Process parameter optimization method and system

Abstract

The disclosed process parameter optimization method and system select key parameters to optimize the process through complex system response functions, and also predict the effects of process optimization. It greatly reduces the number of experiments caused by the selection of factors and the levels of corresponding factors in previous experimental methods. It can also accurately predict the results with a small number of experimental groups. Through the process parameter optimization method and system disclosed in the present disclosure, the number of experiments can be reduced to save time and cost, so as to quickly find the best solution.

Description

[Technical Field]

[0001] This invention relates to a parameter optimization method and system, and more particularly to a process parameter optimization method and system. [Previous Technology]

[0002] With the increasing demand for optoelectronic semiconductor components, the demand for transparent conductive oxide (TCO) films and flexible printed circuits (FPCs) is also growing daily. Taking the TCO process as an example, coating can be performed by combining reactive jetting with lasers. Its main advantage is that it allows for the fabrication of circuit patterns for functional materials on flexible and transparent substrates using an additive process, while also reducing the cost required for other vacuum environment processes.

[0003] However, since reactive jet laser coating technology is carried out in an atmospheric environment, it is more susceptible to complex environmental factors, such as ambient temperature and humidity, which can affect the process results. Therefore, an efficient method that can adapt to complex systems is needed to predict and adjust process parameters, optimize and ensure process stability.

[0004] In known technologies, common process stability optimization methods all require a large number of parameters and experiments to confirm the reliability and reproducibility of the parameters. However, a large number of experiments not only waste materials but also take a long time. In view of this, how to establish optimized process parameters and predict their effects without conducting a large number of experiments is an important issue. [Summary of the Invention]

[0005] In view of this, this disclosure proposes a process parameter optimization method and system, which achieves rapid parameter optimization through Complex System Response (CSR) and adjusts the parameters based on the optimization results. Simultaneously, the CSR method also establishes a physical model corresponding to the complex system, which is used to correspond to actual process conditions and analyze the interaction effects between process parameters.

[0006] This disclosure provides a process parameter optimization method, including: inputting multiple process parameters into a process parameter optimization model and obtaining a first detection value; fitting multiple process parameters according to a first complex system response (CSR); and obtaining multiple key parameters and inputting the multiple key parameters into the process parameter optimization model to obtain a second detection value, wherein the second detection value is better than the first detection value.

[0007] This disclosure also provides a process parameter optimization system, including: a storage device and a processor. The storage device stores multiple models. The processor is coupled to the storage device and configured to: input multiple process parameters into the process parameter optimization model and obtain a first detection value; fit the multiple process parameters according to a first complex system response (CSR); and obtain multiple key parameters and input the multiple key parameters into the process parameter optimization model to obtain a second detection value, wherein the second detection value is better than the first detection value.

[0008] Based on the above, the process parameter optimization method and system disclosed herein selects key parameters through complex system response functions to optimize the process, and also predicts the effect after process optimization. This significantly reduces the large number of experiments required by previous experimental methods due to the selection of factors and their corresponding levels, and allows for accurate prediction of results with a small number of experimental groups. Through the process parameter optimization method and system disclosed herein, the number of experiments can be reduced to save time and costs, and the optimal solution can be found quickly.

Implementation Method

[0009] Examples of the exemplary embodiments disclosed herein will now be described in detail with reference to the accompanying drawings. The terms "first," "second," etc., used throughout this specification (including the claims) are used to name elements or distinguish different embodiments or scopes, and are not intended to limit the upper or lower limit of the number of elements, nor to limit the order of elements. Furthermore, wherever possible, elements / components using the same reference numerals in the drawings and embodiments represent the same or similar parts.

[0010] Figure 1 is a schematic flowchart of a process parameter optimization method provided in this disclosure. In process S110, multiple process parameters are input into the process parameter optimization model and a first detection value is obtained. In process S120, multiple process parameters are fitted according to a first complex system response (CSR) function. In process S130, multiple key parameters are obtained and input into the process parameter optimization model to obtain a second detection value, wherein the second detection value is better than the first detection value.

[0011] Figure 2 is a schematic diagram of a process parameter optimization system provided in this disclosure. The process parameter optimization system 200 may include a processor 210 and a storage device 220.

[0012] In the embodiments disclosed herein, the processor is, for example, a central processing unit (CPU), or other programmable general-purpose or special-purpose microcontrollers (MCUs), microprocessors, digital signal processors (DSPs), programmable controllers, application-specific integrated circuits (ASICs), graphics processing units (GPUs), image signal processors (ISPs), image processing units (IPUs), arithmetic logic units (ALUs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), or other similar elements or combinations thereof. In the process parameter optimization system 200, processor 210 is coupled to memory 220, and processor 210 can execute various models stored in memory 220. It should be understood that all processes shown in FIG. 1 of this disclosure can be executed by processor 210.

[0013] The storage device 220 is, for example, any type of fixed or removable random access memory (RAM), read-only memory (ROM), flash memory, hard disk drive (HDD), solid state drive (SSD), or similar element or combination thereof, and is used to store multiple models or various applications that can be executed by the processor 210. In this embodiment, the storage device 220 may store at least a process parameter optimization model 221.

[0014] Please refer to Figure 3, which is a schematic diagram of the CSR optimization process. In process 310, the optimization process begins. In process 320, the selected factors and corresponding levels are set. In process 330, an experimental design table is established using controllable parameters. In process 340, the experimental response (result) is fitted using CSR. In process 350, the physical meaning is predicted based on the optimization results. If the results in process 350 do not conform to the physical meaning, process 360 is executed to redo the experimental setup based on the initial optimization parameters. After executing process 360, it is necessary to return to process 330 and re-execute processes 330 to 350. If the results in process 350 conform to the physical meaning, process 370 is executed to obtain the current optimization parameters for the complex system. Finally, the optimization process in process 380 ends.

[0015] In the various embodiments disclosed herein, the processor 210 uses design of experiments and complex system response (CSR) methods to optimize the manufacturing process. The following description uses laser-assisted atmospheric plasma coating (LAPPJ) as an example to illustrate process optimization. However, the process parameter optimization processes applicable to this disclosure are not limited to the aforementioned embodiments; any process that can optimize process parameters using the methods disclosed herein falls within the scope of this disclosure.

[0016] In the embodiments disclosed herein, since the deposition of transparent conductive films via reactive jet combined with laser needs to be carried out in an atmospheric environment, complex environmental factors can also affect the process results. Therefore, in this disclosure, multiple environmental factors that may affect the process results can be input into the process parameter optimization model, and key parameters can be adjusted according to the environmental factors to obtain optimized process parameters. In this way, environmental factors can be taken into account in the process without the need for deliberate control of environmental conditions.

[0017] In the embodiments disclosed herein, the Complex Response Surface Method (CSR) function used by the processor 210 can be programmed and provides the minimum number of tests required to optimize the finished product properties. It includes seven forms, each corresponding to a minimum number of experiments. It should be understood that the minimum number of experiments is equivalent to the number of function coefficients. As shown in Table 1, it records the characteristic of adaptively adjusting the number of experiments based on data characteristics during the manufacturing process, and also records the minimum number of experiments corresponding to the selected CSR function. Table 1: Relationship between CSR Function and Number of Experiments form CSR function Minimum number of experiments 1 2 2P+1 3 4 5 P+1 6 P+1 7

[0018] In Table 1, represents the response value of the characteristic of interest. In the embodiments disclosed herein, this is the resistance value, which affects the quality of the coating. represents the process parameter. represents the unknown parameter generated in the complex system. represents the linear term, indicating the response and influence of the process result resistance value on the defined process parameter. represents the quadratic term of . represents the quadratic coefficient, indicating the response of the resistance value to . represents the interaction coefficient, representing the interaction response between the resistance value result and . In the fitting function, represents the level of the parameter, which is an independent variable. The coefficients of the function contain multidimensional response information of the system to the parameters and levels, describing the relationship between the parameter combination and the characteristic of interest. This quantitative relationship can be fitted using the functions listed in Forms 1 to 7 of Table 1.

[0019] In the embodiments disclosed herein, after obtaining experimental values ​​based on the designed experiment, the response relationship between the parameters and the target value can be obtained through fitting the CSR function in Table 1 above. That is, when the processor 210 receives multiple input process parameters, the processor 210 can select one form of CSR function for fitting to choose the most suitable form. After obtaining the response relationship, the relationship between the parameters and the target value response can be presented in the form of a surface plot, thereby obtaining the expected results.

[0020] In the embodiments disclosed herein, the processor 210 may first evaluate the features of interest using Form 1 of the CSR function in Table 1. Since Form 1 includes linear, quadratic, and interaction terms, the processor 210 can initially determine the significance of each feature to the response value by examining the coefficients of each term obtained from Form 1, thereby further selecting other forms in Table 1 for subsequent optimization. For example, if the processor 210 determines that the linear term in the function has no significant impact on the response value, the processor 210 may select Form 4 of the CSR function in Table 1, which does not include linear terms, for subsequent parameter optimization.

[0021] When conducting experiments using previous techniques such as full factorial experimental design, with P adjustable parameters, each parameter has M levels. Therefore, if an exhaustive search is used to find the optimal parameters, MP experimental groups are required to obtain the optimized parameters. However, when conducting experiments using the CSR function disclosed herein, taking Form 1 as an example, the required number of experimental groups is [not specified]. Table 2 below shows a comparison of the number of experimental groups required using the CSR function and the full factorial experimental design, assuming that the number of levels M is 10. Table 2: Comparison of the number of experimental groups required by the CSR function and the full factorial experimental design Number of parameters (P) CSR function() Full factorial experiment (M) P ) 1 3 10 2 6 10 2 3 10 10 3 4 15 10 4 5 twenty one 10 5 6 28 10 6 12 91 10 12

[0022] As shown in Table 2, regardless of the number of parameters, the number of experimental groups required when using the CSR function for parameter optimization is less than the number of groups required for a full factorial experiment.

[0023] In the embodiments disclosed herein, taking the Laser-assisted Atmospheric Plasma Jet (LAPPJ) process as an example, since the process environment is in the atmosphere, it is easily affected by uncontrollable or unknown factors such as temperature and humidity, resulting in significant differences in the resistance value of each manufactured product. The results of the embodiments disclosed herein demonstrate that incorporating CSR into this process can efficiently and quickly obtain the current process equipment characteristics, thereby accurately predicting process results and optimizing parameters.

[0024] As shown in the preceding section, in the embodiments disclosed herein, the processor 210 can select three main factors affecting the process resistance value, namely, the coating speed, the number of coating passes, and the laser power. The parameters used in the experiment and their corresponding responses are shown in Table 3. Table 3: Laser-Assisted Atmospheric Plasma Coating Process Parameter Table Experimental rounds Laser power (Factor 1) Coating speed (Factor 2) Number of coating cycles (Factor 3) resistance value (Response value) 1 14 150 4 176.02 2 14 150 6 115.62 3 14 250 4 533 4 14 250 6 226.44 5 16 150 4 218.32 6 16 150 6 124.33 7 16 250 4 461 8 16 250 6 223.23 9 14 150 4 176.16 10 14 200 5 312 11 14 250 6 229.48 12 15 150 5 133.28 13 15 200 6 171.56 14 15 250 4 475 15 16 150 6 117.7 16 16 200 4 336.6 17 16 250 5 280.12

[0025] The experimental design uses the Orthogonal Array Composite Design (OACD) method, setting each factor (parameter) to correspond to three levels to obtain a global paired combination. The experimental design presents 9 orthogonal arrays plus 8 partial factorial experiments, resulting in 17 experimental groups. However, if a full factorial experiment is used, 27 groups (3+3 groups) are required to obtain the optimized parameters. Thus, the experimental design disclosed herein reduces the number of experimental groups by 10 groups (approximately 37%). If different forms of CSR functions (second complex system response functions) are selected for fitting using Table 1 to analyze the significance (P-value) of the factors, the number of experimental groups can be further reduced. In this embodiment, the minimum number of experimental groups can be reduced to 11, while still maintaining predictive accuracy and stability.

[0026] Continuing from the previous section, the formulas for the aforementioned partial factorial experiments are as follows:

[0027] In the formula for the above partial factorial experiments, N is the required number of experiments, k is the number of factors, and p is the experimental proportion. In the embodiments disclosed herein, a three-factor, two-water full factorial experimental design is used (100% experimental proportion, k=3, p=0), therefore, an additional 8 groups (2+3) of partial factorial experiments are required.

[0028] Continuing from the previous section, the processor 210 reduces the number of experimental groups by using significance analysis to reduce the number of CSR optimization search groups. The significance analysis uses P-value as the degree of rejection of the null hypothesis. By defining the null hypothesis (H0) as the assumption that there is no influence between the factor and the experimental response value (resistance value), the significance analysis yields two possible results: a single factor P-value < 0.05 and a P-value ≥ 0.05. When a single factor P-value < 0.05, it indicates that the factor is significant; when a single factor P-value ≥ 0.05, it indicates that the factor is insignificant in the current experiment. In this embodiment, to reduce the number of experimental groups while ensuring the robustness of the experimental design, the processor 210 retains all experimental groups for significant factors, removes the medium level values ​​from insignificant factors, and retains only the high level and low level experimental groups.

[0029] In this embodiment, in order to verify the impact of reducing the number of experimental groups on the optimization results, the processor 210 reduces the number of experimental groups that have been completed, and the processor 210 verifies the impact of reducing the number of groups on the optimization parameters and prediction results.

[0030] In this embodiment, the processor 210 can first calculate the P-value corresponding to the factor of each group of experiments, and obtain the laser power as the least significant factor among the three factors from the P-value, as shown in Table 4 below: Table 4: P-value greater than or equal to 0.05 indicates an insignificant factor factor parameter P-Value Factor 1 Laser power (%) *0.4263 Factor 2 Processing speed (mm / min) 0.0036 Factor 3 Processing time (minutes) 0.0004

[0031] As shown in Table 4, the P-Value of the laser power is greater than 0.05. Therefore, the processor 210 removes the mid-level laser power in the experimental design. Testing shows that removing the number of duplicate groups and the number of mid-level groups by the processor 210 does not affect the prediction model, and the optimization parameters remain unchanged, as shown in Table 5 below: Table 5: Prediction Results After Removing Duplicate Groups and Mid-Level Groups Number of execution groups Factor 1 Factor 2 Factor 3 Predicting extreme outcomes 17 13.75 175.00 6.00 138 14 13.60 175.00 6.00 138 11 13.57 175.00 6.00 141 9 14.00 192.47 6.00 125

[0032] As shown in Table 5, the original experimental design had 17 groups. Processor 210 obtained 14 experimental groups in Table 5 after removing duplicate experimental groups, 11 experimental groups in Table 5 after removing the number of intermediate-level experimental groups for insignificant factors, and 9 experimental groups in Table 5 after removing the number of high- and low-level experimental groups for insignificant factors. From the extreme prediction results in Table 5, it can be seen that when processor 210 removes the number of intermediate-level experimental groups for insignificant factors, it can reduce the number of experimental groups while maintaining the accuracy of model prediction and optimization.

[0033] Please refer to Figure 4, which is a scatter plot of the CSR experimental results of the first embodiment of this disclosure, showing the CSR experimental results of LAPPJ. In Figure 4, the horizontal axis represents two experiments conducted at four different time periods, namely the first experiment with a larger parameter search range, and the second experiment with a more refined parameter optimization search by appropriately narrowing the parameter search range based on the CSR model prediction results established based on the first experiment data. Therefore, there are eight sets of data. The vertical axis represents the resistance value (Ω), and the black hollow dots represent the experimental resistance values ​​obtained through actual experiments using the 17-set OACD experimental design method, which are used for parameter optimization through CSR. The meaning of these black hollow dots is that the original data is used to construct the quadratic surface model corresponding to the current experiment. The black solid dots are the resistance values ​​applied to the process using the optimized parameters, while the hollow square dots are the resistance values ​​predicted by this disclosure using the optimized parameters.

[0034] Please refer to Figure 4. The error between the predicted resistance value and the actual resistance value is only 4%, and the correlation coefficient of the response model fitted by CSR is as high as 0.98, confirming that CSR technology can be applied to the optimization and prediction of laser process parameters in atmospheric environments. In other words, by using the optimized parameters disclosed in this disclosure in the process, users can be provided with accurate expected results.

[0035] In the embodiments disclosed herein, as can be seen from Table 1, there may be interactions between different parameters. Therefore, this disclosure further presents the response values ​​between each factor in a quadratic surface manner in this embodiment. Taking the embodiment of this disclosure as an example, the three factors can interact with each other, resulting in three graphs. Please refer to Figure 5, which is a parameter response surface diagram of the first embodiment of this disclosure. In Figure 5, the CSR surfaces between different factors are presented respectively. In Figures 5 (A), (B), and (C), the hollow point P represents the predicted response value, and through this predicted response value, the interaction between each factor and its impact on the process results can be determined.

[0036] In the embodiments disclosed herein, the fitting equation used for CSR is the quadratic response surface equation in Table 1. Under the condition of three factors and three levels, the equation has a total of 10 coefficients, as shown in the following equation (1), where C1 represents factor one, C2 represents factor two, C3 represents factor three, and the response is the resistance value of interest in this case. … (1)

[0037] In the second embodiment disclosed herein, taking Aerosol Jet Printing and Laser Direct Synthesis Patterning (AJP-LDSP) as an example, the processing parameters are adjusted using a complex system response function combined with a multi-objective optimization algorithm to obtain the lowest resistance value, linewidth difference, and processing time. Since the process is easily affected by uncontrollable factors such as temperature and humidity, the aerosol atomization and deposition conditions differ in each process, resulting in significant differences in resistance value and linewidth. In this second embodiment, the three main factors are carrier gas flow rate, laser power, and processing rate. The parameters used in the experiment and their corresponding responses are shown in Table 6. Table 6: Parameter Table for Aerosol Jet Printing and Laser Direct Synthesis Patterning Process Experimental rounds Carrier gas flow rate (Factor 1) Laser power (Factor 2) Processing rate (Factor 3) resistance value (Response value 1) Line width difference (Response value 2) 1 50 6.5 60 1.60 87.46 2 50 6.5 120 3.62 95.90 3 50 7.5 60 1.77 82.68 4 50 7.5 120 3.06 87.96 5 100 6.5 60 0.36 6.61 6 100 6.5 120 1.08 37.74 7 100 7.5 60 0.38 18.05 8 100 7.5 120 0.84 35.08 9 50 6.5 60 1.74 90.90 10 50 7 90 2.52 95.21 11 50 7.5 120 2.80 93.25 12 75 6.5 90 1.14 71.76 13 75 7 120 1.15 73.12 14 75 7.5 60 0.40 46.55 15 100 6.5 120 0.98 35.76 16 100 7 60 0.37 13.49 17 100 7.5 90 0.63 31.73

[0038] The second embodiment disclosed herein uses the same experimental design as the first embodiment, namely, the orthogonal array composite design experimental design method, setting each factor (parameter) to correspond to three levels to obtain a full-domain paired combination. The experimental design presents 9 orthogonal arrays plus 8 partial factorial experiments, resulting in 17 experimental groups. However, if a full factorial experiment is used, 27 groups (3+3 groups) of experiments are required to obtain the optimized parameters. This reduces the number of experimental groups by 10 groups (approximately 37%). If different forms of CSR functions (second complex system response functions) are selected for fitting through Table 1 to analyze the significance (P-value) of the factors, the number of experimental groups can be further reduced. The CSR response function used in this embodiment only requires 17 experimental groups, and the correlation coefficients R² of the fitted resistance value and linewidth difference model accuracy can reach 0.992 and 0.993, respectively.

[0039] Please refer to Figures 6 and 7 simultaneously. Figure 6 is a parameter first response surface plot of the second embodiment of this disclosure, and Figure 7 is a parameter second response surface plot of the second embodiment of this disclosure. In Figure 6, the CSR surfaces between different factors are presented to obtain the influence of the response value on the resistance value. In Figures 6(A), (B), and (C), the hollow point P represents the predicted response value, and through this predicted response value, the interaction between the factors and the response value of the resistance group value of the process result can be obtained. In Figure 7, the CSR surfaces between different factors are presented to obtain the influence of the response value on the linewidth difference. In Figures 7(A), (B), and (C), the hollow point P represents the predicted response value, and through this predicted response value, the interaction between the factors and the response value of the linewidth difference of the process result can be obtained.

[0040] In the second embodiment disclosed herein, the fitting equation used by CSR is the quadratic response surface equation in Table 1. Under the condition of three factors and three levels, the equation has a total of 10 coefficients, as shown in equations (2) and (3) below, where C1 represents factor one, C2 represents factor two, C3 represents factor three, response 1 is the resistance value, and response 2 is the linewidth difference. … (2)… (3)

[0041] In the second embodiment of this disclosure, the aforementioned three factors (carrier gas flow rate, laser power, and processing rate) are used as inputs, and the resistance value, linewidth difference, and process time corresponding to different processing rates predicted by the aforementioned CSR model are used as outputs. Using a multi-objective optimization algorithm, the optimal solution set for the multi-objectives, namely the Pareto front, and its corresponding process parameter combination, can be calculated. Through the Pareto front, the process parameter combination with the shortest process time under the condition of meeting the finished product specifications can be selected. Please refer to Figure 8, which is a Pareto front diagram of the second embodiment of this disclosure. In Figure 8, solid black dots represent solutions that do not meet the conditions, hollow black dots represent solutions that meet the conditions, surface A represents the linewidth difference condition, and curve B represents the resistance condition. The predicted values ​​and experimental values ​​of the optimal solutions are compared as shown in Table 5. As shown in Table 7, the experimental results and predicted results are almost identical and have a high correlation. Table 7: Comparison of the optimal solutions and experimental values ​​of the second embodiment. Factor 1 Factor 2 Factor 3 Response 1 Response 2 time Predicted value 99.9 6.6 60 0.3 10.7 39.9 Experimental values 100 6.6 60 0.4 7.4 40

[0042] It should be understood that in the embodiments disclosed herein, after each process parameter optimization model, the processor 210 can perform a detection to obtain a detection value, which is related to the prediction accuracy and stability.

[0043] In the various embodiments disclosed herein, the processor 210 can plan the number of experiments by comparing the forms of the CSR functions in the selected Table 1. The processor 210 can, for example, determine which form to use for planning the number of experiments by analyzing whether the significance of the factor is less than 0.05.

[0044] In the embodiments disclosed herein, after obtaining the optimized key parameters using the CSR function, the processor 210 further predicts whether the process result obtained using the optimized parameters conforms to physical meaning. If it conforms to physical meaning, the processor 210 outputs the key parameters for use in subsequent processes. Conforming to physical meaning includes, for example, a resistance value greater than zero. In other experiments, conforming to physical meaning may, for example, result in a positive quality value for the produced product.

[0045] In summary, the process parameter optimization method and system disclosed herein selects key parameters through complex system response functions to optimize the process, and also predicts the effects after process optimization. This significantly reduces the large number of experiments required by previous experimental methods due to the selection of factors and their corresponding levels, and allows for accurate prediction of results with a small number of experimental groups. Through the process parameter optimization method and system disclosed herein, the number of experiments can be reduced to save time and costs, enabling the rapid finding of the optimal solution. [Simplified Explanation of the Diagram]

[0046] Figure 1 is a flowchart illustrating a process parameter optimization method disclosed herein. Figure 2 is a schematic diagram illustrating a process parameter optimization system disclosed herein. Figure 3 is a flowchart illustrating CSR optimization. Figure 4 is a scatter plot of CSR experimental results of the first embodiment disclosed herein. Figure 5 is a parameter response surface plot of the first embodiment disclosed herein. Figure 6 is a parameter first response surface plot of the second embodiment disclosed herein. Figure 7 is a parameter second response surface plot of the second embodiment disclosed herein. Figure 8 is a Pareto leading edge plot of the second embodiment disclosed herein.

Claims

1. A method for optimizing process parameters, comprising: Input multiple process parameters into the process parameter optimization model and obtain the first detection value; The process parameters are fitted according to the first complex system response (CSR) function, and the number of experiments is planned according to the selected first complex system response function; and multiple key parameters are obtained and input into the process parameter optimization model to obtain a second detection value, wherein the second detection value is better than the first detection value.

2. The process parameter optimization method as described in claim 1 further includes: Input multiple environmental factors into the process parameter optimization model; And adjust these key parameters based on these environmental factors.

3. The process parameter optimization method as described in claim 1 further includes: These process parameters are fitted based on the response function of the second complex system; And obtain these key parameters.

4. The process parameter optimization method as described in claim 3 further includes: The number of experiments is planned based on the selected response function of the second complex system.

5. The process parameter optimization method as described in claim 1 further includes: In response to the second predicted value conforming to physical meaning, these key parameters are output.

6. The process parameter optimization method as described in claim 5, wherein the physical meaning includes a resistance value greater than zero.

7. The process parameter optimization method as described in claim 1 further includes: The Orthogonal Array Composite Design (OACD) is used to obtain a global pairing combination based on these parameters and level combinations, wherein the global pairing combination is related to these key parameters.

8. The process parameter optimization method as described in claim 1 further includes: A number of experiments is obtained based on the response function of the first complex system, wherein the number of experiments is related to the process parameters.

9. The process parameter optimization method as described in claim 1, wherein the process parameters include coating speed, coating times, and laser power.

10. A process parameter optimization system, comprising: Storage, storing multiple models; The processor, coupled to the storage, and configured to: input multiple process parameters into a process parameter optimization model in the models and obtain a first detection value; fit the process parameters according to a first complex system response (CSR) function and plan the number of experiments according to the selected first complex system response function; and obtain multiple key parameters and input the key parameters into the process parameter optimization model to obtain a second detection value, wherein the second detection value is better than the first detection value.

11. The process parameter optimization system as claimed in claim 10, wherein the processor is further configured to: input multiple environmental factors into the process parameter optimization model; and adjust the key parameters based on the environmental factors.

12. The process parameter optimization system as described in claim 10, wherein the processor is further configured to: fit the process parameters according to a second complex system response function; and obtain the key parameters.

13. The process parameter optimization system as described in claim 12, wherein the processor is further configured to: plan the number of experiments based on the selected second complex system response function.

14. The process parameter optimization system as described in claim 10, wherein the processor is further configured to: output the key parameters in response to the second predicted value conforming to physical meaning.

15. The process parameter optimization system as described in claim 14, wherein the physical meaning includes a resistance value greater than zero.

16. The process parameter optimization system as described in claim 10, wherein the processor is further configured to: obtain a global pairing combination based on the parameters and level combination according to Orthogonal Array Composite Design (OACD), wherein the global pairing combination is related to the key parameters.

17. The process parameter optimization system as claimed in claim 10, wherein the processor is further configured to: obtain a number of experiments based on the first complex system response function, wherein the number of experiments is related to the process parameters.

18. The process parameter optimization system as described in claim 10, wherein the process parameters include coating speed, coating times, and laser power.

Images