Optimization Method and System for Structural Parameters of Grooved Schottky Diode

Abstract

The present invention discloses a method and system for optimizing the structural parameters of a trench Schottky diode, which relates to the field of data processing. The method includes: obtaining a set of application scenario features of a target trench Schottky diode; performing key structural parameter matching and historical application record mining; traversing the set of historical application records to calculate the loss amount; retrieving the set of historical application records to obtain a set of historical key structural parameter groups; using the set of historical key structural parameter groups as the initial optimization space, and performing competitive expansion in combination with the set of historical loss amounts and a preset expansion scale to obtain the target optimization space; performing a focused search for key structural parameters to determine the target key structural parameter group as the result of structural parameter optimization. It solves the technical problems of low optimization efficiency and lack of accuracy in the optimization of the existing structural parameters of trench Schottky diodes, and achieves the technical effects of improving the optimization efficiency and ensuring the accuracy and reliability of the optimization results.

Description

Technical Field

[0001] This application relates to the field of data processing, and in particular, to a method and system for optimizing the structural parameters of a trench Schottky diode. Background Art

[0002] In the field of semiconductor devices, as a high-performance power semiconductor device, the trench Schottky diode is widely used in high-frequency, high-voltage, and high-power power electronic systems. The optimization of its performance is crucial for improving the efficiency, stability, and reliability of the entire power electronic system. The performance of the trench Schottky diode is affected by various structural parameters, such as trench depth, trench width, and the contact characteristics between the anode metal and the semiconductor material. In the existing methods for optimizing the structural parameters of the trench Schottky diode, it is usually based on the experience and intuition of engineers, and the optimal solution is found by repeatedly testing and adjusting the parameters. This method is time-consuming and laborious, and the optimization results often depend on the personal experience and knowledge level of engineers, lacking systematicness and accuracy.

[0003] In the current related technologies, there are technical problems such as low optimization efficiency and lack of accuracy in the optimization of the structural parameters of the trench Schottky diode. Summary of the Invention

[0004] This application provides a method and system for optimizing the structural parameters of a trench Schottky diode. By collecting and analyzing the application scenario characteristics of the target trench Schottky diode, based on the application scenario characteristics, matching the key structural parameters and mining the historical application records, obtaining the historical key structural parameter groups and historical loss amount sets similar to the current application scenario, using the loss objective function to calculate the loss amount of the historical application records, and constructing the initial optimization space with the key structural parameter groups as the index, combining the historical loss amount set and the preset expansion scale to competitively expand the initial optimization space to obtain the target optimization space, performing an aggregation search for the key structural parameters in the target optimization space, and finding the target key structural parameter group through iterative optimization as the final optimization result and other technical means, achieving the technical effects of improving the optimization efficiency and ensuring the accuracy and reliability of the optimization results.

[0005] The present application provides a method for optimizing the structural parameters of a trench Schottky diode, including: obtaining a set of application scenario characteristics of a target trench Schottky diode; performing key structural parameter matching and historical application record mining based on the set of application scenario characteristics to obtain a set of key structural parameter groups and a set of historical application records; using a loss objective function to calculate loss amounts by traversing the set of historical application records to obtain a set of historical loss amounts; retrieving the set of historical application records with the set of key structural parameter groups as an index to obtain a set of historical key structural parameter groups; using the set of historical key structural parameter groups as an initial optimization space, and combining the set of historical loss amounts and a preset expansion scale to competitively expand the initial optimization space to obtain a target optimization space; performing a key structural parameter aggregation search on the target optimization space to determine a target set of key structural parameters, and using the target set of key structural parameters as the result of structural parameter optimization.

[0006] In a possible implementation, the following processing is performed: constructing a loss objective function, where the loss objective function is: ; where is the loss amount, is the on-resistance, is the reverse breakdown voltage, is the switching loss, , is the weight parameter of the on-resistance in calculating the loss amount, is the weight parameter of the reverse breakdown voltage in calculating the loss amount, It is a weight parameter for the switching loss in calculating the loss amount. In a possible implementation, taking the set of historical key structure parameter groups as the initial optimization space, combining the set of historical loss amounts and a preset expansion scale to perform competitive expansion on the initial optimization space to obtain a target optimization space, the following processing is executed: Extract the maximum value of the historical loss amounts in the set of historical loss amounts, take the historical key structure parameter group corresponding to the maximum value of the historical loss amounts as the competitive expansion avoidance direction, extract the minimum value of the historical loss amounts in the set of historical loss amounts, and take the historical key structure parameter group corresponding to the minimum value of the historical loss amounts as the competitive expansion movement direction; Based on the competitive expansion avoidance direction and the competitive expansion movement direction, adjust at least one key structure parameter of the set of historical key structure parameter groups in the initial optimization space according to the preset expansion scale to obtain a stage competitive expansion optimization space; Use a competitive expansion evaluation function to evaluate the stage competitive expansion optimization space to obtain a stage competitive expansion evaluation result; Determine whether the stage competitive expansion evaluation result meets the preset requirements. If so, take the stage competitive expansion optimization space as the target optimization space; If not, perform competitive expansion on the set of historical key structure parameter groups in the initial optimization space again and update the stage competitive expansion optimization space according to the expansion result.

[0007] In a possible implementation, using a competitive expansion evaluation function to evaluate the stage competitive expansion optimization space to obtain a stage competitive expansion evaluation result, the following processing is executed: Obtain the competitive expansion evaluation function, where the competitive expansion evaluation function is: ; where is the stage competitive expansion evaluation result, is the stage competitive expansion optimization space, is the number of historical key structure parameter groups in the stage competitive expansion optimization space, is a neighborhood composed of historical key structure parameter groups whose distance from the i-th historical key structure parameter group in the stage competitive expansion optimization space is within a preset distance threshold, is the number of historical key structure parameter groups included in the neighborhood of the i-th historical key structure parameter group in the stage competitive expansion optimization space, is the i-th historical key structure parameter group in the stage competitive expansion optimization space, is the j-th historical key structure parameter group within the neighborhood of the i-th historical key structure parameter group. In a possible implementation, for the target optimization space, a key structure parameter aggregation search is performed to determine the target key structure parameter group, and the following processing is executed: combining the loss objective function, traversing the target optimization space to predict the loss amount, obtaining a set of target optimization loss amounts; extracting the minimum value of the target optimization loss amounts in the set of target optimization loss amounts, and using it as the initial aggregation search target; according to a preset iteration step size, iterating the initial aggregation search target in the set of target optimization loss amounts to obtain an aggregation search target; taking the key structure parameter group corresponding to the aggregation search target as the target key structure parameter group.

[0008] In a possible implementation, the following processing is executed: obtaining the number of loss amounts in the set of target optimization loss amounts whose distance to the initial aggregation search target is less than or equal to the preset iteration step size, to obtain an initial aggregation number; iterating the initial aggregation search target according to the preset iteration step size to obtain an iterative aggregation search target, and obtaining the number of loss amounts in the set of target optimization loss amounts whose distance to the iterative aggregation search target is less than or equal to the preset iteration step size, to obtain an iterative aggregation number; determining whether the initial aggregation number is less than or equal to the iterative aggregation number, and if so, iterating based on the iterative aggregation search target until a preset iteration number is satisfied, and taking the iterative aggregation search target obtained in the last iteration as the aggregation search target. In a possible implementation, the following processing is executed: the key structure parameter group includes the trench oxide layer thickness, the width of the active region between trenches, and the contact angle between the trench oxide layer and the active region.

[0009] This application also provides a trench Schottky diode structure parameter optimization system, including: an application scenario feature set acquisition module, configured to acquire an application scenario feature set of a target trench Schottky diode; a matching and mining module, configured to perform key structure parameter matching and historical application record mining based on the application scenario feature set to obtain a key structure parameter group and a set of historical application records; a loss amount calculation module, configured to use a loss objective function to traverse the set of historical application records to calculate the loss amount and obtain a set of historical loss amounts; a retrieval module, configured to retrieve the set of historical application records with the key structure parameter group as an index to obtain a set of historical key structure parameter groups; a competition and expansion module, configured to use the set of historical key structure parameter groups as an initial optimization space, and combine the set of historical loss amounts and a preset expansion scale to perform competition and expansion on the initial optimization space to obtain a target optimization space; a structure parameter optimization result acquisition module, configured to perform a key structure parameter aggregation search on the target optimization space to determine a target key structure parameter group, and take the target key structure parameter group as the structure parameter optimization result.

[0010] The method and system for optimizing the structural parameters of a trench Schottky diode proposed in this application first obtain the application scenario feature set of the target trench Schottky diode, then perform key structural parameter matching and historical application record mining based on the application scenario feature set to obtain a key structural parameter group and a historical application record set. Then, using the loss objective function, traverse the historical application record set to calculate the loss amount and obtain a historical loss amount set. Then, using the key structural parameter group as an index, retrieve the historical application record set to obtain a historical key structural parameter group set. Furthermore, use the historical key structural parameter group set as the initial optimization space, combine the historical loss amount set and a preset expansion scale to competitively expand the initial optimization space to obtain the target optimization space. Finally, perform a key structural parameter aggregation search on the target optimization space to determine the target key structural parameter group, and use the target key structural parameter group as the result of the structural parameter optimization, achieving the technical effects of improving the optimization efficiency and ensuring the accuracy and reliability of the optimization result. Description of the Drawings

[0011] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings of the embodiments of the present invention will be briefly introduced below. Flowcharts are used in this application to illustrate the operations performed by the systems according to the embodiments of this application. It should be understood that the operations in the front or below do not necessarily need to be executed precisely in sequence. Instead, according to needs, they can be executed in reverse order or simultaneously. At the same time, other operations can also be added to these processes, or one or several operations can be removed from these processes.

[0012] Figure 1 It is a schematic flowchart of the method for optimizing the structural parameters of a trench Schottky diode provided by an embodiment of this application.

[0013] Figure 2 It is a schematic structural diagram of the system for optimizing the structural parameters of a trench Schottky diode provided by an embodiment of this application.

[0014] Description of the reference numerals: Application scenario feature set acquisition module 10, matching and mining module 20, loss amount calculation module 30, retrieval module 40, competitive expansion module 50, structural parameter optimization result acquisition module 60. Detailed Embodiments

[0015] The above description is only an overview of the technical solutions of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features, and advantages of this application more obvious and understandable, the following specifically lists the detailed embodiments of this application.

[0016] To make the objectives, technical solutions, and advantages of this application clearer, the following will further describe this application in detail with reference to the accompanying drawings. The described embodiments should not be construed as limitations on this application. All other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of this application.

[0017] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments. However, it can be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments and can be combined with each other without conflict. The terms "comprising" and "having", and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field of this application. The terms used herein are only for the purpose of describing the embodiments of this application.

[0018] The embodiments of this application provide a method for optimizing the structural parameters of a trench Schottky diode, as Figure 1 shown, the method includes:

[0019] Step S100, obtaining a set of application scenario characteristics of the target trench Schottky diode.

[0020] Specifically, through data collection and preprocessing techniques, such as sensor data collection, historical database query, etc., detailed information about the application scenario of the target trench Schottky diode is obtained, that is, the application scenario characteristics (a series of parameters or conditions that affect the performance and usage conditions of the trench Schottky diode) are determined, such as the operating temperature range, current density, voltage level, package form, etc. Then, the specific values of these characteristics are collected and obtained to form a set of application scenario characteristics.

[0021] Step S200, based on the set of application scenario characteristics, performing key structural parameter matching and historical application record mining to obtain a set of key structural parameters and a set of historical application records.

[0022] Specifically, according to the application scenario characteristics, data mining and machine learning techniques, such as association rule mining, clustering analysis, etc., are used to match key structural parameters (structural characteristics that have an important impact on the performance of the trench Schottky diode, such as the anode trench depth, cathode material type, etc.) from the historical database to form a set of key structural parameters. At the same time, relevant historical application records are mined, that is, the records of using trench Schottky diodes in the same or similar application scenarios in the past are found, including key structural parameters, performance performance, etc.

[0023] In a possible implementation, step S200 further includes step S210. The key structure parameter group includes the trench oxide layer thickness, the width of the active region between trenches, and the contact angle between the trench oxide layer and the active region.

[0024] Specifically, in this example, the key structure parameter group includes the trench oxide layer thickness, the width of the active region between trenches, and the contact angle between the trench oxide layer and the active region. Among them, the trench oxide layer thickness refers to the thickness of the oxide layer inside the trench, which affects the breakdown voltage and leakage current of the diode. The width of the active region between trenches refers to the width of the active region between adjacent trenches, which affects the current capacity and frequency characteristics of the diode. The contact angle between the trench oxide layer and the active region refers to the angle of the contact interface between the trench oxide layer and the active region, which affects the electric field distribution and current flow path of the diode. The selection of the key structure parameters in this implementation is based on a large amount of data mining and professional knowledge. Therefore, it can more accurately reflect the performance requirements of the diode, making the performance of the optimized diode more excellent and improving the pertinence and efficiency of the optimization.

[0025] Step S300: Using the loss objective function, traverse the historical application record set to calculate the loss amount and obtain the historical loss amount set.

[0026] Specifically, the loss objective function is a function used to quantify the gap between the structure parameters and the ideal performance. This function measures the gap between the diode performance and the target performance under the given structure parameters. For each record in the historical application record set, use the loss objective function to calculate its loss amount to form the historical loss amount set. Among them, the loss amount can be calculated based on factors such as the deviation of performance indicators and cost overrun, and is a numerical value measuring the performance deviation.

[0027] In a possible implementation, step S300 further includes step S310: constructing the loss objective function, where the loss objective function is: ; where is the loss amount, is the on-resistance, is the reverse breakdown voltage, is the switching loss, , is the weight parameter of the on-resistance when calculating the loss amount, is the weight parameter of the reverse breakdown voltage when calculating the loss amount, is the weight parameter of the switching loss when calculating the loss amount.

[0028] Specifically, the main performance indicators of the trench Schottky diode include on-resistance, reverse breakdown voltage, and switching loss. These indicators directly reflect the working efficiency, voltage withstand capacity, and dynamic performance of the diode. The loss amount is a comprehensive indicator used to measure the gap between the diode performance and the target performance. The loss amount is the weighted sum of the on-resistance, reverse breakdown voltage, and switching loss, and the weight parameters are determined through statistical analysis of historical data, which are used to measure the relative importance of different performance indicators in the total loss amount. Among them, the on-resistance is the resistance from the anode to the cathode when the diode is forward-biased, which determines the power consumption and efficiency of the diode in the conducting state. The reverse breakdown voltage is the maximum voltage that the diode can withstand without breakdown when it is reverse-biased, reflecting the voltage withstand capacity of the diode. The switching loss is the energy consumed by the diode during the switching process, which affects the dynamic performance and efficiency of the diode. This implementation method provides a comprehensive performance evaluation criterion for the optimization of the structural parameters of the trench Schottky diode by constructing a loss objective function, providing a basis for subsequent optimization steps. This implementation method not only improves the accuracy and efficiency of optimization but also enhances the flexibility and adaptability of the optimization results.

[0029] Step S400: Using the set of key structural parameters as an index, retrieve the historical application record set to obtain a set of historical key structural parameter groups.

[0030] Specifically, use database retrieval techniques such as index query and similarity matching. According to the set of key structural parameters, retrieve the records with similar structural parameters in the historical application record set to form a set of historical key structural parameter groups.

[0031] Step S500: Using the set of historical key structural parameter groups as the initial optimization space, combine the set of historical loss amounts and a preset expansion scale to competitively expand the initial optimization space to obtain a target optimization space.

[0032] Specifically, starting from the set of historical key structural parameter groups, according to the magnitudes of the loss amounts in the set of historical loss amounts and the preset expansion scale (parameters controlling the expansion speed and range of the optimization space, such as search step size, mutation rate, etc.), use an optimization algorithm to expand the initial optimization space, and expand the optimization space by introducing a competition mechanism (such as natural selection, fitness evaluation, etc.) to form a target optimization space containing more potential optimal solutions.

[0033] In a possible implementation, the set of historical key structure parameter groups is used as the initial optimization space, and the initial optimization space is competitively expanded in combination with the set of historical loss amounts and a preset expansion scale to obtain a target optimization space. Step S500 further includes step S510 of extracting the maximum historical loss amount from the set of historical loss amounts, using the historical key structure parameter group corresponding to the maximum historical loss amount as the competitive expansion avoidance direction, extracting the minimum historical loss amount from the set of historical loss amounts, and using the historical key structure parameter group corresponding to the minimum historical loss amount as the competitive expansion movement direction. Specifically, find the record with the largest loss amount in the set of historical loss amounts, and the corresponding historical key structure parameter group is regarded as the direction to be avoided during competitive expansion (i.e., the "avoidance direction"). Similarly, find the record with the smallest loss amount in the set of historical loss amounts, and the corresponding historical key structure parameter group is regarded as the direction that should be approached during the optimization process (i.e., the "movement direction"). Among them, the maximum historical loss amount is the parameter configuration that caused the largest loss in historical applications. The minimum historical loss amount is the parameter configuration that caused the smallest loss in historical applications. The competitive expansion avoidance direction is the parameter configuration direction that should be avoided during the optimization process. The competitive expansion movement direction is the parameter configuration direction that should be pursued during the optimization process.

[0034] Step S520, based on the competitive expansion avoidance direction and the competitive expansion movement direction, adjust at least one key structure parameter of the set of historical key structure parameter groups in the initial optimization space according to the preset expansion scale to obtain a stage competitive expansion optimization space. Specifically, based on the determined competitive expansion avoidance direction and movement direction, and the preset expansion scale (such as the range or step size of parameter adjustment), adjust each historical key structure parameter group in the initial optimization space. For each parameter group, adjust at least one key structure parameter (such as trench depth, width, doping concentration, etc.) to form a new parameter configuration. The new set of parameter configurations formed after parameter adjustment is the stage competitive expansion optimization space.

[0035] Step S530, use the competitive expansion evaluation function to evaluate the stage competitive expansion optimization space to obtain a stage competitive expansion evaluation result. Specifically, use the competitive expansion evaluation function (a function for evaluating the quality of parameter configurations) to evaluate each parameter configuration in the stage competitive expansion optimization space. For each new parameter configuration, calculate its evaluation score to obtain the stage competitive expansion evaluation result. Step S540, determine whether the stage competitive expansion evaluation result meets the preset requirements. If so, use the stage competitive expansion optimization space as the target optimization space. Specifically, compare whether the stage competitive expansion evaluation result meets the preset requirements. If the requirements are met, end the competitive expansion process and use the stage competitive expansion optimization space as the target optimization space. If the requirements are not met, continue to execute the next step.

[0036] Step S550, if not, then compete to expand the set of historical key structure parameter groups in the initial optimization space again, and update the stage competition expansion optimization space according to the expansion result. Specifically, according to the result that does not meet the preset requirements, adjust the expansion strategy (such as changing the expansion scale, etc.), and compete to expand the initial optimization space again, optimize the expansion process, and update the stage competition expansion optimization space. This implementation method, through the method of competitive expansion, while maintaining the direction of the historical optimal parameter configuration, avoids the influence of the worst parameter configuration, thereby effectively searching for a better configuration in the parameter space. Through continuous iteration and adjustment, the global optimal solution can be gradually approximated, improving the performance of the trench Schottky diode.

[0037] In a possible implementation, the stage competition expansion optimization space is evaluated using a competition expansion evaluation function to obtain a stage competition expansion evaluation result. Step S530 further includes step S531 of obtaining the competition expansion evaluation function, where the competition expansion evaluation function is: ; where is the stage competition expansion evaluation result, is the stage competition expansion optimization space, is the number of historical key structure parameter groups in the stage competition expansion optimization space, is a neighborhood composed of historical key structure parameter groups whose distance from the initial point to the i-th historical key structure parameter group in the stage competition expansion optimization space is within a preset distance threshold, is the number of historical key structure parameter groups included in the neighborhood of the i-th historical key structure parameter group in the stage competition expansion optimization space, is the i-th historical key structure parameter group in the stage competition expansion optimization space, is the j-th historical key structure parameter group within the neighborhood of the i-th historical key structure parameter group. Specifically, the competitive expansion evaluation function is used to evaluate the overall performance of the stage competitive expansion optimization space. This function comprehensively considers the relationships among various historical key structure parameter groups (i.e., candidate solutions) within the optimization space, especially their similarities and differences. The competitive expansion evaluation function takes a parameter - the stage competitive expansion optimization space - as input and outputs a scalar value as the evaluation result. For each historical key structure parameter group within the optimization space, its neighborhood is calculated. The neighborhood consists of all other historical key structure parameter groups that are within a preset distance threshold from each historical key structure parameter group. This distance can be the Euclidean distance, Manhattan distance, etc. For each historical key structure parameter group, the number of historical key structure parameter groups contained in its neighborhood is counted. Finally, according to the competitive expansion evaluation function formula, the evaluation result of the entire optimization space is calculated and output. This formula comprehensively considers the information of each historical key structure parameter group and its neighborhood to evaluate the overall performance of the optimization space. This implementation method quantifies and evaluates the overall performance of the optimization space by defining the competitive expansion evaluation function. This function considers the similarities and differences between parameter configurations and helps to identify those sets of parameter configurations with better distribution and diversity. During the optimization process, it is desired to find a set of parameter configurations that can maintain a certain degree of diversity (to avoid premature convergence to local optimal solutions) and can also exhibit good performance. By continuously adjusting and optimizing this set, the global optimal solution can be gradually approximated, improving the accuracy of the optimization and enhancing the robustness and stability of the optimization.

[0038] Step S600, perform a key structure parameter aggregation search on the target optimization space to determine the target key structure parameter group, and use the target key structure parameter group as the structure parameter optimization result.

[0039] Specifically, algorithms such as clustering analysis and local search are used to perform an aggregation search on the key structural parameter groups in the target optimization space, that is, to find parameter groups with similar performance or characteristics. According to the performance evaluation index (such as the minimum loss), the target key structural parameter group is determined, that is, the optimal structural parameter combination determined after optimization, and it is used as the final result of the structural parameter optimization. In the embodiment of the present application, the application scenario characteristics of the target trench Schottky diode are collected and analyzed. Based on the application scenario characteristics, the matching of key structural parameters and the mining of historical application records are carried out to obtain the historical key structural parameter group and the historical loss amount set similar to the current application scenario. The loss objective function is used to calculate the loss amount of the historical application records, and an initial optimization space is constructed with the key structural parameter group as the index. Combining the historical loss amount set and the preset expansion scale, the initial optimization space is competitively expanded to obtain the target optimization space. An aggregation search for key structural parameters is carried out within the target optimization space, and the target key structural parameter group is found through iterative optimization as the final optimization result and other technical means, achieving the technical effects of improving the optimization efficiency and ensuring the accuracy and reliability of the optimization result.

[0040] In a possible implementation manner, when performing an aggregation search for key structural parameters in the target optimization space to determine the target key structural parameter group, step S600 further includes step S610 of traversing the target optimization space for loss prediction in combination with the loss objective function to obtain a target optimization loss amount set. Specifically, the loss objective function is used to traverse each key structural parameter group in the target optimization space, substitute each parameter group into the loss objective function, and calculate the corresponding loss amount. All the calculated loss amounts are combined into a set, that is, the target optimization loss amount set.

[0041] Step S620 is to extract the minimum value of the target optimization loss amount in the target optimization loss amount set and use it as the initial aggregation search target. Specifically, the minimum loss amount value is found in the target optimization loss amount set, and the key structural parameter group corresponding to the minimum loss amount value is used as the initial aggregation search target. This parameter group represents the optimal configuration in the current optimization space.

[0042] Step S630 is to iterate on the initial aggregation search target in the target optimization loss amount set according to a preset iteration step size to obtain an aggregation search target. Specifically, the amplitude or range of parameter adjustment, as well as the search direction and method, are determined in each iteration. Starting from the initial aggregation search target, the parameter configuration is gradually adjusted according to the iteration step size and search strategy, and the loss amount value corresponding to each new configuration is calculated. After each iteration, the loss amount values of the new configuration and the current aggregation search target are compared. If the new configuration is better, the aggregation search target is updated to the new configuration.

[0043] Step S640: Take the key structure parameter group corresponding to the aggregated search target as the target key structure parameter group. Specifically, when the iteration stop condition is met (such as reaching the maximum number of iterations, the loss no longer decreases significantly, etc.), the iterative search process ends. Take the key structure parameter group corresponding to the aggregated search target obtained at the end of the iterative search as the final optimization result, and this parameter group represents the optimal solution found currently. This implementation method selects the minimum value as the initial aggregated search target, that is, starts the search from a configuration with better performance, so that a better configuration can be found more quickly. At the same time, by iterative search to gradually adjust the parameter configuration and update the aggregated search target, it is ensured that the finally obtained parameter group is the optimal or approximately optimal solution under the given conditions, improving the accuracy of optimization.

[0044] In a possible implementation, step S630 further includes step S631: Obtain the number of loss amounts in the target optimization loss amount set whose distance to the initial aggregated search target is less than or equal to the preset iteration step size, to obtain the initial aggregation number. Specifically, use the Euclidean distance to calculate the distance between each loss amount in the target optimization loss amount set and the initial aggregated search target. Traverse the target optimization loss amount set, calculate the distance between each loss amount and the initial aggregated search target. Count the number of loss amounts whose distance is less than or equal to the preset iteration step size, and this number is the initial aggregation number. It indicates how many loss amounts are relatively close around the initial aggregated search target.

[0045] Step S632: Iterate the initial aggregated search target according to the preset iteration step size to obtain an iterative aggregated search target, and obtain the number of loss amounts in the target optimization loss amount set whose distance to the iterative aggregated search target is less than or equal to the preset iteration step size, to obtain the iterative aggregation number. Specifically, based on the preset iteration step size and the random search strategy, iteratively adjust the initial aggregated search target to obtain a new iterative aggregated search target. Similar to step S631, calculate the distance between each loss amount in the target optimization loss amount set and the new iterative aggregated search target. Count the number of loss amounts whose distance is less than or equal to the preset iteration step size to obtain the iterative aggregation number. The iterative aggregated search target is a new parameter configuration obtained through iterative search, and it represents a candidate optimal solution in the current search process.

[0046] Step S633: Determine whether the initial aggregation quantity is less than or equal to the iterative aggregation quantity. If so, perform iteration based on the iterative aggregation search target until the preset number of iterations is satisfied, and use the iterative aggregation search target obtained in the last iteration as the aggregation search target. Specifically, compare the initial aggregation quantity and the iterative aggregation quantity to decide whether to continue the iterative search. If the iterative aggregation quantity is greater than or equal to the initial aggregation quantity, it indicates that a better parameter configuration area has been found through iterative search, and then perform the next iteration based on the current iterative aggregation search target. If the number of iterations reaches the preset value or the iterative aggregation quantity no longer increases significantly, stop the iteration. When the iteration stops, use the iterative aggregation search target obtained in the last iteration as the final aggregation search target. This implementation method guides the iterative search process by comparing the initial aggregation quantity and the iterative aggregation quantity, encourages continued search in this area in order to find a better solution. This strategy not only ensures the efficiency of the search but also increases the possibility of finding the global optimal solution.

[0047] In the above text, reference is made to Figure 1 a detailed description of the method for optimizing the structural parameters of a trench Schottky diode according to an embodiment of the present invention. Next, reference will be made to Figure 2 describe a system for optimizing the structural parameters of a trench Schottky diode according to an embodiment of the present invention.

[0048] The system for optimizing the structural parameters of a trench Schottky diode according to an embodiment of the present invention is used to solve the technical problems of low optimization efficiency and lack of accuracy in the optimization of the structural parameters of existing trench Schottky diodes, and achieve the technical effects of improving the optimization efficiency and ensuring the accuracy and reliability of the optimization results. The system for optimizing the structural parameters of a trench Schottky diode includes: an application scenario feature set acquisition module 10, a matching and mining module 20, a loss quantity calculation module 30, a retrieval module 40, a competitive expansion module 50, and a structural parameter optimization result acquisition module 60.

[0049] An application scenario feature set acquisition module 10 is configured to acquire an application scenario feature set of a target trench Schottky diode; a matching and mining module 20 is configured to perform key structure parameter matching and historical application record mining based on the application scenario feature set to obtain a key structure parameter group and a historical application record set; a loss amount calculation module 30 is configured to use a loss objective function to traverse the historical application record set to calculate a loss amount and obtain a historical loss amount set; a retrieval module 40 is configured to retrieve the historical application record set with the key structure parameter group as an index to obtain a historical key structure parameter group set; a competition and expansion module 50 is configured to use the historical key structure parameter group set as an initial optimization space, and combine the historical loss amount set and a preset expansion scale to perform competition and expansion on the initial optimization space to obtain a target optimization space; a structure parameter optimization result acquisition module 60 is configured to perform key structure parameter aggregation search on the target optimization space to determine a target key structure parameter group, and use the target key structure parameter group as a structure parameter optimization result.

[0050] Next, the specific configuration of the loss amount calculation module 30 will be described in detail. As described above, the loss amount calculation module 30 may further include: a loss objective function construction unit configured to construct a loss objective function, where the loss objective function is: ; where is the loss amount, is the on-resistance, is the reverse breakdown voltage, is the switching loss, , is the weight parameter of the on-resistance when calculating the loss amount, is the weight parameter of the reverse breakdown voltage when calculating the loss amount, is the weight parameter of the switching loss in calculating the loss amount. Next, the specific configuration of the competition expansion module 50 will be described in detail. As described above, taking the set of historical key structure parameters as the initial optimization space, combining the set of historical loss amounts and a preset expansion scale to perform competition expansion on the initial optimization space to obtain the target optimization space, the competition expansion module 50 may further include: A competition expansion direction setting unit is used to extract the maximum historical loss amount in the set of historical loss amounts, take the set of historical key structure parameters corresponding to the maximum historical loss amount as the competition expansion avoidance direction, extract the minimum historical loss amount in the set of historical loss amounts, and take the set of historical key structure parameters corresponding to the minimum historical loss amount as the competition expansion movement direction; A key structure parameter adjustment unit is used to adjust at least one key structure parameter of the set of historical key structure parameters in the initial optimization space according to the preset expansion scale based on the competition expansion avoidance direction and the competition expansion movement direction to obtain a stage competition expansion optimization space; An evaluation unit is used to evaluate the stage competition expansion optimization space by using a competition expansion evaluation function to obtain a stage competition expansion evaluation result; A target optimization space acquisition unit is used to determine whether the stage competition expansion evaluation result meets the preset requirements. If so, take the stage competition expansion optimization space as the target optimization space; A stage competition expansion optimization space update unit is used to, if not, perform competition expansion on the set of historical key structure parameters in the initial optimization space again and update the stage competition expansion optimization space according to the expansion result.

[0051] Among them, when evaluating the stage competition expansion optimization space by using a competition expansion evaluation function to obtain a stage competition expansion evaluation result, the evaluation unit may further include: A competition expansion evaluation function acquisition subunit is used to obtain the competition expansion evaluation function, where the competition expansion evaluation function is: ; where is the stage competition expansion evaluation result, is the stage competition expansion optimization space, is the number of sets of historical key structure parameters in the stage competition expansion optimization space, is a neighborhood composed of sets of historical key structure parameters whose distances from the i-th set of historical key structure parameters in the stage competition expansion optimization space are within a preset distance threshold, is the number of sets of historical key structure parameters included in the neighborhood of the i-th set of historical key structure parameters in the stage competition expansion optimization space, is the i-th set of historical key structure parameters in the stage competition expansion optimization space, is the j-th historical key structure parameter group within the neighborhood of the i-th historical key structure parameter group. Next, the specific configuration of the structure parameter optimization result acquisition module 60 will be described in detail. As described above, a key structure parameter aggregation search is performed on the target optimization space to determine the target key structure parameter group. The structure parameter optimization result acquisition module 60 may further include: a loss amount prediction unit for traversing the target optimization space to predict the loss amount in combination with the loss objective function, and obtaining a set of target optimization loss amounts; a maximum value extraction unit for extracting the minimum value of the target optimization loss amounts in the set of target optimization loss amounts and using it as the initial aggregation search target; an iteration unit for iterating the initial aggregation search target in the set of target optimization loss amounts according to a preset iteration step size to obtain an aggregation search target; and a target key structure parameter group acquisition unit for using the key structure parameter group corresponding to the aggregation search target as the target key structure parameter group.

[0052] Among them, the iteration unit may further include: an initial aggregation number acquisition subunit for obtaining the number of loss amounts in the set of target optimization loss amounts whose distance to the initial aggregation search target is less than or equal to the preset iteration step size, and obtaining an initial aggregation number; an iterative aggregation number acquisition subunit for iterating the initial aggregation search target according to the preset iteration step size to obtain an iterative aggregation search target, and obtaining the number of loss amounts in the set of target optimization loss amounts whose distance to the iterative aggregation search target is less than or equal to the preset iteration step size, and obtaining an iterative aggregation number; and an aggregation search target acquisition subunit for determining whether the initial aggregation number is less than or equal to the iterative aggregation number. If so, iterating based on the iterative aggregation search target until a preset number of iterations is satisfied, and using the iterative aggregation search target obtained in the last iteration as the aggregation search target.

[0053] Next, the specific configuration of the matching and mining module 20 will be described in detail. As described above, the matching and mining module 20 may further include: a key structure parameter group construction unit for constructing a key structure parameter group, which includes the trench oxide layer thickness, the width of the active region between trenches, and the contact angle between the trench oxide layer and the active region.

[0054] The trench Schottky diode structure parameter optimization system provided by the embodiments of the present invention can execute the trench Schottky diode structure parameter optimization method provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the method.

[0055] Although various references are made in this application to certain modules in the system according to embodiments of the present application, however, any number of different modules can be used and run on the user terminal and / or the server. The various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the various functional units are only for the convenience of mutual distinction and do not limit the protection scope of the present invention. The above specific implementation manners do not constitute a limitation on the protection scope of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of this application shall be included within the protection scope of this application. In some cases, the actions or steps recited in this application can be performed in a different order than in the embodiments and still achieve the desired results. Additionally, the processes depicted in the drawings do not necessarily require the particular order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A method for optimizing the structural parameters of a trench Schottky diode, characterized in that: The method comprises: Obtain an application scenario feature set of a target trench Schottky diode; Based on the application scenario feature set, key structural parameter matching and historical application record mining are performed to obtain a key structural parameter group and a historical application record set; Using the loss objective function, traverse the historical application record set to calculate the loss amount, and obtain a historical loss amount set; Using the key structure parameter group as an index, searching the historical application record set to obtain a historical key structure parameter group set; Taking the historical key structural parameter group set as the initial optimization space, and combining the historical loss amount set and the preset expansion scale to competitively expand the initial optimization space, to obtain the target optimization space; Performing a clustered search of key structural parameters on the target optimization space to determine a target key structural parameter group, and using the target key structural parameter group as a structural parameter optimization result; The historical key structural parameter set is used as the initial optimization space, and the historical loss set and the preset expansion scale are combined to competitively expand the initial optimization space to obtain the target optimization space, including: Extracting the maximum value of the historical loss amount in the historical loss amount set, and using the historical key structural parameter group corresponding to the maximum value of the historical loss amount as the competition expansion avoidance direction; extracting the minimum value of the historical loss amount in the historical loss amount set, and using the historical key structural parameter group corresponding to the minimum value of the historical loss amount as the competition expansion movement direction; Based on the competition expansion avoidance direction and the competition expansion movement direction, adjusting at least one key structural parameter of the historical key structural parameter group set in the initial optimization space according to the preset expansion scale to obtain a stage competition expansion optimization space; Using a competition expansion evaluation function to evaluate the stage competition expansion optimization space, and obtaining a stage competition expansion evaluation result; Determine whether the stage competition expansion evaluation result meets the preset requirements, and if so, use the stage competition expansion optimization space as the target optimization space; If not, the historical key structural parameter group set in the initial optimization space is competitively expanded again, and the stage competitive expanded optimization space is updated according to the expansion result.

2. The method for optimizing the structural parameters of a trench Schottky diode according to claim 1, wherein: include: Construct a loss objective function, wherein the loss objective function is: ;in, is the loss amount, is the on-resistance, is the reverse breakdown voltage, Switching losses, , is the weight parameter of on-resistance in calculating the loss, is the weight parameter of reverse breakdown voltage in calculating the loss, It is the weight parameter of switching loss in calculating the loss amount.

3. The method for optimizing the structural parameters of a trench Schottky diode according to claim 1, wherein: The competition expansion optimization space of the stage is evaluated by using the competition expansion evaluation function to obtain the stage competition expansion evaluation result, including: Obtain the competition expansion evaluation function, wherein the competition expansion evaluation function is: ;in, Expand evaluation results for stage competition, Expand optimization space for stage competition, The number of historical key structural parameter groups in the optimization space for stage competition expansion, It is the expansion of the optimization space from stage competition to the The neighborhood of historical key structural parameter groups whose distances to each historical key structural parameter group are within the preset distance threshold. Expand and optimize space for stage competition The number of historical key structural parameter groups contained in the neighborhood of the historical key structural parameter group, Expand and optimize space for stage competition Historical key structural parameter groups, For the The first A historical key structural parameter group.

4. The method for optimizing the structural parameters of a trench Schottky diode according to claim 1, wherein: Performing a clustered search of key structural parameters on the target optimization space to determine a target key structural parameter group includes: In combination with the loss objective function, the target optimization space is traversed to predict the loss amount, and a target optimization loss amount set is obtained; Extracting the minimum value of the target optimization loss amount in the target optimization loss amount set, and using it as the initial clustering search target; Iterating the initial clustered search target in the target optimization loss amount set according to a preset iteration step to obtain a clustered search target; The key structural parameter group corresponding to the clustered search target is used as the target key structural parameter group.

5. The method for optimizing the structural parameters of a trench Schottky diode according to claim 4, wherein: include: Obtain the number of loss amounts in the target optimization loss amount set whose distance to the initial clustering search target is less than or equal to the preset iteration step length, and obtain the initial clustering amount; Iterate the initial clustered search target according to the preset iteration step to obtain an iterative clustered search target, and obtain the number of loss amounts in the target optimization loss amount set whose distance to the iterative clustered search target is less than or equal to the preset iteration step to obtain the iterative clustering number; Determine whether the initial clustering number is less than or equal to the iterative clustering number. If so, iterate based on the iterative clustering search target until a preset number of iterations is met, and use the iterative clustering search target obtained in the last iteration as the clustering search target.

6. The method for optimizing the structural parameters of a trench Schottky diode according to claim 1, wherein: The key structural parameter group includes the thickness of the trench oxide layer, the width of the active area between the trenches and the contact angle between the trench oxide layer and the active area.

7. Trench Schottky diode structural parameter optimization system, characterized in that: The system is used to implement the method for optimizing the structural parameters of a trench Schottky diode according to any one of claims 1 to 6, and the system comprises: An application scenario feature set acquisition module, used to acquire an application scenario feature set of a target trench Schottky diode; A matching and mining module, used to perform key structural parameter matching and historical application record mining based on the application scenario feature set to obtain a key structural parameter group and a historical application record set; A loss amount calculation module, used to use a loss objective function to traverse the historical application record set to calculate the loss amount and obtain a historical loss amount set; A retrieval module, used to retrieve the historical application record set by taking the key structure parameter group as an index to obtain a historical key structure parameter group set; A competitive expansion module, used to use the historical key structural parameter group set as the initial optimization space, and competitively expand the initial optimization space in combination with the historical loss amount set and a preset expansion scale to obtain a target optimization space; The structural parameter optimization result acquisition module is used to perform a clustered search of key structural parameters on the target optimization space, determine a target key structural parameter group, and use the target key structural parameter group as a structural parameter optimization result.

Images