VCSEL (Vertical Cavity Surface Emitting Laser) chip wet oxidation aperture optimization method adopting dynamic adjustment

Through the fitting observation and dynamic adjustment technology of the wet oxidation process of VCSEL chip, the problems of inaccurate control and lack of dynamic adjustment in traditional methods are solved, and the uniformity of oxidation pore size and chip performance are improved.

CN119091154AActive Publication Date: 2024-12-06JIANGSU ETERN

Patent Information

Application Number
CN202411524911.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-30
Publication Date
2024-12-06
Estimated Expiration
2044-10-30

AI Technical Summary

Technical Problem

The control of the wet oxidation process of VCSEL chips is inaccurate, and the lack of detailed division and targeted optimization of different stages of the oxidation process leads to uneven oxidation pore sizes, affecting chip performance and consistency, and lacking a dynamic adjustment mechanism, resulting in stability and reliability of the production process. Insufficient sex.

Method used

By fitting and observing the oxidation process, dividing key stages and setting optimization priorities for control parameters, using monitoring sensors to collect data in real time, making decision generation and parameter linkage prediction based on the global control layer, dynamically adjusting control parameters to ensure that the oxidation process is at its best Perform in an optimal state.

Benefits of technology

It realizes precise control of the wet oxidation process of VCSEL chips, improves the stability and reliability of the production process, ensures uniformity of the oxidation pore size, and improves chip performance and consistency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119091154A_ABST
    Figure CN119091154A_ABST
Patent Text Reader

Abstract

The invention discloses a VCSEL (Vertical Cavity Surface Emitting Laser) chip wet oxidation aperture optimization method adopting dynamic adjustment, and relates to the technical field of semiconductor manufacturing, and the method comprises the following steps: carrying out fitting observation on an oxidation process, dividing key stages, and setting optimization priorities and characteristics of control parameters. And collecting temperature, humidity, aperture change and oxidation time data of the VCSEL chip in the wet oxidation process, and generating a global decision. And the global decision is transmitted to the parameter association control layer for analysis, a global target is established, the global decision is optimized, and the local control layer is called to execute the optimized decision. The technical problems that in the prior art, control over the wet oxidation process of the VCSEL chip is not accurate, and the linkage relation between parameters is not sufficiently considered are solved, and the technical effect of accurate control over the wet oxidation process of the VCSEL chip is achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of semiconductor manufacturing technology, and in particular to a method for optimizing the aperture of a VCSEL chip wet oxidation using dynamic adjustment. Background Art

[0002] In the VCSEL chip manufacturing scenario, the wet oxidation process is crucial to chip performance, and there are many problems in this process. The control of the traditional VCSEL chip wet oxidation process is not precise enough, lacking a detailed division and targeted optimization of the different stages of the oxidation process. Fixed parameter control is often used, which is difficult to adapt to the complex oxidation environment and the requirements of different stages. It may lead to uneven oxidation apertures and affect chip performance and consistency. At the same time, the lack of effective real-time monitoring means and dynamic adjustment mechanisms makes it difficult to obtain key parameters in real time, and it is impossible to adjust control parameters in time, resulting in insufficient stability and reliability of the production process. In addition, the linkage relationship between the various control parameters in the oxidation process is insufficiently considered. The traditional method adjusts the parameters in isolation, and does not fully consider the synergy between the parameters, which limits the optimization space of the VCSEL chip wet oxidation process.

[0003] In the current relevant technologies, there are technical problems such as inaccurate control of the wet oxidation process of VCSEL chips, insufficient consideration of the linkage between parameters, and lack of dynamic adjustment. Summary of the invention

[0004] This application provides a method for optimizing the aperture of wet oxidation of VCSEL chips using dynamic adjustment, adopts fitting observation of the oxidation process, divides the key stages and sets the optimization priority and characteristics of the control parameters for each stage. Use monitoring sensors to collect the temperature, humidity, aperture change and oxidation time data of the VCSEL chip in real time during the wet oxidation process. Through the global control layer, the key stages are matched based on the stage characteristics and real-time data sets, the optimization priority is activated and the global decision is generated. The global decision is passed to the parameter association control layer for analysis, the global goal is established, the global decision is optimized through parameter linkage prediction and synchronous adjustment strategy, and the local control layer is called to execute the optimized decision, so as to achieve the technical effect of accurately controlling the wet oxidation process of VCSEL chips by dynamically adjusting the control parameters according to the characteristics of different stages, ensuring that the oxidation process is always carried out in the optimal state.

[0005] The present application provides a method for optimizing the aperture of a VCSEL chip wet oxidation using dynamic adjustment, comprising: The oxidation process is fitted and observed, and the oxidation process is divided into stages according to the fitting observation results to establish key stage division results, wherein each key stage division result is set with an optimization priority of a control parameter, and the key stage division result is configured with a stage division feature; monitoring sensors are arranged, and the wet oxidation process of the VCSEL chip is monitored by the monitoring sensors to establish a real-time data set, and the real-time data set includes oxidation environment temperature data, humidity data, aperture change data, and oxidation time data; a global control layer is configured, and key stages are matched with the stage division feature and the real-time data set, and the optimization priority of the corresponding key stage division result is activated with the matching result, and the global control layer is initialized with the activated optimization priority, and the real-time data set is processed by the initialized global control layer to generate a global decision; after the global decision is sent to the parameter association control layer, the global decision is analyzed by the parameter association control layer to establish a global goal; the linkage relationship between parameters is predicted with the global goal, and a synchronous adjustment strategy is configured based on the linkage relationship prediction result, and after the global decision is optimized based on the synchronous adjustment strategy, the local control layer is called to execute the optimized global decision.

[0006] The VCSEL chip wet oxidation aperture optimization method using dynamic adjustment proposed in this application is proposed. First, the oxidation process is fitted and observed, the key stages are divided, and the optimization priority and characteristics of the control parameters are set for each stage. The temperature, humidity, aperture change and oxidation time data of the VCSEL chip during the wet oxidation process are collected in real time using monitoring sensors. Through the global control layer, the key stages are matched based on the stage characteristics and real-time data sets, the optimization priority is activated and the global decision is generated. The global decision is passed to the parameter association control layer for analysis, the global goal is established, the global decision is optimized through parameter linkage prediction and synchronous adjustment strategy, and the local control layer is called to execute the optimized decision, so as to achieve the technical effect of accurately controlling the wet oxidation process of the VCSEL chip by dynamically adjusting the control parameters according to the characteristics of different stages to ensure that the oxidation process is always carried out in the optimal state. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0008] Figure 1A schematic diagram of a flow chart of a method for optimizing the aperture of a VCSEL chip wet oxidation using dynamic adjustment provided in an embodiment of the present application; Figure 2 A schematic flow chart of the key phase division results of establishing a method for optimizing the aperture of a VCSEL chip wet oxidation using dynamic adjustment provided in an embodiment of the present application. DETAILED DESCRIPTION

[0009] 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 in accordance with the contents of the specification. 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 listed below.

[0010] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0011] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0012] The present application embodiment provides a method for optimizing the aperture of a VCSEL chip wet oxidation by dynamically adjusting the aperture of the VCSEL chip. Figure 1 As shown, the method includes: Step S100, fitting observation is performed on the oxidation process, and the oxidation process is divided into stages according to the fitting observation results, and a key stage division result is established, wherein each key stage division result is provided with an optimization priority of the control parameters, and the key stage division result is configured with a stage division feature. Specifically, first, fitting observation is performed on the wet oxidation process of the VCSEL chip, and data such as oxidation time, temperature change, humidity change, and pore size change are collected. Mathematical models and data analysis techniques are used to simulate the changing trends of various parameters over time. A large amount of relevant data in the wet oxidation process of the VCSEL chip, including oxidation time, temperature change, humidity change, and pore size change, are collected. The curve fitting method is used to simulate the changing trends of various parameters in the oxidation process over time. By continuously adjusting the parameters of the fitting model, the fitting results are made as close as possible to the actual observed data. Then, the oxidation process is divided into stages according to the fitting observation results, and the key stage division results are determined by comprehensively considering the changes in multiple parameters. Then, the optimization priority of the control parameters is set for each key stage, because different control parameters at different stages have different degrees of influence on the oxidation process. Finally, stage division features are configured for each key stage division result. These features can be specific parameter value ranges, parameter change trends, time intervals, etc., so as to quickly and accurately identify the current stage of the oxidation process and provide a clear basis for subsequent oxidation process control.

[0013] In one possible implementation, Figure 2 As shown, the oxidation process is fitted and observed, and the oxidation process is divided into stages according to the fitting observation results, and a key stage division result is established, wherein each key stage division result is set with an optimization priority of the control parameter, and the key stage division result is configured with a stage division feature, and step S100 further includes step S110, performing time series fitting of the control parameters according to the fitting observation results, and establishing a time series fitting curve. Specifically, after fitting and observing the wet oxidation process of the VCSEL chip, time series fitting is further performed on the control parameters. First, the control parameters that need to be fitted are determined, and the parameters may include temperature, humidity, aperture change, and oxidation time, etc. Then, polynomial fitting is used to fit the observation data of different parameters changing with time. By continuously adjusting the parameters of the fitting model, the fitting curve is made as close as possible to the actual observation data, and a time series fitting curve of the control parameters is established. The curve can intuitively show the change trend of each control parameter in the time dimension.

[0014] Step S120, extracting characteristic points from the time series fitting curve, and establishing clustering results in the time dimension according to the characteristic point extraction results. Specifically, after obtaining the time series fitting curve, extracting characteristic points from it, the characteristic points can be turning points, extreme points or points with obvious change trends in the curve, for example, rapid rising points or falling points in the temperature fitting curve, mutation points in the aperture change curve, etc. can all be used as characteristic points, by extracting characteristic points, the oxidation process can be divided in the time dimension, according to the distribution of characteristic points and time sequence, the time periods with similar characteristics are divided into one category, so as to establish clustering results in the time dimension, and the clustering results can preliminarily reflect the characteristics and change laws of the oxidation process in different time periods.

[0015] Step S130, using the fitting observation results to perform multidimensional clustering of control parameters, and establishing multidimensional clustering results, the features extracted by multidimensional clustering include temperature change rate, humidity volatility, pore size change rate, pore size stability, and oxidation time cumulative effect. Specifically, in addition to the analysis in the time dimension, the relationship between multiple control parameters needs to be considered. Using the previous fitting observation results, the control parameters are multidimensionally clustered. In the multidimensional clustering process, the extracted features include temperature change rate, humidity volatility, pore size change rate, pore size stability, and oxidation time cumulative effect. For the temperature change rate, it can be determined by calculating the ratio of the temperature difference between adjacent time points to the time interval; humidity volatility can be measured by the standard deviation of humidity; pore size change rate and stability can be similarly calculated by the change amount and change trend of pore size; the oxidation time cumulative effect can consider the influence of oxidation time on other parameters. According to the characteristics, clustering algorithms such as K-Means clustering, hierarchical clustering, etc. are used to divide different oxidation states into different categories, thereby establishing multidimensional clustering results, which can more comprehensively reflect the relationship and change pattern between various parameters in the oxidation process.

[0016] Step S140, clustering and merging the clustering results under the time dimension and the multidimensional clustering results to establish the key stage division results. Specifically, clustering and merging the clustering results under the time dimension and the multidimensional clustering results, first, analyzing the characteristics and similarities of each category in the clustering results under the time dimension and the multidimensional clustering results, if some categories under the time dimension have similar parameter characteristics and change trends with the categories in the multidimensional clustering results, they can be merged, for example, a certain fast oxidation time period under the time dimension matches the category with high temperature change rate and fast aperture change rate in the multidimensional clustering results, they can be merged into a key stage, and the key stage division results are established through continuous comparison and merging, and the results integrate the information of the time dimension and multiple control parameters, and can more accurately divide the key stages of the oxidation process, providing a more targeted basis for subsequent optimization control.

[0017] In a possible implementation, clustering fusion is performed on the clustering results under the time dimension and the multi-dimensional clustering results to establish the key stage division results, and step S140 further includes step S141, establishing a timing window according to the clustering results under the time dimension. Specifically, when analyzing the wet oxidation process of the VCSEL chip, the timing window is first established according to the clustering results under the time dimension. Since the clustering results under the time dimension have preliminarily divided the oxidation process in time, each cluster represents a time period with specific characteristics. Through the clustering results, some key time nodes can be determined, and a timing window can be established with the time nodes as the boundary. For example, if the clustering results under the time dimension show that the temperature changes more drastically within a certain time period, then a timing window can be established with the start and end time of this time period as the boundary, which is specifically used to analyze the oxidation process characteristics within this time period.

[0018] Step S142, using the timing window to match the trend consistency of the multidimensional clustering results in the timing window. Specifically, after the timing window is established, the multidimensional clustering results in the window are then matched for trend consistency. The multidimensional clustering results are a classification of the oxidation process from the perspective of multiple control parameters, including characteristics such as temperature change rate, humidity volatility, pore size change rate, pore size stability, and oxidation time cumulative effect. In the timing window, analyze whether the characteristic change trend of the multidimensional clustering results is consistent with the trend represented by the clustering results under the time dimension. For example, if the clustering results under the time dimension show that a certain time period is a rapid oxidation stage, and the multidimensional clustering results in the timing window also show characteristics such as a fast temperature change rate and a high pore size change rate, then it can be considered that the trends of the two are consistent.

[0019] Step S143, if the characteristic change trend of the multidimensional clustering result in the time series window is consistent with the clustering result under the time dimension, the stage division is retained. Specifically, if the characteristic change trend of the multidimensional clustering result in the time series window is consistent with the clustering result under the time dimension, it means that the division of this time period is reasonable and this stage division can be retained, which means that in this time period, the characteristics of the oxidation process can be described by the time dimension and the multidimensional clustering result, which provides a clear basis for stage division for subsequent optimization control. For example, if the multidimensional clustering result in a time series window shows that both the temperature and pore size changes meet the characteristics of rapid oxidation, which is consistent with the time period being divided into the rapid oxidation stage under the time dimension, then it can be determined that this time period is indeed a rapid oxidation stage, and this stage division is retained.

[0020] Step S144, if the characteristic change trend of the multidimensional clustering result in the time series window is inconsistent with the clustering result under the time dimension, the time series window is split and the matching iteration is continued. Specifically, if the characteristic change trend of the multidimensional clustering result in the time series window is inconsistent with the clustering result under the time dimension, it means that there may be a problem with the division of this time period, and the time series window needs to be further adjusted. The time series window can be split into smaller time periods to more finely analyze the characteristics of the oxidation process, and then the iterative process of trend consistency matching is continued until a time division consistent with the multidimensional clustering result is found. For example, if the multidimensional clustering result in a time series window shows that the characteristics of some time points are inconsistent with the clustering result under the time dimension, then the window can be split into two or more smaller windows, and trend consistency matching is performed separately, and the window size and position are continuously adjusted until the most suitable division is found.

[0021] Step S145, cluster fusion is completed according to the matching iteration results and the retained stage division results. Specifically, finally, cluster fusion is completed according to the matching iteration results and the retained stage division results. By continuously adjusting the timing window and matching the trend consistency, a set of stage divisions that are consistent with the clustering results under the time dimension and the multi-dimensional clustering results can be obtained. By fusing the stage divisions, a more accurate and comprehensive key stage division result can be obtained. The result integrates the information of the time dimension and multiple control parameters, and can provide more powerful support for the optimization control of the wet oxidation process of the VCSEL chip. For example, after multiple iterations and adjustments, several key oxidation stages are determined, each stage has a clear time range and feature description, and the stages are fused to obtain a complete key stage division result for guiding the subsequent oxidation process control.

[0022] In a possible implementation, if the feature change trend of the multidimensional clustering result in the time series window is inconsistent with the clustering result under the time dimension, the time series window is split, and the matching iteration is continued. Step S144 further includes step S1441, evaluating the feature change density in the window of the time series window, as follows: ;in, Characterize each time point The eigenvector The corresponding local probability density is, is the total number of multidimensional clustering results, Characterize the bandwidth function, which is determined by the temperature ,humidity , Aperture Adaptive adjustment, Characterize any multidimensional clustering result, is the weight coefficient, representing the The influence of multidimensional clustering results on density estimation, is the kernel function, For time point The eigenvector at For the Specifically, for a given time series window, it is necessary to evaluate the density of feature changes within the window, using the formula To calculate, first, Characterize each time point The eigenvector The corresponding local probability density, the result calculated by the formula can reflect the comprehensive changes of various characteristics in the oxidation process at a specific time point. Then, is the total number of multidimensional clustering results. Multidimensional clustering is a classification of control parameters such as temperature, humidity, pore size, etc. represents all possible clustering situations. Characterize the bandwidth function, which is determined by the temperature ,humidity , Aperture Adaptive adjustment, indicating that the bandwidth function is not fixed, but dynamically adjusted according to the changes in key parameters such as temperature, humidity and pore size during the oxidation process to better adapt to different oxidation states, Characterize any multidimensional clustering result, is the weight coefficient, representing the The influence of multidimensional clustering results on density estimation. Different clustering results play different roles in the evaluation of feature change density. The weight coefficient It can be adjusted according to actual conditions to highlight the contribution of important clustering results to density estimation. is the kernel function, For time point The eigenvector at For the The kernel function is used to map the feature vector to a high-dimensional space for better density estimation.

[0023] Step S1442, perform window segmentation based on the feature change density evaluation result to continue matching iteration. Specifically, once the feature change density evaluation result is obtained, the window segmentation can be performed based on this result. If the feature change density is uneven or inconsistent with the expected feature change trend in a certain timing window, the window segmentation can be considered. For example, if the feature change density in a certain area is high and that in another area is low, it means that there are differences in the oxidation processes of the two areas. The windows need to be segmented and analyzed separately. Through window segmentation, the complex oxidation process can be further refined to more accurately match the trend consistency. After segmentation, the matching iteration process continues, that is, the new window is re-evaluated for feature change density, trend consistency matching and other steps until the best stage division result is found. Continuous iteration can gradually optimize the stage division of the wet oxidation process of the VCSEL chip and improve the accuracy and reliability of the division.

[0024] Step S200, deploy monitoring sensors, use the monitoring sensors to monitor the wet oxidation process of the VCSEL chip, and establish a real-time data set, which includes oxidation environment temperature data, humidity data, aperture change data, and oxidation time data. Specifically, during the wet oxidation process of the VCSEL chip, first select the appropriate type of monitoring sensor according to its characteristics and needs and deploy it reasonably, such as using a temperature sensor to monitor the oxidation environment temperature, a humidity sensor to measure the ambient humidity, a special device to monitor the aperture change, and a timer to record the oxidation time. During the oxidation process, the sensor continues to work and transmits the collected data to the data processing system at a certain frequency, and then integrates the data to establish a real-time data set, including oxidation environment temperature data, humidity data, aperture change data, and oxidation time data, to provide important basic data support for subsequent oxidation process analysis, stage division, and control parameter optimization.

[0025] Step S300, configure the global control layer, match the key stages with the stage division feature and the real-time data set, and activate the optimization priority corresponding to the key stage division result with the matching result, initialize the global control layer with the activated optimization priority, and process the real-time data set through the initialized global control layer to generate a global decision. Specifically, first configure the global control layer, which matches the key stages with the stage division feature and the real-time data set, determines the key stage of the current oxidation process, activates the optimization priority corresponding to the key stage, and initializes the global control layer with this, and processes the real-time data set after the initialization, comprehensively considers the oxidation state, optimization priority and various parameters, and generates a global decision through analysis and calculation, so as to achieve the overall optimization of the wet oxidation process of the VCSEL chip and improve its quality and efficiency.

[0026] In a possible implementation, a global control layer is configured, key stages are matched through the stage division feature and the real-time data set, and the optimization priority corresponding to the key stage division result is activated with the matching result, the global control layer is initialized with the activated optimization priority, and the real-time data set is processed by the initialized global control layer to generate a global decision, step S300 further includes step S310, the global control layer includes a decision module and a feedback module, when the global control layer is initialized, the decision module is called to process the real-time data set to generate a global decision, when any global decision is generated, the global decision is sent to the feedback module, a feedback window is established, and the decision is optimized with the feedback window when the global decision is executed. Specifically, the global control layer is composed of a decision module and a feedback module, after the global control layer is initialized, the decision module processes the real-time data set, comprehensively considers various factors to generate a global decision, and the global decision is sent to the feedback module after it is generated, and the feedback module establishes a feedback window, when the global decision is executed, the global decision is optimized according to the information collected in the feedback window, and the control accuracy and efficiency of the wet oxidation process of the VCSEL chip are continuously improved through the synergy of the decision module and the feedback module.

[0027] Step S400, after sending the global decision to the parameter association control layer, the parameter association control layer analyzes the global decision and establishes a global goal. Specifically, after the global control layer generates the global decision, it sends it to the parameter association control layer. After receiving it, the parameter association control layer carefully analyzes the global decision, extracts key points and integrates them, so as to establish a clear and detailed global goal, which makes the subsequent adjustment of each control parameter more targeted and purposeful, so as to better achieve the optimization control of the wet oxidation process of the VCSEL chip. For example, the global decision proposes to adjust the temperature and humidity parameters at a specific stage to optimize the oxidation effect. After analysis, the parameter association control layer converts it into specific temperature and humidity target values, and takes the realization of these target values ​​as the global goal. Subsequent operations can be carried out around this global goal, so that the adjustment of each control parameter is more targeted and purposeful, so as to better achieve the optimization control of the wet oxidation process of the VCSEL chip.

[0028] Step S500, predict the linkage relationship between parameters based on the global goal, configure a synchronous adjustment strategy based on the prediction result of the linkage relationship, optimize the global decision based on the synchronous adjustment strategy, and call the local control layer to execute the optimized global decision. Specifically, based on the global goal, the linkage relationship between parameters is predicted by analyzing historical data and applying mathematical models and algorithms. Based on this prediction result, a synchronous adjustment strategy is configured, and it is integrated into the global decision for optimization, making the global decision more comprehensive and reasonable. Finally, the local control layer is called to execute the optimized global decision. The local control layer interacts with the oxidation equipment and sensors to accurately adjust each control parameter to ensure that the oxidation process is carried out according to the optimized decision. At the same time, the actual oxidation state is continuously monitored and information is fed back to further optimize the decision.

[0029] In a possible implementation, the linkage relationship between parameters is predicted based on the global goal, a synchronous adjustment strategy is configured based on the linkage relationship prediction result, and after optimizing the global decision based on the synchronous adjustment strategy, the local control layer is called to execute the optimized global decision. Step S500 further includes step S510, extracting multidimensional characteristic parameters, and defining the multidimensional characteristic parameters as parameter nodes, and the parameter nodes include temperature nodes, humidity nodes, aperture change nodes, and time nodes. Specifically, first, multidimensional characteristic parameters are extracted from the wet oxidation process of the VCSEL chip. The parameters reflect different aspects of the oxidation process and have an important influence on the oxidation effect. The multidimensional characteristic parameters are defined as parameter nodes to establish a basic structure for subsequent analysis and prediction, wherein the parameter nodes include temperature nodes, humidity nodes, aperture change nodes, and time nodes. The temperature node represents the change of the ambient temperature during the oxidation process; the humidity node reflects the humidity level of the oxidation environment; the aperture change node reflects the dynamic change of the chip aperture during the oxidation process; and the time node records the time experienced by the oxidation process. By clarifying the parameter nodes, the relationship between the various parameters can be analyzed more systematically.

[0030] Step S520, based on the global goal, historical data is called to establish the initial causal relationship between parameter nodes. Specifically, after the parameter nodes are determined, historical data is called based on the global goal. The historical data contains the values ​​of each parameter in the past oxidation process and the corresponding oxidation results. By analyzing the historical data, an attempt is made to establish the initial causal relationship between the parameter nodes. For example, the aperture change and oxidation time are observed under different temperature and humidity conditions, and a preliminary judgment is made as to whether the temperature and humidity have a causal effect on the aperture change and oxidation time. If it is found in the historical data that the aperture change tends to accelerate when the temperature rises, then a preliminary causal relationship between the temperature node and the aperture change node can be established.

[0031] Step S530, construct a conditional probability table between each parameter node and its parent node according to the initial causal relationship, and perform causal relationship graph iteration. Specifically, according to the established initial causal relationship, construct a conditional probability table between each parameter node and its parent node. For example, if it is determined that the temperature node is the parent node of the aperture change node, then the probability of aperture change under different temperature conditions can be calculated according to historical data to form a conditional probability table between the temperature node and the aperture change node. By constructing the conditional probability table, the relationship between the parameter nodes can be described more quantitatively. Then, perform causal relationship graph iteration. In each iteration, according to the new observation data and the existing conditional probability table, the causal relationship between the parameter nodes is continuously adjusted and optimized. For example, if the new observation data shows that in some cases, humidity also has an important impact on aperture change, then it is necessary to adjust the relationship between the humidity node and the aperture change node, and update the corresponding conditional probability table.

[0032] Step S540, completing the prediction of the linkage relationship between parameters according to the iteration result of the causal relationship graph. Specifically, according to the iteration result of the causal relationship graph, the prediction of the linkage relationship between parameters is completed. After multiple iterations, the causal relationship graph will gradually stabilize. At this time, the linkage relationship between each parameter node can be predicted more accurately. For example, the trend of pore size change and the possible range of oxidation time under specific temperature and humidity conditions can be predicted. The linkage relationship prediction result can provide an important basis for the optimization control of the oxidation process and help formulate a more reasonable control strategy to achieve efficient and stable wet oxidation process of VCSEL chips.

[0033] In a possible implementation, the linkage relationship between parameters is predicted based on the global target, and a synchronous adjustment strategy is configured based on the linkage relationship prediction result. After the global decision is optimized based on the synchronous adjustment strategy, the local control layer is called to execute the optimized global decision. Step S500 further includes step S550, establishing an abnormal observation layer, which is used to monitor the execution results and perform abnormal monitoring accumulation, and the abnormal monitoring accumulation includes frequency accumulation and abnormal value accumulation. Specifically, first, an abnormal observation layer is established, and the abnormal observation layer is used to closely monitor the execution results to ensure that the oxidation process is carried out according to the expected global decision and optimization strategy. One of the main functions of the abnormal observation layer is to monitor the execution results. During the oxidation process, the actual changes of various key parameters and the execution effect of the global decision are continuously paid attention to. For example, whether the parameters such as temperature, humidity, aperture change and oxidation time are within a reasonable range in real time, and whether the local control layer executes the optimized global decision accurately, by continuously collecting and analyzing these data, the abnormal observation layer can timely discover any situation that does not meet the expected results, and the abnormal observation layer performs abnormal monitoring accumulation, including frequency accumulation, which refers to the abnormal situation. The frequency of occurrence of abnormal conditions is counted. For example, if the temperature exceeds the normal range many times within a certain period of time, the frequency of occurrence of this abnormal situation will be recorded. Through frequency accumulation, we can understand which abnormal conditions occur more frequently, so as to focus on these problem areas and take corresponding measures to adjust and optimize them. In addition to frequency accumulation, abnormal monitoring accumulation also includes abnormal value accumulation. Abnormal value accumulation mainly records and accumulates parameter values ​​that exceed the normal range. For example, if a large abnormal value suddenly appears in the aperture change, the abnormal observation layer will record the abnormal value and add it to the abnormal value accumulation. Through abnormal value accumulation, we can understand the severity of the abnormal situation and which parameters have a greater impact on the oxidation process.

[0034] Step S560, abnormality reporting is performed according to the triggering result of the accumulated abnormal monitoring result. Specifically, finally, abnormality reporting is performed according to the triggering result of the accumulated abnormal monitoring result. When the accumulated abnormal monitoring reaches a certain threshold or meets a specific triggering condition, the abnormal observation layer will issue an abnormal report. The triggering result can be set according to the actual situation. For example, when the abnormal frequency exceeds a certain number of times or the accumulated abnormal value exceeds a certain value, the abnormality reporting will be triggered. The abnormality reporting can be performed in a variety of ways, such as sounding an alarm, sending an email or text message to notify relevant personnel, etc., so as to take timely measures to deal with it and avoid the abnormal situation from causing a greater impact on the oxidation process.

[0035] In a possible implementation, an exception is reported according to a trigger result of the abnormal monitoring accumulation result, and step S560 further includes step S561, establishing an abnormal trigger threshold, triggering and verifying the abnormal monitoring accumulation result through the abnormal trigger threshold, and generating a trigger result. Specifically, first, an abnormal trigger threshold is established. The threshold is used to determine whether the accumulated results of abnormal monitoring have reached the standard for abnormal reporting. The setting of the abnormal trigger threshold needs to comprehensively consider the characteristics of the wet oxidation process of the VCSEL chip, historical data, and tolerance for abnormal situations. For example, the appropriate threshold can be determined based on the frequency and severity of abnormalities in the previous oxidation process, as well as the degree of impact on chip quality. Then, the established abnormal trigger threshold is used to trigger and verify the accumulated results of abnormal monitoring. The accumulated results of abnormal monitoring include information such as frequency accumulation and abnormal value accumulation. The accumulated results are compared with the abnormal trigger threshold to determine whether the threshold is exceeded. If the accumulated results exceed the threshold, it means that the abnormal situation is more serious and corresponding measures may need to be taken to deal with it. By comparing the accumulated results of abnormal monitoring with the abnormal trigger threshold, a trigger result is generated. There are two possible trigger results: a verification pass result and a verification fail result. If the accumulated results of abnormal monitoring do not exceed the abnormal trigger threshold, then the generated trigger result is a verification fail result, indicating that the current abnormal situation is still within an acceptable range and no abnormal reporting is required. If the accumulated results of abnormal monitoring exceed the abnormal trigger threshold, then the generated trigger result is a verification pass result.

[0036] Step S562, if the trigger result is a verification pass result, an exception is reported. Specifically, when the trigger result is a verification pass result, an exception is reported, and the method of reporting the exception can be selected according to the actual situation, such as sounding an alarm, displaying abnormal information on the monitoring interface, sending an email or SMS to notify relevant personnel, etc. The purpose of reporting the exception is to promptly remind relevant personnel to pay attention to the abnormal situation so that appropriate measures can be taken to deal with it, so as to avoid the abnormal situation from causing greater impact on the oxidation process and chip quality.

[0037] The embodiment of the present application adopts fitting observation of the oxidation process, divides the key stages and sets the optimization priority and characteristics of the control parameters for each stage. The temperature, humidity, aperture change and oxidation time data of the VCSEL chip during the wet oxidation process are collected in real time using monitoring sensors. Through the global control layer, the key stages are matched based on the stage characteristics and the real-time data set, the optimization priority is activated and the global decision is generated. The global decision is passed to the parameter association control layer for analysis, the global goal is established, the global decision is optimized through parameter linkage prediction and synchronous adjustment strategy, the local control layer is called to execute the optimized decision, and the control parameters are dynamically adjusted according to the characteristics of different stages to ensure that the oxidation process is always carried out in the optimal state, so as to achieve the technical effect of precise control of the wet oxidation process of the VCSEL chip.

[0038] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art 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 principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A method for optimizing the aperture of a VCSEL chip by wet oxidation using dynamic adjustment, characterized in that: The method comprises: Perform fitting observation on the oxidation process, divide the oxidation process into stages according to the fitting observation results, and establish key stage division results, wherein each key stage division result is provided with an optimization priority of a control parameter, and the key stage division result is configured with a stage division feature; Deploy monitoring sensors, use the monitoring sensors to monitor the wet oxidation process of the VCSEL chip, and establish a real-time data set, wherein the real-time data set includes oxidation environment temperature data, humidity data, aperture change data, and oxidation time data; Configure the global control layer, match the key stages with the stage division feature and the real-time data set, activate the optimization priority of the corresponding key stage division result with the matching result, initialize the global control layer with the activated optimization priority, and process the real-time data set through the initialized global control layer to generate a global decision; After the global decision is sent to the parameter association control layer, the parameter association control layer analyzes the global decision and establishes a global goal; The linkage relationship between parameters is predicted based on the global goal, a synchronization adjustment strategy is configured based on the linkage relationship prediction result, and after the global decision is optimized based on the synchronization adjustment strategy, the local control layer is called to execute the optimized global decision.

2. The method for optimizing the aperture of a VCSEL chip by wet oxidation using dynamic adjustment according to claim 1, characterized in that: The step of dividing the oxidation process into stages according to the fitting observation results and establishing key stage division results also includes: Perform time series fitting of control parameters according to fitting observation results and establish a time series fitting curve; Extracting feature points from the time series fitting curve, and establishing clustering results in the time dimension according to the feature point extraction results; Using the fitting observation results, multidimensional clustering of control parameters is performed to establish multidimensional clustering results, and the features extracted by multidimensional clustering include temperature change rate, humidity volatility, pore size change rate, pore size stability, and cumulative effect of oxidation time; The clustering results under the time dimension and the multi-dimensional clustering results are clustered and fused to establish the key stage division results.

3. The method for optimizing the aperture of a VCSEL chip by wet oxidation using dynamic adjustment according to claim 2, characterized in that: The clustering fusion of the clustering results under the time dimension and the multi-dimensional clustering results also includes: Establish a time series window based on the clustering results in the time dimension; Using the time series window to match the trend consistency of the multidimensional clustering results within the time series window; If the characteristic change trend of the multidimensional clustering results within the time series window is consistent with the clustering results under the time dimension, the stage division is retained; If the characteristic change trend of the multidimensional clustering result in the time series window is inconsistent with the clustering result under the time dimension, the time series window is split and the matching iteration is continued; The clustering fusion is completed according to the matching iteration results and the division results of the retention stage.

4. The method for optimizing the aperture of a VCSEL chip by wet oxidation using dynamic adjustment according to claim 3, characterized in that: The window segmentation of the timing window and the continued matching iteration also include: The feature change density within the time series window is evaluated as follows: ; in, Characterize each time point The eigenvector The corresponding local probability density is, is the total number of multidimensional clustering results, Characterize the bandwidth function, which is determined by the temperature ,humidity , Aperture Adaptive adjustment, Characterize any multidimensional clustering result, is the weight coefficient, representing the The influence of multidimensional clustering results on density estimation, is the kernel function, For time point The eigenvector at For the The feature vector of the multidimensional clustering results; The window is segmented based on the feature change density evaluation result to continue the matching iteration.

5. The method for optimizing the aperture of a VCSEL chip by wet oxidation using dynamic adjustment according to claim 1, characterized in that: The global control layer includes a decision module and a feedback module. When the global control layer is initialized, the decision module is called to process the real-time data set and generate a global decision. When any global decision is generated, the global decision is sent to the feedback module to establish a feedback window. When the global decision is executed, the feedback window is used to optimize the decision.

6. The method for optimizing the aperture of a VCSEL chip by wet oxidation using dynamic adjustment according to claim 1, characterized in that: The predicting of linkage relationship between parameters based on the global objective also includes: Extracting multidimensional characteristic parameters and defining the multidimensional characteristic parameters as parameter nodes, the parameter nodes include temperature nodes, humidity nodes, aperture change nodes, and time nodes; Call historical data based on global goals to establish initial causal relationships between parameter nodes; Construct a conditional probability table between each parameter node and its parent node based on the initial causal relationship, and perform causal relationship graph iteration; The linkage relationship prediction between parameters is completed according to the iterative results of the causal relationship diagram.

7. The method for optimizing the aperture of a VCSEL chip by wet oxidation using dynamic adjustment according to claim 1, characterized in that: The method further comprises: Establishing an abnormal observation layer, the abnormal observation layer is used to monitor the execution results and perform abnormal monitoring accumulation, the abnormal monitoring accumulation includes frequency accumulation and abnormal value accumulation; The abnormality is reported according to the trigger result of the accumulated results of abnormal monitoring.

8. The method for optimizing the aperture of a VCSEL chip by wet oxidation using dynamic adjustment according to claim 7, characterized in that: The method further comprises: Establishing an abnormal trigger threshold, triggering and verifying the abnormal monitoring accumulation result through the abnormal trigger threshold, and generating a trigger result; If the trigger result is a verification pass result, an exception is reported.

Citation Information

Patent Citations

  • Open-type packaging structure used for sensor chip and manufacturing method thereof

    CN102838079A

  • Pressure sensor chip based on MEMS (Micro Electro Mechanical System) technology and manufacturing method thereof

    CN105000529A

  • LED chip and manufacturing method thereof

    CN110379899A

  • Vacuum heat insulation pipeline temperature and pressure double-closed-loop control method

    CN118838457A

  • Heat treatment process optimization method for carbon dioxide steel cylinder

    CN118861582A

Cited By

  • Chip test method and test system based on temperature response

    CN119471327A

  • Accurate control method of VCSEL chip oxidation pore channel

    CN121332291A

  • A method for precisely controlling oxidation channels of a VCSEL chip

    CN121332291B