Heat dissipation control method and device for ultra-high voltage silicon-based MOSFET structure

Abstract

The present application provides a heat dissipation control method and device for a UHV silicon-based MOSFET structure, which relates to the field of semiconductor device technology, including: performing integrated simulation generation of drift zone temperature distribution based on monitored temperature information, obtaining multiple simulated drift zone temperature distributions, and performing accuracy analysis and screening the simulated drift zone temperature distribution with the highest accuracy to obtain the drift zone temperature distribution; performing heat dissipation control optimization based on the drift zone temperature distribution to obtain the optimal heat dissipation control parameters, wherein the heat dissipation control optimization includes power adjustment optimization and radiator control optimization; using the optimal heat dissipation control parameters for heat dissipation control. This application can solve the technical problem that traditional heat dissipation solutions cannot effectively handle the local hot spot phenomenon in the drift zone inside the UHV MOSFET, and there is a poor flexibility and accuracy in the heat dissipation control of the drift zone, and can achieve the technical goal of efficient heat dissipation control, ensuring the stability and reliability of the device when working at high power.

Description

Technical Field

[0001] The present application relates to the field of semiconductor device technology, and in particular to a heat dissipation control method and device for an ultra-high voltage silicon-based MOSFET structure. Background Art

[0002] The heat dissipation solution of UHV silicon-based MOSFET usually relies on external heat sinks, fans or liquid cooling systems, mainly reducing the temperature by dissipating heat outside the device. Due to the uneven distribution of current in the drift region, some areas will have higher temperatures than other areas. Traditional heat dissipation methods cannot respond to this local thermal effect in a timely and accurate manner, resulting in the inability to effectively solve the problem of local overheating. The resulting local overheating will not only accelerate the aging of the device, but may even cause thermal failure of the device.

[0003] In addition, traditional heat dissipation solutions usually adopt static and passive heat dissipation control methods, which are difficult to adjust in real time according to actual temperature changes. This heat dissipation control strategy is not only inefficient, but may also lead to increased power consumption and decreased device performance.

[0004] In summary, traditional heat dissipation solutions cannot effectively deal with the local hot spot phenomenon in the drift region inside the UHV MOSFET, and there are technical problems such as poor flexibility and accuracy in heat dissipation control in the drift region. Summary of the invention

[0005] The purpose of this application is to provide a heat dissipation control method and device for a UHV silicon-based MOSFET structure, so as to solve the technical problem that traditional heat dissipation solutions cannot effectively deal with the local hot spot phenomenon in the drift region inside the UHV MOSFET, and there is poor flexibility and accuracy in heat dissipation control in the drift region.

[0006] In view of the above problems, the present application provides a heat dissipation control method and device for an ultra-high voltage silicon-based MOSFET structure.

[0007] In the first aspect, the present application provides a heat dissipation control method for an ultra-high voltage silicon-based MOSFET structure, which is implemented by a heat dissipation control device for an ultra-high voltage silicon-based MOSFET structure, comprising: monitoring and obtaining monitoring temperature information of the drift region during operation through a temperature sensor configured on a drift region within the ultra-high voltage silicon-based MOSFET structure, wherein a plurality of temperature sensors are configured on the drift region; performing integrated simulation generation of the drift region temperature distribution based on the monitoring temperature information to obtain a plurality of simulated drift region temperature distributions, and performing accuracy analysis and screening of the simulated drift region temperature distribution with the highest accuracy to obtain the drift region temperature distribution; when the drift region temperature distribution meets the heat dissipation requirements, performing heat dissipation control optimization of the ultra-high voltage silicon-based MOSFET structure based on the drift region temperature distribution to obtain optimal heat dissipation control parameters, wherein the heat dissipation control optimization includes power adjustment optimization and radiator control optimization; and performing heat dissipation control on the ultra-high voltage silicon-based MOSFET structure using the optimal heat dissipation control parameters.

[0008] In a second aspect, the present application also provides a heat dissipation control device for an ultra-high voltage silicon-based MOSFET structure, which is used to execute a heat dissipation control method for an ultra-high voltage silicon-based MOSFET structure as described in the first aspect, including: a temperature information monitoring module, which is used to monitor and obtain monitoring temperature information of the drift region during operation through a temperature sensor configured on the drift region within the ultra-high voltage silicon-based MOSFET structure, wherein a plurality of temperature sensors are configured on the drift region; a temperature distribution acquisition module, which is used to perform integrated simulation generation of the drift region temperature distribution according to the monitoring temperature information, obtain a plurality of simulated drift region temperature distributions, and perform accuracy analysis and screen the simulated drift region temperature distribution with the highest accuracy to obtain the drift region temperature distribution; a heat dissipation control optimization module, which is used to perform heat dissipation control optimization of the ultra-high voltage silicon-based MOSFET structure according to the drift region temperature distribution when the drift region temperature distribution meets the heat dissipation requirements, and obtain optimal heat dissipation control parameters, wherein the heat dissipation control optimization includes power adjustment optimization and radiator control optimization; a heat dissipation control module, which is used to use the optimal heat dissipation control parameters to perform heat dissipation control on the ultra-high voltage silicon-based MOSFET structure.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0010] The invention discloses a method for controlling the temperature distribution of the drift region of a UHV silicon-based MOSFET structure. The method comprises the following steps: configuring a plurality of temperature sensors on the drift region within the UHV silicon-based MOSFET structure, and using the plurality of temperature sensors to monitor and obtain the monitoring temperature information of the drift region during operation; then, based on the monitoring temperature information, performing integrated simulation generation of the temperature distribution of the drift region, obtaining a plurality of simulated temperature distributions of the drift region, and performing accuracy analysis on the plurality of simulated temperature distributions of the drift region, and then screening the simulated temperature distribution of the drift region with the greatest accuracy and setting it as the temperature distribution of the drift region; when the temperature distribution of the drift region meets the heat dissipation requirement, performing heat dissipation control optimization of the UHV silicon-based MOSFET structure according to the temperature distribution of the drift region, wherein the heat dissipation control optimization includes power adjustment optimization and heat sink control optimization, and obtaining the optimal heat dissipation control parameters; finally, using the optimal heat dissipation control parameters to control the heat dissipation of the UHV silicon-based MOSFET structure; the method can accurately obtain the temperature distribution of the drift region, improve the flexibility and accuracy of the heat dissipation control strategy setting, achieve the technical goal of efficient heat dissipation control, and ensure the stability and reliability of the UHV silicon-based MOSFET structure when working at high power.

[0011] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0013] Figure 1 A schematic flow chart of a heat dissipation control method for a UHV silicon-based MOSFET structure in this application;

[0014] Figure 2 A schematic diagram of a process for obtaining monitoring temperature information of a drift region in a heat dissipation control method for a UHV silicon-based MOSFET structure of the present application;

[0015] Figure 3 This is a structural schematic diagram of a heat dissipation control device for an ultra-high voltage silicon-based MOSFET structure in the present application.

[0016] Description of reference numerals:

[0017] Temperature information monitoring module 11, temperature distribution acquisition module 12, heat dissipation control optimization module 13, heat dissipation control module 14. DETAILED DESCRIPTION

[0018] This application provides a heat dissipation control method and device for a UHV silicon-based MOSFET structure, which solves the technical problem that the traditional heat dissipation solution cannot effectively deal with the local hot spot phenomenon in the drift region inside the UHV MOSFET, and has poor flexibility and accuracy in heat dissipation control in the drift region. The temperature distribution in the drift region can be accurately obtained, the flexibility and accuracy of the heat dissipation control strategy setting can be improved, the technical goal of efficient heat dissipation control can be achieved, and the stability and reliability of the UHV silicon-based MOSFET structure when working at high power can be guaranteed.

[0019] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0020] For example, please refer to the attached Figure 1 The present application provides a heat dissipation control method for a UHV silicon-based MOSFET structure, which is applied to a heat dissipation control device for a UHV silicon-based MOSFET structure, and specifically includes the following steps:

[0021] Step 1: During operation, the temperature information of the drift region is monitored and obtained by configuring a temperature sensor on the drift region in the UHV silicon-based MOSFET structure, wherein a plurality of temperature sensors are configured on the drift region.

[0022] Specifically, in the UHV silicon-based MOSFET structure, the drift region is designed to be long in order to withstand high voltage and prevent the device from breaking down under high voltage conditions. However, due to the uneven distribution of the drift region length and internal current density, local hot spots may occur in certain areas, which in turn affects the overall performance of the device. In order to effectively monitor and control the temperature distribution of the drift region, multiple temperature sensors can be arranged on the drift region. These sensors can collect temperature information at different locations in the drift region in real time.

[0023] First, multiple temperature sensors are configured at three key positions of the drift region in the UHV silicon-based MOSFET structure, including the area near the source, the middle of the drift region, and the area near the drain. Due to the long drift region and uneven current distribution, the heat accumulation at different positions may be quite different. By arranging sensors at these three key positions, the temperature changes in the drift region can be fully and carefully monitored, especially the local hotspot areas. The areas near the source and drain usually bear different voltages and currents, so the temperature monitoring of these positions is particularly important. The middle position, as the transition area between current and electric field, also needs to be monitored. Then, during the operation of the UHV silicon-based MOSFET structure, the temperature information of the three key positions of the drift region is obtained by monitoring multiple temperature sensors, and the temperature information of the three key positions is integrated to obtain the monitoring temperature information. By configuring multiple temperature sensors at key positions of the drift region (near the source, the middle, and the drain), the temperature information of different positions can be monitored and integrated. This method can monitor the temperature changes of different positions in the drift region in real time, provide detailed temperature distribution information, and provide an accurate basis for subsequent heat dissipation control.

[0024] Step 2: Based on the monitored temperature information, perform integrated simulation generation of the drift zone temperature distribution to obtain multiple simulated drift zone temperature distributions, perform accuracy analysis and screen the simulated drift zone temperature distribution with the highest accuracy to obtain the drift zone temperature distribution.

[0025] Specifically, first, a generative adversarial network is constructed, where the generative adversarial network consists of two parts: a generator and a discriminator. The generator generates different drift zone temperature distributions based on the temperature information of multiple monitoring points; the discriminator evaluates the authenticity of the generated temperature distribution by learning the real temperature distribution. Through adversarial training, the generator gradually improves its ability to generate temperature distribution, so that the generated temperature distribution is close to the actual monitoring data. This method can improve the accuracy of the drift zone temperature distribution simulation and ensure that the generated simulation distribution is more in line with the actual working conditions. Then, the generative adversarial network is used to perform an integrated simulation of the drift zone temperature distribution based on the monitored temperature information, that is, the generator generates the drift zone temperature distribution multiple times based on the monitored temperature information, and the discriminator makes judgments by learning the real temperature distribution, gradually improving the generator's generation ability, and finally obtaining multiple simulated drift zone temperature distributions.

[0026] Then, the accuracy of multiple simulated drift zone temperature distributions is analyzed. Since the drift zone is long, the temperature distributions in different areas may be significantly different. Therefore, for each simulated drift zone temperature distribution, the accuracy of each simulated distribution is judged by comparing its consistency with the actual monitoring data. The analysis includes multiple dimensions such as temperature deviation and reliability of the simulation path. Further based on the results of the accuracy analysis, the simulated drift zone temperature distribution with the highest accuracy is selected as the drift zone temperature distribution. By analyzing the accuracy of multiple simulated drift zone temperature distributions and selecting the result with the highest accuracy, the accuracy of the drift zone temperature distribution can be improved. At the same time, since the drift zone is long and the temperature distribution is complex, selecting the optimal simulated temperature distribution helps to accurately reflect the actual temperature changes in the drift zone, providing a reliable basis for subsequent heat dissipation control and optimization decisions.

[0027] Step three: When the temperature distribution in the drift region meets the heat dissipation requirements, the heat dissipation control optimization of the UHV silicon-based MOSFET structure is performed according to the temperature distribution in the drift region to obtain optimal heat dissipation control parameters, wherein the heat dissipation control optimization includes power adjustment optimization and radiator control optimization.

[0028] Specifically, when configuring the heat dissipation requirements, the heat dissipation requirements can be set according to actual scenarios to adapt to the needs of ultra-high voltage silicon-based MOSFETs in different working environments. Different temperature thresholds can be set according to the material properties and working environment of the MOSFET device. For example, in high-power operation scenarios, the temperature threshold of the drift zone is set to a higher value (such as 150°C), while in low-power or stable operation scenarios, a lower threshold (such as 100°C) is set to ensure that the device always operates within a safe range.

[0029] When the temperature distribution of the drift region reaches the heat dissipation requirement, the heat dissipation control optimization of the UHV silicon-based MOSFET structure is performed according to the temperature distribution of the drift region, wherein the heat dissipation control optimization includes power adjustment optimization and heat sink control optimization, wherein the heat sink (such as a fan, a liquid cooling system) is an important part of MOSFET heat dissipation, and according to the temperature distribution, the working state of the heat sink can be dynamically adjusted, such as increasing the fan speed, increasing the coolant flow, or enabling a more efficient heat dissipation mode, and this dynamic adjustment can be customized according to the temperature requirements of different regions, so as to better control the temperature of the hot spot area. On the other hand, the heat generation of MOSFET is closely related to its power output. By reducing the power output, the heat generation can be effectively reduced. When the temperature of the drift region is high, the working current or voltage of the MOSFET is appropriately reduced to reduce the risk of local overheating. At the same time, the power adjustment needs to be performed without affecting the system operation requirements, so as to ensure that the performance of the MOSFET is balanced with the heat dissipation requirements. Then, after the heat sink and power optimization, the effects of the two can be comprehensively analyzed to determine a set of optimal heat dissipation control parameters, which can minimize power loss while meeting the temperature requirements, ensure the stable and efficient operation of the MOSFET, and obtain the optimal heat dissipation control parameters, which include the optimal heat dissipation parameters and the optimal power adjustment parameters.

[0030] By simultaneously optimizing power regulation and heat sink control, the temperature of the MOSFET can be effectively reduced to prevent overheating while ensuring the stability and efficiency of power output.

[0031] Step 4: Using the optimal heat dissipation control parameters to control the heat dissipation of the ultra-high voltage silicon-based MOSFET structure.

[0032] Specifically, finally, the heat dissipation control of the UHV silicon-based MOSFET structure is performed according to the optimal heat dissipation control parameters, that is, the heat sink is adjusted according to the optimal heat dissipation parameters, such as increasing the speed of the fan or increasing the flow of the liquid cooling system. This dynamic adjustment ensures that the heat sink can always maintain the optimal operating state to cope with the different power load requirements of the MOSFET and ensure the heat dissipation efficiency; the power output of the UHV silicon-based MOSFET structure is controlled according to the optimal power adjustment parameters. For example, when it is detected that the temperature of the MOSFET is close to the preset threshold, the power output can be appropriately reduced to reduce the heat generation. The goal of power adjustment is to ensure that the MOSFET maintains the best working performance while meeting the heat dissipation requirements. By dynamically adjusting the heat sink and power output, it is possible to respond to temperature changes in real time, maintain the best working state of the device, and prevent overheating damage. At the same time, through continuous monitoring and optimization, the heat dissipation control system can maintain flexibility and stability under different load and environmental conditions, ensuring the efficient operation of the UHV silicon-based MOSFET structure when working at high power.

[0033] The heat dissipation control method of a UHV silicon-based MOSFET structure is applied to a heat dissipation control device of a UHV silicon-based MOSFET structure, which can solve the technical problems that traditional heat dissipation solutions cannot effectively deal with the local hot spot phenomenon in the drift region inside the UHV MOSFET, and have poor flexibility and accuracy in heat dissipation control in the drift region. The invention discloses a method for controlling the temperature distribution of the drift region of a UHV silicon-based MOSFET structure. The method comprises the following steps: configuring a plurality of temperature sensors on the drift region within the UHV silicon-based MOSFET structure, and using the plurality of temperature sensors to monitor and obtain the monitoring temperature information of the drift region during operation; then, based on the monitoring temperature information, performing integrated simulation generation of the temperature distribution of the drift region, obtaining a plurality of simulated temperature distributions of the drift region, and performing accuracy analysis on the plurality of simulated temperature distributions of the drift region, and then screening the simulated temperature distribution of the drift region with the greatest accuracy and setting it as the temperature distribution of the drift region; when the temperature distribution of the drift region meets the heat dissipation requirement, performing heat dissipation control optimization of the UHV silicon-based MOSFET structure according to the temperature distribution of the drift region, wherein the heat dissipation control optimization includes power adjustment optimization and heat sink control optimization, and obtaining the optimal heat dissipation control parameters; finally, using the optimal heat dissipation control parameters to control the heat dissipation of the UHV silicon-based MOSFET structure; the method can accurately obtain the temperature distribution of the drift region, improve the flexibility and accuracy of the heat dissipation control strategy setting, achieve the technical goal of efficient heat dissipation control, and ensure the stability and reliability of the UHV silicon-based MOSFET structure when working at high power.

[0034] Further, as attached Figure 2 As shown, a temperature sensor is configured on the drift region in the UHV silicon-based MOSFET structure, and during operation, the monitoring temperature information of the drift region is monitored and obtained. The application includes:

[0035] By configuring temperature sensors at a first position, a second position and a third position on a drift region within a UHV silicon-based MOSFET structure, during operation, first temperature information, second temperature information and third temperature information of the drift region are monitored and obtained, wherein the second position is a middle position of the drift region, the first position is a position of the drift region close to a source, and the third position is a position of the drift region close to a drain; the first temperature information, the second temperature information and the third temperature information are integrated to obtain monitoring temperature information.

[0036] Specifically, first, a first position, a second position, and a third position are selected in the drift region within the UHV silicon-based MOSFET structure, wherein the first position is located at the center of the half of the drift region close to the source, and is mainly responsible for monitoring the temperature information of the drift region on the source side; the second position is located in the middle of the drift region, and is responsible for monitoring the middle temperature information of the entire drift region. The second position is the core area of ​​the drift region, and the temperature distribution is usually more critical; the third position is located at the center of the half of the drift region close to the drain, and monitors the temperature changes of the drift region on the drain side. Then, a first temperature sensor, a second temperature sensor and a third temperature sensor are arranged at the first position, the second position and the third position on the drift region, and during the operation of the ultra-high voltage silicon-based MOSFET structure, the first temperature sensor, the second temperature sensor and the third temperature sensor are used to monitor and obtain the first temperature information, the second temperature information and the third temperature information of the drift region, wherein the first temperature information is obtained by the sensor arranged at the center position of the drift region close to the source half, reflecting the heat accumulation in the region; the second temperature information is obtained by the sensor arranged at the middle position of the drift region, providing the temperature data of the core position of the drift region; the third temperature information is obtained by the sensor arranged at the center position of the drift region close to the drain half, reflecting the temperature change near the drain.

[0037] By integrating the first temperature information, the second temperature information and the third temperature information, the monitoring temperature information of the entire drift zone is obtained. Since the temperature information of each sensor covers the heat changes in different parts of the drift zone, by integrating these temperature data, the heat distribution in the drift zone can be understood more accurately, thereby identifying local overheating problems or temperature imbalance areas.

[0038] Further, based on the monitored temperature information, integrated simulation generation of drift zone temperature distribution is performed, and the present application includes:

[0039] Pre-training a drift zone temperature distribution generation channel, wherein the drift zone temperature distribution generation channel includes a plurality of drift zone temperature distribution generation paths; inputting the monitored temperature information into the plurality of drift zone temperature distribution generation paths respectively, and integrating simulation generation to obtain a plurality of simulated drift zone temperature distributions.

[0040] Specifically, in the UHV silicon-based MOSFET, a drift region temperature distribution generation channel is designed and pre-trained, wherein the drift region temperature distribution generation channel includes multiple drift region temperature distribution generation paths, and the drift region temperature distribution generation path is constructed based on a generative adversarial network, including a generator and an adversary. The generator is used to generate the drift region temperature distribution according to the input temperature monitoring data, and the adversary evaluates whether the generated data is close to the real temperature distribution, ensuring that the generated result has high accuracy. These paths are based on different workloads, ambient temperatures, currents, voltages and other factors, and independently generate drift region temperature distributions, wherein each generation path simulates the temperature changes of MOSFET under specific working conditions. By integrating the simulation results of multiple paths, a more comprehensive temperature distribution prediction can be formed to ensure that the heat distribution of the drift region can be accurately simulated under various operating conditions. Then the monitoring temperature information is input into the multiple drift region temperature distribution generation paths for simulation, wherein the temperature distribution generated by each generation path only reflects the temperature changes under specific conditions, and the integrated simulation outputs multiple simulated drift region temperature distributions.

[0041] Further, the pre-training drift zone temperature distribution generation channel, the present application includes:

[0042] According to the historical operating temperature test data of the ultra-high voltage silicon-based MOSFET structure, a sample temperature monitoring information set and a sample drift region temperature distribution set are collected; part of the sample data is randomly extracted multiple times with replacement from the sample temperature monitoring information set and the sample drift region temperature distribution set to obtain multiple groups of temperature distribution generation supervision training data; based on the generative adversarial network, the generators and adversaries in the multiple drift region temperature distribution generation paths are constructed; the multiple groups of temperature distribution generation supervision training data are respectively used to perform supervised training on the generators and adversaries in the multiple drift region temperature distribution generation paths until convergence; the multiple drift region temperature distribution generation paths that have completed training are combined to obtain the drift region temperature distribution generation channel.

[0043] Specifically, first, from the historical operating temperature test data of ultra-high voltage silicon-based MOSFET, a sample temperature monitoring information set and a corresponding sample drift zone temperature distribution set are collected. These sample data reflect the temperature monitoring results of MOSFET under different working conditions, as well as the detailed temperature distribution in the drift zone. These data sets will be used for subsequent training of the Generative Adversarial Network (GAN) and are the key source of supervised training data. Then, the sample temperature monitoring information set and the sample drift zone temperature distribution set are randomly extracted with replacement for multiple times. That is, in order to increase data diversity and avoid overfitting problems, a random sampling method with replacement is used to extract multiple subsets from historical data, each of which contains different temperature monitoring data and corresponding drift zone temperature distribution. These extracted data subsets are used for supervised training of the Generative Adversarial Network to obtain multiple groups of temperature distribution generation supervised training data.

[0044] Then, based on the generative adversarial network, the generator and the adversary in the plurality of drift zone temperature distribution generation paths are constructed, wherein the generator is responsible for generating the temperature distribution of the drift zone according to the input temperature monitoring information, and the adversary judges whether the result output by the generator is close enough to the real data by learning the real temperature distribution data. The adversarial training of the generator and the adversary ensures that the generator can gradually generate results closer to the real drift zone temperature distribution. The generator is constructed by a neural network, and the model architecture includes multiple hidden layers for extracting deep features from the temperature monitoring information. The generator combines these features to generate a temperature distribution simulation result similar to the historical temperature distribution. As the training proceeds, the generator will gradually improve its ability to generate the drift zone temperature distribution, making the generated distribution closer to the real situation. The input of the generator is the temperature data monitored from the drift zone of the ultra-high voltage silicon-based MOSFET, including the real-time temperature of different positions collected by multiple temperature sensors. The generator generates the corresponding drift zone temperature distribution by learning the features of these input data. The adversary builds a classification network to compare the generated data with the real data and learn the difference between the two. The architecture of the adversary is also a deep neural network. Through continuous training, the adversary will gradually enhance its ability to distinguish between generated data and real data. The feedback of the adversary will be used to guide the generator to continuously improve its generation results. The input of the adversary is two types of data, one is the real drift zone temperature distribution data, and the other is the temperature distribution data generated by the generator. The goal of the adversary is to learn how to distinguish between these two types of data.

[0045] Then, the multiple groups of temperature distributions are used to generate supervised training data, and the generators and adversaries in the multiple drift zone temperature distribution generation paths are supervised and trained respectively. During the training process, the generator and the adversary are repeatedly antagonistic. The generator generates a temperature distribution based on the temperature monitoring data to try to deceive the adversary. The adversary helps the generator generate a more accurate simulated temperature distribution by continuously improving its ability to distinguish between real data and generated data until the training converges. When the adversarial training converges, it means that the generator can reliably generate a temperature distribution close to the real data.

[0046] When the training of the generators and adversaries in each generation path is completed, the multiple drift zone temperature distribution generation paths that have been trained are combined to obtain the drift zone temperature distribution generation channel, where each generation path is trained based on different data and conditions. By combining multiple trained paths, a comprehensive temperature distribution generation channel can be formed. This channel can generate drift zone temperature distribution under different conditions based on real-time temperature monitoring information, providing more accurate temperature simulation results.

[0047] Further, accuracy analysis is performed and the simulated drift region temperature distribution with the highest accuracy is screened to obtain the drift region temperature distribution. The present application includes:

[0048] Calculate the average temperatures of the multiple simulated drift region temperature distributions to obtain the average drift region temperature, and respectively calculate the average temperatures within the multiple simulated drift region temperature distributions to obtain multiple distribution average temperatures; respectively calculate the deviation percentages between the multiple distribution average temperatures and the average drift region temperature, and calculate and obtain multiple temperature accuracy parameters by subtracting the deviation percentage from 1; test the multiple drift region temperature distribution generation paths to obtain multiple simulation accuracy parameters, and multiply the multiple temperature accuracy parameters to obtain multiple accuracy parameters; and screen the simulated drift region temperature distribution with the largest accuracy parameter to obtain the drift region temperature distribution.

[0049] Specifically, the average temperature of the multiple simulated drift region temperature distributions is calculated to obtain the average drift region temperature, that is, the global average temperature value; on the other hand, the average temperature of each simulated drift region temperature distribution is calculated to obtain multiple distribution average temperatures. Each simulated drift region temperature distribution may have a different temperature change pattern. By calculating the independent average temperature of each distribution, the overall temperature trend of the distribution can be evaluated for subsequent deviation and accuracy calculations.

[0050] Then, the deviation percentages of the average temperatures of the multiple distributions and the average drift region temperature are calculated respectively, wherein the deviation percentage is the ratio of the difference between the average temperature of the distribution and the average drift region temperature to the average drift region temperature, and the deviation percentage reflects the difference between each simulated distribution and the average temperature of the entire drift region. The greater the deviation, the more serious the deviation of the distribution from the overall temperature. And the temperature accuracy parameter of each simulated distribution is calculated by subtracting the deviation percentage of each distribution from 1, and the temperature accuracy parameter is the difference between 1 and the deviation percentage, wherein the closer the temperature accuracy parameter is to 1, the smaller the deviation of the simulated distribution from the overall temperature, and the higher the accuracy.

[0051] On the other hand, the plurality of drift zone temperature distribution generation paths are tested, that is, the model accuracy test is performed on each drift zone temperature distribution generation path, and the accuracy of the generation path is evaluated. For example, by verifying the matching degree between the generated data and the real data, the simulation accuracy of each path is obtained, and a plurality of simulation accuracy parameters are obtained. Then, the simulation accuracy parameter of each generation path is multiplied by the corresponding temperature accuracy parameter, and the overall accuracy parameter of each simulated drift zone temperature distribution is calculated, wherein the product of the simulation accuracy parameter and the temperature accuracy parameter represents the comprehensive accuracy of each simulated distribution. The higher the overall accuracy parameter, the more reliable the model of the generation process is, and the more accuracy parameters are obtained. Finally, the accuracy parameters of all simulated drift zone temperature distributions are screened, and the simulation distribution with the largest accuracy parameter is selected as the final drift zone temperature distribution. By screening the simulation distribution with the largest accuracy parameter, the most accurate drift zone temperature distribution can be obtained, while ensuring that the selected temperature distribution can reflect the actual situation and providing a reliable basis for optimizing the heat dissipation strategy.

[0052] Further, when the temperature distribution of the drift region meets the heat dissipation requirements, the heat dissipation control optimization of the ultra-high voltage silicon-based MOSFET structure is performed according to the temperature distribution of the drift region to obtain the optimal heat dissipation control parameters. The present application includes:

[0053] The average temperature in the temperature distribution of the drift zone is calculated to determine whether the heat dissipation temperature threshold is reached. If not, heat dissipation is not performed; if so, the current power information of the ultra-high voltage silicon-based MOSFET structure is retrieved, a power adjustment space is constructed, and a heat sink control space of the heat sink of the ultra-high voltage silicon-based MOSFET structure is retrieved; a first power and a first heat sink parameter are randomly generated in the power adjustment space and the heat sink control space, and a heat dissipation simulation prediction is performed in combination with the temperature distribution of the drift zone to obtain a first heat dissipation parameter and a first loss power, and a first heat dissipation fitness is calculated, wherein the first heat dissipation parameter is the highest temperature of the drift zone after simulated heat dissipation; the power and heat sink parameters are continuously adjusted and optimized in the power adjustment space and the heat sink control space until convergence, and the optimal power and optimal heat sink parameters with the maximum heat dissipation fitness are output as the optimal heat dissipation control parameters.

[0054] Specifically, the temperature distribution data of the current drift region is calculated to obtain the average temperature of the entire drift region. Then, it is determined whether the average temperature reaches the heat dissipation temperature threshold, where each ultra-high voltage silicon-based MOSFET structure has a set heat dissipation temperature threshold. When the temperature of the drift region exceeds this value, it indicates that the device has generated a large amount of heat, which may affect performance and life, and heat dissipation control is required. If the temperature is lower than the threshold, it means that the heat dissipation demand is low and there is no need to turn on the heat dissipation system. If the average temperature is less than the heat dissipation temperature threshold, no heat dissipation is performed, that is, when the temperature does not reach the heat dissipation threshold, the MOSFET is still within the safe operating temperature range, and no additional energy is required for heat dissipation. At this time, it can continue to work, and temperature monitoring is performed to ensure that the temperature is always within a safe range.

[0055] If the average temperature is greater than or equal to the heat dissipation temperature threshold, the current power information of the UHV silicon-based MOSFET structure is retrieved, where power is closely related to temperature, and high power is usually accompanied by high heat; and based on the current power information, a power adjustment space is established to evaluate the range in which the power can be reduced. The power adjustment space refers to the power range that can be reduced without affecting the stability and performance of the system. The power adjustment space is constructed to optimize the balance between heat dissipation and performance. In some cases, reducing power can effectively reduce the temperature rise rate, but it is necessary to ensure that the power adjustment does not affect the normal operation of the system. On the other hand, the radiator control space of the UHV silicon-based MOSFET structure radiator is retrieved, where the radiator control space refers to the working parameters that the radiator can adjust, such as fan speed, liquid cooling flow, etc. These parameters determine the working efficiency of the radiator. By retrieving the radiator control space, the system can dynamically adjust the operating status of the radiator according to the current temperature and power conditions, thereby improving the heat dissipation effect.

[0056] Then, any power is randomly selected in the power adjustment space and set as the first power, and any radiator parameter is randomly selected in the radiator control space and set as the first radiator parameter. Then, based on the first power and the first radiator parameter, combined with the temperature distribution of the drift zone, a heat dissipation simulation prediction is performed, and the first heat dissipation parameter and the first power loss are obtained through the simulation prediction results, wherein the first heat dissipation parameter refers to the maximum value of the drift zone temperature after the heat dissipation simulation under the current power and radiator settings. The first power loss refers to the system power loss caused by the power adjustment during the heat dissipation process. The first heat dissipation fitness is calculated based on the first heat dissipation parameter and the first power loss. The heat dissipation fitness is a key indicator for measuring the heat dissipation effect, wherein the heat dissipation parameter, power loss and heat dissipation fitness are negatively correlated, that is, the smaller the heat dissipation parameter, the smaller the power loss and the better the heat dissipation fitness.

[0057] Then, the second power and the second heat sink parameter are randomly selected in the power adjustment space and the heat sink control space; then, the heat dissipation simulation prediction is performed based on the second power, the second heat sink parameter and the temperature distribution in the drift zone to obtain the second heat dissipation parameter and the second loss power, and the second heat dissipation fitness is calculated based on the second heat dissipation parameter and the second loss power. The same method is used to continue to adjust and optimize the power and heat sink parameters in the power adjustment space and the heat sink control space. When the adjustment of the power and heat sink parameters makes the heat dissipation fitness no longer significantly improved, it is judged that the parameter optimization process has reached convergence. For example, the judgment of the convergence condition is usually based on the change amplitude of the heat dissipation fitness. If after several iterations, the improvement amplitude of the fitness is less than the preset threshold (for example, less than 0.01%), it is considered that the system has reached the optimal state and no longer continues to optimize. At this time, the optimal power and optimal heat sink parameters with the largest heat dissipation fitness are output as the optimal heat dissipation control parameters.

[0058] By optimizing parameters in the power adjustment space and radiator control space, continuously adjusting the power and radiator parameters, and calculating the heat dissipation fitness, the optimal solution is gradually found. When the heat dissipation fitness reaches a convergence state, the optimal power and optimal radiator parameters with the maximum heat dissipation fitness are output as the final optimal heat dissipation control parameters, which can not only ensure the heat dissipation effect, but also ensure the lowest possible power loss, thereby optimizing the overall performance and heat dissipation efficiency.

[0059] Further, heat dissipation simulation prediction is performed in combination with the temperature distribution of the drift region to obtain a first heat dissipation parameter and a first power loss, and a first heat dissipation fitness is calculated. The present application includes:

[0060] According to the historical data of the heat dissipation test of the ultra-high voltage silicon-based MOSFET structure, a sample drift region temperature distribution set, a sample power set, a sample radiator parameter set, and a maximum temperature of the drift region after a preset heat dissipation time are collected to obtain a sample heat dissipation parameter set; the sample drift region temperature distribution set, the sample power set, and the sample radiator parameter set are used as input supervised training data, and the sample heat dissipation parameter set is used as output supervised training data to train a heat dissipation simulator until convergence; the drift region temperature distribution, the first power, and the first heat dissipation parameter are input into the heat dissipation simulator for heat dissipation simulation to obtain a first heat dissipation parameter; the difference between the first power and the current power information of the ultra-high voltage silicon-based MOSFET structure is calculated to obtain a first loss power; combined with the first heat dissipation parameter and the first loss power, a first heat dissipation fitness is calculated and obtained as follows: ; Among them, HDF is the heat dissipation fitness, and is the weight, is the heat dissipation parameter, is the power loss, This is the current power information of the UHV silicon-based MOSFET structure.

[0061] Specifically, based on the historical data of heat dissipation tests on ultra-high voltage silicon-based MOSFET structures, a sample drift zone temperature distribution set, a sample power set, and a sample radiator parameter set are collected. On the other hand, the maximum temperature of the drift zone after a preset heat dissipation time is collected and set as the sample heat dissipation parameter. For example, the highest temperature in the drift zone after 1 hour of heat dissipation is collected as the heat dissipation parameter to obtain a sample heat dissipation parameter set.

[0062] On the other hand, a heat dissipation simulator is constructed based on machine learning, for example, a heat dissipation simulator is constructed based on a BP neural network, wherein the heat dissipation simulator is a neural network model that can be iteratively optimized in machine learning, including an input layer, multiple hidden layers, and an output layer, wherein the input data of the input layer are the temperature distribution, power, and heat sink parameters of the drift zone, and the output data of the output layer are the heat dissipation parameters. The sample drift zone temperature distribution set, the sample power set, and the sample heat sink parameter set are further used as input supervised training data, and the sample heat dissipation parameter set is used as output supervised training data to perform supervised training on the heat dissipation simulator, and the gradient of the loss function to the network weight is calculated by the back propagation algorithm, and the weight and bias are adjusted using the gradient descent algorithm to gradually reduce the value of the loss function until the loss function of the model converges to a smaller value, or the training process reaches the set maximum number of iterations, and a heat dissipation simulator that meets the expected conditions is obtained.

[0063] Then, the temperature distribution of the drift region, the first power and the first heat sink parameter are input into the heat dissipation simulator for heat dissipation simulation, and the first heat dissipation parameter is output; on the other hand, the difference between the first power and the current power information of the ultra-high voltage silicon-based MOSFET structure is calculated to obtain the first loss power, which is the power difference of the current power minus the first power.

[0064] Construct a heat dissipation fitness calculation function: ; Among them, HDF is the heat dissipation fitness. The greater the heat dissipation fitness, the better the heat dissipation effect. The smaller the heat dissipation parameter, the smaller the loss power, and the greater the heat dissipation fitness; is the heat dissipation parameter weight, is the loss power weight, where the heat dissipation parameter weight and the loss power weight can be set according to the degree of influence of the indicator on the heat dissipation fitness. The greater the influence of the indicator, the greater the corresponding weight; is the heat dissipation parameter, is the power loss, is the current power information of the UHV silicon-based MOSFET structure. Then, based on the heat dissipation fitness calculation function, the first heat dissipation parameter and the first loss power are calculated to obtain the first heat dissipation fitness.

[0065] In summary, the heat dissipation control method of a UHV silicon-based MOSFET structure provided in this application has the following technical effects:

[0066] The invention discloses a method for controlling the temperature distribution of the drift region of a UHV silicon-based MOSFET structure. The method comprises the following steps: configuring a plurality of temperature sensors on the drift region within the UHV silicon-based MOSFET structure, and using the plurality of temperature sensors to monitor and obtain the monitoring temperature information of the drift region during operation; then, based on the monitoring temperature information, performing integrated simulation generation of the temperature distribution of the drift region, obtaining a plurality of simulated temperature distributions of the drift region, and performing accuracy analysis on the plurality of simulated temperature distributions of the drift region, and then screening the simulated temperature distribution of the drift region with the greatest accuracy and setting it as the temperature distribution of the drift region; when the temperature distribution of the drift region meets the heat dissipation requirement, performing heat dissipation control optimization of the UHV silicon-based MOSFET structure according to the temperature distribution of the drift region, wherein the heat dissipation control optimization includes power adjustment optimization and heat sink control optimization, and obtaining the optimal heat dissipation control parameters; finally, using the optimal heat dissipation control parameters to control the heat dissipation of the UHV silicon-based MOSFET structure; the method can accurately obtain the temperature distribution of the drift region, improve the flexibility and accuracy of the heat dissipation control strategy setting, achieve the technical goal of efficient heat dissipation control, and ensure the stability and reliability of the UHV silicon-based MOSFET structure when working at high power.

[0067] Embodiment 2: Based on the heat dissipation control method of a UHV silicon-based MOSFET structure in the above embodiment, the present application also provides a heat dissipation control device for a UHV silicon-based MOSFET structure, see the attached Figure 3 ,include:

[0068] A temperature information monitoring module 11 is used to monitor and obtain the monitoring temperature information of the drift region during operation through a temperature sensor configured on the drift region within the UHV silicon-based MOSFET structure, wherein a plurality of temperature sensors are configured on the drift region; a temperature distribution acquisition module 12 is used to perform integrated simulation generation of the drift region temperature distribution according to the monitoring temperature information, obtain a plurality of simulated drift region temperature distributions, and perform accuracy analysis and screen the simulated drift region temperature distribution with the highest accuracy to obtain the drift region temperature distribution; a heat dissipation control optimization module 13 is used to perform heat dissipation control optimization of the UHV silicon-based MOSFET structure according to the drift region temperature distribution when the drift region temperature distribution meets the heat dissipation requirements, and obtain optimal heat dissipation control parameters, wherein the heat dissipation control optimization includes power adjustment optimization and radiator control optimization; a heat dissipation control module 14 is used to use the optimal heat dissipation control parameters to perform heat dissipation control on the UHV silicon-based MOSFET structure.

[0069] Furthermore, the heat dissipation control device of the UHV silicon-based MOSFET structure is also used for:

[0070] By configuring temperature sensors at a first position, a second position and a third position on a drift region within a UHV silicon-based MOSFET structure, during operation, first temperature information, second temperature information and third temperature information of the drift region are monitored and obtained, wherein the second position is a middle position of the drift region, the first position is a position of the drift region close to a source, and the third position is a position of the drift region close to a drain; the first temperature information, the second temperature information and the third temperature information are integrated to obtain monitoring temperature information.

[0071] Furthermore, the heat dissipation control device of the UHV silicon-based MOSFET structure is also used for:

[0072] Pre-training a drift zone temperature distribution generation channel, wherein the drift zone temperature distribution generation channel includes a plurality of drift zone temperature distribution generation paths; inputting the monitored temperature information into the plurality of drift zone temperature distribution generation paths respectively, and integrating simulation generation to obtain a plurality of simulated drift zone temperature distributions.

[0073] Furthermore, the heat dissipation control device of the UHV silicon-based MOSFET structure is also used for:

[0074] According to the historical operating temperature test data of the ultra-high voltage silicon-based MOSFET structure, a sample temperature monitoring information set and a sample drift region temperature distribution set are collected; part of the sample data is randomly extracted multiple times with replacement from the sample temperature monitoring information set and the sample drift region temperature distribution set to obtain multiple groups of temperature distribution generation supervision training data; based on the generative adversarial network, the generators and adversaries in the multiple drift region temperature distribution generation paths are constructed; the multiple groups of temperature distribution generation supervision training data are respectively used to perform supervised training on the generators and adversaries in the multiple drift region temperature distribution generation paths until convergence; the multiple drift region temperature distribution generation paths that have completed training are combined to obtain the drift region temperature distribution generation channel.

[0075] Furthermore, the heat dissipation control device of the UHV silicon-based MOSFET structure is also used for:

[0076] Calculate the average temperatures of the multiple simulated drift region temperature distributions to obtain the average drift region temperature, and respectively calculate the average temperatures within the multiple simulated drift region temperature distributions to obtain multiple distribution average temperatures; respectively calculate the deviation percentages between the multiple distribution average temperatures and the average drift region temperature, and calculate and obtain multiple temperature accuracy parameters by subtracting the deviation percentage from 1; test the multiple drift region temperature distribution generation paths to obtain multiple simulation accuracy parameters, and multiply the multiple temperature accuracy parameters to obtain multiple accuracy parameters; and screen the simulated drift region temperature distribution with the largest accuracy parameter to obtain the drift region temperature distribution.

[0077] Furthermore, the heat dissipation control device of the ultra-high voltage silicon-based MOSFET structure is also used to: calculate the average temperature in the temperature distribution of the drift zone, determine whether the heat dissipation temperature threshold is reached, and if not, no heat dissipation is performed; if so, retrieve the current power information of the ultra-high voltage silicon-based MOSFET structure, construct a power adjustment space, and retrieve the heat sink control space of the heat sink of the ultra-high voltage silicon-based MOSFET structure; randomly generate a first power and a first heat sink parameter in the power adjustment space and the heat sink control space, perform heat dissipation simulation prediction in combination with the drift zone temperature distribution, obtain a first heat dissipation parameter and a first loss power, and calculate to obtain a first heat dissipation fitness, wherein the first heat dissipation parameter is the highest temperature of the drift zone after simulated heat dissipation; continue to adjust and optimize the power and heat sink parameters in the power adjustment space and the heat sink control space until convergence, and output the optimal power and optimal heat sink parameters with the maximum heat dissipation fitness as the optimal heat dissipation control parameters.

[0078] Furthermore, the heat dissipation control device of the UHV silicon-based MOSFET structure is also used for:

[0079] According to the historical data of the heat dissipation test of the ultra-high voltage silicon-based MOSFET structure, a sample drift region temperature distribution set, a sample power set, a sample radiator parameter set, and a maximum temperature of the drift region after a preset heat dissipation time are collected to obtain a sample heat dissipation parameter set; the sample drift region temperature distribution set, the sample power set, and the sample radiator parameter set are used as input supervised training data, and the sample heat dissipation parameter set is used as output supervised training data to train a heat dissipation simulator until convergence; the drift region temperature distribution, the first power, and the first heat dissipation parameter are input into the heat dissipation simulator for heat dissipation simulation to obtain a first heat dissipation parameter; the difference between the first power and the current power information of the ultra-high voltage silicon-based MOSFET structure is calculated to obtain a first loss power; combined with the first heat dissipation parameter and the first loss power, a first heat dissipation fitness is calculated and obtained as follows: ; Among them, HDF is the heat dissipation fitness, and is the weight, is the heat dissipation parameter, is the power loss, This is the current power information of the UHV silicon-based MOSFET structure.

[0080] Each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The heat dissipation control method and specific examples of a UHV silicon-based MOSFET structure in the aforementioned embodiment 1 are also applicable to a heat dissipation control device of a UHV silicon-based MOSFET structure in this embodiment. Through the aforementioned detailed description of the heat dissipation control method of a UHV silicon-based MOSFET structure, those skilled in the art can clearly know the heat dissipation control device of a UHV silicon-based MOSFET structure in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0081] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0082] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. A heat dissipation control method for a UHV silicon-based MOSFET structure, characterized in that: Methods include: By configuring a temperature sensor on a drift region in a UHV silicon-based MOSFET structure, during operation, monitoring temperature information of the drift region is obtained, wherein a plurality of temperature sensors are configured on the drift region; According to the monitored temperature information, integrated simulation generation of drift region temperature distribution is performed to obtain multiple simulated drift region temperature distributions, and accuracy analysis is performed and the simulated drift region temperature distribution with the greatest accuracy is screened to obtain the drift region temperature distribution; When the temperature distribution of the drift region meets the heat dissipation requirement, heat dissipation control optimization of the UHV silicon-based MOSFET structure is performed according to the temperature distribution of the drift region to obtain optimal heat dissipation control parameters, wherein the heat dissipation control optimization includes power adjustment optimization and heat sink control optimization; Using the optimal heat dissipation control parameters to control the heat dissipation of the ultra-high voltage silicon-based MOSFET structure; When the temperature distribution of the drift region meets the heat dissipation requirement, heat dissipation control optimization of the ultra-high voltage silicon-based MOSFET structure is performed according to the temperature distribution of the drift region to obtain optimal heat dissipation control parameters, including: Calculating the average temperature in the temperature distribution of the drift region to determine whether the heat dissipation temperature threshold is reached, and if not, not performing heat dissipation; If yes, then retrieve the current power information of the UHV silicon-based MOSFET structure, construct a power adjustment space, and retrieve the heat sink control space of the heat sink of the UHV silicon-based MOSFET structure; Randomly generate a first power and a first heat sink parameter in the power adjustment space and the heat sink control space, perform heat dissipation simulation prediction in combination with the temperature distribution of the drift zone, obtain a first heat dissipation parameter and a first power loss, and calculate to obtain a first heat dissipation fitness, wherein the first heat dissipation parameter is the highest temperature of the drift zone after simulated heat dissipation; Continue to adjust and optimize the power and radiator parameters in the power adjustment space and radiator control space until convergence, and output the optimal power and optimal radiator parameters with the maximum heat dissipation adaptability as the optimal heat dissipation control parameters.

2. The heat dissipation control method of the UHV silicon-based MOSFET structure according to claim 1, characterized in that: A temperature sensor is configured on the drift region within the UHV silicon-based MOSFET structure. During operation, the temperature information of the drift region is monitored and obtained, including: By configuring temperature sensors at a first position, a second position, and a third position on a drift region in a UHV silicon-based MOSFET structure, first temperature information, second temperature information, and third temperature information of the drift region are monitored and obtained during operation, wherein the second position is a middle position of the drift region, the first position is a position of the drift region close to a source, and the third position is a position of the drift region close to a drain; The first temperature information, the second temperature information and the third temperature information are integrated to obtain monitoring temperature information.

3. The heat dissipation control method of the UHV silicon-based MOSFET structure according to claim 1, characterized in that: Based on the monitored temperature information, integrated simulation generation of drift zone temperature distribution is performed, including: Pre-training a drift region temperature distribution generation channel, wherein the drift region temperature distribution generation channel includes a plurality of drift region temperature distribution generation paths; The monitored temperature information is input into the plurality of drift region temperature distribution generation paths respectively, and a plurality of simulated drift region temperature distributions are obtained by integrated simulation generation.

4. The heat dissipation control method of the UHV silicon-based MOSFET structure according to claim 3, characterized in that: Pre-trained drift zone temperature distribution generation channel, including: Based on the historical operating temperature test data of the UHV silicon-based MOSFET structure, a sample temperature monitoring information set is collected, and a sample drift region temperature distribution set is collected; Randomly extracting part of the sample data with replacement from the sample temperature monitoring information set and the sample drift zone temperature distribution set multiple times to obtain multiple groups of temperature distributions to generate supervised training data; Based on a generative adversarial network, construct generators and adversaries within the plurality of drift zone temperature distribution generation paths; The plurality of groups of temperature distributions are respectively used to generate supervised training data, and the generators and adversaries in the plurality of drift region temperature distribution generation paths are respectively trained in a supervised manner until convergence; The multiple drift zone temperature distribution generation paths that have been trained are combined to obtain the drift zone temperature distribution generation channel.

5. The heat dissipation control method of the UHV silicon-based MOSFET structure according to claim 4, characterized in that: Perform accuracy analysis and screen the most accurate simulated drift zone temperature distribution to obtain the drift zone temperature distribution, including: Calculating the average temperature of the plurality of simulated drift region temperature distributions to obtain an average drift region temperature, and respectively calculating the average temperature within the plurality of simulated drift region temperature distributions to obtain a plurality of distribution average temperatures; Calculating the deviation percentages of the plurality of distributed average temperatures and the average drift region temperature respectively, and subtracting the deviation percentages from 1 to obtain a plurality of temperature accuracy parameters; Testing the plurality of drift region temperature distribution generation paths to obtain a plurality of simulation accuracy parameters, and multiplying the plurality of temperature accuracy parameters to obtain a plurality of accuracy parameters; The simulated drift zone temperature distribution with the largest accuracy parameter is screened to obtain the drift zone temperature distribution.

6. The heat dissipation control method of the UHV silicon-based MOSFET structure according to claim 1, characterized in that: Performing heat dissipation simulation prediction in combination with the temperature distribution of the drift region to obtain a first heat dissipation parameter and a first power loss, and calculating to obtain a first heat dissipation fitness includes: According to the historical data of heat dissipation test of UHV silicon-based MOSFET structure, a sample drift region temperature distribution set, a sample power set, a sample heat sink parameter set, and the maximum temperature of the drift region after the preset heat dissipation time are collected to obtain a sample heat dissipation parameter set; Using the sample drift region temperature distribution set, the sample power set, and the sample radiator parameter set as input supervised training data, and using the sample heat dissipation parameter set as output supervised training data, to train the heat dissipation simulator until convergence; Inputting the drift region temperature distribution, the first power and the first heat sink parameter into the heat dissipation simulator to perform heat dissipation simulation to obtain the first heat dissipation parameter; Calculate the difference between the first power and the current power information of the UHV silicon-based MOSFET structure to obtain a first loss power; The first heat dissipation fitness is calculated by combining the first heat dissipation parameter and the first power loss, as shown in the following formula: ； Among them, HDF is the heat dissipation fitness, and is the weight, T is the heat dissipation parameter, U is the loss power, This is the current power information of the UHV silicon-based MOSFET structure.

7. A heat dissipation control device for a UHV silicon-based MOSFET structure, characterized in that: The steps for implementing the heat dissipation control method of the UHV silicon-based MOSFET structure according to any one of claims 1 to 6 include: A temperature information monitoring module is used to monitor and obtain monitoring temperature information of the drift region during operation through a temperature sensor configured on the drift region in the UHV silicon-based MOSFET structure, wherein a plurality of temperature sensors are configured on the drift region; A temperature distribution acquisition module is used to perform integrated simulation generation of drift zone temperature distribution according to the monitored temperature information, obtain multiple simulated drift zone temperature distributions, and perform accuracy analysis and screen the simulated drift zone temperature distribution with the greatest accuracy to obtain the drift zone temperature distribution; A heat dissipation control optimization module, used to perform heat dissipation control optimization of the UHV silicon-based MOSFET structure according to the temperature distribution of the drift region when the temperature distribution of the drift region meets the heat dissipation requirements, and obtain optimal heat dissipation control parameters, wherein the heat dissipation control optimization includes power adjustment optimization and radiator control optimization; The heat dissipation control module is used to adopt the optimal heat dissipation control parameters to control the heat dissipation of the ultra-high voltage silicon-based MOSFET structure.

Images