performance status parameters corresponding to each stage are extracted, including sensible heat transfer, latent heat exchange of phase change and interfacial thermal resistance loss.

[0061] These quantities are used as multi-domain coupled heat flux constraints. The heat flux deviations in different phase transition domains are weighted (e.g., 0.6 for hot spot domain, 0.3 for transition domain, and 0.1 for heat dissipation domain) to obtain the heat flux correction amount for each region. The correction results are then filtered in conjunction with the phase hysteresis response relationship to finally generate the heat flux feedback value.

[0062] In the test cases during the high-load phase, the heat flow feedback value was stable with small error, providing reliable input data for subsequent closed-loop optimization.

[0063] Closed-loop optimization and control module; This module uses the heat flow feedback value as an empowerment sample for the phase change thermal management surface, establishes closed-loop management optimization logic, and retrieves the phase change thermal evolution stage and response mode through the phase change thermal management surface state reconstruction mechanism to adjust the configuration path of the bidirectional phase change thermal management model.

[0064] The specific execution process of the empowered sample is as follows: introduce phase change heat load sensitivity and heat conduction path information, perform feature enhancement on the empowered sample, and establish a mapping relationship between the empowered sample and the node state in the phase change heat management surface evolution network.

[0065] Based on the phase change thermal management surface state reconstruction mechanism, the thermal management strategy parameters corresponding to the matching results are extracted and fed back to the phase change thermal management optimization sample library to iterate the bidirectional phase change thermal management model.

[0066] In a case where the system ran continuously for 72 hours, the module completed 28 iterations, improving the model's prediction accuracy compared to the initial iteration and reducing system energy consumption.

[0067] The closed-loop management optimization logic includes a strategy decision-making layer and an execution feedback layer.

[0068] The strategy decision layer is used to: analyze and determine the thermal management needs of the current heat exchange scenario based on the empowered samples and the graded phase change thermal management parameters, and generate corresponding phase change control strategies (such as "prioritizing the strengthening of hot spot paths and delaying the phase change in the middle zone").

[0069] The execution feedback layer is used to: perform actual control execution on the phase change heat exchange surface based on the phase change control strategy (achieved through solenoid valves and micro electric field controllers), and calculate the heat flow feedback value in combination with the multi-domain energy heat balance function, and then transmit the heat flow feedback value back to the strategy decision layer.

[0070] In the fault simulation case (where a heat conduction path was manually cut off), the execution feedback layer detected the deviation within a specified time, and the strategy decision layer immediately generated a backup path strategy. The system restored normal temperature control within a safe time, demonstrating extremely strong robustness.

[0071] The process of retrieving the phase change thermal management surface state reconstruction mechanism to retrieve the phase change thermal evolution stage and response mode is as follows: Based on the operating data in the phase change thermal management optimization sample library, the different phase change evolution curves are normalized in intervals, and the strengthening demand level and phase change thermal management surface state are used as auxiliary alignment factors to perform operating condition labeling, thereby obtaining a multi-dimensional strengthening feature vector.

[0072] The multidimensional enhancement feature vector is flexibly registered on the time axis. The phase change risk changes of different enhancement operations under the same phase change thermal management surface state conditions are matched in time sequence. Common enhancement evolution node features are extracted, and the corresponding phase change thermal evolution stage and response mode are retrieved.

[0073] In cases involving long-term alternating high and low loads, the mechanism successfully identified the common node of "rapid melting - slow solidification," advancing the phase change risk warning time to before it actually occurs and avoiding material overload.

[0074] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A phase change enhanced thermal management system based on an artificial diamond superconducting thermal layer, characterized in that, include: Phase change heat management parameter acquisition module: acquires heat flux data of phase change heat transfer surface, constructs phase change heat management surface evolution network, and detects local hot spots by combining phase change heat load sensitivity, and acquires graded phase change heat management parameters; Regional heat transfer enhancement scheme output module: Based on the graded phase change heat management parameters, it responds to the heat exchange scenario requirements and takes local hot spots as the driving basis. It defines the enhanced heat management trigger threshold through a two-way phase change heat management model and outputs a multi-region phase change heat transfer efficiency enhancement scheme. Heat flow feedback calculation module: Executes the multi-region phase change heat transfer efficiency improvement scheme, models the heat flow guidance between regions according to the heat flow density distribution, constructs a forced convection path, and obtains the heat flow feedback value that satisfies the enhanced response constraint relationship by combining the multi-domain performance status at different load stages. Closed-loop optimization and control module: The heat flow feedback value is used as the empowerment sample of the phase change thermal management surface to establish closed-loop management optimization logic. The phase change thermal evolution stage and response mode are retrieved through the phase change thermal management surface state reconstruction mechanism to adjust the configuration path of the bidirectional phase change thermal management model.

2. The system according to claim 1, characterized in that, The method for constructing the phase change thermal management surface evolution network is as follows: Based on the heat flux data of the phase change heat surface, the phase change heat surface is divided into heat flux density clusters to obtain thermal region association units; The thermal region association unit is abstracted as a network node, and the basic thermal state data is used as the node attribute. The heat flow transfer relationship between regions is used as the node edge connection. A phase change state transition mechanism is introduced to expand the connection relationship between the network nodes, and a phase change thermal management surface evolution network is constructed.

3. The system according to claim 1, characterized in that, The process of local hotspot detection includes the following steps: The heat source contact surface is divided into multiple thermal response coupling sub-regions. An independent thermal response state vector is extracted in each coupling sub-region, and the heat flow increase change and phase change accumulation change are mapped into local hot spot constraint functions. The intersection time between the hotspot critical curve and the actual heat flow boundary is calculated using the local hotspot constraint function. The reciprocal of the intersection time is then weighted and coupled with the hotspot constraint growth gradient. Combined with the phase change heat load sensitivity, the hotspot region of the heat source contact surface is defined to detect local hotspots.

4. The system according to claim 1, characterized in that, The process by which the staged phase change thermal management parameters respond to the needs of the heat exchange scenario is as follows: Based on the aforementioned graded phase change heat management parameters, the heat flux density distribution of the phase change heat exchange surface is analyzed to extract the scene characteristics of the current heat exchange scenario. Based on the characteristics of the scenario, the parameters are matched with a pre-established hierarchical phase change heat management parameter library, and with the core objectives of local hot spot suppression and overall temperature field equilibrium, parameter combinations suitable for the current heat exchange scenario are selected.

5. The system according to claim 1, characterized in that, The construction process of the bidirectional phase change thermal management model includes the following steps: Based on local hotspots, the diffusion path of heat flow within the interface is reconstructed and described. Using the hierarchical phase change thermal management parameters as the configuration path, the thermal response characteristics of different spatial regions are mapped to construct a bidirectional coupled feedback architecture. Based on the bidirectional coupled feedback architecture, the phase change heat in both directions is cross-boundary empowered and linked in the time and space domains, and the heat conduction path weights and regional heat flow distribution coefficients are archived periodically to construct a bidirectional phase change heat management model.

6. The system according to claim 1, characterized in that, The output process of the multi-region phase change heat transfer efficiency improvement scheme includes the following steps: The phase change heat surface is discretized, the heat flux density of each region is extracted, and the local hot spot regions are classified and calibrated according to the heat flux density to determine the thermal management priority of different regions. Based on the aforementioned thermal management priority, the graded phase change thermal management parameters are mapped to each region's phase change unit to construct the corresponding phase change control path; Based on the phase change control path, the enhanced thermal management trigger threshold is defined, and combined with the heat flow distribution relationship of each region, the phase change execution strategy is adaptively set to output a multi-region phase change heat transfer efficiency improvement scheme.

7. The system according to claim 5, characterized in that, The bidirectional coupled feedback architecture includes: a forward phase change driving structure and a reverse heat flow feedback control structure; The execution process of the forward phase change driving structure is as follows: based on the graded phase change thermal management parameters and the enhanced thermal management trigger threshold, the phase change heat transfer surface is divided into regions and prioritized. Combined with the anisotropic thermal conductivity of the artificial diamond superconducting thermal layer, the heat flow conduction path is configured in a directional manner to construct an enhanced thermal conduction channel for hot spot areas, driving the rapid diffusion and redistribution of heat between multiple regions. The execution process of the reverse heat flow feedback control structure is as follows: receiving the heat flow feedback value and regional heat flow distribution coefficient output by the heat flow feedback calculation module, performing inverse analysis on the thermal response state of the current phase change heat surface, and adjusting the heat conduction path weight update and regional heat flow distribution ratio in the bidirectional phase change heat management model in reverse.

8. The system according to claim 1, characterized in that, The process of modeling the inter-regional heat flow orientation and constructing forced convection paths includes the following steps: Each discrete region is abstracted as a network node, and the heat flow transfer intensity and thermal conductivity efficiency are used as edge weights to weight and modulate the heat flow transmission priority of different paths to construct a multi-region heat flow guiding topology network. Based on the heat flow guiding topology network, the heat flow transfer relationship and path weight between each node are globally analyzed to identify the optimal heat transfer path between the high heat flow density region and the low temperature sink region. By adjusting the thermal conductivity directionality, interfacial thermal resistance, and local phase transition triggering state of the artificial diamond superconducting layer, a forced convection path is constructed by enhancing the configuration of the selected path.

9. The system according to claim 1, characterized in that, The process of obtaining the heat flow feedback value by combining the multi-domain performance states at different load stages includes the following steps: The operation process of the forced convection path is divided into load stages, and the multi-domain performance state parameters corresponding to each stage are extracted. Sensible heat transfer, latent heat exchange during phase change, and interfacial thermal resistance loss are used as multi-domain coupled heat flow constraints. The heat flow deviations in different phase change domains are weighted and the heat flow corrections for each region are obtained. The correction results are then filtered in conjunction with the phase lag response relationship to generate heat flow feedback values.

10. The system according to claim 1, characterized in that, The specific execution process of the empowered sample includes the following steps: By introducing phase change heat load sensitivity and heat conduction path information, feature enhancement is performed on the empowered sample to establish a mapping relationship between the empowered sample and the node state in the phase change heat management surface evolution network. Based on the phase change thermal management surface state reconstruction mechanism, the thermal management strategy parameters corresponding to the matching results are extracted and fed back to the phase change thermal management optimization sample library to iterate the bidirectional phase change thermal management model.

11. The system according to claim 1, characterized in that, The closed-loop management optimization logic includes: a strategy decision-making layer and an execution feedback layer; The strategy decision layer is used to: analyze and determine the thermal management requirements of the current heat exchange scenario based on the empowered sample and the graded phase change thermal management parameters, and generate corresponding phase change control strategies; The execution feedback layer is used to: perform actual control execution on the phase change heat exchange surface based on the phase change control strategy, calculate the heat flow feedback value in combination with the multi-domain energy heat balance function, and transmit the heat flow feedback value back to the strategy decision layer.

12. The system according to claim 1, characterized in that, The process by which the phase change thermal management surface state reconstruction mechanism retrieves the phase change thermal evolution stage and response mode includes the following steps: Based on the operational data in the phase change thermal management optimization sample library, the different phase change evolution curves are normalized in intervals, and the strengthening demand level and phase change thermal management surface state are used as auxiliary alignment factors for working condition labeling, resulting in a multi-dimensional strengthening feature vector. The multidimensional enhancement feature vector is flexibly registered on the time axis, and the phase change risk changes of different enhancement operations under the same phase change thermal management surface state conditions are matched in time sequence to retrieve the corresponding phase change thermal evolution stage and response mode.

Images