Diamond single crystal directional heat dissipation monitoring method and system based on artificial intelligence

Abstract

The invention relates to the technical field of heat dissipation monitoring control, and discloses a diamond single crystal directional heat dissipation monitoring method and system based on artificial intelligence. The method comprises the steps of collecting temperature data of the diamond single crystal heat dissipation system through a temperature sensor network, dynamically modeling to generate a heat flow control model, calculating and outputting fan and flow control instructions in a closed loop mode, adjusting the cooling system in real time and obtaining feedback, analyzing temperature deviation to generate correction, automatically optimizing a heat dissipation control device, and achieving intelligent and efficient heat dissipation. Active prediction and dynamic regulation and control of the heat dissipation system are realized through an artificial intelligence technology, and the heat dissipation efficiency and the system reliability are improved.

Description

Technical Field

[0001] This application relates to the technical field of heat dissipation monitoring and control, and particularly to an artificial intelligence-based diamond single crystal directional heat dissipation monitoring method and system. Background Art

[0002] With the rapid development of electronic information technology, electronic components are highly integrated, and the feature size is continuously reduced, resulting in an increasing demand for heat dissipation. Due to the limitations of their own thermal conductivity and thermal expansion coefficient, traditional metal materials and second-generation electronic packaging materials cannot achieve efficient heat dissipation, which in turn leads to a relatively high junction temperature of power devices, affecting the performance and service life of the devices themselves. Diamond is the substance with the highest known thermal conductivity in nature and has a low thermal expansion coefficient, making it have broad application potential in the fields of electronic equipment, aerospace, semiconductors, etc. The diamond / Cu composite material with diamond as the reinforcement is regarded as a new generation of electronic packaging material due to its good thermal conductivity and low thermal expansion coefficient.

[0003] In the current diamond / Cu composite material preparation process, the obtained composite materials exhibit obvious differences in diamond size, low diamond content, and uneven distribution of diamond and copper. These characteristics affect the heat dissipation effect and thermal expansion coefficient of the diamond / Cu composite material. At the same time, traditional heat dissipation monitoring methods lack an accurate grasp of the anisotropic heat dissipation characteristics of diamond single crystals and cannot make full use of the high thermal conductivity characteristics of the 100 crystal plane for directional heat dissipation optimization. Existing monitoring technologies mainly use temperature sensors at fixed positions for measurement, making it difficult to capture the key hot spot changes on complex heat flow paths. In addition, conventional monitoring methods do not have the ability to predict material properties and cannot evaluate and warn of the long-term performance degradation of heat dissipation materials, resulting in passive and lagging maintenance, reducing the reliability and service life of high-power electronic devices. Summary of the Invention

[0004] This application provides an artificial intelligence-based diamond single crystal directional heat dissipation monitoring method and system for realizing the active prediction and dynamic regulation of the heat dissipation system through artificial intelligence technology, improving the heat dissipation efficiency and system reliability.

[0005] In a first aspect, the present application provides an artificial intelligence-based diamond single crystal directional heat dissipation monitoring method. The artificial intelligence-based diamond single crystal directional heat dissipation monitoring method includes: collecting temperature parameter data of the diamond single crystal heat dissipation system through a temperature sensor network to obtain heat dissipation temperature distribution data; dynamically modeling the heat dissipation characteristics of the diamond single crystal based on the heat dissipation temperature distribution data to generate a real-time heat flow control model; performing a closed-loop calculation on heat dissipation control parameters according to the real-time heat flow control model and outputting fan speed and flow control instructions; performing real-time adjustment on the cooling system actuator based on the fan speed and flow control instructions to form heat dissipation control feedback data; performing real-time analysis on the system temperature deviation according to the heat dissipation control feedback data to generate a cooling system parameter correction amount, and automatically adjusting the heat dissipation control device according to the cooling system parameter correction amount.

[0006] In a second aspect, the present application provides an artificial intelligence-based diamond single crystal directional heat dissipation monitoring system. The artificial intelligence-based diamond single crystal directional heat dissipation monitoring system includes: A collection module for collecting temperature parameter data of the diamond single crystal heat dissipation system through a temperature sensor network to obtain heat dissipation temperature distribution data; A modeling module for dynamically modeling the heat dissipation characteristics of the diamond single crystal based on the heat dissipation temperature distribution data to generate a real-time heat flow control model; An output module for performing a closed-loop calculation on heat dissipation control parameters according to the real-time heat flow control model and outputting fan speed and flow control instructions; An adjustment module for performing real-time adjustment on the cooling system actuator based on the fan speed and flow control instructions to form heat dissipation control feedback data; An analysis module for performing real-time analysis on the system temperature deviation according to the heat dissipation control feedback data to generate a cooling system parameter correction amount, and automatically adjusting the heat dissipation control device according to the cooling system parameter correction amount.

[0007] In a third aspect, there is provided an artificial intelligence-based diamond single crystal directional heat dissipation monitoring device, including: a memory and at least one processor, wherein instructions are stored in the memory; the at least one processor invokes the instructions in the memory so that the artificial intelligence-based diamond single crystal directional heat dissipation monitoring device executes the above-mentioned artificial intelligence-based diamond single crystal directional heat dissipation monitoring method.

[0008] In a fourth aspect, there is provided a computer-readable storage medium, in which instructions are stored, and when it runs on a computer, it causes the computer to execute the above-mentioned artificial intelligence-based diamond single crystal directional heat dissipation monitoring method.

[0009] In the technical solution provided by this application, a temperature sensor network is arranged at multiple points on the diamond single crystal heat dissipation system, achieving high-resolution dynamic monitoring of the heat distribution. Compared with the traditional heat dissipation evaluation method that only relies on limited temperature measurement points, it can capture more detailed heat diffusion changes, improving the detection accuracy and response speed of thermal anomalies. Secondly, using dynamic modeling technology, based on the collected heat dissipation temperature distribution data, a real-time heat flow control model is constructed, which can effectively cope with the thermal disturbances brought by system load changes, significantly enhancing the system's adaptability and the real-time performance of heat control. In the closed-loop calculation link, a multi-layer domain adaptation neural network structure is introduced. Through sub-modules such as feature extraction, domain adaptation, and classification, deep control features are extracted from historical control data and the real-time model, and distribution differences are eliminated. This algorithm mechanism not only improves the system's learning ability of the relationship between control input and thermal response but also ensures the robustness of the algorithm model to new working condition environments. Especially in intelligent heat dissipation optimization, the AI model does not just provide a simple regression prediction method but makes causal reasoning-based optimization decisions on heat dissipation control parameters through methods such as domain-invariant feature mining and heat flow path modeling, and then forms an optimal control strategy for fan speed and coolant flow rate. The cooling actuator is driven through PWM control, and fine control is achieved by combining the PID feedback adjustment mechanism. In real-time feedback, the intelligent analysis and correction of the system operating state can also be carried out through the abnormal temperature difference identification mechanism, enabling the cooling system to have a highly automated and intelligent dynamic regulation ability. This invention not only reflects the coordinated control ability of sensors and actuators but also completes the intelligent closed-loop control of the entire process from perception, modeling, decision-making to execution in the thermal management system by introducing artificial intelligence algorithms. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0011] Figure 1 It is a schematic diagram of an embodiment of the diamond single crystal directional heat dissipation monitoring method based on artificial intelligence in the embodiments of this application; Figure 2 It is a schematic diagram of an embodiment of the diamond single crystal directional heat dissipation monitoring system based on artificial intelligence in the embodiments of this application; Figure 3 It is a schematic block diagram of the structure of the diamond single crystal directional heat dissipation monitoring device based on artificial intelligence in the embodiments of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0012] The embodiments of the present application provide a method and system for monitoring the directional heat dissipation of diamond single crystals based on artificial intelligence. The terms "first", "second", "third", "fourth", etc. (if any) in the specification, claims, and the above-mentioned drawings of the present application are used to distinguish similar objects and do not necessarily need to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances so that the embodiments described here can be implemented in an order different from that illustrated or described here. In addition, the term "including" or "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0013] For ease of understanding, the specific process of the embodiments of the present application will be described below. Please refer to Figure 1 , an embodiment of the method for monitoring the directional heat dissipation of diamond single crystals based on artificial intelligence in the embodiments of the present application includes: Step S101: Collect temperature parameter data of the diamond single crystal heat dissipation system through a temperature sensor network to obtain heat dissipation temperature distribution data; Step S102: Dynamically model the heat dissipation characteristics of the diamond single crystal based on the heat dissipation temperature distribution data to generate a real-time heat flow control model; Step S103: Perform a closed-loop calculation on the heat dissipation control parameters according to the real-time heat flow control model, and output fan speed and flow control instructions; Step S104: Based on the fan speed and flow control instructions, perform real-time adjustment on the cooling system actuator to form heat dissipation control feedback data; Step S105: Analyze the system temperature deviation in real time according to the heat dissipation control feedback data to generate a correction amount for the cooling system parameters, and automatically adjust the heat dissipation control device according to the correction amount of the cooling system parameters.

[0014] It can be understood that the execution subject of the present application can be a system for monitoring the directional heat dissipation of diamond single crystals based on artificial intelligence, or a terminal or a server. Specifically, no limitation is made here. The embodiments of the present application will be described by taking the server as the execution subject as an example.

[0015] Specifically, temperature parameter acquisition of the diamond single crystal heat dissipation system is carried out through a temperature sensor network to obtain heat dissipation temperature distribution data. Platinum resistance temperature sensors are arranged at preset positions on the diamond single crystal radiator grown in the 100 crystal plane orientation to form a temperature monitoring point array. The platinum resistance temperature sensor has a measurement range of -50°C to 250°C and an accuracy of ±0.1°C, which is suitable for accurately monitoring the temperature change of the diamond radiator. At the same time, a heat flux density sensor is installed at the interface between the diamond single crystal and the metal to obtain interface heat flux data. The heat flux density sensor adopts the thermopile principle and can directly measure the magnitude of the heat flux passing through a unit area, with a measurement range of 0 - 500 W / cm². The signals collected by the temperature monitoring point array and the heat flux density sensor are synchronously collected and processed to generate an original data stream. The original data stream contains temperature and heat flux data with a sampling frequency of 100 Hz, and each data point is attached with a timestamp and a position identifier. The original data stream is subjected to a three-stage cascaded filtering process, successively performing Butterworth low-pass filtering, median filtering, and Kalman filtering to obtain noise-reduced temperature data. The Butterworth low-pass filter removes high-frequency interference above 50 Hz, the median filter removes burst noise points, and the Kalman filter smooths the data through two stages of prediction and correction. The temperature field spatial distribution matrix, heat flux density vector, and cooling medium parameter vector are calculated based on the noise-reduced temperature data to generate multi-modal heat dissipation parameters. The multi-modal heat dissipation parameters are fused with the sensor position information to output heat dissipation temperature distribution data.

[0016] Dynamically model the heat dissipation characteristics of single-crystal diamond based on heat dissipation temperature distribution data to generate a real-time heat flux control model. Extract the thermal conductivity values of single-crystal diamond at different temperature points from the heat dissipation temperature distribution data and construct a thermal conductivity temperature response matrix. This matrix contains the thermal conductivity values at 36 different temperature points, with the temperature range from 25°C to 200°C and an interval of 5°C. Perform polynomial regression fitting on the thermal conductivity temperature response matrix to obtain continuous thermal conductivity data. The polynomial regression fitting uses the least squares method to determine the coefficients and express the discrete thermal conductivity data points as a continuous function. Decompose the continuous thermal conductivity data according to the crystal orientation for the 100 crystal plane to obtain the principal direction thermal conductivity component and the secondary direction thermal conductivity component. The 100 crystal plane of single-crystal diamond has obvious heat conduction anisotropy, with the principal direction thermal conductivity being approximately 2120 W / (m·K) and the secondary direction thermal conductivity being reduced by 8 - 15%. Discretize the heat conduction characteristics of the diamond / Cu interface region spatially through a three-dimensional interpolation algorithm to form a set of heat conduction boundary conditions. The three-dimensional interpolation algorithm divides the interface region into grids and calculates the local thermal contact conductivity for each grid point to form a set of heat conduction boundary conditions. Integrate the principal direction thermal conductivity component, the secondary direction thermal conductivity component with the set of heat conduction boundary conditions to establish a three-dimensional heat conduction mathematical model for the diamond heat dissipation material. Solve the temperature field distribution of the three-dimensional heat conduction mathematical model under a standard heat load through the finite difference method to generate a real-time heat flux control model. The finite difference method discretizes the continuous region into grids, uses difference approximations for the time and space partial derivatives, and obtains the steady-state temperature distribution through iterative calculations.

[0017] The heat dissipation control parameters are calculated in a closed loop according to the real-time heat flux control model, and the fan speed and flow control commands are output. The real-time heat flux control model and the historical temperature control data are used as the source domain data and the target domain data respectively, and input into a multi-layer domain adaptation neural network structure including a feature extraction layer, a domain adaptation layer and a classification layer. The multi-layer domain adaptation neural network can handle the distribution difference problem between the source domain data and the target domain data, and improve the heat dissipation control accuracy. The source domain data and the target domain data are processed by the wavelet packet decomposition and reconstruction algorithm to generate temperature control feature vectors. The wavelet packet decomposition and reconstruction algorithm reduces signal redundancy through multi-scale analysis while retaining key heat dissipation feature information. The temperature control feature vectors are input into the three-layer feature extraction layer of the multi-layer domain adaptation neural network, with 64, 128 and 256 neurons in each layer in turn. After being processed by the non-linear activation function, deep feature representation data is obtained. For the deep feature representation data, the multi-kernel maximum mean discrepancy algorithm is used in the domain adaptation layer to calculate the distribution difference between the source domain and the target domain, and the network weights are adjusted through backpropagation to obtain domain-invariant control features. The multi-kernel maximum mean discrepancy algorithm quantifies the degree of distribution difference by calculating the distance between two distributions in the reproducing kernel Hilbert space, and helps the network extract domain-invariant features. The domain-invariant control features are input into a three-layer classification network, which includes 128, 64 and 32 neurons respectively, and the maximum probability value is used as the pseudo-label for the target domain data to form a heat dissipation control optimization scheme. Based on the heat dissipation control optimization scheme, the optimal fan speed curve and flow control curve are calculated, and the fan speed and flow control commands are output.

[0018] Based on the fan speed and flow control commands, the actuators of the cooling system are adjusted in real time to form heat dissipation control feedback data. The fan speed and flow control commands are converted into PWM control signals to drive the fan motor and the coolant pump, and the actuators of the cooling system are started. The speed of the fan motor is adjusted in a closed loop by a PID controller. If the deviation between the actual speed and the target speed is greater than 5%, the PWM duty cycle is adjusted until the speed reaches the set value. The flow rate of the coolant pump is adjusted proportionally. When the detected flow deviation exceeds the set threshold, it is judged whether there is a pipeline blockage, and the pump pressure is increased accordingly or a warning signal is issued. The real-time operating parameters of the actuators of the cooling system are collected, including the fan speed value, the flow rate value, the temperature value and the power consumption value, to construct an execution status matrix. The execution status matrix is compared and analyzed with the cooling command to judge the degree of execution deviation and generate an execution effect evaluation value. The execution effect evaluation value is subjected to time series correlation analysis with the temperature monitoring point data to form heat dissipation control feedback data.

[0019] Perform real-time analysis on the system temperature deviation based on the heat dissipation control feedback data, generate the correction amount of the cooling system parameters, and automatically adjust the heat dissipation control device according to the correction amount of the cooling system parameters. Extract the actual temperature distribution characteristics from the heat dissipation control feedback data and construct a real-time temperature feature vector. Calculate the ideal temperature distribution value under the corresponding working conditions according to the theoretical heat dissipation characteristics of the diamond single crystal 100 crystal plane, and generate an ideal temperature reference vector. Calculate the difference between the real-time temperature feature vector and the ideal temperature reference vector to obtain a temperature deviation matrix. If there are three consecutive points in the temperature deviation matrix with a deviation exceeding 10°C, it is determined as an abnormal temperature distribution. Conduct statistical analysis on the temperature deviation matrix, calculate the temperature deviation rate and the proportion of the temperature abnormal area. When the temperature deviation rate exceeds 20%, trigger an emergency cooling program. Based on the temperature deviation rate and the proportion of the temperature abnormal area, perform correction calculation on the working parameters of the cooling system to generate the fan speed correction amount, the flow rate correction amount, and the radiator angle correction amount. Convert the fan speed correction amount, the flow rate correction amount, and the radiator angle correction amount into control signals through the parameter adjustment control unit, and automatically adjust the heat dissipation control device until the temperature deviation rate drops below 5%.

[0020] In the embodiment of the present application, a temperature sensor network is arranged at multiple points on the diamond single crystal heat dissipation system, realizing high-resolution dynamic monitoring of the heat distribution. Compared with the traditional heat dissipation evaluation method that only relies on limited temperature measurement points, it can capture more detailed heat diffusion changes, improving the detection accuracy and response speed of heat anomalies. Secondly, using dynamic modeling technology, based on the collected heat dissipation temperature distribution data, a real-time heat flow control model is constructed, which can effectively cope with the heat disturbance caused by system load changes, significantly enhancing the system's adaptive ability and the real-time performance of heat control. Introduce a multi-layer domain adaptation neural network structure in the closed-loop calculation link. Through sub-modules such as feature extraction, domain adaptation, and classification, extract deep control features from historical control data and the real-time model and eliminate distribution differences. This algorithm mechanism not only improves the system's learning ability of the relationship between control input and heat response, but also ensures the robustness of the algorithm model to the new working condition environment. Especially in intelligent heat dissipation optimization, the AI model does not just provide a simple regression prediction method, but makes causal reasoning-based optimization decisions on heat dissipation control parameters through methods such as domain-invariant feature mining and heat flow path modeling, and then forms an optimal control strategy for the fan speed and coolant flow rate. Drive the cooling actuator through PWM control and combine it with the PID feedback adjustment mechanism to achieve fine control. In real-time feedback, the system operation state can also be intelligently analyzed and corrected through the abnormal temperature difference identification mechanism, enabling the cooling system to have a highly automated and intelligent dynamic regulation ability. The present invention not only reflects the coordinated control ability of sensors and actuators, but also completes the intelligent closed-loop control of the entire process from perception, modeling, decision-making to execution in the thermal management system by introducing artificial intelligence algorithms.

[0021] In a specific embodiment, the process of performing step S101 may specifically include the following steps: Arrange platinum resistance temperature sensors at preset positions on a diamond single crystal heat sink grown with a (100) crystal plane orientation to form a temperature monitoring point array; Install a heat flux density sensor at the interface between the diamond single crystal grown with a (100) crystal plane orientation and the metal to obtain interface heat flux data; Synchronously collect and process the signals collected by the temperature monitoring point array and the heat flux density sensor to generate an original data stream; Perform three-level cascaded filtering on the original data stream, successively perform Butterworth low-pass filtering, median filtering, and Kalman filtering to obtain noise-reduced temperature data; Calculate the temperature field spatial distribution matrix, heat flux density vector, and cooling medium parameter vector based on the noise-reduced temperature data to generate multi-modal heat dissipation parameters; Fuse the multi-modal heat dissipation parameters with the sensor position information and output the heat dissipation temperature distribution data.

[0022] Specifically, arrange platinum resistance temperature sensors at preset positions on a diamond single crystal heat sink grown with a (100) crystal plane orientation to form a temperature monitoring point array. The diamond single crystal grown with a (100) crystal plane orientation refers to a diamond single crystal that preferentially grows along the (100) crystal plane direction of the diamond crystal during the crystal growth process. This crystal plane-oriented diamond has excellent thermal conductivity, with a thermal conductivity as high as 2000 - 2200 W / (m·K). When arranging the temperature sensors, a non-uniform distribution strategy is adopted, and they are deployed according to a density ratio of 4:2:1 in the heat flow convergence area, the heat source contact surface, and the heat dissipation boundary area. Specifically, 8 sensors are densely arranged in the area near the heat source, 4 sensors are arranged on the main heat transfer path, and 4 sensors are arranged in the edge heat dissipation area, with a total of 16 monitoring points forming a temperature monitoring network. The platinum resistance temperature sensor is of the PT100 type, with a wide measurement range of -50°C to 250°C and a high precision of ±0.1°C, and is installed using a micro-drilling and high-thermal-conductivity silver glue fixing method to ensure good thermal contact with the diamond single crystal.

[0023] A heat flux density sensor is installed at the interface between a diamond single crystal grown in the <100> crystal plane and a metal to obtain interface heat flux data. The heat flux density sensor operates based on the thermopile principle and consists of multiple pairs of thermocouples connected in series, capable of directly measuring the magnitude of the heat flux passing through a unit area. An ultra-thin flexible heat flux density sensor with a thickness less than 200 μm, an area of 2 mm × 2 mm, a measurement range of 0 - 500 W / cm², and an accuracy of ±3% is selected. Four heat flux density sensors are installed at the key positions of the interface between the diamond single crystal and the copper substrate to focus on monitoring the interface area through which the heat flux channel passes and capture the interface heat conduction condition. The data of these heat flux density sensors are crucial for identifying changes in interface thermal resistance and early detection of interface degradation problems.

[0024] The signals collected by the temperature monitoring point array and the heat flux density sensors are synchronously collected and processed to generate an original data stream. The synchronous collection is achieved through an industrial-grade data acquisition unit, and the sampling frequency is set to 100 Hz to ensure capturing the transient heat response process. Each data point is attached with a time stamp with millisecond-level accuracy and an accurate spatial position identifier to form a spatio-temporally correlated original data stream. The original data stream contains a temperature data matrix T(x, y, z, t) and a heat flux density matrix q(x, y, z, t), where x, y, and z are spatial coordinates and t is the time coordinate. Due to factors such as environmental electromagnetic interference, sensor self-noise, and measurement random errors, the original data stream usually contains noise components with different frequencies and amplitudes and needs further processing.

[0025] The original data stream is subjected to a three-stage cascaded filtering process, successively performing Butterworth low-pass filtering, median filtering, and Kalman filtering to obtain noise-reduced temperature data. The first-stage Butterworth low-pass filtering mainly deals with high-frequency interference, with a cut-off frequency set to 50 Hz, which can effectively filter out power frequency interference of 50 Hz and above. The Butterworth filter is characterized by a flat amplitude-frequency characteristic in the passband and a moderate transition band width, having a good suppression effect on high-frequency noise in the temperature signal. The second-stage median filtering is aimed at impulse noise and sudden outliers. With a window width of 5 data points, the data sequence is slid, and the median value within the window is taken to replace the central point value, effectively removing outliers and spike noise. The third-stage Kalman filtering estimates the true temperature optimally through two stages of prediction and correction, based on historical data and current measurement values. The Kalman filter is set with a process noise covariance of 0.01 and a measurement noise covariance of 0.1, which can smooth the temperature curve and retain the true temperature change trend. After the three-stage cascaded filtering process, the noise level of the temperature data is reduced from the original ±0.8 °C to ±0.15 °C, significantly improving the signal-to-noise ratio of the temperature data.

[0026] Calculate the spatial distribution matrix of the temperature field, the heat flux density vector, and the cooling medium parameter vector based on the noise-reduced temperature data to generate multi-modal heat dissipation parameters. The spatial distribution matrix of the temperature field is obtained by performing three-dimensional interpolation on the temperature data of 16 monitoring points. The radial basis function interpolation method is used to construct a continuous temperature field within the entire volume of the single-crystal diamond heat sink. The heat flux density vector is calculated from the measured interface heat flux data combined with the temperature gradient information, representing the direction and intensity of heat propagation in the single-crystal diamond. The cooling medium parameter vector includes key parameters such as the cooling fan speed, coolant flow rate, and ambient temperature, describing the working state of the heat dissipation system. The multi-modal heat dissipation parameters integrate various data such as the temperature field distribution matrix T(x, y, z), the temperature gradient matrix ∇T(x, y, z), the heat flux density vector q(x, y, z), and the cooling medium parameter vector C(t) to form a comprehensive description of the heat dissipation state.

[0027] Fuse the multi-modal heat dissipation parameters with the sensor position information and output the heat dissipation temperature distribution data. The data fusion process uses the weighted average and covariance intersection verification methods to assign weights according to the accuracy and position importance of each sensor to generate more accurate temperature distribution data. The sensor position information includes three-dimensional spatial coordinates and the identification of the region type where the sensor is located, which is used to map the data of discrete measurement points into a continuous space. The heat dissipation temperature distribution data after fusion processing is output in JSON format, including metadata fields (time, location, sensor type, etc.) and data fields (temperature value, heat flux density value, certainty evaluation, etc.). At the same time, calculate the temperature change rate ∂T / ∂t and the spatial temperature gradient ∇T(x, y, z) as supplementary fields of the data packet.

[0028] In a specific embodiment, the process of executing step S102 may specifically include the following steps: Extract the thermal conductivity values of the single-crystal diamond at different temperature points from the heat dissipation temperature distribution data and construct a thermal conductivity temperature response matrix; Perform polynomial regression fitting on the thermal conductivity temperature response matrix to obtain continuous thermal conductivity data; According to the anisotropic characteristics of heat conduction in the 100 crystal plane orientation, decompose the continuous thermal conductivity data by crystal direction to obtain the principal direction thermal conductivity component and the secondary direction thermal conductivity component; Perform spatial discretization on the heat conduction characteristics of the diamond / Cu interface region through a three-dimensional interpolation algorithm to form a set of heat conduction boundary conditions; Integrate the principal direction thermal conductivity component, the secondary direction thermal conductivity component, and the set of heat conduction boundary conditions to establish a three-dimensional heat conduction mathematical model of the diamond heat dissipation material; Solve the temperature field distribution of the three-dimensional heat conduction mathematical model under the standard heat load by the finite difference method to generate a real-time heat flow control model.

[0029] Specifically, the thermal conductivity values of single-crystal diamond at different temperature points are extracted from the heat dissipation temperature distribution data to construct a thermal conductivity temperature response matrix. This is achieved through reverse calculation, where the thermal conductivity of the material is deduced from the known temperature field distribution and heat flux data. In specific operations, using Fourier's law of heat conduction, the thermal conductivity at each point is calculated under the conditions of known temperature gradient and heat flux density. Fourier's law of heat conduction states that the heat flux density is equal to the product of the thermal conductivity and the temperature gradient. By performing finite difference calculations on the temperature field around the temperature monitoring points, the temperature gradient value is obtained, and then combined with the data measured by the heat flux density sensor, the thermal conductivity at the corresponding position is calculated. This calculation is carried out in the temperature range from 25°C to 200°C, with 36 temperature points set at intervals of 5°C. Each temperature point is calculated three times and the average value is taken to form a 36×2 thermal conductivity temperature response matrix, with the first column being the temperature value and the second column being the corresponding thermal conductivity value.

[0030] Polynomial regression fitting is performed on the thermal conductivity temperature response matrix to obtain continuous thermal conductivity data. Polynomial regression fitting uses the least squares method to fit discrete thermal conductivity data points into a continuous function. First, the order of the polynomial is determined, generally choosing from 3 to 5 orders to balance the fitting accuracy and computational complexity. Then, a least squares problem is constructed to find the polynomial coefficients that minimize the sum of the squares of the residuals. The sum of the squares of the residuals is calculated as the sum of the squares of the differences between the measured values and the fitted values. In specific operations, the temperature values are standardized to avoid numerical instability problems caused by overly large coefficients of high-order terms, and then the QR decomposition or singular value decomposition is used to solve the least squares problem to obtain the polynomial coefficients. The finally obtained polynomial function can calculate the corresponding thermal conductivity at any temperature point, forming continuous thermal conductivity data. The continuous thermal conductivity data is represented in the form of a function, which can more accurately describe the continuous variation law of the thermal conductivity with temperature compared to the original discrete data points.

[0031] According to the anisotropic characteristics of heat conduction based on the 100 crystal plane orientation, the continuous heat conductivity data is decomposed by crystal orientation to obtain the main direction heat conductivity component and the secondary direction heat conductivity component. Diamond single crystal is a typical anisotropic material, that is, the heat conduction performance varies in different directions. The diamond single crystal with a 100 crystal plane orientation has the best heat conduction performance in the direction perpendicular to the crystal plane, which is called the main direction; while the heat conduction performance in the direction within the crystal plane is lower, which is called the secondary direction. By analyzing the heat conductivity data measured in different directions, a relationship model between crystal orientation and heat conductivity is established. In specific operations, using the principle of crystal symmetry, it is assumed that there is a certain proportional relationship between the heat conductivity in the main direction and the secondary direction. By solving the decoupled equations, the measured heat conductivity value is decomposed into the main direction component and the secondary direction component. The decomposition process needs to consider the angle of the measurement point relative to the crystal principal axis. Through coordinate transformation and mathematical transformation, the main direction heat conductivity component and the secondary direction heat conductivity component at different temperatures are obtained. This decomposition makes the heat conduction model more accurate and can reflect the heat conduction characteristics of the material in different directions.

[0032] The spatial discretization of the heat conduction characteristics in the diamond / Cu interface region is carried out by a three-dimensional interpolation algorithm to form a set of heat conduction boundary conditions. The diamond / Cu interface is a key region in the heat dissipation system, and the interface thermal resistance has a significant impact on the overall heat dissipation performance. The three-dimensional interpolation algorithm is used to construct a continuous distribution of the interface heat conduction characteristics from limited measurement points. First, the interface region is divided into a three-dimensional grid, with a typical grid size of 80×80×30 and a total number of nodes of 192,000. Then, for each grid point, based on the values of the surrounding measurement points, interpolation calculation is used to obtain the heat conduction characteristics of this point. The interpolation method usually adopts trilinear interpolation or radial basis function interpolation, considering the spatial distance and measurement value difference between points. During the calculation process, the physical characteristics of the interface, such as roughness, wetting angle, and bonding strength, etc., also need to be considered, as these factors will affect the local heat conduction performance. The finally obtained set of heat conduction boundary conditions contains the thermal contact heat conductivity of each grid point in the interface region, forming a complete description of the boundary conditions.

[0033] Integrate the main-direction thermal conductivity component, the sub-direction thermal conductivity component, and the set of heat conduction boundary conditions to establish a three-dimensional heat conduction mathematical model for diamond heat dissipation materials. This integration process needs to consider the physical properties, geometric structure, and boundary conditions of the material to construct a complete heat conduction equation. First, set the geometric boundaries of the computational domain, including single-crystal diamond, copper matrix, and their interface regions. Then, use the corresponding material properties within each region. For example, in the diamond region, use the decomposed main-direction and sub-direction thermal conductivity components, in the copper matrix region, use the thermal conductivity of copper, and in the interface region, use the thermal contact conductivity in the set of heat conduction boundary conditions. For the unsteady heat conduction case, a partial differential equation of heat conduction containing a time term needs to be constructed to describe the variation of the temperature field with time and space. For the case with heat sources, a heat source term also needs to be added to represent the heat generation rate per unit volume. After integrating these conditions, a complete three-dimensional heat conduction mathematical model is formed, which can accurately describe the heat conduction behavior of diamond heat dissipation materials under actual working conditions.

[0034] Solve the temperature field distribution of the three-dimensional heat conduction mathematical model under a standard heat load by the finite difference method to generate a real-time heat flow control model. The finite difference method is a commonly used method in numerical calculations. Its basic idea is to discretize the continuous region into a grid and use the difference approximation to replace the derivative terms in the differential equation. In the specific implementation, grid division is adopted to divide the computational domain into a large number of discrete grid cells, and the center of each grid cell is the computational node. For the time derivative in the heat conduction equation, forward difference is used, and for the spatial derivative, central difference is used to construct the solution format. For the boundary conditions, corresponding difference forms are adopted according to their types. For example, for the first kind of boundary condition, the temperature of the boundary node is directly specified; for the second kind of boundary condition, symmetric difference is used to handle the heat flow boundary; and for the third kind of boundary condition, the boundary heat exchange coefficient is considered. During the calculation process, first set the initial temperature field, and then iterate and solve according to the time step until the steady state condition or the specified termination time is reached. The steady state judgment criterion is usually that the maximum difference between the temperature fields of two adjacent iterations is less than a preset threshold. The final solution result generates a real-time heat flow control model, which can accurately predict the temperature distribution and heat flow path inside the diamond heat dissipation materials under the given heat load conditions.

[0035] In a specific embodiment, the process of executing step S103 may specifically include the following steps: Use the real-time heat flow control model and the historical temperature control data as the source domain data and the target domain data respectively, and input them into a multi-layer domain adaptation neural network structure including a feature extraction layer, a domain adaptation layer, and a classification layer; Process the source domain data and the target domain data through the wavelet packet decomposition and reconstruction algorithm to generate temperature control feature vectors; Input the temperature control feature vector into the three-layer feature extraction layer of the multi-layer domain adaptation neural network. Each layer contains 64, 128, and 256 neurons in sequence. After being processed by the non-linear activation function, the deep feature representation data is obtained; For the deep feature representation data, calculate the distribution difference between the source domain and the target domain using the multi-kernel maximum mean discrepancy algorithm in the domain adaptation layer, and adjust the network weights through backpropagation to obtain the domain-invariant control feature; Input the domain-invariant control feature into the three-layer classification network, which contains 128, 64, and 32 neurons respectively, and use the maximum probability value as the pseudo-label for the target domain data to form the heat dissipation control optimization scheme; Based on the heat dissipation control optimization scheme, calculate the optimal fan speed curve and flow control curve, and output the fan speed and flow control instructions.

[0036] Specifically, input the real-time heat flow control model and historical temperature control data as the source domain data and the target domain data respectively into the multi-layer domain adaptation neural network structure including the feature extraction layer, the domain adaptation layer, and the classification layer. The source domain data refers to the ideal heat flow distribution and temperature field data generated by the theoretical model, including the anisotropic heat conduction characteristics, interface thermal resistance characteristics, and ideal heat dissipation effect of single-crystal diamond. The target domain data is the historical temperature control data collected during the operation of the actual heat dissipation system, which reflects the heat dissipation performance in the actual working environment. The multi-layer domain adaptation neural network is a deep learning architecture specifically used to handle the problem of inconsistent data distributions between the source domain and the target domain, and can learn domain-invariant features to improve the applicability of the model in the actual environment. This network structure includes three main parts: the feature extraction layer is used to extract deep features from the original data; the domain adaptation layer is responsible for reducing the distribution difference between the source domain and the target domain; the classification layer makes control decisions based on the extracted features.

[0037] Process the source domain data and the target domain data through the wavelet packet decomposition and reconstruction algorithm to generate the temperature control feature vector. Wavelet packet decomposition is an extension of the traditional wavelet transform, which can provide a finer frequency division. The processing process first determines the decomposition level, usually 3 to 4 layers, and then performs multi-layer decomposition on the original signal to obtain the coefficients of different frequency sub-bands. For temperature data, this decomposition can effectively separate different scale features of temperature changes, such as rapid fluctuations, medium-term changes, and long-term trends. After decomposition, according to indicators such as energy distribution and information entropy, select the sub-bands containing the main feature information for reconstruction. The reconstruction process is the inverse process of decomposition, recombining the retained sub-band coefficients to generate a temperature signal with significant features but reduced redundant information. The data processed in this way forms the temperature control feature vector, which retains the key dynamic characteristics of the original data while reducing the computational amount and noise interference.

[0038] Input the temperature control feature vector into the three-layer feature extraction layer of the multi-layer domain adaptation neural network. Each layer contains 64, 128, and 256 neurons in sequence. After being processed by the non-linear activation function, the deep feature representation data is obtained. The feature extraction layer adopts the structure of a deep convolutional neural network. The 64 neurons in the first layer are responsible for extracting low-level features such as temperature gradient and heat flow direction; the 128 neurons in the second layer combine low-level features to form middle-level features such as heat conduction paths and temperature distribution patterns; the 256 neurons in the third layer extract high-level abstract features such as the state of the heat dissipation system and abnormal patterns. Neurons between each layer are connected through non-linear activation functions. Commonly used activation functions include ReLU or Leaky ReLU, which can introduce non-linear transformations and enhance the expressive power of the network. The data processing path of the feature extraction layer is: original feature vector → first layer convolution + activation → second layer convolution + activation → third layer convolution + activation → deep feature representation data. After layers of transformation, the original temperature control feature vector is transformed into a representation in the high-dimensional abstract feature space, containing the essential characteristics of the heat dissipation system.

[0039] For the deep feature representation data, the multi-kernel maximum mean discrepancy algorithm is used in the domain adaptation layer to calculate the distribution difference between the source domain and the target domain. The network weights are adjusted through backpropagation to obtain domain-invariant control features. The multi-kernel maximum mean discrepancy algorithm is a method for measuring the difference between two distributions. Its basic principle is to calculate the distance between two distributions in the reproducing kernel Hilbert space. The algorithm first maps the deep features to a high-dimensional space, and then calculates the mean difference between the source domain and target domain features in this space. By using a combination of multiple kernel functions (such as Gaussian kernel, polynomial kernel, etc.), it can capture distribution differences at different scales and forms. The calculated distribution difference is used as one of the loss functions of the network. Through the backpropagation algorithm, the network parameters are adjusted with the goal of minimizing the distribution difference between the source domain and the target domain. During the backpropagation process, the network weights are updated in the gradient direction, gradually reducing the distribution difference. After multiple rounds of training iterations, the features learned by the network have similar distributions between the source domain and the target domain, that is, domain-invariant control features. These features can be applied to both the theoretical model and the actual environment. The domain-invariant control features are input into a three-layer classification network, which contains 128, 64, and 32 neurons respectively. The maximum probability value is used as the pseudo-label for the target domain data to form a heat dissipation control optimization scheme. The classification network adopts a fully connected neural network structure. Starting from the domain-invariant control features, through three layers of non-linear transformations, it finally outputs the heat dissipation control decision. The first layer of 128 neurons processes the domain-invariant features, the second layer of 64 neurons performs intermediate representation, and the third layer of 32 neurons generates the control output. Since the target domain data usually lacks control labels, the pseudo-label strategy is used to solve the unsupervised learning problem. The specific approach is to first train the initial model with the labeled data in the source domain, and then apply the model to the target domain data. For each target domain sample, the classification network will output multiple possible control schemes and their probability values, and the scheme with the highest probability is selected as the pseudo-label for this sample. With the pseudo-labels, the target domain data can participate in the network training to further improve the model performance. The training process uses cross-validation, dividing the dataset into training set, validation set, and test set to ensure the generalization ability of the model. The finally trained network can directly output the heat dissipation control optimization scheme according to the input heat dissipation state data.

[0040] Calculate the optimal fan speed curve and flow control curve based on the heat dissipation control optimization scheme, and output the fan speed and flow control commands. The heat dissipation control optimization scheme includes the control parameters of each actuator in the heat dissipation system, and needs to be further processed and converted into specific control commands. First, according to the control objectives and constraints in the heat dissipation control optimization scheme, construct the optimization problems for fan speed and flow control. The optimization objectives are usually to minimize the temperature deviation and energy consumption, and the constraints include the fan speed range, flow limit, noise level, etc. Then, by solving the optimization problems, obtain the fan speed curve and flow control curve that vary with time. The curve generation process takes into account the inertial characteristics of the thermal system to avoid frequent jumps in the control output. Finally, discretize the continuous control curves into an instruction sequence that can be executed by the actual control system, including the fan PWM control signal and the flow valve opening signal. These control commands are sent to the actuators of the heat dissipation system through the control interface to achieve precise control of the heat dissipation process.

[0041] In a specific embodiment, the process of performing the step of calculating the optimal fan speed curve and flow control curve based on the heat dissipation control optimization scheme may specifically include the following steps: Convert the heat dissipation control optimization scheme into a thermal resistance network topology diagram, and mark the heat source points and heat dissipation boundary points of the diamond single crystal. According to the anisotropic characteristics of the 100 crystal plane of the diamond single crystal, assign heat transfer weight values to each node in the thermal resistance network topology diagram. Starting from the heat source point, traverse the thermal resistance network topology diagram using the heat path tracing algorithm to calculate all possible heat transfer paths. Perform cumulative thermal resistance calculations on all possible heat transfer paths to obtain a set of path thermal resistance values. Select the heat transfer channel with the minimum thermal resistance from the set of path thermal resistance values as the main heat dissipation channel, and at the same time select the channel with the second-lowest thermal resistance as the auxiliary heat dissipation channel. According to the distribution ratio that the main heat dissipation channel undertakes 70% of the heat load and the auxiliary heat dissipation channel undertakes 30% of the heat load, calculate the required heat dissipation power for each channel, and combine the temperature gradient value and the required heat flow rate on each channel to determine the fan speed value and the coolant flow rate value, and output the fan speed and flow control commands.

[0042] Specifically, the temperature distribution and heat flux data in the optimization solution are mapped onto the thermal resistance network through data structure conversion. The thermal resistance network topology diagram is a graphical structure representing the heat dissipation path, composed of nodes and connections. Each node represents a spatial position in the diamond single crystal, and the connections represent the paths of heat transfer. The conversion process uses a grid division method to divide the diamond single crystal heat sink into fine grids. A typical grid size is 80×80×30, with a total of 192,000 nodes. The center of each grid cell serves as a node in the thermal resistance network, and a connection relationship is established between adjacent nodes. For the diamond single crystal heat dissipation system, two types of special nodes need to be clearly marked: the heat source point and the heat dissipation boundary point. The heat source point refers to the area where the high-power electronic device contacts the diamond single crystal, which is the position of heat input; the heat dissipation boundary point is the outer surface point where the diamond single crystal contacts the heat sink or the environment, which is the position of heat output. These special nodes are identified through geometric boundary conditions and temperature threshold criteria. For example, the area where the temperature exceeds the set threshold is marked as the heat source point, and the outer surface area where the temperature is lower than another threshold is marked as the heat dissipation boundary point. According to the anisotropic characteristics of the 100 crystal plane of the diamond single crystal, a heat transfer weight value is assigned to each node in the thermal resistance network topology diagram. The diamond single crystal is a typical anisotropic thermal conductive material, and there are significant differences in its thermal conductivity in different crystal orientations. The 100 crystal plane refers to a specific crystal plane of the diamond crystal, which is characterized by the best thermal conductivity along the direction perpendicular to the crystal plane and lower thermal conductivity in the direction within the crystal plane. For each node in the thermal resistance network, it is necessary to calculate the direction angle of its position relative to the main axis of the diamond crystal, and then assign a heat transfer weight value according to this angle. The weight assignment uses the cosine square function, that is, the weight value is proportional to the square of the cosine of the angle between the crystal orientation and the main direction. In specific operations, first, the crystal orientation of the diamond single crystal is determined through X-ray diffraction or electron backscatter diffraction technology to obtain the relationship between the crystal main axis and the heat sink coordinate system. Then, the angle between the heat transfer direction and the crystal main axis is calculated for each node. The weight value along the main direction is set to 1.0, and the weight value perpendicular to the main direction is set according to the measured anisotropy ratio, usually about 0.8. In this way, each node in the thermal resistance network has a directional weight, reflecting the anisotropic thermal conductive characteristics of the diamond single crystal.

[0043] Starting from the heat source point, the thermal resistance network topology diagram is traversed using the thermal path tracing algorithm to calculate all possible heat transfer paths. The thermal path tracing algorithm is a special algorithm based on graph traversal and is applicable to finding heat transfer paths in a heat dissipation system. This algorithm is similar to the traditional Dijkstra shortest path algorithm but has improvements specifically for heat conduction characteristics. The algorithm starts from the marked heat source point and gradually explores the paths to each heat dissipation boundary point. For each current node, the algorithm considers all adjacent nodes, calculates the thermal resistance values for heat transfer to each adjacent node, and then selects the node with the minimum thermal resistance as the next exploration point. To avoid the algorithm falling into local optima, a simulated annealing strategy is also adopted, that is, there is a certain probability of selecting non-optimal paths during the exploration process. Due to the complexity of the thermal resistance network, the computational effort for a complete traversal of all possible paths is huge. In actual operation, a pruning strategy is usually adopted, and only the paths with thermal resistance less than a certain threshold are retained. After traversal and calculation, multiple possible heat transfer paths from the heat source point to each heat dissipation boundary point are obtained. The thermal resistance cumulative calculation is performed on all possible heat transfer paths to obtain the set of path thermal resistance values. The thermal resistance cumulative calculation is the process of adding up the thermal resistance values between all nodes on each heat transfer path. For any two adjacent nodes on the path, the thermal resistance between the nodes is determined by the distance between the nodes, the average thermal conductivity, and the effective heat transfer cross-sectional area. The calculation formula is that the thermal resistance between nodes is equal to the distance between nodes divided by (the average thermal conductivity multiplied by the effective heat transfer cross-sectional area). Among them, the distance between nodes is the Euclidean distance between the two nodes in space; the average thermal conductivity takes into account the temperatures and crystal orientations of the two nodes and is obtained through interpolation calculation; the effective heat transfer cross-sectional area is the cross-sectional area perpendicular to the heat flow direction. For the heat transfer path passing through the interface, the interface thermal resistance also needs to be added, and the interface thermal resistance is determined by the thermal contact conductivity and the contact area of the interface. Through this cumulative calculation, the total thermal resistance value of each heat transfer path is obtained, forming the set of path thermal resistance values.

[0044] The heat transfer channel with the minimum thermal resistance is selected from the set of path thermal resistance values as the main heat dissipation channel, and at the same time, the channel with the second-lowest thermal resistance is selected as the auxiliary heat dissipation channel. In the screening process, the set of path thermal resistance values is first sorted in ascending order of thermal resistance, and the path with the minimum thermal resistance is selected as the candidate for the main heat dissipation channel. However, simply selecting the path with the minimum thermal resistance may cause the heat flow to be overly concentrated, resulting in local hot spots. Therefore, when selecting the auxiliary heat dissipation channel, not only the thermal resistance value but also the spatial distribution factor needs to be considered. In specific operations, the spatial distance between the candidate path and the main heat dissipation channel is calculated, and a path with both a relatively low thermal resistance and an appropriate spatial separation from the main heat dissipation channel is selected as the auxiliary heat dissipation channel. The purpose of spatial separation is to balance the heat dissipation load and avoid local overheating. Usually, 1 - 3 main heat dissipation channels and 3 - 5 auxiliary heat dissipation channels are selected to form a multi-path parallel heat dissipation path network.

[0045] According to the distribution ratio that the main heat dissipation channel undertakes 70% of the heat load and the auxiliary heat dissipation channel undertakes 30% of the heat load, calculate the required heat dissipation power for each channel. Combine the temperature gradient value and the required heat flux on each channel to determine the fan speed value and the coolant flow rate value, and output the fan speed and flow control commands. The heat load distribution is based on the thermal resistance ratio of the main heat dissipation channel and the auxiliary heat dissipation channel. The main heat dissipation channel has a lower thermal resistance and is allocated more heat load; the auxiliary heat dissipation channel has a higher thermal resistance and is allocated less heat load. Calculate the required heat dissipation power for each channel through heat flow distribution. The formula is that the channel heat dissipation power is equal to the total heat load multiplied by the load distribution ratio of this channel. The temperature gradient value on each channel is obtained by dividing the temperature difference between the heat source point and the heat dissipation boundary point by the path length. Combine the required heat flux and the temperature gradient of each channel to calculate the required heat dissipation capacity. There is a mapping relationship between the heat dissipation capacity and the fan speed and the coolant flow rate, which is usually obtained through a look-up table established by experimental measurement. The input of the look-up table is the required heat dissipation capacity, and the output is the corresponding fan speed value and coolant flow rate value. The fan speed value is usually expressed in RPM and is converted into a PWM control signal; the coolant flow rate value is expressed in L / min and is converted into the opening signal of the flow control valve. These control commands are sent to the corresponding actuators to achieve precise control of the heat dissipation system.

[0046] In a specific embodiment, the process of executing step S104 may specifically include the following steps: Convert the fan speed and flow control commands into PWM control signals to drive the fan motor and the coolant pump, and start the cooling system actuator; Perform closed-loop regulation of the fan motor speed by a PID controller. If the deviation between the actual speed and the target speed is greater than 5%, adjust the PWM duty cycle until the speed reaches the set value; Perform proportional regulation on the coolant pump flow rate. When it is detected that the flow deviation exceeds the set threshold, judge whether there is a pipeline blockage, and increase the pump pressure or issue a warning signal; Collect the real-time operating parameters of the cooling system actuator, including the fan speed value, the flow rate value, the temperature value, and the power consumption value, and construct an execution status matrix; Compare and analyze the execution status matrix with the cooling command to judge the degree of execution deviation and generate an execution effect evaluation value; Perform time-series correlation analysis on the execution effect evaluation value and the data of the temperature monitoring point to form heat dissipation control feedback data.

[0047] Specifically, the process of converting the fan speed and flow control commands into PWM control signals involves signal conversion and modulation. The fan speed and flow control commands need to be converted into specific execution signals to drive the actuators of the cooling system. This conversion process is achieved through a control signal mapping table, which is the correspondence between the fan speed and the PWM duty cycle, and between the flow value and the pump speed. For the fan speed, the typical mapping relationship is linear or quadratic function, and the speed range is generally 1000 - 4000 RPM, with the corresponding PWM duty cycle range of 20% - 100%. The calculation formula is: PWM duty cycle = (current set speed - minimum speed) / (maximum speed - minimum speed) × 80% + 20%. For the coolant pump, the control methods include variable frequency speed regulation and PWM modulation, and the flow range is usually 0.5 - 3.0 L / min, and the corresponding control signal is also given by the mapping table. The converted PWM signal is output through a drive circuit, which includes a power amplification and level conversion part to ensure that the control signal can correctly drive the fan motor and the coolant pump. The frequency selection of the PWM signal needs to consider the response characteristics of the actuator. The PWM frequency for fan control is usually 25 kHz, avoiding the audible range of the human ear and meeting the motor control requirements at the same time; the control frequency of the coolant pump is determined according to the pump type, and the variable frequency pump is generally adjusted in the range of 0 - 60 Hz.

[0048] The closed-loop regulation of the fan motor speed is carried out by a PID controller. The PID controller consists of three parts: proportional, integral, and derivative, and is used to adjust the control output according to the feedback information. During the fan control process, first, the actual speed value is obtained from the fan speed sensor, and then compared with the target speed value to calculate the speed deviation. When the deviation is greater than the set threshold (5%), the PID controller calculates the increment of the PWM duty cycle to be adjusted according to the magnitude and change trend of the deviation. The control formula is: PWM duty cycle increment = Kp × deviation + Ki × integral of deviation + Kd × rate of change of deviation. Where Kp, Ki, and Kd are the proportional, integral, and derivative parameters respectively, and need to be adjusted according to the dynamic characteristics of the fan. For most cooling fans, the typical PID parameter values are: Kp = 0.8, Ki = 0.05, Kd = 0.1. For example, when the target speed is 3000 RPM and the actually measured speed is 2700 RPM, the deviation is 300 RPM, accounting for 10%, exceeding the 5% threshold, triggering the PID regulation. Assuming the current PWM duty cycle is 60%, the calculated duty cycle increment is 5%, then the adjusted PWM duty cycle is 65%. The PID control process is iterative, with sampling and adjustment performed every 100 ms until the actual speed reaches the target value or the deviation is less than the threshold.

[0049] The proportional regulation of the coolant pump flow rate adopts a feedback control strategy that is simpler than motor speed control. Flow control first obtains the actual flow rate value through a flow sensor and compares it with the target flow rate. When the detected flow deviation exceeds the set threshold, the control system analyzes the cause of the deviation. Different from the fan, abnormal coolant flow often means that there may be faults such as pipeline blockage in the system. The method to judge whether there is pipeline blockage is to monitor the flow rate and pressure data simultaneously and analyze through the flow rate-pressure relationship curve. Under normal circumstances, the flow rate is proportional to the pump speed and inversely proportional to the pipeline resistance. If the flow rate does not increase significantly after increasing the pump speed while the pressure increases significantly, it is determined that there may be pipeline blockage. At this time, there are two control strategies: for slight blockage, appropriately increase the pump pressure to maintain the target flow rate; for severe blockage, issue a warning signal to avoid pump overload damage. The method of increasing the pump pressure varies according to the pump type. For variable-frequency pumps, the pump pressure is increased by increasing the frequency, and for PWM control pumps, it is achieved by increasing the duty cycle. The warning signal is sent to the control center through the system bus to trigger the maintenance process.

[0050] Collecting the real-time operating parameters of the cooling system actuator is to comprehensively monitor the working state of the heat dissipation system. The real-time operating parameters include fan speed value, flow rate value, temperature value, and power consumption value, and these data are collected through dedicated sensors. The fan speed is measured by a Hall sensor or an optical encoder with an accuracy of ±1%; the flow rate value is obtained through a turbine flowmeter or an ultrasonic flowmeter with an accuracy of ±2%; the temperature value comes from temperature sensors distributed throughout the system; the power consumption is calculated through a precision current detection circuit. All parameters are sampled according to a unified time reference with a sampling frequency of 10Hz to ensure the time consistency of the data. These parameters are processed to form an execution status matrix, and the matrix structure is n rows and 4 columns, where n is the number of sampling points, and the 4 columns correspond to the fan speed, flow rate, temperature, and power data respectively. The execution status matrix is a digital mapping of the system operating state and provides a data basis for subsequent analysis.

[0051] Comparing and analyzing the execution status matrix with the cooling instruction to determine the degree of execution deviation is an important step in evaluating the effectiveness of heat dissipation control. The comparison and analysis first calculate the differences between the execution status and the control instruction, including fan speed deviation, flow rate deviation, temperature deviation, and power consumption deviation. Each deviation has a corresponding weight coefficient, reflecting its importance to the heat dissipation effect. The deviation calculation uses the weighted mean square formula to comprehensively consider the differences of various parameters. The calculated deviation value is normalized to obtain an execution deviation metric value between 0 and 1. According to the size of the deviation metric value, the execution effect is divided into four grades: excellent, good, medium, and poor, corresponding to the cases where the deviation value is less than 0.1, 0.1 - 0.2, 0.2 - 0.3, and greater than 0.3 respectively. The execution effect evaluation not only considers the static deviation but also includes dynamic response characteristics, such as rise time, overshoot, and settling time, etc. These indicators are obtained by analyzing the time series characteristics of the execution status matrix. The finally generated execution effect evaluation value is a comprehensive index, comprehensively reflecting the accuracy and responsiveness of the heat dissipation control.

[0052] Performing time - series correlation analysis on the execution effect evaluation value and the temperature monitoring point data to form heat dissipation control feedback data is the data basis for closed - loop control. The time - series correlation analysis first determines the analysis window. The typical window width is 60 seconds, and the sliding step is 10 seconds. Within each window, calculate the correlation between the execution effect evaluation value and the temperature change, including lag correlation and prediction correlation. The lag correlation analyzes the response delay of the temperature after the control action, usually calculating the time delay corresponding to the maximum correlation through the cross - correlation function. The prediction correlation analyzes the influence of the current control state on the future temperature, using the autoregressive moving average model to predict the future temperature change. The correlation analysis results are combined with the actual temperature data to form complete heat dissipation control feedback data. The feedback data structure includes fields such as the current temperature state, temperature change trend, control execution state, execution effect evaluation, temperature - control correlation, and system anomaly flag, etc. These data are integrated according to the unified timestamp and stored in JSON format.

[0053] In a specific embodiment, the process of executing step S105 may specifically include the following steps: Extract the actual temperature distribution characteristics from the heat dissipation control feedback data to construct a real - time temperature feature vector; According to the theoretical heat dissipation characteristics of the 100 crystal plane of the diamond single crystal, calculate the ideal temperature distribution value under the corresponding working conditions to generate an ideal temperature reference vector; Perform a difference calculation on the real - time temperature feature vector and the ideal temperature reference vector to obtain a temperature deviation matrix. If there are three consecutive points in the temperature deviation matrix with a deviation exceeding 10°C, it is determined as an abnormal temperature distribution; Perform statistical analysis on the temperature deviation matrix, calculate the temperature deviation rate and the proportion of the temperature anomaly area. When the temperature deviation rate exceeds 20%, trigger the emergency cooling procedure; Based on the temperature deviation rate and the proportion of the temperature anomaly area, perform correction calculations on the working parameters of the cooling system to generate the fan speed correction amount, the flow rate correction amount, and the radiator angle correction amount; Through the parameter adjustment control unit, convert the fan speed correction amount, the flow rate correction amount, and the radiator angle correction amount into control signals to automatically adjust the heat dissipation control device.

[0054] Specifically, extract features from the original temperature data. The heat dissipation control feedback data includes the sequential temperature data collected by the temperature sensor network, and these data come from the temperature sensors distributed at each key position of the single-crystal diamond radiator. The feature extraction process uses the principal component analysis method to reduce the dimensionality of the high-dimensional temperature data and extract the main temperature distribution patterns. In specific operations, first organize the temperature data into a matrix form, where each row represents a time point and each column represents a sensor position. Then calculate the covariance matrix of this matrix, and solve the eigenvalues and eigenvectors of the covariance matrix. The magnitude of the eigenvalues represents the importance of the corresponding eigenvectors. Usually, select the eigenvectors corresponding to the first k largest eigenvalues as the principal components. A typical value of k is 4 - 6, which can capture more than 90% of the variability in the original data. By projecting the original temperature data onto these k principal components, a low-dimensional feature representation is obtained. In addition to principal component analysis, it is also necessary to calculate the time derivative and spatial gradient of the temperature field, which represent the rate of change of temperature and the non-uniformity of spatial distribution, respectively. Finally, the real-time temperature feature vector includes multiple features such as principal component coefficients, temperature change rate, and temperature gradient, comprehensively describing the temperature state of the current heat dissipation system.

[0055] According to the theoretical heat dissipation characteristics of the 100 crystal plane of a diamond single crystal, calculate the ideal temperature distribution value under the corresponding working conditions, and generate an ideal temperature reference vector. As a material with extremely high thermal conductivity, the heat dissipation characteristics of a diamond single crystal are closely related to the crystal orientation. The 100 crystal plane is a specific crystal plane of the diamond crystal, and the thermal conductivity is the highest in this direction. The calculation of the ideal temperature distribution is based on the heat conduction theory model, considering the anisotropic thermal conductivity, interface thermal resistance, and boundary conditions of the diamond single crystal. The calculation process first determines the current working load conditions, including the heat source power, ambient temperature, and cooling conditions. Then, by solving the heat conduction equation, the ideal temperature values at each monitoring point under steady-state conditions are calculated. The heat conduction equation takes into account the differences in thermal conductivity of the diamond single crystal in different directions, setting the thermal conductivity in the main direction of the 100 crystal plane as the highest value and the secondary direction as a relatively lower value. The equation is solved using the finite element method or the finite difference method, with the grid size being fine enough to capture the details of the heat flow. The boundary conditions include the heat flux boundary at the heat source interface and the convective boundary at the heat dissipation boundary. The calculated ideal temperature distribution values form an ideal temperature reference vector, which has the same data structure as the real-time temperature eigenvector, facilitating subsequent comparative analysis.

[0056] Calculate the difference between the real-time temperature feature vector and the ideal temperature reference vector to obtain the temperature deviation matrix. If there are three consecutive points in the temperature deviation matrix with a deviation exceeding 10°C, it is determined as an abnormal temperature distribution. The difference calculation is performed element by element. For each corresponding element in the vector, calculate the difference between the actual value and the ideal value. The obtained temperature deviation matrix maintains the dimension and structure of the original vector, and each element represents the deviation value of the corresponding feature. For temperature values, the deviation represents the difference between the actual temperature and the ideal temperature; for the temperature change rate, the deviation represents the difference between the actual change rate and the ideal change rate; for the temperature gradient, the deviation represents the difference between the actual gradient and the ideal gradient. The temperature deviation matrix reflects the difference between the actual operating state and the ideal state of the heat dissipation system. The anomaly detection uses the continuous threshold method to monitor the consecutive abnormal points in the deviation matrix. The determination criterion of three consecutive points with a deviation exceeding 10°C takes into account occasional noise and random fluctuations, avoiding misjudgment caused by single-point anomalies and being able to detect systematic heat dissipation anomalies in a timely manner. Such an abnormal temperature distribution usually means that there are potential problems in the heat dissipation system, such as interface deterioration, channel blockage, or cooling failure, etc. Conduct a statistical analysis on the temperature deviation matrix, calculate the temperature deviation rate and the proportion of the abnormal temperature area. When the temperature deviation rate exceeds 20%, trigger the emergency cooling procedure. The statistical analysis first calculates the temperature deviation rate, that is, the relative deviation between the actual temperature and the ideal temperature. The calculation formula is that the deviation rate is equal to the deviation value divided by the ideal value and then multiplied by 100%. Calculate the deviation rate for each monitoring point, and then take the maximum value of the absolute value as the overall temperature deviation rate index. The proportion of the abnormal temperature area is defined as the percentage of the number of monitoring points with a temperature deviation exceeding the preset threshold in the total number of monitoring points. The preset threshold is usually set to 5%, that is, if the difference between the actual temperature and the ideal temperature exceeds 5%, it is regarded as abnormal. The statistical analysis also includes calculating statistics such as the mean, standard deviation, skewness, and kurtosis of the deviation matrix. These statistics reflect the overall distribution characteristics of the temperature deviation. When the temperature deviation rate exceeds 20%, the system determines that the heat dissipation anomaly is serious and automatically triggers the emergency cooling procedure, including measures such as adjusting the cooling fan to the maximum speed and the coolant pump to the maximum flow rate to prevent equipment damage due to overheating.

[0057] Based on the temperature deviation rate and the proportion of the temperature anomaly area, the working parameters of the cooling system are corrected and calculated to generate the fan speed correction amount, the flow rate correction amount, and the radiator angle correction amount. The correction calculation is based on fuzzy control rules. According to the combination of the temperature deviation rate and the proportion of the anomaly area, the correction amplitude of different parameters is determined. The fuzzy control rules are expressed in the form of "IF-THEN", for example, "IF the temperature deviation rate is large AND the proportion of the anomaly area is large THEN the fan speed correction amount is large". The rule base usually contains 9-16 rules, covering various possible combinations of situations. In the fuzzy inference process, first, the temperature deviation rate and the proportion of the anomaly area are converted into the membership degrees of fuzzy sets, then the fuzzy set of the control output is deduced through fuzzy rules, and finally, the specific correction amount value is obtained through defuzzification calculation. The fan speed correction amount represents the percentage of the rotational speed that needs to be adjusted, usually in the range of -20% to +50%; the flow rate correction amount represents the percentage of the flow rate that needs to be adjusted, with a similar range; the radiator angle correction amount represents the value of the angle change that needs to be adjusted, usually from -15° to +15°. These correction amounts comprehensively consider the severity and distribution range of the temperature anomaly and take corresponding adjustment measures for different situations.

[0058] The parameter adjustment control unit converts the fan speed correction amount, the flow rate correction amount, and the radiator angle correction amount into control signals to automatically adjust the heat dissipation control device. The parameter adjustment control unit is the interface connecting the analysis and decision-making with the actuator, and is responsible for converting the results of the correction calculation into actual control signals. In the conversion process, the current system state is first considered. Based on the current fan rotational speed, flow rate, and radiator angle values, the target value is calculated using the correction amount. For example, if the current fan rotational speed is 3000 RPM and the correction amount is +20%, then the target rotational speed is 3600 RPM. Then, the control unit converts the target value into the corresponding control signal format: the fan rotational speed is converted into the PWM duty cycle, the flow rate is converted into the pump speed or the valve opening degree, and the radiator angle is converted into the number of steps of the stepper motor. The control signals are sent to the corresponding drive circuits, and the drive circuits amplify and adjust the signal levels, and finally drive the actuators to achieve parameter adjustment. A gradual change strategy is adopted during the adjustment process, especially for the adjustment of the fan rotational speed and the radiator angle, to avoid mechanical shocks and noises caused by sudden changes. The control unit also has a self-checking function to monitor the output state of the control signals and the response of the actuators to ensure that the control commands are correctly executed.

[0059] The above describes the method for monitoring the diamond single crystal directional heat dissipation based on artificial intelligence in the embodiments of the present application. Next, the system for monitoring the diamond single crystal directional heat dissipation based on artificial intelligence in the embodiments of the present application will be described. Please refer to Figure 2 In one embodiment, the system for monitoring the diamond single crystal directional heat dissipation based on artificial intelligence in the embodiments of the present application includes: The acquisition module 201 is used to collect temperature parameter data of the diamond single crystal heat dissipation system through a temperature sensor network, and obtain heat dissipation temperature distribution data. The modeling module 202 is used to dynamically model the heat dissipation characteristics of the diamond single crystal according to the heat dissipation temperature distribution data, and generate a real-time heat flow control model. The output module 203 is used to perform a closed-loop calculation on the heat dissipation control parameters according to the real-time heat flow control model, and output fan speed and flow control instructions. The adjustment module 204 is used to adjust the cooling system actuator in real time based on the fan speed and flow control instructions, and form heat dissipation control feedback data. The analysis module 205 is used to analyze the system temperature deviation in real time according to the heat dissipation control feedback data, generate a correction amount for the cooling system parameters, and automatically adjust the heat dissipation control device according to the correction amount of the cooling system parameters.

[0060] Through the collaborative cooperation of the above-mentioned components, multiple points of the diamond single crystal heat dissipation system are arranged through a temperature sensor network, realizing high-resolution dynamic monitoring of the heat distribution. Compared with the traditional heat dissipation evaluation method that only relies on limited temperature measurement points, it can capture more detailed heat diffusion changes, improve the detection accuracy and response speed of heat anomalies. Secondly, using dynamic modeling technology, a real-time heat flow control model is constructed based on the collected heat dissipation temperature distribution data, which can effectively cope with the heat disturbance caused by system load changes, and significantly enhance the adaptive ability of the system and the real-time performance of heat control. In the closed-loop calculation link, a multi-layer domain adaptation neural network structure is introduced. Through sub-modules such as feature extraction, domain adaptation, and classification, deep control features are extracted from historical control data and the real-time model, and the distribution differences are eliminated. This algorithm mechanism not only improves the system's learning ability of the relationship between control input and heat response, but also ensures the robustness of the algorithm model to new working conditions. Especially in intelligent heat dissipation optimization, the AI model not only provides a simple regression prediction method, but also makes causal reasoning-based optimization decisions on heat dissipation control parameters through methods such as domain-invariant feature mining and heat flow path modeling, and then forms an optimal control strategy for fan speed and coolant flow. The cooling actuator is driven by PWM control and combined with a PID feedback regulation mechanism to achieve fine control. In real-time feedback, the system operation state can also be intelligently analyzed and corrected through an abnormal temperature difference recognition mechanism, enabling the cooling system to have a highly automated and intelligent dynamic regulation ability. The present invention not only reflects the coordinated control ability of sensors and actuators, but also completes the intelligent closed-loop control of the entire process from perception, modeling, decision-making to execution in the heat management system by introducing artificial intelligence algorithms.

[0061] Above Figure 2From the perspective of modular functional entities, the artificial intelligence-based diamond single crystal directional heat dissipation monitoring system in the embodiments of the present invention is described in detail. Next, the artificial intelligence-based diamond single crystal directional heat dissipation monitoring device in the embodiments of the present invention is described in detail from the perspective of hardware processing.

[0062] Figure 3 FIG. 4 is a schematic structural diagram of an artificial intelligence-based diamond single crystal directional heat dissipation monitoring device provided by an embodiment of the present invention. The artificial intelligence-based diamond single crystal directional heat dissipation monitoring device 300 may vary greatly due to different configurations or performances, and may include one or more processors (central processing units, CPUs) 310 (for example, one or more processors) and a memory 320, and one or more storage media 330 for storing application programs 333 or data 332 (for example, one or more mass storage device terminals). Among them, the memory 320 and the storage media 330 may be transient storage or persistent storage. The program stored in the storage media 330 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations on the artificial intelligence-based diamond single crystal directional heat dissipation monitoring device 300. Further, the processor 310 may be configured to communicate with the storage media 330 and execute a series of instruction operations in the storage media 330 on the artificial intelligence-based diamond single crystal directional heat dissipation monitoring device 300 to implement the steps of the above-mentioned artificial intelligence-based diamond single crystal directional heat dissipation monitoring method.

[0063] The artificial intelligence-based diamond single crystal directional heat dissipation monitoring device 300 may further include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input / output interfaces 360, and / or one or more operating systems 331, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, and so on. Those skilled in the art can understand that Figure 3 the shown structure of the artificial intelligence-based diamond single crystal directional heat dissipation monitoring device does not limit the artificial intelligence-based diamond single crystal directional heat dissipation monitoring device provided by the present invention, and may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.

[0064] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions run on a computer, the computer is caused to execute the steps of the artificial intelligence-based diamond single crystal directional heat dissipation monitoring method.

[0065] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described systems, systems, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0066] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing an artificial intelligence-based diamond single crystal directional heat dissipation monitoring device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0067] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. An artificial intelligence-based monitoring method for the directional heat dissipation of single-crystal diamond, characterized in that, The method includes: Collecting temperature parameter data of the diamond single crystal heat dissipation system through a temperature sensor network to obtain heat dissipation temperature distribution data; Dynamically modeling the heat dissipation characteristics of the diamond single crystal according to the heat dissipation temperature distribution data to generate a real-time heat flux control model; Performing closed-loop calculation on the heat dissipation control parameters according to the real-time heat flux control model and outputting fan speed and flow control instructions; Based on the fan speed and flow control instructions, performing real-time adjustment on the cooling system actuator to form heat dissipation control feedback data; According to the heat dissipation control feedback data, performing real-time analysis on the system temperature deviation to generate a correction amount for the cooling system parameters, and automatically adjusting the heat dissipation control device according to the correction amount of the cooling system parameters.

2. The method for monitoring the directional heat dissipation of diamond single crystals based on artificial intelligence according to claim 1, wherein The collecting temperature parameter data of the diamond single crystal heat dissipation system through a temperature sensor network to obtain heat dissipation temperature distribution data includes: Arranging platinum resistance temperature sensors at preset positions of the diamond single crystal radiator grown in the 100 crystal plane orientation to form a temperature monitoring point array; Installing a heat flux density sensor at the interface between the diamond single crystal grown in the 100 crystal plane orientation and the metal to obtain interface heat flux data; Performing synchronous acquisition and processing on the signals collected by the temperature monitoring point array and the heat flux density sensor to generate an original data stream; Performing three-stage cascaded filtering processing on the original data stream, sequentially performing Butterworth low-pass filtering, median filtering, and Kalman filtering to obtain noise-reduced temperature data; Calculating a temperature field spatial distribution matrix, a heat flux density vector, and a cooling medium parameter vector according to the noise-reduced temperature data to generate multi-modal heat dissipation parameters; Performing fusion processing on the multi-modal heat dissipation parameters and the sensor position information and outputting heat dissipation temperature distribution data.

3. The method for monitoring the directional heat dissipation of diamond single crystals based on artificial intelligence according to claim 1, wherein, The dynamically modeling the heat dissipation characteristics of the diamond single crystal according to the heat dissipation temperature distribution data to generate a real-time heat flux control model includes: Extracting the thermal conductivity values of the diamond single crystal at different temperature points from the heat dissipation temperature distribution data to construct a thermal conductivity temperature response matrix; Performing polynomial regression fitting on the thermal conductivity temperature response matrix to obtain continuous thermal conductivity data; According to the anisotropic characteristics of heat conduction in the 100 crystal plane orientation, decomposing the continuous thermal conductivity data according to the crystal direction to obtain a main direction thermal conductivity component and a secondary direction thermal conductivity component; Performing spatial discretization processing on the heat conduction characteristics of the diamond / Cu interface region through a three-dimensional interpolation algorithm to form a set of heat conduction boundary conditions; Integrating the main direction thermal conductivity component, the secondary direction thermal conductivity component, and the set of heat conduction boundary conditions to establish a three-dimensional heat conduction mathematical model of the diamond heat dissipation material; Solving the temperature field distribution of the three-dimensional heat conduction mathematical model under a standard heat load through the finite difference method to generate a real-time heat flux control model.

4. The method for monitoring the directional heat dissipation of diamond single crystals based on artificial intelligence according to claim 1, wherein The performing closed-loop calculation on the heat dissipation control parameters according to the real-time heat flux control model and outputting fan speed and flow control instructions includes: Taking the real-time heat flux control model and the historical temperature control data as source domain data and target domain data respectively, and inputting them into a multi-layer domain adaptation neural network structure including a feature extraction layer, a domain adaptation layer, and a classification layer; Process the source domain data and target domain data through the wavelet packet decomposition and reconstruction algorithm to generate temperature control feature vectors; Input the temperature control feature vectors into the three-layer feature extraction layer of the multi-layer domain adaptation neural network, with each layer containing 64, 128, and 256 neurons in sequence. After being processed by the non-linear activation function, deep feature representation data is obtained; Calculate the distribution difference between the source domain and the target domain for the deep feature representation data in the domain adaptation layer using the multi-kernel maximum mean discrepancy algorithm, and adjust the network weights through backpropagation to obtain domain-invariant control features; Input the domain-invariant control features into a three-layer classification network, which contains 128, 64, and 32 neurons respectively, and use the maximum probability value as the pseudo-label for the target domain data to form a heat dissipation control optimization scheme; Calculate the optimal fan speed curve and flow control curve based on the heat dissipation control optimization scheme, and output fan speed and flow control instructions.

5. The method for monitoring the directional heat dissipation of diamond single crystals based on artificial intelligence according to claim 4, wherein The calculation of the optimal fan speed curve and flow control curve based on the heat dissipation control optimization scheme and the output of fan speed and flow control instructions include: Convert the heat dissipation control optimization scheme into a thermal resistance network topology diagram, and mark the heat source points and heat dissipation boundary points of the single crystal diamond; According to the anisotropic characteristics of the 100 crystal plane of the single crystal diamond, assign heat transfer weight values to each node in the thermal resistance network topology diagram; Starting from the heat source point, traverse the thermal resistance network topology diagram using the heat path tracing algorithm to calculate all possible heat transfer paths; Perform thermal resistance cumulative calculation on all possible heat transfer paths to obtain a set of path thermal resistance values; Select the heat transfer channel with the minimum thermal resistance from the set of path thermal resistance values as the main heat dissipation channel, and at the same time select the channel with the second lowest thermal resistance value as the auxiliary heat dissipation channel; According to the distribution ratio that the main heat dissipation channel undertakes 70% of the heat load and the auxiliary heat dissipation channel undertakes 30% of the heat load, calculate the required heat dissipation power for each channel, and combine the temperature gradient value and the required heat flow on each channel to determine the fan speed value and the coolant flow value, and output fan speed and flow control instructions.

6. The method for monitoring the directional heat dissipation of diamond single crystals based on artificial intelligence according to claim 1, wherein, The real-time adjustment of the cooling system actuator based on the fan speed and flow control instructions to form heat dissipation control feedback data includes: Convert the fan speed and flow control instructions into PWM control signals to drive the fan motor and the coolant pump, and start the cooling system actuator; Perform closed-loop regulation of the fan motor speed using a PID controller. If the deviation between the actual speed and the target speed is greater than 5%, adjust the PWM duty cycle until the speed reaches the set value; Perform proportional regulation on the coolant pump flow. When the detected flow deviation exceeds the set threshold, judge whether there is a pipeline blockage, and increase the pump pressure or issue a warning signal; Collect the real-time operating parameters of the cooling system actuator, including fan speed value, flow value, temperature value, and power consumption value, and construct an execution status matrix; Compare and analyze the execution status matrix with the cooling instructions to judge the degree of execution deviation and generate an execution effect evaluation value; Perform time-series correlation analysis on the execution effect evaluation value and the temperature monitoring point data to form heat dissipation control feedback data.

7. The method for monitoring the directional heat dissipation of diamond single crystals based on artificial intelligence according to claim 1, wherein Performing real-time analysis on the system temperature deviation according to the heat dissipation control feedback data, generating a correction amount for the cooling system parameters, and automatically adjusting the heat dissipation control device based on the correction amount of the cooling system parameters, including: Extracting the actual temperature distribution characteristics from the heat dissipation control feedback data and constructing a real-time temperature feature vector; Calculating the ideal temperature distribution value under the corresponding working conditions according to the theoretical heat dissipation characteristics of the diamond single crystal 100 crystal plane and generating an ideal temperature reference vector; Performing a difference calculation between the real-time temperature feature vector and the ideal temperature reference vector to obtain a temperature deviation matrix. If there are three consecutive points in the temperature deviation matrix with a deviation exceeding 10°C, it is determined as an abnormal temperature distribution; Performing statistical analysis on the temperature deviation matrix, calculating the temperature deviation rate and the proportion of the temperature abnormal area. When the temperature deviation rate exceeds 20%, an emergency cooling program is triggered; Performing a correction calculation on the working parameters of the cooling system based on the temperature deviation rate and the proportion of the temperature abnormal area, and generating a fan speed correction amount, a flow rate correction amount, and a radiator angle correction amount; Converting the fan speed correction amount, the flow rate correction amount, and the radiator angle correction amount into control signals through a parameter adjustment control unit to automatically adjust the heat dissipation control device.

8. An artificial intelligence-based diamond single crystal directional heat dissipation monitoring system, characterized in that, For implementing the artificial intelligence-based diamond single crystal directional heat dissipation monitoring method described in any one of claims 1-7, the artificial intelligence-based diamond single crystal directional heat dissipation monitoring system includes: An acquisition module for collecting temperature parameter data of the diamond single crystal heat dissipation system through a temperature sensor network to obtain heat dissipation temperature distribution data; A modeling module for dynamically modeling the heat dissipation characteristics of the diamond single crystal according to the heat dissipation temperature distribution data to generate a real-time heat flow control model; An output module for performing a closed-loop calculation on the heat dissipation control parameters according to the real-time heat flow control model and outputting fan speed and flow rate control instructions; An adjustment module for performing real-time adjustment on the cooling system actuator based on the fan speed and flow rate control instructions to form heat dissipation control feedback data; An analysis module for performing real-time analysis on the system temperature deviation according to the heat dissipation control feedback data, generating a correction amount for the cooling system parameters, and automatically adjusting the heat dissipation control device based on the correction amount of the cooling system parameters.

9. An artificial intelligence-based diamond single crystal directional heat dissipation monitoring device, characterized in that, Including a memory and a processor, the memory stores a computer program that can run on the processor, and when the processor executes the computer program, it implements the artificial intelligence-based diamond single crystal directional heat dissipation monitoring method described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is run by the processor, the processor is caused to execute the artificial intelligence-based diamond single crystal directional heat dissipation monitoring method described in any one of claims 1 to 7.

Images