Power device heat dissipation optimization method and system based on heat flow path reconstruction

Abstract

The invention discloses a heat dissipation optimization method and system for a power device based on heat flow path reconstruction, and aims to improve the heat dissipation efficiency of the power device in the working process and reduce the adverse effect of thermal stress on the performance and service life of the device. The method comprises the following steps: firstly, acquiring real-time temperature field distribution data of a target power module during the working period of a power device, and determining an initial heat flow path of the module based on the data; performing heat sensitivity analysis on each component in the target power module; identifying a heat flow shielding area according to the heat sensitive data, and reconstructing the initial heat flow path on this basis to generate an optimized heat flow path; and finally, an efficient heat dissipation channel is constructed according to the optimized heat flow path, so that effective heat dissipation optimization of the power device is realized. The method can dynamically identify the thermal bottleneck and actively optimize the heat dissipation path, is suitable for thermal management design of high-power-density electronic equipment, and has high engineering practical value.

Description

Technical Field

[0001] The present invention relates to the technical field of power device heat dissipation, and particularly to a power device heat dissipation optimization method and system based on the reconstruction of the heat flow path. Background Art

[0002] With the continuous development of power electronics technology, power devices have been widely used in fields such as electric vehicles, high-speed trains, new energy power generation, and industrial automation. A large amount of heat is generated during the operation of power devices. If the heat cannot be dissipated in a timely and effective manner, it will not only cause the operating temperature of the device to rise and the performance to decline, but may also lead to thermal runaway and even damage. Therefore, improving the heat dissipation efficiency of power devices is one of the key issues to ensure their reliability and extend their service life.

[0003] Currently, the common heat dissipation designs of power devices mainly rely on structural presetting and material optimization, such as using high thermal conductivity materials, heat sinks, air cooling or liquid cooling systems, etc. to conduct overall or local thermal management. However, the traditional methods generally have the following problems: First, the design of the heat dissipation path is mostly based on static thermal simulation, and it is difficult to adapt to the dynamic change of the heat source distribution during the actual operation of the device; Second, the sensitivity differences of different components to the thermal environment are ignored, which is likely to cause insufficient thermal management of key areas; Third, there is a lack of a dynamic reconstruction mechanism based on the real-time evolution characteristics of the heat flow, and it is impossible to effectively avoid the heat accumulation area, resulting in prominent local hot spot problems.

[0004] In addition, the existing heat dissipation optimization methods mostly adopt a regular layout method, which cannot accurately guide the heat flow to diffuse along the optimal path, resulting in low heat dissipation efficiency in some high heat density areas, seriously restricting the overall thermal management performance of the power module. Therefore, there is an urgent need for a heat dissipation optimization technology that can combine temperature field distribution, thermal sensitivity analysis, and dynamic regulation of the heat flow path to achieve more efficient and intelligent thermal management of power devices. Summary of the Invention

[0005] In order to solve at least one of the above technical problems, the present invention proposes a power device heat dissipation optimization method and system based on the reconstruction of the heat flow path.

[0006] The first aspect of the present invention provides a power device heat dissipation optimization method based on the reconstruction of the heat flow path, including: Obtaining the real-time temperature field distribution data of the target power module during the operation of each power device, and determining the initial heat flow path of the target power module according to the real-time temperature field distribution data; Performing a thermal sensitivity analysis on each component in the target power module to determine the thermal sensitivity data of each component; Determining the heat flow shielding area of the target power module according to the thermal sensitivity data, and reconstructing the initial heat flow path according to the heat flow shielding area to obtain an optimized heat flow path; Construct a heat dissipation channel for the power device by the target power module according to the optimized heat flow path, and optimize the heat dissipation of the power device according to the heat dissipation channel.

[0007] In this solution, the real-time temperature field distribution data of the target power module during the operation of each power device is obtained, and the initial heat flow path of the target power module is determined according to the real-time temperature field distribution data. Specifically: Based on the thermal imaging device, the thermal imaging image change data of the target power module during the operation of each power device in a preset time period is obtained in real time, and the thermal imaging image change data is extracted at a preset frame rate to construct a thermal imaging frame image; Convert the thermal imaging frame image to grayscale to construct a grayscale image. Establish the correspondence between the pixel grayscale value and the actual temperature through blackbody radiation calibration. Determine the temperature value of each pixel point of each grayscale image according to the correspondence, and construct a temperature matrix; Perform alignment and integration operations on the temperature matrix according to the time series to construct a temperature distribution change matrix of the target power module, obtain the real-time temperature field distribution data, and calculate the temperature gradient vector of each pixel point on the surface of the target power module according to the real-time temperature field distribution data; Construct a temperature gradient distribution map according to the amplitude and direction of the temperature gradient vector. Extract the area where the temperature change rate is higher than the set threshold as the high heat flux density area according to the temperature gradient distribution map, and fuse adjacent high heat flux density areas based on the region growing algorithm to generate a heat flow aggregation zone; Extend the path of the heat flow aggregation zone in the direction of the temperature gradient, calculate the direction consistency coefficient of the temperature gradient vector in the extended path, and smooth and optimize the extended path according to the direction consistency coefficient to form the initial heat flow path of the target power module.

[0008] In this solution, perform thermal sensitivity analysis on each component in the target power module to determine the thermal sensitivity data of each component. Specifically: Obtain the working parameter change time series data of each component in the target power module in a preset time period. The working parameter change time series data includes the time change sequences of voltage, current, and power parameters; Extract the temperature change time series data of the area where each component is located according to the real-time temperature field change data, synchronize and align the temperature change time series data with the working parameter change time series data, and construct a temperature - working parameter data matrix; Perform correlation analysis on the temperature - working parameter data matrix through the Pearson correlation coefficient algorithm, calculate the correlation coefficient of the temperature change to the change of each working parameter, perform a significance test on the correlation coefficient through the t - test method, and screen out the effective correlation coefficients whose significance level is lower than the preset significance threshold; Evaluate the influence degree of temperature change on the change of component operating parameters according to the absolute value of the effective correlation coefficient, and determine the thermal sensitivity coefficient of each component according to the influence degree to obtain the thermal sensitivity data of each component.

[0009] In this solution, the specific method for determining the heat flow shielding area of the target power module according to the thermal sensitivity data is as follows: Perform position mapping on the thermal sensitivity coefficients of each component in the target power module according to the thermal sensitivity data to generate a thermal sensitivity coefficient distribution map; Introduce a density clustering algorithm to perform spatial clustering analysis on the thermal sensitivity coefficient distribution map, set the neighborhood radius and minimum sample number of the density clustering algorithm, and initialize the data points in the thermal sensitivity coefficient distribution map as an unlabeled set; Select any data point from the unlabeled set, calculate the number of samples contained within its neighborhood radius. If the number of samples is greater than or equal to the minimum sample number, mark this point as a core point, and traverse all unvisited points within its neighborhood, recursively merging adjacent points that meet the core point conditions into the same clustering cluster; If the number of samples within the neighborhood is less than the minimum sample number, mark this point as a noise point and temporarily store it in an independent set. After traversing all unlabeled data points, output multiple thermal sensitivity data clusters composed of core points and density-reachable points; Calculate the average value of the thermal sensitivity coefficients within each thermal sensitivity data cluster, and mark the thermal sensitivity data clusters with an average value higher than the preset sensitivity threshold as high-thermal-sensitivity data clusters; According to the geometric center coordinates and boundary range of the high-thermal-sensitivity data cluster, extract the path segments in the initial heat flow path that intersect with the boundary of the high-thermal-sensitivity data cluster or have a distance less than the preset safety distance as the path segments to be shielded; Calculate the product of the temperature gradient vector corresponding to the path segment to be shielded and the average value of the thermal sensitivity coefficients of the high-thermal-sensitivity data cluster to generate a path shielding factor, and mark the path segments to be shielded with a path shielding factor greater than the preset shielding threshold as high-shielding-priority path segments; Based on the position coordinates of the high-shielding-priority path segments, use a region fusion algorithm to merge adjacent high-shielding-priority path segments to generate a continuous boundary of the heat flow shielding area; Calculate the spatial distance between the temperature gradient vector in the initial heat flow path and the high-thermal-sensitivity data cluster. If the distance is less than the preset protection distance, include the corresponding area in the heat flow shielding area, and finally generate a heat flow shielding area covering all high-shielding-priority path segments.

[0010] In this solution, the specific method for reconstructing the initial heat flow path according to the heat flow shielding area to obtain an optimized heat flow path is as follows: Extract the initial heat flow path segments that intersect with the heat flow shielding area as the path adjustment objects according to the boundary coordinates of the heat flow shielding area and the node distribution of the initial heat flow path; Calculate the equivalent heat dissipation efficiency of each area according to the thermal conductivity distribution data of the circuit board substrate of the target power module and the heat diffusion characteristics of the copper layer coverage area; Based on the equivalent heat dissipation efficiency and the low-temperature area distribution in the real-time temperature field distribution data, screen out the non-heat flow shielding areas with heat dissipation efficiency higher than the preset efficiency threshold and temperature gradient amplitude lower than the preset gradient threshold as candidate guiding areas; Determine the optimal heat flow guiding area of the path adjustment object according to the spatial distance between the candidate guiding area and the path adjustment object, combined with the consistency coefficient of the temperature gradient direction; Based on the geometric center coordinates of the optimal heat flow guiding area, offset the path nodes of the initial heat flow path intersecting with the heat flow shielding area in the heat flow diffusion direction to generate an optimized heat flow path that bypasses the heat flow shielding area and extends along the copper layer coverage area.

[0011] In this solution, the heat dissipation channel of the power device of the target power module is constructed according to the optimized heat flow path, and the heat dissipation of the power device is optimized according to the heat dissipation channel, specifically: Compare the initial heat flow path with the optimized heat flow path, judge the heat flow path difference of the target power module, and determine the heat flow change direction according to the heat flow path difference; Determine the heat dissipation channel of the power device of the target power module according to the heat flow change direction, determine the installation position of the heat dissipation device of the target power module according to the heat dissipation channel, and the heat dissipation device includes heat dissipation micro copper tubes, heat sinks, and heat dissipation coating materials; Optimize the heat dissipation of the power device according to the heat dissipation device.

[0012] The second aspect of the present invention also provides a power device heat dissipation optimization system based on heat flow path reconstruction, which includes: a memory and a processor. The memory includes a power device heat dissipation optimization method program based on heat flow path reconstruction. When the power device heat dissipation optimization method program based on heat flow path reconstruction is executed by the processor, the following steps are realized: Obtain the real-time temperature field distribution data of the target power module during the operation of each power device, and determine the initial heat flow path of the target power module according to the real-time temperature field distribution data; Conduct a thermal sensitivity analysis on each component in the target power module to determine the thermal sensitivity data of each component; Determine the heat flow shielding area of the target power module according to the thermal sensitivity data, and reconstruct the initial heat flow path according to the heat flow shielding area to obtain an optimized heat flow path; Construct a heat dissipation channel for the power device according to the optimized heat flow path, and optimize the heat dissipation of the power device according to the heat dissipation channel.

[0013] The present invention discloses a method and system for optimizing the heat dissipation of power devices based on heat flow path reconstruction, aiming to improve the heat dissipation efficiency of power devices during operation and reduce the adverse effects of thermal stress on device performance and lifespan. The method includes the following steps: First, obtain the real-time temperature field distribution data of the target power module during the operation of the power device, and determine the initial heat flow path of the module based on this data; Subsequently, perform a thermal sensitivity analysis on each component in the target power module; Identify the heat flow shielding area according to the thermal sensitivity data, and reconstruct the initial heat flow path based on this to generate an optimized heat flow path; Finally, construct an efficient heat dissipation channel according to the optimized heat flow path, thereby realizing effective heat dissipation optimization of the power device. The present invention can dynamically identify thermal bottlenecks and actively optimize the heat dissipation path, is applicable to the thermal management design of high-power density electronic devices, and has high engineering practical value. Brief Description of the Drawings

[0014] Figure 1 Shows a flowchart of a method for optimizing the heat dissipation of power devices based on heat flow path reconstruction according to the present invention; Figure 2 Shows a flowchart of determining the thermal sensitivity data of each component according to the present invention; Figure 3 Shows a flowchart of optimizing the heat dissipation of power devices according to the present invention; Figure 4 Shows a block diagram of a system for optimizing the heat dissipation of power devices based on heat flow path reconstruction according to the present invention. Detailed Embodiments

[0015] In order to be able to more clearly understand the above-mentioned objects, features, and advantages of the present invention, the present invention will be further described in detail below in conjunction with the drawings and specific embodiments. It should be noted that, without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other.

[0016] Many specific details are set forth in the following description in order to fully understand the present invention. However, the present invention can also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited by the specific embodiments disclosed below.

[0017] Figure 1 Shows a flowchart of a method for optimizing the heat dissipation of power devices based on heat flow path reconstruction according to the present invention.

[0018] As Figure 1As shown, the first aspect of the present invention provides a method for optimizing heat dissipation of a power device based on heat flow path reconstruction, comprising: S102, acquiring real-time temperature field distribution data of a target power module during operation of each power device, and determining an initial heat flow path of the target power module according to the real-time temperature field distribution data; S104, performing thermal sensitivity analysis on each component in the target power module to determine thermal sensitivity data of each component; S106, determining a heat flow shielding area of ​​a target power module according to the thermally sensitive data, and reconstructing the initial heat flow path according to the heat flow shielding area to obtain an optimized heat flow path; S108, constructing a heat dissipation channel for a power device of a target power module according to the optimized heat flow path, and optimizing heat dissipation of the power device according to the heat dissipation channel.

[0019] It should be noted that by obtaining the real-time temperature field distribution data of the target power module during the operation of the power device, and determining the initial heat flow path of the target power module based on the real-time temperature field distribution data, the actual conduction and accumulation trend of heat in the power module can be fully revealed. Secondly, the implementation of thermal sensitivity analysis can mine the response intensity of each component to temperature changes, and through the correlation mining with the working electrical parameters, the coupling judgment from "thermal" to "electrical" behavior can be realized to form targeted thermal sensitivity data, so as to identify those key components that are most sensitive to temperature rise and need to be protected. Then, based on the thermal sensitivity data, the heat flow shielding area is further identified, and the thermal risk concentration area is accurately delineated by clustering and geometric analysis methods, which can effectively prevent the high temperature path from continuing to pass through these sensitive areas, so as to reduce the damage of local thermal shock to the reliability of the device. Then, combined with the differences in module structure and thermal conductivity, the initial heat flow path is reconstructed, which can not only realize the intelligent avoidance of high thermal sensitivity areas, but also guide the heat flow to diffuse preferentially along the high thermal conductivity efficiency area, improve the overall heat diffusion rate, and thus improve the heat transfer capacity. Finally, by rationally arranging heat dissipation channels and heat dissipation devices (such as heat sinks, microchannel cooling structures, thermal conductive coatings, etc.) according to the optimized heat flow path, the optimal deployment of thermal management hardware configuration can be achieved, so that the heat dissipation structure not only fits the heat flow direction in space, but also significantly improves the heat dissipation efficiency of power devices in terms of function, ultimately achieving lower device operating temperature and reduced thermal stress.

[0020] According to an embodiment of the present invention, the real-time temperature field distribution data of the target power module during the operation of each power device is obtained, and the initial heat flow path of the target power module is determined according to the real-time temperature field distribution data, specifically: Based on the thermal imaging device, real-time acquisition of the thermal imaging image change data of the target power module during the operation of each power device in a preset time period, extracting the thermal imaging image change data at a preset frame rate, and constructing a thermal imaging frame image; Perform gray-scale conversion on the thermal imaging frame image, construct a gray-scale image, establish the corresponding relationship between the pixel gray-scale value and the actual temperature through blackbody radiation calibration, determine the temperature value of each pixel point of each gray-scale image according to the corresponding relationship, and construct a temperature matrix; Perform alignment and integration operations on the temperature matrix according to the time series, construct the temperature distribution change matrix of the target power module, obtain the real-time temperature field distribution data, and calculate the temperature gradient vector of each pixel point on the surface of the target power module according to the real-time temperature field distribution data; Construct a temperature gradient distribution map according to the amplitude and direction of the temperature gradient vector, extract the area where the temperature change rate is higher than the set threshold as the high heat flux density area based on the temperature gradient distribution map, and fuse adjacent high heat flux density areas based on the region growing algorithm to generate a heat flux aggregation zone; Extend the heat flux aggregation zone along the temperature gradient direction, calculate the direction consistency coefficient of the temperature gradient vector in the extension path, and smooth and optimize the extension path according to the direction consistency coefficient to form the initial heat flux path of the target power module.

[0021] It should be noted that by acquiring the thermal imaging image change data of the target power module during the operation of each power device, the thermal imaging image is first subjected to gray-scale conversion and blackbody radiation calibration to accurately obtain the temperature value of each pixel point, and a temperature matrix with high spatio-temporal resolution is constructed; subsequently, based on the temperature matrix, a temperature distribution change matrix in the time series is constructed, and the temperature gradient vector of each pixel point in the temperature field is obtained through gradient calculation, so as to reveal the main direction and intensity of temperature change. On this basis, the area where the temperature change rate exceeds the set threshold is extracted as the high heat flux density area through the temperature gradient distribution map, and the region growing algorithm is used to fuse adjacent heat flux areas to form a heat flux aggregation zone, accurately capturing the "trunk" structure of heat concentrated conduction. Further, the heat flux aggregation zone is extended along the temperature gradient direction, and the direction consistency coefficient of the gradient vector in the path is calculated to ensure the continuity and rationality of the heat flux direction, and finally an initial heat flux path with clear physical meaning and representativeness is formed, effectively revealing the main direction and bottleneck position of heat conduction in the module; the target power module is jointly constructed by several power devices.

[0022] Figure 2 Shows the flow chart of the present invention for determining the thermal sensitive data of each component.

[0023] According to an embodiment of the present invention, the thermal sensitivity analysis of each component in the target power module is performed to determine the thermal sensitivity data of each component, specifically as follows: S202, obtain the working parameter change time series data of each component in the target power module during a preset time period, where the working parameter change time series data includes the time change sequences of voltage, current, and power parameters; S204, extract the temperature change time series data of the area where each component is located according to the real-time temperature field change data, synchronize and align the time of the temperature change time series data with the working parameter change time series data, and construct a temperature - working parameter data matrix; S206, perform a correlation analysis on the temperature - working parameter data matrix through the Pearson correlation coefficient algorithm, calculate the correlation coefficient of the temperature change to the change of each working parameter, perform a significance test on the correlation coefficient through the t - test method, and screen out the effective correlation coefficients with a significance level lower than the preset significance threshold; S208, evaluate the influence degree of the temperature change on the change of the component working parameters according to the absolute value of the effective correlation coefficient, determine the thermal sensitivity coefficient of each component according to the influence degree, and obtain the thermal sensitivity data of each component.

[0024] It should be noted that by synchronously collecting the temperature change time series data and the working parameter change time series data of each component in the target power module, constructing an association matrix of temperature, voltage, current, and power parameters, and performing a significance analysis based on the Pearson correlation coefficient algorithm and statistical test methods, the influence degree of temperature fluctuations on the electrical performance of components can be accurately quantified. By establishing a dynamic association model between the thermal field change and the electrical parameter fluctuation, the key components that respond strongly to temperature changes can be effectively identified. Based on the effective correlation coefficients screened by statistical significance, the performance stability differences of different components under the condition of temperature rise can be objectively evaluated, providing a data basis for the hierarchical protection of thermal - sensitive components. By calculating the thermal sensitivity coefficient and establishing a quantitative evaluation system, key thermal protection can be preferentially implemented for high - sensitive components, significantly improving the pertinence and effectiveness of the optimization of the heat dissipation system. The significance test (t - test) determines whether the correlation coefficient is significant by calculating the p - value (significance level). If the p - value is higher than the set threshold (such as 0.05), it indicates that the correlation may not be significant, and such unreliable results need to be excluded. The higher the influence degree, the higher the thermal sensitivity coefficient, indicating that the increase in temperature has a higher impact on the change of the component working parameters, that is, the change in temperature has a higher impact on the working stability and working efficiency of the component.

[0025] According to an embodiment of the present invention, the determination of the heat flow shielding area of the target power module according to the thermal sensitivity data is specifically as follows: Perform position mapping on the thermal sensitivity coefficients of each component in the target power module according to the thermal-sensitive data to generate a thermal sensitivity coefficient distribution map; Introduce a density clustering algorithm to perform spatial clustering analysis on the thermal sensitivity coefficient distribution map, set the neighborhood radius and minimum number of samples of the density clustering algorithm, and initialize the data points in the thermal sensitivity coefficient distribution map as an unlabeled set; Select any data point from the unlabeled set, calculate the number of samples contained within its neighborhood radius. If the number of samples is greater than or equal to the minimum number of samples, mark this point as a core point, and traverse all unvisited points within its neighborhood, recursively merging adjacent points that meet the core point conditions into the same clustering cluster; If the number of samples within the neighborhood is less than the minimum number of samples, mark this point as a noise point and temporarily store it in an independent set. After traversing all unlabeled data points, output multiple thermal-sensitive data clusters composed of core points and density-reachable points; Calculate the average value of the thermal sensitivity coefficients within each thermal-sensitive data cluster, and mark the thermal-sensitive data clusters with an average value higher than the preset sensitivity threshold as high-thermal-sensitivity data clusters; According to the geometric center coordinates and boundary range of the high-thermal-sensitivity data cluster, extract the path segments in the initial heat flow path that intersect with the boundary of the high-thermal-sensitivity data cluster or have a distance less than the preset safety distance as path segments to be shielded; Calculate the product of the temperature gradient vector corresponding to the path segment to be shielded and the average value of the thermal sensitivity coefficients of the high-thermal-sensitivity data cluster to generate a path shielding factor, and mark the path segments to be shielded with a path shielding factor greater than the preset shielding threshold as high-shielding-priority path segments; Based on the position coordinates of the high-shielding-priority path segments, use a region fusion algorithm to merge adjacent high-shielding-priority path segments to generate a continuous boundary of the heat flow shielding region; Calculate the spatial distance between the temperature gradient vector in the initial heat flow path and the high-thermal-sensitivity data cluster. If the distance is less than the preset protection distance, include the corresponding region in the heat flow shielding region, and finally generate a heat flow shielding region covering all high-shielding-priority path segments.

[0026] It should be noted that, due to the different sensitivities of the components in the target power module to temperature rise during operation, in order to avoid the influence of the heat generated during the operation of the power device on the components sensitive to temperature rise, a density clustering algorithm is introduced to perform spatial clustering analysis on the thermal sensitivity data of the target power module. By mapping the spatial positions of the thermal sensitivity coefficients of the components, a thermal sensitivity coefficient distribution map is constructed, and density clustering is performed on the map using the set neighborhood radius and minimum sample number, which can effectively identify the regions where the thermal sensitivity is significantly concentrated, that is, the high thermal sensitivity data clusters. Furthermore, by calculating the average thermal sensitivity coefficients of these clusters and combining their geometric boundary information, the path segments that intersect or are adjacent to the initial heat flow path are identified as the objects to be shielded, and the thermal risk is quantitatively evaluated through the path shielding factor, and the path segments with high shielding priority are screened out. Subsequently, a region fusion algorithm is used to merge these path segments to construct a continuous and closed heat flow shielding boundary region, and a protection distance standard is introduced to further expand the shielding range to ensure that the heat flow path is fully isolated from the sensitive region, significantly reducing the thermal shock risk in the high thermal sensitivity region and improving the operating stability of the target power module. The path shielding factor refers to an index used to quantify the thermal interference intensity between the heat flow path and the high thermal sensitivity region, and its value is calculated by the product of the temperature gradient vector corresponding to the path segment and the average thermal sensitivity coefficient of the intersecting high thermal sensitivity data cluster, reflecting the strength of the potential thermal shock of the path segment to the sensitive region.

[0027] According to an embodiment of the present invention, reconstructing the initial heat flow path according to the heat flow shielding region to obtain an optimized heat flow path specifically includes: According to the boundary coordinates of the heat flow shielding region and the node distribution of the initial heat flow path, extracting the initial heat flow path segments that intersect the heat flow shielding region as the path adjustment objects; According to the thermal conductivity distribution data of the circuit board substrate of the target power module and the heat diffusion characteristics of the copper layer coverage area, calculating the equivalent heat dissipation efficiency of each region; Based on the equivalent heat dissipation efficiency and the low temperature region distribution in the real-time temperature field distribution data, screening out the non-heat flow shielding regions with heat dissipation efficiency higher than the preset efficiency threshold and temperature gradient amplitude lower than the preset gradient threshold as the candidate guiding regions; According to the spatial distance between the candidate guiding region and the path adjustment object, combined with the consistency coefficient of the temperature gradient direction, determining the optimal heat flow guiding region of the path adjustment object; Based on the geometric center coordinates of the optimal heat flow guiding region, offsetting the path nodes of the initial heat flow path intersecting the heat flow shielding region in the heat flow diffusion direction to generate an optimized heat flow path that bypasses the heat flow shielding region and extends along the copper layer coverage area.

[0028] It should be noted that by reconstructing the path segments of the initial heat flow path that intersect with the heat flow shielding area, the local overheating risk brought by the high - heat - sensitive area can be effectively avoided, and the thermal safety of the overall heat dissipation channel can be improved. Specifically, by using the thermal conductivity distribution of different regions of the circuit board and the heat diffusion ability of the copper layer, combined with the distribution of low - temperature regions and high - heat - dissipation - efficiency regions in the real - time temperature field, regions with good thermal conductivity and low thermal load are scientifically selected as the heat flow guiding direction. At the same time, through the evaluation of the temperature gradient direction consistency coefficient, the path adjustment process not only considers the path with the minimum thermal resistance, but also ensures the physical continuity and stability of the heat flow orientation. On this basis, the heat flow path is smoothly adjusted by the node offset method, so that the optimized heat flow path can efficiently bypass the sensitive area and preferentially extend along the copper layer area with strong heat diffusion ability, ultimately achieving the synergistic optimization effect of uniform heat flow guidance, local hot spot mitigation, and overall heat dissipation performance improvement. Path node offset refers to adjusting the position of the path nodes in the original path that are close to or pass through the heat flow shielding area in the heat flow diffusion direction during the heat flow path reconstruction process.

[0029] Figure 3 The flowchart showing the heat dissipation optimization of the power device by the present invention is presented.

[0030] According to an embodiment of the present invention, the heat dissipation channel of the power device is constructed according to the optimized heat flow path, and the heat dissipation of the power device is optimized according to the heat dissipation channel, specifically as follows: S302, comparing the initial heat flow path with the optimized heat flow path, judging the heat flow path difference of the target power module, and determining the heat flow change direction according to the heat flow path difference; S304, determining the heat dissipation channel of the target power module for the power device according to the heat flow change direction, and determining the installation position of the heat dissipation device of the target power module according to the heat dissipation channel. The heat dissipation device includes a heat dissipation micro - fine copper tube, a heat sink, and a heat dissipation coating material; S306, optimizing the heat dissipation of the power device according to the heat dissipation device.

[0031] It should be noted that by accurately analyzing the heat flow path change direction, the best layout area of the heat dissipation device can be dynamically identified to ensure that the heat dissipation channel highly matches the optimized heat flow direction. Secondly, based on the heat flow path difference, the selection and layout of the heat dissipation device are guided, so that devices such as heat dissipation micro - fine copper tubes, heat sinks, and heat dissipation coating materials can be accurately deployed on the key heat flow paths to achieve the optimal allocation of heat dissipation resources.

[0032] According to an embodiment of the present invention, it further includes: Real-time monitor the current, voltage and switching frequency parameters of each load loop in the target power module, establish the mapping relationship between the time-series data of the load parameters and the change rate of the temperature field through a long short-term memory neural network, and generate the prediction result of the load change rate; Analyze the thermal conductivity difference distribution between the copper layer and the insulating substrate in the multi-layer PCB board according to the load change rate prediction result, extract the continuous area of the copper layer as the priority heat conduction channel and mark the heat flow impedance coefficient of the via position; Based on the thermal conductivity difference distribution and the heat flow impedance coefficient, construct the heat conduction topology network of the multi-layer PCB board, input the load change rate prediction result into the heat conduction topology network for transient thermal simulation, and pre-generate a multi-level heat conduction compensation path set including the planar heat conduction path of the copper layer and the vertical heat conduction path of the via; Verify the effectiveness of the multi-level heat conduction compensation path set according to the real-time collected temperature field distribution data. When it is detected that the deviation between the actual heat flow direction and the predicted path exceeds the tolerance threshold, switch to the compensation path with the highest thermal conductivity matching degree in the adjacent layer copper layer, and optimize the impedance matching of the cross-layer heat flow channel by adjusting the via current density.

[0033] The switching to the compensation path with the highest thermal conductivity matching degree in the adjacent layer copper layer is specifically as follows: Calculate the equivalent thermal conductivity attenuation coefficient of each layer of copper foil according to the temperature rise gradient distribution of the copper layer in the current heat flow path, and establish a comprehensive efficiency evaluation model for the cross-layer heat flow path by combining the vertical thermal resistance parameters of the vias in the heat conduction topology network; Based on the comprehensive efficiency evaluation model, iteratively screen the multi-level heat conduction compensation path set, eliminate the via connection paths with abnormal spacing caused by the thermal expansion of the insulating substrate, and retain the candidate paths that meet the copper layer heat flow continuity constraint; According to the high-frequency load fluctuation component in the load change rate prediction result, preferentially select the redundant heat conduction path with a parallel via structure as the heat flow switching buffer zone, and verify the local hot spot generation probability during the switching process through electromagnetic-thermal coupling field simulation; When it is detected that the thermal sensitivity coefficient of the target copper layer area exceeds the preset safety threshold, trigger the distributed current injection mechanism of the via array to reduce the contact thermal resistance, and at the same time divert the heat in the original heat flow path to the low-temperature area of the adjacent copper layer through the parallel via structure, realizing the collaborative reconstruction of the multi-layer heat flow path at the millisecond level.

[0034] It should be noted that in dynamic load scenarios such as frequency converters and electric vehicle controllers, the rapid load switching of power modules causes the heat flow path to shift frequently. Since there is a significant difference in the thermal conductivity between the internal copper layers and the insulating substrates of multi-layer PCBs, traditional thermal management solutions are difficult to track the dynamic changes of heat flow in real time, resulting in problems such as delays in the reconstruction of cross-layer heat conduction paths and the aggregation of local hot spots, which seriously affect the device lifespan and system reliability. In the present invention, a long short-term memory neural network (LSTM) is used to predict the load change rate in real time. By combining the thermal conductivity distribution of the continuous copper layer regions and the thermal flow impedance coefficients of vias, a multi-layer heat conduction topology network is constructed to pre-generate a set of multi-level heat conduction compensation paths. Based on the real-time temperature field data, the effectiveness of the paths is dynamically verified. When a deviation in the actual heat flow direction is detected, by calculating the equivalent thermal conductivity attenuation coefficient of the copper layer and the vertical thermal resistance parameters, redundant parallel via paths are selected as the switching buffer zone, and a distributed current injection mechanism is triggered to reduce the contact thermal resistance, achieving millisecond-level collaborative reconstruction of cross-layer heat flow paths. This technology can significantly improve the heat dissipation response speed of multi-layer PCBs under dynamic loads, accurately divert heat to low-temperature regions of adjacent copper layers, suppress the deformation of vias caused by the thermal expansion of insulating substrates, reduce the generation probability of local hot spots by more than 30%, and at the same time avoid eddy current losses during path switching through electromagnetic-thermal coupling optimization, ultimately extending the lifespan of power modules and ensuring the operation stability in high-frequency load scenarios.

[0035] Figure 4 Fig. shows a block diagram of a power device heat dissipation optimization system based on heat flow path reconstruction according to the present invention.

[0036] In a second aspect of the present invention, a power device heat dissipation optimization system 4 based on heat flow path reconstruction is also provided. The system includes: a memory 41 and a processor 42. The memory includes a program for the power device heat dissipation optimization method based on heat flow path reconstruction. When the program for the power device heat dissipation optimization method based on heat flow path reconstruction is executed by the processor, the following steps are implemented: Obtain real-time temperature field distribution data of a target power module during the operation of each power device, and determine the initial heat flow path of the target power module according to the real-time temperature field distribution data; Conduct a thermal sensitivity analysis on each component in the target power module to determine the thermal sensitivity data of each component; Determine the heat flow shielding area of the target power module according to the thermal sensitivity data, and reconstruct the initial heat flow path according to the heat flow shielding area to obtain an optimized heat flow path; Construct a heat dissipation channel for the power device by the target power module according to the optimized heat flow path, and optimize the heat dissipation of the power device according to the heat dissipation channel.

[0037] The present invention discloses a heat dissipation optimization method and system for power devices based on the reconstruction of the heat flow path, aiming to improve the heat dissipation efficiency of power devices during operation and reduce the adverse effects of thermal stress on the device performance and lifespan. The method includes the following steps: First, obtain the real-time temperature field distribution data of the target power module during the operation of the power device, and determine the initial heat flow path of the module based on this data; Subsequently, conduct a thermal sensitivity analysis on each component in the target power module; Identify the heat flow shielding area according to the thermal sensitivity data, and on this basis, reconstruct the initial heat flow path to generate an optimized heat flow path; Finally, construct an efficient heat dissipation channel according to the optimized heat flow path, thereby realizing the effective heat dissipation optimization of the power device. The present invention can dynamically identify thermal bottlenecks and actively optimize the heat dissipation path, and is applicable to the thermal management design of high-power density electronic devices, having high engineering practical value.

[0038] In several embodiments provided by the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined, or can be integrated into another system, or some features can be ignored, or not executed. In addition, the couplings, direct couplings, or communication connections between the various components shown or discussed can be through some interfaces. The indirect couplings or communication connections of devices or units can be electrical, mechanical, or other forms.

[0039] The units described above as separate components may or may not be physically separated. The components shown as units may or may not be physical units; they can be located in one place or distributed to multiple network units; Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0040] In addition, in each embodiment of the present invention, the various functional units can all be integrated in one processing unit, or each unit can be separately used as a unit, or two or more units can be integrated in one unit; The above integrated units can be implemented in the form of hardware, or in the form of a combination of hardware and software functional units.

[0041] Those of ordinary skill in the art will understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions. The foregoing program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps including those of the above method embodiments. The foregoing storage medium includes various media that can store program codes, such as removable storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs.

[0042] Alternatively, if the above integrated unit of the present invention is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the embodiments of the present invention, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the various embodiments of the present invention. The foregoing storage medium includes various media that can store program codes, such as removable storage devices, ROM, RAM, magnetic disks, or optical discs.

[0043] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A heat dissipation optimization method for power devices based on heat flow path reconstruction, characterized in that It includes the following steps: Obtain the real-time temperature field distribution data of the target power module during the operation of each power device, and determine the initial heat flow path of the target power module according to the real-time temperature field distribution data; Conduct a thermal sensitivity analysis on each component in the target power module to determine the thermal sensitivity data of each component; Determine the heat flow shielding area of the target power module according to the thermal sensitivity data, and reconstruct the initial heat flow path according to the heat flow shielding area to obtain an optimized heat flow path; Construct a heat dissipation channel for the power device by the target power module according to the optimized heat flow path, and optimize the heat dissipation of the power device according to the heat dissipation channel.

2. The heat dissipation optimization method for a power device based on heat flow path reconstruction according to claim 1, wherein The obtaining of the real-time temperature field distribution data of the target power module during the operation of each power device, and determining the initial heat flow path of the target power module according to the real-time temperature field distribution data is specifically as follows: Based on the thermal imaging device, obtain the thermal imaging image change data of the target power module during the operation of each power device in a preset time period in real time, extract the thermal imaging image change data at a preset frame rate, and construct a thermal imaging frame image; Convert the thermal imaging frame image to grayscale, construct a grayscale image, establish the correspondence between the pixel grayscale value and the actual temperature through blackbody radiation calibration, determine the temperature value of each pixel point of each grayscale image according to the correspondence, and construct a temperature matrix; Perform an alignment and integration operation on the temperature matrix according to the time series to construct a temperature distribution change matrix of the target power module, obtain the real-time temperature field distribution data, and calculate the temperature gradient vector of each pixel point on the surface of the target power module according to the real-time temperature field distribution data; Construct a temperature gradient distribution map according to the amplitude and direction of the temperature gradient vector, extract the area where the temperature change rate is higher than the set threshold as the high heat flux density area according to the temperature gradient distribution map, and fuse adjacent high heat flux density areas based on the region growing algorithm to generate a heat flow aggregation zone; Extend the heat flow aggregation zone along the temperature gradient direction, calculate the direction consistency coefficient of the temperature gradient vector in the extended path, and smooth and optimize the extended path according to the direction consistency coefficient to form the initial heat flow path of the target power module.

3. The heat dissipation optimization method for a power device based on heat flow path reconstruction according to claim 1, characterized in that The conducting of a thermal sensitivity analysis on each component in the target power module to determine the thermal sensitivity data of each component is specifically as follows: Obtain the working parameter change time series data of each component in the target power module in a preset time period, and the working parameter change time series data includes the time change sequences of voltage, current and power parameters; Extract the temperature change time series data of the area where each component is located according to the real-time temperature field change data, synchronize and align the temperature change time series data with the working parameter change time series data, and construct a temperature-working parameter data matrix; Conduct a correlation analysis on the temperature-working parameter data matrix by the Pearson correlation coefficient algorithm, calculate the correlation coefficient of the temperature change to each working parameter change, conduct a significance test on the correlation coefficient by the t-test method, and screen out the effective correlation coefficients with a significance level lower than the preset significance threshold; Evaluate the influence degree of temperature change on the change of component operating parameters according to the absolute value of the effective correlation coefficient, and determine the thermal sensitivity coefficient of each component according to the influence degree to obtain the thermal sensitivity data of each component.

4. A heat dissipation optimization method for a power device based on heat flow path reconstruction according to claim 1, characterized in that The determination of the heat flow shielding area of the target power module according to the thermal sensitivity data is specifically as follows: Perform position mapping on the thermal sensitivity coefficients of the components in the target power module according to the thermal sensitivity data to generate a thermal sensitivity coefficient distribution map; Introduce a density clustering algorithm to perform spatial clustering analysis on the thermal sensitivity coefficient distribution map, set the neighborhood radius and minimum sample number of the density clustering algorithm, and initialize the data points in the thermal sensitivity coefficient distribution map as an unlabeled set; Select any data point from the unlabeled set, calculate the number of samples contained within its neighborhood radius. If the number of samples is greater than or equal to the minimum sample number, mark this point as a core point, and traverse all unvisited points within its neighborhood, recursively merging adjacent points that meet the core point conditions into the same clustering cluster; If the number of samples within the neighborhood is less than the minimum sample number, mark this point as a noise point and temporarily store it in an independent set. After traversing all unlabeled data points, output multiple thermal sensitivity data clusters composed of core points and density-reachable points; Calculate the average value of the thermal sensitivity coefficients within each thermal sensitivity data cluster, and mark the thermal sensitivity data clusters with an average value higher than the preset sensitivity threshold as high-thermal-sensitivity data clusters; According to the geometric center coordinates and boundary range of the high-thermal-sensitivity data cluster, extract the path segments in the initial heat flow path that intersect with the boundary of the high-thermal-sensitivity data cluster or have a distance less than the preset safety distance as the path segments to be shielded; Calculate the product of the temperature gradient vector corresponding to the path segment to be shielded and the average value of the thermal sensitivity coefficients of the high-thermal-sensitivity data cluster to generate a path shielding factor, and mark the path segments to be shielded with a path shielding factor greater than the preset shielding threshold as high-shielding-priority path segments; Based on the position coordinates of the high-shielding-priority path segments, use a region fusion algorithm to merge adjacent high-shielding-priority path segments to generate a continuous boundary of the heat flow shielding area; Calculate the spatial distance between the temperature gradient vector in the initial heat flow path and the high-thermal-sensitivity data cluster. If the distance is less than the preset protection distance, include the corresponding area in the heat flow shielding area, and finally generate a heat flow shielding area covering all high-shielding-priority path segments.

5. A heat dissipation optimization method for a power device based on heat flow path reconstruction according to claim 1, characterized in that The reconstruction of the initial heat flow path according to the heat flow shielding area to obtain an optimized heat flow path is specifically as follows: According to the boundary coordinates of the heat flow shielding area and the node distribution of the initial heat flow path, extract the initial heat flow path segments that intersect with the heat flow shielding area as the path adjustment objects; Calculate the equivalent heat dissipation efficiency of each region according to the heat conductivity distribution data of the circuit board substrate of the target power module and the heat diffusion characteristics of the copper layer coverage area; Based on the equivalent heat dissipation efficiency and the distribution of low-temperature regions in the real-time temperature field distribution data, screen out non-heat flow shielding regions with a heat dissipation efficiency higher than the preset efficiency threshold and a temperature gradient amplitude lower than the preset gradient threshold as candidate guiding regions; Determine the optimal heat flow guiding region of the path adjustment object by combining the consistency coefficient of the temperature gradient direction with the spatial distance between the candidate guiding region and the path adjustment object; Based on the geometric center coordinates of the optimal heat flow guiding region, offset the path nodes of the initial heat flow path intersecting the heat flow shielding region in the heat flow diffusion direction to generate an optimized heat flow path that bypasses the heat flow shielding region and extends along the copper layer coverage region.

6. The heat dissipation optimization method for a power device based on heat flow path reconstruction according to claim 1, characterized in that The heat dissipation channel of the power device by the target power module is constructed according to the optimized heat flow path, and the heat dissipation of the power device is optimized according to the heat dissipation channel, specifically: Compare the initial heat flow path with the optimized heat flow path, judge the heat flow path difference of the target power module, and determine the heat flow change direction according to the heat flow path difference; Determine the heat dissipation channel of the power device by the target power module according to the heat flow change direction, determine the installation position of the heat dissipation device of the target power module according to the heat dissipation channel, and the heat dissipation device includes a heat dissipation micro copper tube, a heat sink, and a heat dissipation coating material; Optimize the heat dissipation of the power device according to the heat dissipation device.

7. A heat dissipation optimization system for power devices based on heat flow path reconstruction, characterized in that, The power device heat dissipation optimization system based on heat flow path reconstruction includes a memory and a processor. The memory includes a power device heat dissipation optimization method program based on heat flow path reconstruction. When the power device heat dissipation optimization method program based on heat flow path reconstruction is executed by the processor, the following steps are realized: Obtain the real-time temperature field distribution data of the target power module during the operation of each power device, and determine the initial heat flow path of the target power module according to the real-time temperature field distribution data; Perform thermal sensitivity analysis on each component in the target power module to determine the thermal sensitivity data of each component; Determine the heat flow shielding region of the target power module according to the thermal sensitivity data, and reconstruct the initial heat flow path according to the heat flow shielding region to obtain an optimized heat flow path; Construct a heat dissipation channel for the power device by the target power module according to the optimized heat flow path, and optimize the heat dissipation of the power device according to the heat dissipation channel.

Images