Method, device, electronic device and storage medium for determining equipment operating parameters

Abstract

The present application provides a method, device, electronic device and storage medium for determining equipment operating parameters, and relates to the field of semiconductor technology. The method includes: obtaining a first set of operating parameters of the production equipment, a plurality of first production result parameters corresponding to the first set of operating parameters within a preset historical time period, and evaluation scores for the plurality of first production result parameters, and then, based on the first set of operating parameters, the plurality of first production result parameters and the evaluation scores, using a preset supervised learning data model to obtain a second set of operating parameters for the production equipment, so as to enable the production equipment to operate with the second set of operating parameters. Compared with the manual equipment parameter debugging and determination method, by introducing a supervised learning model in parameter determination, manual participation is reduced, so that the parameter determination of the equipment does not depend on the knowledge level and industry experience of the operator. By using the acquired data information, a second set of operating parameters that can significantly improve the production efficiency of the equipment can be quickly and effectively obtained.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor technology, and in particular to a method and apparatus for determining device operating parameters, an electronic device, and a storage medium. Background Art

[0002] Currently, in the semiconductor industry, when adjusting the parameters of semiconductor equipment such as strippers, etchers, CVD machines, etc., it is usually done by continuously running the equipment through related processes, manually analyzing and comparing different parameters and related operating results, and inferring a reasonable combination of equipment parameters. Then, the new equipment parameter combination is used to run the related processes, and this operation is repeated continuously until a set of parameter combinations that can make the equipment operate most efficiently is selected.

[0003] This parameter adjustment method requires a large number of running tests to obtain the final result, the adjustment speed is slow, and the accuracy of the calculated optimal parameter combination is low. Summary of the Invention

[0004] The purpose of the present invention is to address the deficiencies in the above-mentioned prior art and provide a method, device, electronic device and storage medium for determining equipment operating parameters so that the equipment parameters are infinitely close to reasonable values, thereby enabling the equipment to better perform its performance.

[0005] To achieve the above objectives, the technical solutions adopted in the embodiments of the present application are as follows:

[0006] In a first aspect, an embodiment of the present application provides a method for determining device operating parameters, the method comprising:

[0007] Obtaining a first set of operating parameters of a production device and a plurality of first production result parameters corresponding to the first set of operating parameters within a preset historical time period;

[0008] Obtaining evaluation scores for the plurality of first production result parameters;

[0009] Based on the first set of operating parameters, the multiple first production result parameters and the evaluation scores, a preset supervised learning data model is used to obtain a second set of operating parameters for the production equipment, so that the production equipment operates with the second set of operating parameters.

[0010] Optionally, obtaining the second set of operating parameters of the production equipment using a preset supervised learning data model based on the first set of operating parameters, the plurality of first production result parameters, and the evaluation scores includes:

[0011] generating a three-dimensional graph according to the first set of operating parameters, the plurality of first production result parameters, and the equipment operating time corresponding to the first set of operating parameters;

[0012] performing denoising processing on the three-dimensional graphics;

[0013] The second set of operating parameters is obtained using the supervised learning data model based on the first set of operating parameters, the production result parameters of the production result dimension in the denoised three-dimensional graphics, and the corresponding evaluation scores.

[0014] Optionally, the performing denoising on the three-dimensional graphics includes:

[0015] performing wavelet decomposition on the three-dimensional graphics to obtain a multi-layer wavelet signal;

[0016] performing threshold processing on the multi-layer wavelet signal according to a preset wavelet coefficient threshold to obtain a denoised multi-layer wavelet signal;

[0017] The denoised three-dimensional graphics are obtained according to the denoised multi-layer wavelet signal.

[0018] Optionally, performing wavelet decomposition on the three-dimensional graphics to obtain a multi-layer wavelet signal includes:

[0019] The three-dimensional graphics are subjected to wavelet grading using a preset orthogonal wavelet basis and a preset number of decomposition layers to obtain multi-layer wavelet signals whose number is equal to the preset number of decomposition layers, where the preset number of decomposition layers is four.

[0020] Optionally, performing threshold processing on the multi-layer wavelet signal to obtain a denoised multi-layer wavelet signal includes:

[0021] According to the preset wavelet coefficient threshold, the wavelet coefficients of the first wavelet signal in the multi-layer wavelet signal that are greater than the preset wavelet coefficient threshold are retained, and the wavelet coefficients of the second wavelet signal in the multi-layer wavelet signal that are less than the preset wavelet coefficient threshold are set to zero.

[0022] Optionally, performing threshold processing on the multi-layer wavelet signal according to a preset wavelet coefficient threshold to obtain a denoised multi-layer wavelet signal includes:

[0023] According to the preset wavelet coefficient threshold, the wavelet coefficients of the first wavelet signal in the multi-layer wavelet signal that are greater than the preset wavelet coefficient threshold are shrunk toward zero, and the wavelet coefficients of the second wavelet signal in the multi-layer wavelet signal that are less than the preset wavelet coefficient threshold are set to zero.

[0024] Optionally, before obtaining the second set of operating parameters of the production equipment using a preset supervised learning data model based on the first set of operating parameters, the plurality of first production result parameters, and the evaluation scores, the method further includes:

[0025] The supervised learning data model is constructed using parameter rules of the production equipment, wherein the parameter rules include parameter ranges corresponding to multiple types of operating parameters of the production equipment.

[0026] In a second aspect, an embodiment of the present application further provides a device for determining equipment operating parameters, including: an acquisition module and a processing module;

[0027] The acquisition module is configured to acquire a first set of operating parameters of the production equipment and a plurality of first production result parameters corresponding to the first set of operating parameters within a preset historical time period; and to acquire evaluation scores for the plurality of first production result parameters;

[0028] The processing module is used to obtain a second set of operating parameters for the production equipment based on the first set of operating parameters, the multiple first production result parameters and the evaluation score, using a preset supervised learning data model, so as to enable the production equipment to operate with the second set of operating parameters.

[0029] In a third aspect, an embodiment of the present application further provides an electronic device comprising: a processor, a storage medium and a bus, wherein the storage medium stores program instructions executable by the processor. When the electronic device is running, the processor communicates with the storage medium through the bus, and the processor executes the program instructions to perform the steps of the device operating parameter determination method as described in any one of the first aspects.

[0030] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the device operating parameter determination method as described in any one of the first aspects are executed.

[0031] The beneficial effects of the present application are as follows: the embodiment of the present application provides a method for determining equipment operating parameters, after obtaining a first set of operating parameters of a production equipment, a plurality of first production result parameters corresponding to the first set of operating parameters within a preset historical time period, and evaluation scores for the plurality of first production result parameters, a preset supervised learning data model is used based on the first set of operating parameters, the plurality of first production result parameters, and the evaluation scores to obtain a second set of operating parameters for the production equipment, so that the production equipment operates with the second set of operating parameters. Compared with the manual equipment parameter debugging and determination method, the introduction of the supervised learning model in the parameter determination reduces manual participation, so that the parameter determination of the equipment does not depend on the knowledge level and industry experience of the operator. By using the acquired data information, a second set of operating parameters that can significantly improve the production efficiency of the equipment can be obtained quickly and effectively. In addition, the determination of equipment parameters by the supervised learning model is much better than manual adjustment in terms of accuracy and actual effect. By continuously learning and summarizing historical data, the second set of operating parameters can be infinitely approximated to the optimal combination of parameters, thereby enabling the production equipment to achieve maximum performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0033] Figure 1 A flowchart of a method for determining equipment operating parameters provided in one embodiment of the present application;

[0034] Figure 2 A flowchart of a method for determining equipment operating parameters provided in another embodiment of the present application;

[0035] Figure 3 A schematic diagram of a three-dimensional graphic generated for an embodiment of the present application;

[0036] Figure 4 A flowchart of a method for determining equipment operating parameters provided in another embodiment of the present application;

[0037] Figure 5 A schematic diagram of a device for determining equipment operating parameters provided in one embodiment of the present application;

[0038] Figure 6 A schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments.

[0040] In this application, unless otherwise expressly specified and limited, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, the features defined as "first" and "second" may explicitly or implicitly include at least one feature. In the description of the present invention, the meaning of "multiple" is at least two, such as two or three, unless otherwise clearly and specifically limited. The terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or equipment. In the absence of further restrictions, the elements defined by the sentence "comprise a..." do not exclude the presence of other identical elements in the process, method, article or equipment including the elements.

[0041] The present application is applied to production equipment (such as semiconductor equipment such as degumming machines, etc., CVD machines, etc.); or to the control equipment corresponding to the production equipment (for example, the electronic equipment corresponding to the production equipment is implemented, and the electronic equipment can be, for example, a terminal device or a server). The production equipment and the corresponding control equipment can be connected through a bus or the like, and the control equipment adjusts the equipment operating parameters of the production equipment by executing the equipment operating parameter determination method of the present application.

[0042] During equipment operation, operators set various parameters, such as voltage, current, power, speed, pressure, and temperature. The equipment automatically adjusts to these settings during production to produce the desired product. The combination of different operating parameters often determines the quality of the product. For example, during the heating process of wafer production equipment, if the temperature is too high, the product will melt and be scrapped, while if the temperature is too low, the process requirements cannot be met. Only when all parameters are set to the appropriate values ​​can the equipment operate normally at the highest efficiency and the product quality be guaranteed.

[0043] Manually adjusting equipment operating parameters to achieve optimal performance takes a very long time. The accuracy and effectiveness of these parameters depend on factors such as the knowledge and industry experience of the operator, and errors can occur, creating significant limitations. Furthermore, even a small adjustment to a parameter can significantly impact production results.

[0044] To address the shortcomings of current methods for determining device operating parameters, the present application provides multiple possible implementations to ensure that device parameters are as close to reasonable values ​​as possible, thereby enabling the device to perform better. This is explained below using multiple examples with reference to the accompanying figures. Figure 1 This is a flow chart of a method for determining equipment operating parameters provided in one embodiment of the present application. Figure 1 As shown, the method includes:

[0045] Step 101: Acquire a first set of operating parameters of a production device and a plurality of first production result parameters corresponding to the first set of operating parameters within a preset historical time period.

[0046] It should be noted that the production equipment in each production link will have its own operating parameters. Depending on the type of production equipment, the type and quantity of the obtained operating parameters may vary. The first set of operating parameters may include one or more specific parameter types. In one possible implementation, the first set of operating parameters of the production equipment may include one or more of the following: voltage, current, power, speed, pressure, temperature, etc. This application does not limit this. In addition, the first production result parameter is a digital description of the specific production result after the production equipment is operated according to the first set of operating parameters. Depending on the type of production equipment, the type of the obtained production result parameter may vary. The first production result parameter may include one or more specific parameter types. In one possible implementation, the first production result parameter of the production equipment may include one or more of the following: yield, quantity, production time, etc.

[0047] It should also be noted that for each first set of operating parameters, the specific time length of the preset historical time period may be different. The user can select the preset historical event segment corresponding to the first set of operating parameters based on the actual operating conditions of the production equipment or the data requirements of the required first production result parameters.

[0048] In a specific implementation, if the production equipment is a wafer production equipment, the first set of operating parameters set are 20Tor air pressure, 30°C temperature, and 500W power. 8-inch wafers (8'wafer) are produced according to this first set of operating parameters. It takes 1.5 hours to produce 100 pics, and the yield is 90%. That is, the first production result parameters corresponding to the first set of operating parameters within the preset historical event period (1.5 hours) are yield: 90%, quantity: 100 pics, and production time: 1.5 hours.

[0049] It should also be noted that the same production equipment may involve the production of multiple products. Therefore, when implementing the equipment operating parameter determination method of the present application, for each product, the first set of operating parameters and first production result parameters corresponding to the product will be obtained through the above method. In the subsequent equipment operating parameter determination, the subsequent second set of operating parameters will be confirmed based on the correlation between the production products. For example, when there is a strong correlation between each product and the operating parameters in the production equipment (that is, no matter which product is produced, the obtained production result parameters are similar or obey the same rules), the first set of operating parameters and first production result parameters of each product can be processed as a whole to determine a second set of operating parameters corresponding to the production equipment; for another example, when there is a correlation between each product and the operating parameters in the production equipment (that is, for each product, the obtained production result parameters have large differences or no rules), the first set of operating parameters and first production result parameters of each product can be processed separately to determine the second set of operating parameters corresponding to each product in the production equipment.

[0050] In a specific implementation, the production equipment is a wafer production equipment that can produce 8-inch wafers (8'wafers), 10-inch wafers (10'wafers), and 12-inch wafers (12'wafers). The first production result parameters obtained for each type of wafer are shown in the following table. Table 1 is a production result parameter table of the wafer production equipment provided in one embodiment of the present application, as shown in Table 1:

[0051] Table 1 Production result parameter table of the wafer production equipment provided by one embodiment of the present application

[0052] Product Name yield quantity Production time 8'wafer 90％ 100pic 1.5 hours 10'wafer 92％ 200pic 2.7 hours 12'wafer 95％ 300pic 3.2 hours

[0053] Step 102: Obtain evaluation scores for a plurality of first production result parameters.

[0054] In step 101, the production equipment obtains multiple first production result parameters within a preset historical time period based on the first set of operating parameters. These multiple first production result parameters, as a set of production result parameters, are a digitized expression of the production results, and an evaluation score is assigned to each set of production result parameters. It should be noted that the evaluation score can be a manual evaluation performed by analyzing the multiple first production result parameters, or it can be calculated using a preset scoring scheme. This application does not limit the specific scoring method for the evaluation score, and users can select it according to actual needs.

[0055] Step 103: Based on the first set of operating parameters, multiple first production result parameters and evaluation scores, a preset supervised learning data model is used to obtain a second set of operating parameters for the production equipment, so as to enable the production equipment to operate with the second set of operating parameters.

[0056] It should be noted that supervised learning is the use of labeled data for learning. Labels are used to guide and strengthen the machine's learning process, constantly correct the machine's errors, and iterate repeatedly until the machine no longer makes the same mistakes. For example, weather forecasting uses a supervised learning model, which takes historical data (such as atmospheric, circulation, air pressure, humidity, wind speed and other data indicators) and corresponding weather condition labels as input, and trains the algorithm model through data-intensive computing to ultimately obtain a robust and accurate prediction output. This application forms an artificial intelligence framework by establishing a supervised learning data model, which realizes the processing of the first set of operating parameters, the multiple first production result parameters and the evaluation scores, and generates a second set of operating parameters for the production equipment by processing the multiple first sets of operating parameters and their corresponding first production result parameters and the evaluation scores.

[0057] It's also important to note that artificial intelligence (AI) can be understood as the implementation of human intelligence on computers. Some have even described it as simulating or reproducing the processes of a "carbon-based brain" with a "silicon-based brain." AI can be divided into two categories: weak AI (top-down AI) and strong AI (bottom-up AI). The former lacks true reasoning and problem-solving capabilities and merely resembles a human; the latter is capable of true reasoning and problem-solving, becoming a self-aware intelligent entity. With the rise of technologies like cloud computing, big data, and deep neural networks, AI technology, centered on ultra-high computing power, vast data resources, and powerful AI algorithms, has advanced rapidly. Specifically, the various rules, protocols, or algorithm-implementing programs running on computers can be considered the epitome of AI, though scientists didn't call them that at the time. These programs, based on human programming concepts, continuously implement human ideas, endowing them with intelligence that can partially replace or assist humans, capabilities not inherent in machines. It can be said that the immense computing power enabled by technologies like cloud computing, the vast amounts of data samples available in the big data era, and the superior performance of deep neural networks have collectively contributed to the current success of AI.

[0058] In one specific implementation, if the production equipment is a wafer production device, and the first set of operating parameters set is 20 Torr pressure, 30°C temperature, and 500W power, then 8-inch wafer production is performed using this first set of operating parameters, resulting in first production result parameters of 90% yield, 100 pics, and 1.5 hours of production time. The evaluation score corresponding to these first production result parameters is 90. After obtaining this data, the data is analyzed using a preset supervised learning data model to obtain a second set of operating parameters for the production equipment: 21 Torr pressure, 30°C temperature, and 500W power. The production equipment will then operate using this new second set of operating parameters in the next run.

[0059] In one possible implementation, it is difficult to quickly determine an effective parameter combination for a production device based solely on a first set of operating parameters and their corresponding multiple first production result parameters and evaluation scores. Therefore, multiple sets of operating parameters from historical production activities of the production device, along with the corresponding production result parameters and evaluation scores for each set of operating parameters, can be pre-entered into a preset supervised learning data model. This helps the preset supervised learning data model quickly determine the direction of improvement for the operating parameters, thereby quickly and effectively obtaining a second set of operating parameters that significantly improves the production efficiency of the device. For example, the historical production activities include the following data: a first set of operating parameters is 20 Torr pressure, 30°C temperature, and 500W power. An 8-inch wafer is produced based on this first set of operating parameters, resulting in first production result parameters of 90% yield, 100 pic, and 1.5 hours of production time, with a corresponding evaluation score of 90. The first set of operating parameters obtained this time was 21 Torr pressure, 30°C temperature, and 500W power. 8-inch wafer production was conducted using this first set of operating parameters, resulting in a first production result with a yield of 85%, a quantity of 80 pics, and a production time of 1 hour, corresponding to an evaluation score of 80. By comparing the two evaluation scores, the first set of operating parameters, and the first production result parameters, the pre-set supervised learning data model can identify the strengths and weaknesses of the operating parameter combinations and determine how the second set of operating parameters should be improved based on the first set of operating parameters.

[0060] In another possible implementation, if there is no historical data in the preset supervised learning data model, after obtaining the first set of operating parameters, the multiple first production result parameters and the evaluation scores for the first time, due to the small amount of data, it is impossible to determine the direction of improvement of the operating parameters. At this time, the second set of operating parameters generated can be a set of randomly generated operating parameters to increase the data amount of the preset supervised learning data model for subsequent processing.

[0061] The above method is only an example. In actual implementation, there may be other ways to obtain the second set of operating parameters, and this application does not limit this.

[0062] In summary, the embodiment of the present application provides a method for determining equipment operating parameters, which obtains a first set of operating parameters of a production equipment, a plurality of first production result parameters corresponding to the first set of operating parameters within a preset historical time period, and an evaluation score for the plurality of first production result parameters, and then, based on the first set of operating parameters, the plurality of first production result parameters, and the evaluation score, a preset supervised learning data model is used to obtain a second set of operating parameters for the production equipment, so that the production equipment operates with the second set of operating parameters. Compared with the manual equipment parameter debugging and determination method, the introduction of the supervised learning model in the parameter determination reduces manual participation, so that the parameter determination of the equipment does not depend on the knowledge level and industry experience of the operator. By using the acquired data information, a second set of operating parameters that can significantly improve the production efficiency of the equipment can be obtained quickly and effectively. In addition, the determination of equipment parameters by the supervised learning model is much better than manual adjustment in terms of accuracy and actual effect. By continuously learning and summarizing historical data, the second set of operating parameters can be infinitely approximated to the optimal combination of parameters, thereby enabling the production equipment to achieve maximum performance.

[0063] Optional, in the above Figure 1 Based on this, the present application also provides a possible implementation of a method for determining equipment operating parameters. Figure 2 A flow chart of a method for determining equipment operating parameters provided in another embodiment of the present application; Figure 2 As shown, according to the first set of operating parameters, the multiple first production result parameters and the evaluation scores, a preset supervised learning data model is used to obtain a second set of operating parameters of the production equipment, including:

[0064] Step 201: Generate a three-dimensional graph according to a first set of operating parameters, a plurality of first production result parameters, and equipment operating time corresponding to the first set of operating parameters.

[0065] In one possible implementation, in order to facilitate the subsequent processing of the preset supervised learning data model, the first set of operating parameters, multiple first production result parameters, and the equipment operating time corresponding to the first set of operating parameters can be processed into the form of a three-dimensional graphic. This application does not limit the specific generation method of the three-dimensional graphic, the establishment rules of the three-dimensional graphic coordinate system, etc., and users can set it according to actual needs.

[0066] In a specific implementation, the x-axis can be set to be the first set of operating parameters, the y-axis can be set to be the equipment operating time corresponding to the first set of operating parameters, and the z-axis can be set to be the multiple first production result parameters to obtain a three-dimensional graph. Figure 3 A three-dimensional graphic diagram generated for an embodiment of the present application, such as Figure 3As shown, the x-axis is the first set of operating parameters. If there are multiple specific parameter types in the first set of operating parameters of the production equipment, a three-dimensional graph is generated for each operating parameter (for example, if the first set of operating parameters of the production equipment includes temperature and pressure, a three-dimensional graph is generated for temperature, and a three-dimensional graph is generated for pressure). In other words, in a three-dimensional graph, the meaning of the operating parameters represented by the x-axis is determined. The y-axis is the equipment operating time corresponding to the first set of operating parameters, and the extension of the y-axis coordinate reflects the change of events. The z-axis is a plurality of first production result parameters, which can be a specific characterization value obtained after processing of the plurality of first production result parameters, or it can be the numerical value of a production result parameter of the plurality of first production result parameters. This application does not limit this. Overall, the three-dimensional graph represents the change in production results as the operating time increases under the operating parameters represented by x weeks.

[0067] Step 202: De-noising the three-dimensional graphics.

[0068] It should be noted that due to the influence of various factors in the production environment, monitoring data, especially time series (such as Figure 3 There is a lot of noise in the data (production equipment noise, environmental noise, etc.). These noises seriously affect further quantitative analysis and data mining, and it is necessary to denoise them in advance. It should also be noted that denoising of three-dimensional graphics can be performed using non-local denoising methods (Non-local method), transform denoising methods (Transform method), Gaussian denoising, median filtering, etc. This application does not limit the specific denoising processing method, and users can choose according to actual needs.

[0069] Step 203: Based on the first set of operating parameters, the production result parameters of the production result dimension in the denoised three-dimensional graph, and the corresponding evaluation scores, a supervised learning data model is used to obtain a second set of operating parameters.

[0070] After denoising the three-dimensional graphics, the production result parameters of the production result dimension in the processed three-dimensional graphics are obtained. Then, according to the supervised learning data model, the first set of operating parameters, the production result parameters of the production result dimension in the denoised three-dimensional graphics, and the corresponding evaluation scores are processed to obtain the second set of operating parameters.

[0071] By generating a three-dimensional graph of the first set of operating parameters, the multiple first production result parameters, and the equipment operating time corresponding to the first set of operating parameters, and denoising the three-dimensional graph, noise interference in the supervised learning data model during subsequent processing is eliminated, thereby making the adjustment direction, amplitude, etc. of the obtained second set of operating parameters more accurate and efficient.

[0072] Optional, in the above Figure 2 Based on this, the present application also provides a possible implementation of a method for determining equipment operating parameters. Figure 4 This is a flow chart of a method for determining equipment operating parameters provided by another embodiment of the present application; Figure 4 As shown, performing denoising on the three-dimensional graphics includes:

[0073] Step 401: Perform wavelet decomposition on the three-dimensional graphics to obtain multi-layer wavelet signals.

[0074] It should be noted that due to the inherent nonlinearity and high signal-to-noise ratio of time series, traditional methods such as Gaussian denoising and median filtering often have many drawbacks. Therefore, wavelet filtering can be used for denoising. Because wavelet theory was developed based on the requirements of time-frequency localization, it has adaptive and mathematical microscopic properties, making it particularly suitable for processing non-stationary and nonlinear signals. As the name suggests, a wavelet is a very small wave whose integral is approximately 0. Wavelets are divided into six types based on their shape. Wavelet decomposition is the process of breaking down an existing wave into many smaller waveforms.

[0075] One possible implementation method for time series denoising is nonlinear thresholding. Its principle is that after wavelet transform, the energy of the useful signal is concentrated in a small number of wavelet coefficients, while white noise remains dispersed across a large number of wavelet coefficients in the wavelet transform domain. Therefore, the wavelet coefficients of the useful signal are necessarily larger than those of the noise, which has dispersed energy and smaller amplitudes. Therefore, the useful signal and noise can be separated based on the amplitude of the spectrum. Therefore, when denoising a three-dimensional image, the first step is to perform wavelet decomposition on the image to obtain a multi-layer wavelet signal.

[0076] Step 402: performing threshold processing on the multi-layer wavelet signal according to a preset wavelet coefficient threshold to obtain a denoised multi-layer wavelet signal.

[0077] Step 403: Obtain a denoised three-dimensional graphic based on the denoised multi-layer wavelet signal.

[0078] Threshold processing is performed on the multi-layer wavelet signal obtained by decomposition. The wavelet coefficients of each layer of wavelet signal can be adjusted according to the preset wavelet coefficient threshold, thereby obtaining the denoised multi-layer wavelet signal. The wavelet coefficient threshold is used to distinguish the threshold processing method for the wavelet signal. For example, according to the wavelet coefficient threshold, the multi-layer wavelet signal can be divided into wavelet signals with wavelet coefficients greater than the wavelet coefficient threshold and wavelet signals with wavelet coefficients less than the wavelet coefficient threshold. Different threshold processing methods can be set for these two categories. It should be noted that when threshold processing is performed on the multi-layer wavelet signal, the wavelet coefficients can be increased, reduced, set to zero, etc., and this application does not limit this. After denoising is completed, the processed three-dimensional graphics are obtained based on the denoised multi-layer wavelet signal.

[0079] Denoising is achieved by using wavelet decomposition, which has the advantages of low entropy, multi-resolution, decorrelation, and flexibility in basis selection.

[0080] Optional, in the above Figure 4 Based on the above, the present application also provides a possible implementation of a method for determining equipment operating parameters, wherein the three-dimensional graphics are subjected to wavelet decomposition to obtain a multi-layer wavelet signal, including:

[0081] The three-dimensional graphics are subjected to wavelet grading using a preset orthogonal wavelet basis and a preset number of decomposition layers to obtain multi-layer wavelet signals whose number is equal to the preset number of decomposition layers, where the preset number of decomposition layers is four.

[0082] Before wavelet decomposition, an appropriate orthogonal wavelet basis and a preset number of decomposition layers can be set as needed to perform wavelet transform on the three-dimensional image and decompose it into the preset number of layers, wherein the preset number of decomposition layers refers to the number of layers of the wavelet signal obtained by decomposition.

[0083] In addition, it should be noted that, according to the theory of multi-resolution analysis, the wavelet coefficients of high-level decomposition correspond to the low-frequency part, and the low-frequency part is mainly composed of signals. Therefore, the higher the decomposition level, the more low-frequency components are removed, the more obvious the denoising effect is, but the distortion also increases. To be conservative, in actual implementation, in order to achieve better denoising without distortion, the decomposition level should not be too high, and the maximum should not exceed 5 layers. For sequences with stronger volatility (such as CPU-time sequences), it is generally not more than 4 layers. In a specific implementation method, according to the practice of the actual denoising effect of this application, decomposition based on the db4 wavelet function can be used, and decomposition to 3-4 layers can achieve good results, so the preset number of decomposition layers can be set to four.

[0084] Optional, in the above Figure 4 On the basis of the above, the present application also provides a possible implementation of a method for determining equipment operating parameters, which performs threshold processing on the multi-layer wavelet signal according to a preset wavelet coefficient threshold to obtain a denoised multi-layer wavelet signal, including:

[0085] According to the preset wavelet coefficient threshold, the wavelet coefficients of the first wavelet signal in the multi-layer wavelet signal that are greater than the preset wavelet coefficient threshold are retained, and the wavelet coefficients of the second wavelet signal in the multi-layer wavelet signal that are less than the preset wavelet coefficient threshold are set to zero.

[0086] In one possible implementation, the multi-layer wavelet signal can be divided into a wavelet signal whose wavelet coefficient is greater than the wavelet coefficient threshold (i.e., the first wavelet signal) and a wavelet signal whose wavelet coefficient is less than the wavelet coefficient threshold (i.e., the second wavelet signal) according to the wavelet coefficient threshold. For the first wavelet signal, its wavelet coefficient is retained; for the second wavelet signal, its wavelet coefficient is set to zero.

[0087] In one specific implementation, a wavelet coefficient threshold is set. Wavelet signals with wavelet coefficients greater than the threshold are considered valid signals, and their wavelet coefficients are retained. Wavelet signals with wavelet coefficients less than the threshold are considered noise, and the interference is eliminated by setting the wavelet coefficients to zero. The above is merely an example. In actual implementation, setting different wavelet coefficient thresholds can also be used to perform other types of processing on multi-layer wavelet signals, and this application does not limit this.

[0088] The threshold processing of multi-layer wavelet signals is realized by the above method, which is simple to implement and has fast calculation speed.

[0089] Optional, in the above Figure 4 On the basis of the above, the present application also provides a possible implementation of a method for determining equipment operating parameters, which performs threshold processing on the multi-layer wavelet signal according to a preset wavelet coefficient threshold to obtain a denoised multi-layer wavelet signal, including:

[0090] According to the preset wavelet coefficient threshold, the wavelet coefficients of the first wavelet signal in the multi-layer wavelet signal that are greater than the preset wavelet coefficient threshold are shrunk toward zero, and the wavelet coefficients of the second wavelet signal in the multi-layer wavelet signal that are less than the preset wavelet coefficient threshold are set to zero.

[0091] In one possible implementation, the multi-layer wavelet signal can be divided into a wavelet signal whose wavelet coefficient is greater than the wavelet coefficient threshold (i.e., the first wavelet signal) and a wavelet signal whose wavelet coefficient is less than the wavelet coefficient threshold (i.e., the second wavelet signal) according to the wavelet coefficient threshold. For the first wavelet signal, its wavelet coefficient is shrunk toward zero; for the second wavelet signal, its wavelet coefficient is set to zero.

[0092] In a specific implementation, when performing denoising, the goal is to remove the low-frequency part. Therefore, wavelet signals with wavelet coefficients less than the wavelet coefficient threshold are considered to be noise, and the interference is eliminated by setting the wavelet coefficients to zero. However, for wavelet signals with wavelet coefficients greater than the wavelet coefficient threshold, they may also contain some noise. Therefore, their wavelet coefficients are shrunk to zero to achieve better denoising and fidelity.

[0093] The denoising is performed according to the above method, and the denoised multi-layer wavelet signal obtained is an approximate optimal estimation of the original signal (ie, the noise-free signal) and has wider applicability.

[0094] Optionally, based on the above embodiment, the present application further provides a possible implementation of a method for determining equipment operating parameters, the method comprising: before obtaining a second set of operating parameters for the production equipment using a preset supervised learning data model based on the first set of operating parameters, the multiple first production result parameters, and the evaluation score, the method further comprises:

[0095] The supervised learning data model is constructed using parameter rules of the production equipment, wherein the parameter rules include parameter ranges corresponding to multiple types of operating parameters of the production equipment.

[0096] It should be noted that each type of production equipment has specific operating parameters, most of which are expressed in data form. For example, air pressure = 10 Torr, temperature = 8°C, power = 100W, etc. When constructing the supervised learning data model, the model must be processed to obtain a second set of operating parameters to guide subsequent production practices. To ensure the rationality and safety of the parameters, parameter rules for the production equipment must be pre-set in the supervised learning data model. These parameter rules include the parameter ranges and parameter accuracy corresponding to the various operating parameters of the production equipment.

[0097] In a specific implementation, Table 2 is a parameter rule for a production device provided in an embodiment of the present application, as shown in Table 2:

[0098] Table 2 Parameter rules for a production device provided in one embodiment of the present application

[0099]

[0100]

[0101] The parameter rules in Table 2 illustrate the parameter ranges corresponding to the three operating parameters of the production equipment, namely, air pressure, temperature, and power. In subsequent use, the supervised learning data model can determine the equipment operating parameters within the parameter range corresponding to the parameter rules based on the first set of operating parameters, the multiple first production result parameters, and the evaluation scores.

[0102] The above is only an example. In actual implementation, other parameter rules can be set according to usage needs to construct the supervised learning data model, and this application does not limit this.

[0103] The following describes the device for determining the device operating parameters, electronic equipment, and storage medium provided in this application. The specific implementation process and technical effects are described above and will not be repeated below.

[0104] An embodiment of the present application provides a possible implementation example of an apparatus for determining equipment operating parameters, which can execute the method for determining equipment operating parameters provided in the above embodiment. Figure 5 This is a schematic diagram of a device for determining equipment operating parameters provided in one embodiment of the present application. Figure 5 As shown, the device 100 for determining equipment operating parameters includes: an acquisition module 51 and a processing module 53;

[0105] An acquisition module 51 is configured to acquire a first set of operating parameters of a production device and a plurality of first production result parameters corresponding to the first set of operating parameters within a preset historical time period; and to acquire evaluation scores for the plurality of first production result parameters;

[0106] The processing module 53 is used to obtain a second set of operating parameters for the production equipment based on the first set of operating parameters, multiple first production result parameters and evaluation scores using a preset supervised learning data model, so as to enable the production equipment to operate with the second set of operating parameters.

[0107] Optionally, the processing module 53 is used to generate a three-dimensional graph based on the first set of operating parameters, multiple first production result parameters, and the equipment operating time corresponding to the first set of operating parameters; denoise the three-dimensional graph; and obtain the second set of operating parameters using a supervised learning data model based on the first set of operating parameters, the production result parameters of the production result dimension in the denoised three-dimensional graph, and the corresponding evaluation scores.

[0108] Optionally, the processing module 53 is used to perform wavelet decomposition on the three-dimensional graphics to obtain a multi-layer wavelet signal; perform threshold processing on the multi-layer wavelet signal according to a preset wavelet coefficient threshold to obtain a denoised multi-layer wavelet signal; and obtain a denoised three-dimensional graphics based on the denoised multi-layer wavelet signal.

[0109] Optionally, the processing module 53 is configured to perform wavelet grading on the three-dimensional graphics using a preset orthogonal wavelet basis and a preset number of decomposition layers to obtain multi-layer wavelet signals having a number equal to the preset number of decomposition layers, where the preset number of decomposition layers is four.

[0110] Optionally, the processing module 53 is used to retain the wavelet coefficients of the first wavelet signal in the multi-layer wavelet signal that are greater than the preset wavelet coefficient threshold according to the preset wavelet coefficient threshold, and to set the wavelet coefficients of the second wavelet signal in the multi-layer wavelet signal that are less than the preset wavelet coefficient threshold to zero.

[0111] Optionally, the processing module 53 is used to shrink the wavelet coefficients of the first wavelet signal in the multi-layer wavelet signal that are greater than the preset wavelet coefficient threshold toward zero according to a preset wavelet coefficient threshold, and to set the wavelet coefficients of the second wavelet signal in the multi-layer wavelet signal that are less than the preset wavelet coefficient threshold to zero.

[0112] Optionally, the processing module 53 is configured to construct a supervised learning data model using parameter rules of the production equipment, where the parameter rules include parameter ranges corresponding to multiple types of operating parameters of the production equipment.

[0113] The above-mentioned device is used to execute the method provided in the above-mentioned embodiment. Its implementation principle and technical effect are similar and will not be repeated here.

[0114] The above modules can be one or more integrated circuits configured to implement the above methods, such as one or more application specific integrated circuits (ASICs), one or more digital singnal processors (DSPs), or one or more field programmable gate arrays (FPGAs). For another example, when a module is implemented by scheduling program code through a processing element, the processing element can be a general-purpose processor, such as a central processing unit (CPU) or other processor that can call program code. For another example, these modules can be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0115] An embodiment of the present application provides a possible implementation example of an electronic device, which can execute the device operating parameter determination method provided in the above embodiment. Figure 6 A schematic diagram of an electronic device provided in an embodiment of the present application, which can be integrated into a terminal device or a chip of a terminal device, and the terminal can be a computing device with data processing capabilities.

[0116] The electronic device includes a processor 601, a storage medium 602, and a bus. The storage medium stores program instructions executable by the processor. When the control device is operating, the processor and the storage medium communicate via the bus, and the processor executes the program instructions to perform the steps of the method for determining device operating parameters. The specific implementation methods and technical effects are similar and will not be repeated here.

[0117] An embodiment of the present application provides a possible implementation example of a computer-readable storage medium, which can execute the device operating parameter determination method provided in the above embodiment. A computer program is stored on the storage medium, and when the computer program is executed by a processor, the steps of the above device operating parameter determination method are executed.

[0118] A computer program stored in a storage medium may include instructions for causing a computer device (such as a personal computer, server, or network device) or a processor to execute some of the steps of the methods of various embodiments of the present invention. Such storage media include various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0119] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be through some interface, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0120] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0121] In addition, the functional units in various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or hardware plus software functional units.

[0122] The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) or a processor (English: processor) to perform some steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: a USB flash drive, a mobile hard disk, a read-only memory (English: Read-Only Memory, abbreviated: ROM), a random access memory (English: Random Access Memory, abbreviated: RAM), a magnetic disk or an optical disk, and other media that can store program code.

[0123] The above are only specific embodiments of the present application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A method for determining equipment operating parameters, characterized in that: The method comprises: Obtaining a first set of operating parameters for a production device, and a plurality of first production result parameters corresponding to the first set of operating parameters within a preset historical time period; wherein the first production result parameters are digital descriptions of specific production results after the production device is operated according to the first set of operating parameters; wherein the production device is a wafer production device; the same production device is used to produce multiple products, the first set of operating parameters includes: a first set of operating parameters corresponding to each product, and the first production result parameters include: a first production result parameter corresponding to each product; Obtaining evaluation scores for the plurality of first production result parameters; A supervised learning data model is constructed using parameter rules of the production equipment, wherein the parameter rules include: parameter ranges corresponding to multiple types of operating parameters of the production equipment; generating a three-dimensional graph based on the first set of operating parameters, the plurality of first production result parameters, and the equipment operating time corresponding to the first set of operating parameters; the first set of operating parameters includes: a plurality of types of operating parameters, and the three-dimensional graph includes: a three-dimensional graph for the plurality of types of operating parameters; performing denoising processing on the three-dimensional graphics; Based on the first set of operating parameters, the production result parameters of the production result dimension in the three-dimensional graphics after denoising, and the evaluation score, the preset supervised learning data model is used to obtain a second set of operating parameters for the production equipment, so as to enable the production equipment to operate with the second set of operating parameters.

2. The method according to claim 1, wherein The denoising process on the three-dimensional graphics includes: performing wavelet decomposition on the three-dimensional graphics to obtain a multi-layer wavelet signal; performing threshold processing on the multi-layer wavelet signal according to a preset wavelet coefficient threshold to obtain a denoised multi-layer wavelet signal; The denoised three-dimensional graphics are obtained according to the denoised multi-layer wavelet signal.

3. The method according to claim 2, wherein The step of performing wavelet decomposition on the three-dimensional graphics to obtain a multi-layer wavelet signal includes: The three-dimensional graphics are subjected to wavelet grading using a preset orthogonal wavelet basis and a preset number of decomposition layers to obtain multi-layer wavelet signals whose number is equal to the preset number of decomposition layers, where the preset number of decomposition layers is four.

4. The method according to claim 2, wherein The step of performing threshold processing on the multi-layer wavelet signal according to a preset wavelet coefficient threshold to obtain a denoised multi-layer wavelet signal comprises: According to the preset wavelet coefficient threshold, the wavelet coefficients of the first wavelet signal in the multi-layer wavelet signal that are greater than the preset wavelet coefficient threshold are retained, and the wavelet coefficients of the second wavelet signal in the multi-layer wavelet signal that are less than the preset wavelet coefficient threshold are set to zero.

5. The method according to claim 2, wherein The step of performing threshold processing on the multi-layer wavelet signal according to a preset wavelet coefficient threshold to obtain a denoised multi-layer wavelet signal comprises: According to the preset wavelet coefficient threshold, the wavelet coefficients of the first wavelet signal in the multi-layer wavelet signal that are greater than the preset wavelet coefficient threshold are shrunk toward zero, and the wavelet coefficients of the second wavelet signal in the multi-layer wavelet signal that are less than the preset wavelet coefficient threshold are set to zero.

6. A device for determining equipment operating parameters, characterized in that: include: Acquisition module, processing module; The acquisition module is configured to acquire a first set of operating parameters for a production device, and a plurality of first production result parameters corresponding to the first set of operating parameters within a preset historical time period; wherein the first production result parameters are digital descriptions of specific production results after the production device is operated according to the first set of operating parameters; wherein the production device is a wafer production device; the same production device is used to produce multiple products, the first set of operating parameters includes: a first set of operating parameters corresponding to each product, and the first production result parameters include: a first production result parameter corresponding to each product; and obtain evaluation scores for the plurality of first production result parameters; The processing module is used to construct a supervised learning data model using the parameter rules of the production equipment, wherein the parameter rules include: parameter ranges corresponding to multiple types of operating parameters of the production equipment; generating a three-dimensional graph based on the first group of operating parameters, the multiple first production result parameters, and the equipment operating time corresponding to the first group of operating parameters; the first group of operating parameters includes: multiple types of operating parameters, and the three-dimensional graph includes: three-dimensional graphs for the multiple types of operating parameters; denoising the three-dimensional graph; and obtaining a second group of operating parameters of the production equipment using the preset supervised learning data model based on the first group of operating parameters, the production result parameters of the production result dimension in the denoised three-dimensional graph, and the evaluation score, so as to enable the production equipment to operate with the second group of operating parameters.

7. An electronic device, characterized in that: include: A processor, a storage medium and a bus, wherein the storage medium stores program instructions executable by the processor. When the electronic device is running, the processor and the storage medium communicate via the bus, and the processor executes the program instructions to perform the steps of the device operating parameter determination method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that The storage medium stores a computer program, which, when executed by a processor, executes the steps of the method for determining device operating parameters according to any one of claims 1 to 5.

Images