Circuit board device thermal analysis system based on target detection

Abstract

The invention relates to the technical field of electronic device thermal management, in particular to a circuit board device thermal analysis system based on target detection, which comprises a thermal processor, and the thermal processor is in communication connection with a thermal map generation module, a thermal model construction analysis module, an improved evaluation module and a thermal response module. An image acquisition unit and an image processing unit are arranged in the heat map generation module; according to the method, the device on the PCB is accurately identified and positioned through image processing and target detection technologies, and the thermal infrared image is accurately associated with the device, so that thermal analysis more accurately reflects the actual heat production condition of the PCB, a more targeted basis is provided for thermal management through model construction and analysis, and the thermal management efficiency is improved. Meanwhile, the thermal analysis label is obtained based on information feedback, targeted management on the PCB based on the thermal analysis label is facilitated, simulation evaluation is conducted on improvement of the improvement scheme, whether the improvement scheme of the PCB reaches the standard or not is judged, and the improvement effect and feasibility of the improvement scheme are visually known.

Description

Technical Field

[0001] The invention relates to the technical field of electronic device thermal management, and in particular to a circuit board device thermal analysis system based on target detection. Background Art

[0002] As electronic devices develop towards miniaturization and high integration, the density of devices on circuit boards continues to increase, and thermal management issues become increasingly prominent. If the heat generated by electronic devices during operation cannot be dissipated in a timely and effective manner, the device temperature will rise, affecting its performance and reliability, and even causing system failures. Therefore, accurate thermal analysis of circuit board devices and timely detection of hot spots and thermal problems are of great significance to ensure the stable operation of electronic equipment. The existing patent CN110942458A proposes a method and system for detecting and locating temperature anomaly defects. The system includes an infrared thermal imaging temperature measurement module, a visible light imaging module, a parameter setting module, a result display module, an alarm module, and an image analysis and positioning module. The temperature anomaly detection and positioning results are obtained through image analysis and processing. However, the existing technology still has the following problems in the thermal analysis of circuit board devices: it is difficult to accurately associate the hot zone with the specific electronic device, it is difficult to locate the specific source of the thermal problem, resulting in reduced positioning accuracy of the electronic device, and it is difficult to accurately identify the overheated device and provide feedback on the cause type, which is not conducive to the management of the overheated device. At the same time, the existing technology lacks a thermal management optimization suggestion mechanism and cannot provide targeted improvement solutions for circuit board design; In view of the above technical defects, a solution is now proposed. Summary of the invention

[0003] The purpose of the present invention is to provide a circuit board device thermal analysis system based on target detection to solve the above-mentioned technical defects. The present invention accurately identifies and locates devices on the PCB board through image processing and target detection technology, and accurately associates thermal infrared images with devices to obtain temperature data of each device, so that thermal analysis can more accurately reflect the actual heat generation of the PCB board. In the association process, the image association matching accuracy is analyzed to improve the image registration accuracy and reliability. Through model construction and analysis, the temperature data of each device can be analyzed in detail, and local hot spots on the PCB board can be found in time, that is, the local hot spot detection capability is enhanced, and a more targeted basis is provided for thermal management. At the same time, a thermal analysis label is obtained based on information feedback, which helps to intuitively understand the thermal analysis feedback results of the PCB board on the one hand, and on the other hand, the PCB board is managed in a targeted manner based on the thermal analysis label. By performing simulation evaluation on the improvement of the improvement plan, it is judged whether the improvement plan of the PCB board meets the standards. Information feedback helps to intuitively understand the improvement effect and feasibility of the improvement plan.

[0004] The object of the present invention can be achieved by the following technical solutions: A circuit board device thermal analysis system based on object detection, including a thermal processor, which is communicatively connected to a thermal map generation module, a thermal model construction and analysis module, an improvement evaluation module, and a thermal response module. An image acquisition unit and an image processing unit are provided in the thermal map generation module; The image acquisition unit is used to acquire the optical feature image and the thermal infrared feature image of the PCB board. The image processing unit is used to perform image matching and fusion analysis on the optical feature image and the thermal infrared feature image to obtain a device-temperature map; The thermal processor is used to store and respond to the device-temperature map, and display the device-temperature map on the display panel; The thermal model construction and analysis module is used to construct a three-dimensional heat conduction model of the PCB board and perform thermal management analysis and calibration operations to obtain overheated devices + defect types + defect hazard scores, and set the overheated devices + defect types + defect hazard scores as thermal analysis labels; The improvement evaluation module is used to perform improvement effect evaluation and feedback analysis on the improvement parameters in the improvement plan pre-generated for the PCB board, and perform discrimination processing on the obtained improvement effect scores to determine whether the improvement plan of the PCB board meets the standard.

[0005] Preferably, the image matching and fusion analysis process is as follows: S1: Place the PCB board to be analyzed at a fixed position, and obtain the optical feature image and the thermal infrared feature image of the PCB board through an optical camera and a thermal infrared camera; S2: Input the optical feature image into a pre-set object detection model. The object detection model processes the optical feature image to identify the category information and position coordinates of each device; S3: Preprocess the optical feature image and the thermal infrared feature image, and register the preprocessed optical feature image and thermal infrared feature image through feature point matching and transformation matrix calculation; S4: Extract the temperature data of each device in the PCB board. The temperature data includes the highest temperature, the lowest temperature, and the average temperature. Obtain the device-temperature map of the PCB board according to the temperature data and position coordinates of each device.

[0006] Preferably, S3: The process of realizing registration is as follows: S31: Set the pads, vias, and device corners in the PCB board as feature points; S32: Obtain the visual description features of each feature point. The visual description features include the temperature gradient direction and the texture pattern; S33: Calculate the Hamming distance between the visual description features of the optical feature image and the thermal infrared feature image based on the visual description features of each feature point and the pre-set nearest neighbor algorithm, and filter out the feature point matching pairs with the Hamming distance less than the pre-set Hamming distance threshold.

[0007] Preferably, it further includes S34: Randomly select n groups of feature point matching pairs, where n is a natural number greater than 3. Respectively construct the spatial coordinates of the thermal infrared feature image and the optical feature image, and select a point to be mapped in the thermal infrared feature image. Map the spatial coordinates of the thermal infrared feature image to the coordinate system of the optical feature image through the existing affine transformation technology to obtain the mapped point. Obtain the ratio of the number of feature point matching pairs corresponding to the error value between the mapped point and the corresponding actual point being less than the pre-set error threshold to n, and set the ratio of the number of feature point matching pairs corresponding to the error value between the mapped point and the actual point being less than the pre-set error threshold to n as the matching accuracy, and perform discriminant processing on the matching accuracy to obtain an adjustment signal or a matching signal.

[0008] Preferably, the thermal management analysis and calibration operation process is as follows: Based on the design drawing of the PCB board and the device-temperature diagram, construct a three-dimensional heat conduction model of the PCB board. Divide the PCB board into multiple grid units in the three-dimensional heat conduction model, set the thermal conductivity and specific heat capacity of each grid unit, and at the same time set the environmental conditions and constraint conditions. The environmental conditions include environmental temperature and air flow rate, and the constraint conditions include device power consumption and interface contact thermal resistance; Set the simulation duration, obtain the power consumption of each device within the simulation duration based on the three-dimensional heat conduction model, and set the power consumption of the device as the internal heat source, and then obtain the heat conduction process of the heat in the PCB board during the simulation process of each internal heat source; Based on the three-dimensional heat conduction model, obtain the temperature distribution characteristic image and the heat flux density distribution characteristic image of the PCB board at different time points through simulation.

[0009] Preferably, use the temperature distribution characteristic image and the heat flux density distribution characteristic image as the input layer and input them into the pre-set thermal defect recognition model to obtain the output result of the thermal defect recognition model. The output result includes defect type and defect hazard score; At the same time, obtain the temperature values of each device in the three-dimensional heat conduction model within the simulation duration, obtain the devices corresponding to the temperature values exceeding the pre-set temperature value threshold, and set the devices corresponding to the temperature values exceeding the pre-set temperature value threshold as risk devices. At the same time, obtain the duration corresponding to the temperature values of the risk devices exceeding the pre-set temperature value threshold, and set the risk devices corresponding to the duration exceeding the pre-set duration threshold as overheated devices; Set the overheated device + defect type + defect hazard score as the thermal analysis label.

[0010] Preferably, the improvement effect evaluation and feedback analysis process is as follows: Based on the thermal analysis tags, the improvement parameters in the improvement plan pre-generated for the PCB board are obtained. The improvement parameters in the improvement plan pre-generated for the PCB board are reset in the three-dimensional heat conduction model, and the improvement effect parameters of the overheated devices before and after optimization are obtained. The improvement effect parameters include the temperature drop value and the peak value reduction of the heat flux density.

[0011] Preferably, a pre-set weight factor coefficient is assigned to each parameter in the improvement effect parameters, and each parameter in the simulation effect parameters is multiplied by the corresponding weight factor coefficient. The sum value obtained after multiplying each parameter in the simulation effect parameters by the corresponding weight factor coefficient is set as the improvement effect score, and a determination process is performed on the improvement effect score to obtain a non-conforming signal or a conforming signal.

[0012] The beneficial effects of the present invention are as follows: (1) The present invention accurately identifies and locates the devices on the PCB board through image processing and target detection technologies, and precisely associates the thermal infrared image with the devices to obtain the temperature data of each device, enabling the thermal analysis to more accurately reflect the actual heat generation situation of the PCB board. During the association process, the image association matching accuracy is analyzed to improve the registration accuracy and reliability of the image; (2) Through model construction and analysis, the present invention can analyze the temperature data of each device in detail, promptly discover local hot spots on the PCB board, that is, enhance the local hot spot detection ability, provide a more targeted basis for thermal management, and at the same time obtain thermal analysis tags based on information feedback. On the one hand, it helps to intuitively understand the thermal analysis feedback results of the PCB board, and on the other hand, it conducts targeted management of the PCB board based on the thermal analysis tags. By performing simulation evaluation on the improvement of the improvement plan, it is judged whether the improvement plan of the PCB board meets the standards, and through information feedback, it helps to intuitively understand the improvement effect and feasibility of the improvement plan. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The present invention will be further described below with reference to the accompanying drawings; Figure 1 is the system flowchart of the present invention; Figure 2 is the local analysis reference diagram of the present invention; Figure 3 is the analysis step diagram of Embodiment 1 of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0014] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0015] Referring to "embodiments" herein means that the specific features, structures, or characteristics described in connection with the embodiments may be included in at least one embodiment of the present invention. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein may be combined with other embodiments;

[0016] Embodiment 1: Please refer to Figures 1 to 3 As shown, the present invention is a circuit board device thermal analysis system based on object detection, including a thermal processor, which is communicatively connected to a thermal image generation module, a thermal model construction and analysis module, an improved evaluation module, and a thermal response module. An image acquisition unit and an image processing unit are provided in the thermal image generation module; The thermal processor is in a bidirectional communication connection with the thermal image generation module, the thermal processor is in a unidirectional communication connection with the thermal model construction and analysis module, the thermal model construction and analysis module is in a unidirectional communication connection with both the improved evaluation module and the thermal response module, and the improved evaluation module is in a unidirectional communication connection with the thermal response module; The image acquisition unit is used to acquire the optical feature image and the thermal infrared feature image of the PCB board; The image processing unit is used to perform image matching and fusion analysis on the optical feature image and the thermal infrared feature image to obtain a device-temperature map. The specific image matching and fusion analysis process is as follows: S1: Place the PCB board to be analyzed in a fixed position, and obtain the optical feature image and the thermal infrared feature image of the PCB board through an optical camera and a thermal infrared camera. It should be noted that the image resolution should meet the requirements of device recognition and temperature analysis. In actual applications, first, under the normal working state of the PCB board, collect the infrared thermal imaging image and the visible light image simultaneously. During collection, ensure that the PCB board has been operating stably for at least 30 minutes to obtain the steady-state thermal distribution; S2: Input the optical feature image into a pre-set object detection model. The object detection model processes the optical feature image to identify the category information and position coordinates of each device. The category information includes various devices such as resistors, capacitors, inductors, chips, etc. That is, process the visible light image, and identify the positions and types of each device on the PCB board through a deep learning network. The recognition results include device types and position coordinates; For example: The function of the target detection model is to split the optical feature image into pixel points, and layer by layer extract "features" through a convolutional neural network (CNN) - from small device edges and corners to the overall shape and texture of the device (such as the circular shape of a capacitor and the rectangular outline of a chip); Train the model with a large amount of labeled data (circuit board images with device positions and categories) to let the model remember "what features correspond to what devices", such as "rectangle + pin features = chip" and "circle + color rings = resistor"; S3: Preprocess the optical feature image and the thermal infrared feature image. The preprocessing includes operations such as noise reduction, enhancement, and size unification. Register the preprocessed optical feature image and thermal infrared feature image through feature point matching and transformation matrix calculation; S4: Extract the temperature data of each device in the PCB board. The temperature data includes the maximum temperature, minimum temperature, average temperature, etc. Obtain the device-temperature map of the PCB board based on the temperature data and position coordinates of each device; The thermal processor is used to store the response device-temperature map and display the device-temperature map on the display panel to intuitively understand the temperature of each device in the PCB board and the temperature distribution of the PCB board; In the embodiment of the present invention, through the target detection technology, the hot area is accurately associated with specific electronic devices, and the source of the thermal problem is accurately located; device-level thermal characteristic analysis is provided to evaluate the contribution of each device to the overall thermal distribution; the thermal analysis results are combined with the PCB board design data to provide accurate guidance for thermal design optimization; the degree of automation is high, and the efficiency and accuracy of the thermal analysis of the PCB board can be significantly improved; S3: The process of registering the preprocessed optical feature image and thermal infrared feature image through feature point matching and transformation matrix calculation is as follows: S31: Set the pads, vias, device corners, etc. in the PCB board as feature points; S32: Obtain the visual description features of each feature point. The visual description features include the temperature gradient direction, texture pattern, etc.; S33: Based on the visual description features of each feature point and the pre-set nearest neighbor algorithm, calculate the Hamming distance between the feature points in the optical feature image and the thermal infrared feature image, and screen out the feature point matching pairs with the Hamming distance less than the pre-set Hamming distance threshold; That is, using the similarity measure between visual description features, match the feature points in the optical feature image and the thermal infrared feature image, and calculate the Hamming distance through the pre-set nearest neighbor algorithm. The smaller the Hamming distance, the more similar the two feature points are; The nearest neighbor algorithm is a machine learning method based on instances, and its core is to achieve classification or regression prediction by calculating the distance between samples; S34: Randomly select n groups of feature point matching pairs, where n is a natural number greater than 3. Respectively construct the spatial coordinates of the thermal infrared feature image and the spatial coordinates of the optical feature image. And select a point to be mapped in the thermal infrared feature image, and map the spatial coordinates of the thermal infrared feature image to the coordinate system of the optical feature image through the existing affine transformation technology to obtain the mapped point. Obtain the ratio of the number of feature point matching pairs with the error value between the mapped point and the corresponding actual point less than the preset error threshold to n, and set the ratio of the number of feature point matching pairs with the error value between the mapped point and the actual point less than the preset error threshold to the matching accuracy. And perform discrimination processing on the matching accuracy. If the matching accuracy is less than the preset matching accuracy threshold, generate an adjustment signal. If the matching accuracy is greater than or equal to the preset matching accuracy threshold, generate a matching signal; The thermal response module is used to respond to the adjustment signal or the matching signal, and immediately display the preset warning text corresponding to the adjustment signal or the matching signal, so as to adjust and optimize the registration process of the optical feature image and the thermal infrared feature image according to the text feedback situation, so as to improve the registration accuracy and reliability.

[0017] Embodiment 2: The thermal model construction and analysis module is used to construct a three-dimensional heat conduction model of the PCB board and perform thermal management analysis and calibration operations to obtain a thermal analysis label. The specific process of the thermal management analysis and calibration operation is as follows: Based on the design drawing of the PCB board and the device-temperature diagram, construct a three-dimensional heat conduction model of the PCB board. Divide the PCB board into multiple grid units in the three-dimensional heat conduction model, and set the thermal physical properties parameters such as the thermal conductivity and specific heat capacity of each grid unit. At the same time, set the environmental conditions and constraint conditions. The environmental conditions include environmental temperature, air flow rate, etc. The constraint conditions include device power consumption, interface contact thermal resistance, etc.; Set the simulation duration, obtain the power consumption of each device within the simulation duration based on the three-dimensional heat conduction model, and set the power consumption of the device as the internal heat source. Furthermore, obtain the heat conduction process of the heat in the PCB board during the simulation process of each internal heat source; Based on the three-dimensional heat conduction model, obtain the temperature distribution characteristic image and the heat flux density distribution characteristic image of the PCB board at different time points through simulation; Take the temperature distribution characteristic image and the heat flux density distribution characteristic image as the input layer and input them into the pre-set thermal defect recognition model to obtain the output result of the thermal defect recognition model. The output result includes defect type, defect hazard score (0-100), and the defect type includes layout defect, material defect, etc.; Meanwhile, the temperature values of each device in the three-dimensional heat conduction model within the simulation duration are obtained. Devices with temperature values exceeding the preset temperature value threshold are obtained, and the devices with temperature values exceeding the preset temperature value threshold are set as risk devices. Meanwhile, the duration during which the temperature value of the risk device exceeds the preset temperature value threshold is obtained, and the risk device with a duration exceeding the preset duration threshold is set as an overheated device; The overheated device + defect type + defect hazard score is set as the thermal analysis label. The thermal response module is used to respond to the thermal analysis label and immediately display the preset warning text corresponding to the thermal analysis label. On the one hand, it helps to intuitively understand the thermal analysis feedback results of the PCB board. On the other hand, it conducts targeted management of the PCB board based on the thermal analysis label; The improvement evaluation module is used to conduct an improvement effect evaluation and feedback analysis on the improvement parameters in the improvement plan pre-generated for the PCB board, and conduct a discriminant process on the obtained improvement effect score to determine whether the improvement plan of the PCB board meets the standard. The specific improvement effect evaluation and feedback analysis process is as follows: Based on the thermal analysis label, the improvement parameters in the improvement plan pre-generated for the PCB board are obtained. The improvement parameters in the improvement plan pre-generated for the PCB board are reset in the three-dimensional heat conduction model, and the improvement effect parameters of the overheated device before and after optimization are obtained. The improvement effect parameters include the temperature drop value, the peak reduction of the heat flux density, etc.; For example: the improvement parameter is to increase the size of the heat sink (such as heat sink material = copper alloy, size = 30mm × 30mm × 5mm); Pre-set weight factor coefficients are assigned to each parameter in the improvement effect parameters. Each parameter in the simulation effect parameters is multiplied by the corresponding weight factor coefficient. The sum value obtained after multiplying each parameter in the simulation effect parameters by the corresponding weight factor coefficient is set as the improvement effect score, and a determination process is conducted on the improvement effect score. If the improvement effect score is less than the preset improvement effect score threshold, an unqualified signal is generated. If the improvement effect score is greater than or equal to the preset improvement effect score threshold, a qualified signal is generated. The thermal response module is used to respond to the unqualified signal or the qualified signal and immediately display the preset warning text corresponding to the unqualified signal or the qualified signal. On the one hand, it helps to intuitively understand the improvement effect and feasibility of the improvement plan. On the other hand, it helps to timely adjust the improvement parameters, such as replacing a higher thermal conductivity material, thereby improving the thermal management effect of the PCB board; In summary, the present invention accurately identifies and locates the devices on the PCB through image processing and object detection technologies, precisely associates the thermal infrared image with the devices, obtains the temperature data of each device, enables the thermal analysis to more accurately reflect the actual heat generation situation of the PCB. During the association process, the image association matching accuracy is analyzed to improve the registration accuracy and reliability of the image. Through model construction and analysis, the temperature data of each device can be analyzed in detail, and local hotspots on the PCB can be detected in a timely manner, that is, the local hotspot detection ability is enhanced, providing a more targeted basis for thermal management. At the same time, based on the information feedback, thermal analysis labels are obtained. On the one hand, it helps to intuitively understand the thermal analysis feedback results of the PCB. On the other hand, targeted management of the PCB is carried out based on the thermal analysis labels. Through the simulation evaluation of the improvement of the improvement scheme, it is judged whether the improvement scheme of the PCB meets the standard. The information feedback helps to intuitively understand the improvement effect and feasibility of the improvement scheme.

[0018] The setting of the threshold is for result comparison and analysis to determine whether it is good or bad. Regarding the value of the threshold, it is set and stored by combining the large model analysis of sample data and manual experience, and can also be appropriately adjusted according to seasonal or regular influencing conditions.

[0019] The magnitude of the coefficient is a specific value obtained by quantifying each parameter, which is convenient for subsequent comparison. Regarding the magnitude of the coefficient, it depends on the amount of sample data and the corresponding operating coefficients initially set by those skilled in the art for each group of sample data; as long as the proportional relationship between the parameter and the quantified value is not affected.

[0020] The above is only a preferred specific embodiment 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, according to the technical solution of the present invention and its inventive concept, makes equivalent substitutions or changes, and all should be covered within the protection scope of the present invention.

Claims

1. A circuit board device thermal analysis system based on object detection, characterized in that It includes a thermal processor, which is communicatively connected to a thermal map generation module, a thermal model construction and analysis module, an improvement evaluation module, and a thermal response module. An image acquisition unit and an image processing unit are provided in the thermal map generation module; The image acquisition unit is used to acquire the optical feature image and the thermal infrared feature image of the PCB board, and the image processing unit is used to perform image matching and fusion analysis on the optical feature image and the thermal infrared feature image to obtain a device-temperature map; The thermal processor is used to store the response device-temperature map and display the device-temperature map on the display panel; The thermal model construction and analysis module is used to construct a three-dimensional heat conduction model of the PCB board and perform thermal management analysis and calibration operations to obtain an overheated device + defect type + defect hazard score, and set the overheated device + defect type + defect hazard score as a thermal analysis label; The improvement evaluation module is used to perform improvement effect evaluation and feedback analysis on the improvement parameters in the pre-generated improvement plan of the PCB board, and perform discrimination processing on the obtained improvement effect score to determine whether the improvement plan of the PCB board meets the standard.

2. The thermal analysis system of circuit board devices based on object detection according to claim 1, wherein The image matching and fusion analysis process is as follows: S1: Place the PCB board to be analyzed at a fixed position, and acquire the optical feature image and the thermal infrared feature image of the PCB board through an optical camera and a thermal infrared camera; S2: Input the optical feature image into a pre-set target detection model, and the target detection model processes the optical feature image to identify the category information and position coordinates of each device; S3: Preprocess the optical feature image and the thermal infrared feature image, and register the preprocessed optical feature image and thermal infrared feature image through feature point matching and transformation matrix calculation; S4: Extract the temperature data of each device in the PCB board. The temperature data includes the highest temperature, the lowest temperature, and the average temperature. Obtain the device-temperature map of the PCB board according to the temperature data and position coordinates of each device.

3. The thermal analysis system for circuit board devices based on object detection according to claim 2, wherein S3: The process of realizing registration is as follows: S31: Set the pads, vias, and device corners in the PCB board as feature points; S32: Obtain the visual description features of each feature point. The visual description features include the temperature gradient direction and the texture pattern; S33: Based on the visual description features of each feature point and the pre-set nearest neighbor algorithm, calculate the Hamming distance of the visual description features of the optical feature image and the thermal infrared feature image, and screen out the feature point matching pairs with the Hamming distance less than the preset Hamming distance threshold.

4. The thermal analysis system for circuit board devices based on object detection according to claim 3, wherein It also includes S34: Randomly select n groups of feature point matching pairs, where n is a natural number greater than 3. Respectively construct the spatial coordinates of the thermal infrared feature image and the optical feature image, and select a point to be mapped in the thermal infrared feature image. Map the spatial coordinates of the thermal infrared feature image to the optical feature image coordinate system through the existing affine transformation technology to obtain a mapped point. Obtain the ratio of the number of feature point matching pairs with the error value between the mapped point and the corresponding actual point less than the preset error threshold to n, and set the ratio of the number of feature point matching pairs with the error value between the mapped point and the actual point less than the preset error threshold to n as the matching accuracy, and perform discrimination processing on the matching accuracy to obtain an adjustment signal or a matching signal.

5. The thermal analysis system of circuit board devices based on object detection according to claim 1, wherein The process of thermal management analysis and calibration is as follows: Based on the design drawing of the PCB board and the device-temperature diagram, a three-dimensional heat conduction model of the PCB board is constructed. In the three-dimensional heat conduction model, the PCB board is divided into multiple grid units, and the thermal conductivity and specific heat capacity of each grid unit are set. At the same time, the environmental conditions and constraint conditions are set. The environmental conditions include the ambient temperature and air flow rate, and the constraint conditions include the device power consumption and the interface contact thermal resistance; Set the simulation duration, obtain the power consumption of each device within the simulation duration based on the three-dimensional heat conduction model, and set the power consumption of the device as the internal heat source, and then obtain the conduction process of heat in the PCB board during the simulation of each internal heat source; Based on the three-dimensional heat conduction model, obtain the temperature distribution characteristic image and heat flux density distribution characteristic image of the PCB board at different time points through simulation.

6. The thermal analysis system for circuit board devices based on object detection according to claim 5, characterized in that, Take the temperature distribution characteristic image and heat flux density distribution characteristic image as the input layer and input them into the pre-set thermal defect recognition model to obtain the output result of the thermal defect recognition model. The output result includes the defect type and defect hazard score; At the same time, obtain the temperature values of each device in the three-dimensional heat conduction model within the simulation duration, obtain the devices corresponding to the temperature values exceeding the preset temperature value threshold, and set the devices corresponding to the temperature values exceeding the preset temperature value threshold as risk devices. At the same time, obtain the duration corresponding to the temperature value of the risk device exceeding the preset temperature value threshold, and set the risk device corresponding to the duration exceeding the preset duration threshold as an overheated device; Set the overheated device + defect type + defect hazard score as the thermal analysis label.

7. The thermal analysis system for circuit board devices based on object detection according to claim 1, characterized in that, The process of the improvement effect evaluation and feedback analysis is as follows: Obtain the improvement parameters in the pre-generated improvement plan of the PCB board based on the thermal analysis label, re-set the improvement parameters in the pre-generated improvement plan of the PCB board in the three-dimensional heat conduction model, and obtain the improvement effect parameters of the overheated device before and after optimization. The improvement effect parameters include the temperature drop value and the peak value reduction of the heat flux density.

8. The thermal analysis system for circuit board devices based on object detection according to claim 7, wherein Assign the pre-set weight factor coefficients to each parameter in the improvement effect parameters, multiply each parameter in the simulation effect parameters by the corresponding weight factor coefficient, set the sum value obtained by multiplying each parameter in the simulation effect parameters by the corresponding weight factor coefficient as the improvement effect score, and perform a determination process on the improvement effect score to obtain a non-conforming signal or a conforming signal.

Images