Photoetching model correction method, equipment, medium and product
By introducing feature parameters and weight coefficients of the initial lithography model into the genetic algorithm, the problem of lithography model degradation after feature adjustment is solved, thereby improving the quality and efficiency of model correction and ensuring stability and fast convergence.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-05
- Publication Date
- 2026-03-10
AI Technical Summary
After adjusting the number of feature terms, existing lithography models are prone to unpredictable degradation, leading to a decrease in correction quality and efficiency.
A genetic algorithm is used to correct the lithography model to be corrected. By introducing the feature parameter values and weight coefficient values of the initial lithography model into the initial population of the genetic algorithm, and combining the global exploration capability, the adjusted model is optimized.
It improves the correction quality and efficiency of the lithography model after adjusting the number of feature terms, maintains the stability and consistency of the model, avoids degradation, and significantly accelerates the convergence process of the genetic algorithm.
Smart Images
Figure CN121634728A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of semiconductor technology, and in particular to a method, apparatus, medium, and product for correcting photolithography models. Background Technology
[0002] A lithography model is a mathematical model that describes the actual lithography process using abstract formulas. It is used to simulate and predict the lithographic imaging results of a designed pattern, thereby assisting in the correction of the design pattern. A lithography model contains multiple terminology. These terminology exist in functional form, representing a type of physical or chemical effect in the lithography process. Each terminology has a parameter part and a weighting coefficient part. Correcting the lithography model is the process of determining the optimal parameter values and weighting coefficient values for the terminology.
[0003] In actual model building, it is often necessary to adjust the number of feature terms in the lithography model to balance model accuracy and computational resource consumption. However, after adjusting the number of feature terms in the lithography model, the recalibrated model is prone to unpredictable degradation. Summary of the Invention
[0004] One object of the present invention is to provide a correction method, apparatus, medium and product for a lithographic model that helps to improve the correction quality and efficiency of the model after feature adjustment.
[0005] Specifically, the present invention provides a method for correcting a photolithography model, comprising: An initial lithography model is obtained, which is constructed based on a first feature group. The initial lithography model has been preliminarily corrected for each feature in the first feature group and the corresponding initial parameter values and initial weight coefficient values have been determined. A lithography model to be corrected is constructed based on a second feature group. The second feature group is obtained from the first feature group through a first type of operation or a second type of operation. The first type of operation is to add at least one new feature to the first feature group, and the second type of operation is to delete at least one original feature from the first feature group. A genetic algorithm is used to correct the lithography model to be corrected, resulting in a corrected lithography model in which the final parameter value and the final weight coefficient value are determined for each feature item of the lithography model to be corrected. In the initial population generated by the genetic algorithm, each feature item of an individual belonging to the first feature item group is assigned the initial parameter value and the initial weight coefficient value.
[0006] Optionally, the step of using a genetic algorithm to correct the lithography model to be corrected includes: Obtain the initial population generated by the genetic algorithm, wherein the initial population includes multiple initial random individuals generated by assigning random parameter values to each feature item in the lithography model to be corrected and setting preset weight coefficient values; Select one of the initial random individuals as the replacement individual, and replace the parameter value and weight coefficient value of each feature item belonging to the first feature item group of the replacement individual with the initial parameter value and initial weight coefficient value of the same feature item.
[0007] Optionally, the step of using a genetic algorithm to correct the lithography model to be corrected further includes: When the second feature group is obtained from the first feature group through the first operation type, the weight coefficient values of all the newly added feature items of the replacement individual are replaced with 0, and the parameter values are replaced with the middle value of the corresponding parameter range.
[0008] Optionally, the step of obtaining the initial photolithography model is followed by: By using different sets of second feature terms, the step of constructing a lithographic model to be corrected based on the second feature term set is performed respectively, resulting in multiple lithographic models to be corrected. The multiple sets of second feature terms are obtained by the first feature term set through the same type of operation. The step of using a genetic algorithm to correct the lithography models to be corrected is performed on each of the multiple lithography models to be corrected, resulting in multiple corrected lithography models.
[0009] Optionally, after obtaining multiple calibration lithography models, one of the multiple calibration lithography models is selected as the final calibration lithography model for this calibration.
[0010] Optionally, the step of selecting one of the plurality of calibration lithography models as the final output of this calibration includes: Determine whether the final output corrected lithography model meets the preset requirements. If not, use the final output corrected lithography model as the new initial lithography model and continue the correction.
[0011] Optionally, the step of selecting one of the plurality of calibration lithography models as the final output of this calibration includes: Select all calibration lithography models whose preset evaluation indicators meet preset standards from the plurality of calibration lithography models; The calibration lithography model with the lowest root mean square error is selected from all calibration lithography models that meet the preset criteria as the final calibration lithography model output for this calibration.
[0012] The photolithography model calibration method of the present invention adds an initial photolithography model to the initial population generated by the genetic algorithm for the photolithography model to be calibrated. This is equivalent to adding a better "gene" (i.e., a combination of parameter values and coefficient values) to the initial population. This helps to effectively guide the genetic algorithm for the photolithography model to be calibrated to the solution space region near the initial photolithography model for optimization. This avoids the optimization of the photolithography model to be calibrated from getting stuck in local optima that deviate significantly from the initial photolithography model after adjusting the number of feature terms of the initial photolithography model.
[0013] Moreover, this scheme ensures that the obtained calibrated lithography model maintains good inheritance and consistency with the initial lithography model in some key evaluation indicators (such as contour morphology, offset uniformity, etc.). It avoids the new model from falling into a local optimum that is completely different from the initial lithography model after adjusting the number of feature terms of the initial lithography model, which would lead to unpredictable deterioration of other evaluation indicators. This is conducive to improving the stability of model iteration and the predictability of result quality.
[0014] Furthermore, by incorporating a superior initial lithography model into the calibration of the lithography model to be calibrated, its superior "genes" can be rapidly transmitted during the iterative optimization process of the genetic algorithm, significantly accelerating the convergence process and improving its convergence speed. Simultaneously, because the initial population still contains a large number of original random individuals, the genetic algorithm retains a strong global exploration capability. Therefore, even when the initial lithography model is not ideal, the genetic algorithm can overcome this inadequacy and find a more suitable solution.
[0015] In particular, for cases where new feature terms are added to the initial lithography model, this scheme helps to effectively ensure that the fitting accuracy of the corrected lithography model does not deteriorate compared to the initial lithography model, effectively avoiding the problem that the increase in parameter space dimension and computational load due to the addition of new feature terms leads to an increase in model complexity but a decrease in accuracy.
[0016] Therefore, this scheme can effectively improve the model correction quality and efficiency of the lithography model after adjusting the number of feature terms.
[0017] The above and other objects, advantages and features of the present invention will become more apparent to those skilled in the art from the following detailed description of specific embodiments of the invention in conjunction with the accompanying drawings. Attached Figure Description
[0018] The following sections will describe some specific embodiments of the invention in detail by way of example and not limitation, with reference to the accompanying drawings. The same reference numerals in the drawings denote the same or similar parts or portions. Those skilled in the art should understand that these drawings are not necessarily drawn to scale. In the drawings: Figure 1 This is a schematic flowchart of a photolithography model correction method according to an embodiment of the present invention; Figure 2 This is a schematic flowchart illustrating the steps of using a genetic algorithm to correct the lithography model to be corrected in a lithography model correction method according to an embodiment of the present invention. Figure 3 This is a schematic flowchart of a photolithography model correction method according to another embodiment of the present invention; Figure 4 This is a schematic flowchart illustrating the step of selecting the final output corrected lithography model in a lithography model correction method according to another embodiment of the present invention; Figure 5 This is a simplified schematic diagram of a model calibration process according to another embodiment of the present invention; Figure 6 This is a schematic diagram of a computer device according to an embodiment of the present invention; Figure 7 This is a schematic diagram of a computer-readable storage medium according to an embodiment of the present invention; Figure 8 This is a schematic diagram of a computer program product according to an embodiment of the present invention. Detailed Implementation
[0019] Those skilled in the art should understand that the embodiments described below are merely a part of the embodiments of the present invention, and not all of the embodiments of the present invention. These partial embodiments are intended to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention. Based on the embodiments provided by the present invention, all other embodiments obtained by those skilled in the art without creative effort should still fall within the scope of protection of the present invention.
[0020] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus or device (such as a computer-based system, a processor-included system or other system that can fetch and execute instructions from, an instruction execution system, apparatus or device).
[0021] The flowchart provided in this embodiment is not intended to indicate that the operations of the method will be performed in any particular order, or that all operations of the method are included in every case. Furthermore, the method may include additional operations. Within the scope of the technical concept provided by the method in this embodiment, additional variations can be made to the above method.
[0022] like Figure 1As shown, in one embodiment, the method for correcting the photolithographic model generally includes: Step S101: Obtain the initial lithography model. The initial lithography model is constructed based on the first feature term group, and the initial lithography model has been preliminarily corrected for each feature term in the first feature term group and the corresponding initial parameter values and initial weight coefficient values have been determined.
[0023] First, let's illustrate the form of a photolithography model with an example. Specifically, a photolithography model can be represented by the following formula: ; Here, RI stands for Resist Image, used to describe the pattern formed on the wafer predicted by the lithography model; AI stands for Aerial Image, used to describe the optical distribution on the wafer; each coeff*term(params) represents a feature term, where coeff represents the weight coefficient of the feature term, term(params) represents the function part of the feature term, and params represents the parameters in the function, i.e., the independent variables of the function. Each feature term is used to describe a type of physical or chemical effect on the wafer. The process of simulating the generation of the photoresist image by superimposing the feature terms onto the spatial image is to describe the actual lithography process using abstract formulas.
[0024] Subsequently, the initial lithography model is a model for each feature term that has been preliminarily corrected and whose specific initial parameter values and initial weight coefficient values have been determined.
[0025] Step S102: Construct the lithography model to be corrected based on the second feature term group. The second feature term group is obtained from the first feature term group through a first type of operation or a second type of operation. The first type of operation is to add at least one new feature term to the first feature term group. The second type of operation is to delete at least one original feature term from the first feature term group.
[0026] Specifically, this involves constructing a lithographic model to be corrected based on the parameter values and weight coefficient values of the feature terms using the second feature term group. Furthermore, the feature terms in the second feature term group must contain at least some of the feature terms from the first feature term group.
[0027] For example, the first feature group includes three feature terms, term1, term2 and term3, and the initial lithography model is constructed based on term1, term2 and term3.
[0028] In one embodiment, the second feature group comprises four feature terms, term1, term2, term3, and term4, including all feature terms, term1, term2, and term3 of the first feature group, and has a newly added feature term4. That is, the second feature group adds a new feature term4 to the first feature group, obtained by performing a first type of operation on the first feature group. The lithography model to be corrected is constructed based on term1, term2, term3, and term4.
[0029] In another embodiment, the second feature group contains two feature terms, term1 and term2, and the original feature term3 is removed from the first feature group. That is, the second feature group is obtained from the first feature group through a second type of operation. The lithographic model to be corrected is constructed based on term1 and term2.
[0030] Step S103: A genetic algorithm is used to calibrate the lithography model to be calibrated, resulting in a calibrated lithography model where the final parameter values and final weight coefficient values are determined for each feature item of the lithography model to be calibrated. In the initial population generated by the genetic algorithm, each feature item belonging to the first feature item group of one individual is assigned an initial parameter value and an initial weight coefficient value.
[0031] Specifically, the genetic algorithm is a commonly used algorithm for finding the optimal parameter values of a formula. In the genetic algorithm's solution process, the parameters in the formula are first randomly assigned values, and the weight coefficients are set to preset values, typically 0, resulting in multiple individuals. Each individual represents a formula with specific parameter and weight values. These individuals constitute the initial population of the genetic algorithm. Then, based on the initial population, iterative processes such as evaluation, selection, crossover, and mutation are performed until the algorithm obtains what it considers the optimal solution. This results in a corrected lithography model where each feature has a determined final parameter and weight value.
[0032] This embodiment does not limit the iterative process of the genetic algorithm for the initial population; the change is that one individual in the initial population of the genetic algorithm is replaced.
[0033] Reference Figure 2 As shown, in one embodiment, this step includes: Step S201: Obtain the initial population generated by the genetic algorithm. The initial population includes multiple initial random individuals generated by assigning random parameter values to each feature item in the lithography model to be corrected and setting preset weight coefficient values.
[0034] Specifically, an initial population is obtained by the genetic algorithm for the lithography model to be calibrated. This initial population consists of multiple initial random individuals, also known as random solution vectors. Each initial random individual is generated by the genetic algorithm to assign random parameter values and preset weight coefficient values to each feature of the lithography model to be calibrated. In other words, each initial random individual is generated by the genetic algorithm to randomly assign values to the parameter portion of each feature of the lithography model to be calibrated and to assign weight coefficient values according to preset weight coefficient values.
[0035] Step S202: Select one of the initial random individuals as the replacement individual, and replace the parameter value and weight coefficient value of each feature item belonging to the first feature item group of the replacement individual with the initial parameter value and initial weight coefficient value of the same feature item.
[0036] For example, referring to the previous example, the lithography model to be calibrated is constructed based on term1, term2, term3, and term4. The genetic algorithm initially assigns random values to the parameter parts of term1, term2, term3, and term4 of the lithography model to be calibrated, resulting in 10 initial random individuals. One of these 10 initial random individuals is used as a replacement individual, and the parameter values and weight coefficients of each feature item belonging to the first feature item group are replaced with the initial parameter values and initial weight coefficients of the same feature item. That is, the parameter values and weight coefficients of term1, term2, and term3 of the replacement individual are replaced with the initial parameter values and initial weight coefficients corresponding to term1, term2, and term3 in the initial lithography model.
[0037] For example, referring to the previous example, the lithography model to be calibrated is constructed based on term1 and term2. The genetic algorithm initially assigns random values to the parameter parts of term1 and term2 of the lithography model to be calibrated, resulting in 10 initial random individuals. One of these 10 initial random individuals is used as a replacement individual, and the parameter values and weight coefficients of each feature item belonging to the first feature item group are replaced with the initial parameter values and initial weight coefficients of the same feature item. That is, the parameter values and weight coefficients of term1 and term2 of the replacement individual are replaced with the initial parameter values and initial weight coefficients corresponding to term1 and term2 in the initial lithography model.
[0038] In this embodiment, an initial lithography model is obtained based on a first feature group, where the initial parameter values and initial weight coefficients for each feature in the first feature group are determined. A lithography model to be calibrated is then constructed based on a second feature group. A genetic algorithm is used to calibrate the lithography model to be calibrated, resulting in a calibrated lithography model where the final parameter values and final weight coefficients for each feature of the calibrated lithography model are determined. Furthermore, each feature belonging to the first feature group in an individual in the initial population generated by the genetic algorithm is assigned an initial parameter value and an initial weight coefficient value.
[0039] In other words, this scheme adds an initial lithography model to the initial population generated by the genetic algorithm for the lithography model to be calibrated. This is equivalent to adding a better "gene" (i.e., a combination of parameter and coefficient values) to the initial population. This helps to effectively guide the genetic algorithm for the lithography model to be calibrated to the solution space region near the initial lithography model for optimization. This avoids the optimization of the lithography model to be calibrated from getting stuck in local optima that deviate significantly from the initial lithography model after adjusting the number of feature terms of the initial lithography model.
[0040] Moreover, the above method ensures that the obtained calibrated lithography model maintains good inheritance and consistency with the initial lithography model in some key evaluation indicators. It avoids the new model from falling into a local optimum that is completely different from the initial lithography model after adjusting the number of feature terms of the initial lithography model, which would lead to unpredictable deterioration of other evaluation indicators (such as contour morphology, offset uniformity, etc.). This is conducive to improving the stability of model iteration and the predictability of result quality.
[0041] Furthermore, by incorporating a superior initial lithography model into the calibration of the lithography model to be calibrated, its superior "genes" can be rapidly transmitted during the iterative optimization process of the genetic algorithm, significantly accelerating the convergence process and improving its convergence speed. Simultaneously, because the initial population still contains a large number of original random individuals, the genetic algorithm retains a strong global exploration capability. Therefore, even when the initial lithography model is not ideal, the genetic algorithm can overcome this inadequacy and find a more suitable solution.
[0042] In particular, for cases where new feature terms are added to the initial lithography model, this method helps to effectively ensure that the fitting accuracy of the corrected lithography model does not deteriorate compared to the initial lithography model, effectively avoiding the problem that the addition of new feature terms leads to a surge in the dimension of the parameter space and the amount of computation, resulting in an increase in model complexity but a decrease in accuracy.
[0043] Therefore, this scheme can effectively improve the model correction quality and efficiency of the lithography model after adjusting the number of feature terms.
[0044] This method incorporates an initial lithography model into the initial population generated by the genetic algorithm for the lithography model to be calibrated, thus constructing a mechanism that combines guidance and global exploration. This allows the genetic algorithm to converge quickly using prior knowledge without being completely bound by the solution of the initial lithography model. It has the potential to escape the suboptimal local solutions that may exist in the initial lithography model and discover better global solutions, thereby achieving an excellent balance between solution speed and the quality of the final solution.
[0045] Furthermore, by using an initial random individual from the initial population as a replacement individual—that is, without changing the original number of individuals in the initial population generated by the genetic algorithm—the stability of the genetic algorithm's population size is maintained. This ensures that the core parameters of the genetic algorithm, such as selection pressure, crossover probability, and computational cost, always conform to the preset logic during the iteration process, avoiding anomalies that may be caused by changes in population size and guaranteeing the stability of the correction process.
[0046] Furthermore, in one embodiment, the step of using a genetic algorithm to correct the lithography model to be corrected further includes: When the second feature group is obtained from the first feature group through the first operation type, the weight coefficient values of all newly added feature items of the replacement individual are replaced with 0, and the parameter values are replaced with the median value of the corresponding parameter range.
[0047] Specifically, for the replacement individuals in the initial population, not only are the parameter values and weight coefficients of each feature item belonging to the first feature item group replaced with the initial parameter values and initial weight coefficients of the same feature item, but the weight coefficients of all newly added feature items are also replaced with 0, and the parameter values are replaced with the median value of the corresponding parameter range. The parameter range is a pre-configured range of values for each feature item.
[0048] Referring to the previous example, the lithography model to be corrected is constructed based on term1, term2, term3, and term4. The parameter values and weight coefficient values of term1, term2, and term3 of the replacement individual are replaced with the initial parameter values and initial weight coefficient values corresponding to term1, term2, and term3 in the initial lithography model. At the same time, the parameter value corresponding to term4 is replaced with the middle value of the corresponding parameter range, and the weight coefficient value is replaced with 0.
[0049] The above configuration ensures that the replacement individual is essentially equivalent to the initial lithographic model, meaning the contribution of the newly added feature terms in the replacement individual is zero. This helps guarantee that the genetic iteration process is triggered from a stable and reliable starting point, avoiding the risk of excessive fluctuations at the beginning of the iteration due to the randomness introduced by the newly added feature terms. Furthermore, a relatively safe configuration is set for the parameter values of the newly added feature terms. Even if the parameter values of the newly added feature terms in the replacement individual are activated subsequently, the setting of the intermediate value of the parameter range helps to ensure the reasonableness of their contribution and improves the stability of the algorithm correction.
[0050] like Figure 3 As shown, in one embodiment, the method for correcting the photolithographic model generally includes: Step S301: Obtain the initial lithography model. The initial lithography model is constructed based on the first feature term group, and the initial lithography model has been preliminarily corrected for each feature term in the first feature term group and the corresponding initial parameter values and initial weight coefficient values have been determined.
[0051] Step S302 involves performing the step of constructing a lithographic model to be corrected based on different sets of second feature terms, resulting in multiple lithographic models to be corrected. Furthermore, the multiple sets of second feature terms are obtained from the first set of feature terms through the same type of operation. That is, the multiple sets of second feature terms are either all obtained from the first set of feature terms through a first type of operation (adding new feature terms), or all obtained from the first set of feature terms through a second type of operation (deleting original feature terms).
[0052] Specifically, when the second feature group is obtained from the first feature group through a first type of operation, multiple different second feature groups are formed by adding different new feature items to the first feature group. In other words, the new feature items added to multiple different second feature groups are different from each other.
[0053] For example, taking two second feature term groups as an example, the first feature term group contains three feature terms term1, term2 and term3, and the initial lithography model is constructed based on term1, term2 and term3.
[0054] A second feature term group contains four feature terms: term1, term2, term3, and term4. It is obtained by adding a new feature term4 to the first feature term group. The corresponding lithography model to be corrected is constructed based on term1, term2, term3, and term4.
[0055] Another second feature term group contains four feature terms: term1, term2, term3, and term5. It is obtained by adding a new feature term5 to the first feature term group. The corresponding lithography model to be corrected is constructed based on term1, term2, term3, and term5.
[0056] The newly added features in the two second feature term groups are term4 and term5, which are different newly added features.
[0057] Specifically, when the second feature group is obtained from the first feature group through a second type of operation, multiple different second feature groups are formed by deleting different original feature groups from the first feature group. In other words, the original feature groups deleted from multiple different second feature groups are different from each other.
[0058] For example, taking two second feature term groups as an example, the first feature term group contains three feature terms term1, term2 and term3, and the initial lithography model is constructed based on term1, term2 and term3.
[0059] A second feature term group contains two feature terms, term1 and term2, obtained by deleting the original feature term3 from the first feature term group. The corresponding lithographic model to be corrected is constructed based on term1 and term2.
[0060] Another second feature term group contains two feature terms, term2 and term3, obtained by deleting the original feature term1 from the first feature term group. The corresponding lithographic model to be corrected is constructed based on term2 and term3.
[0061] The original features deleted from the two second feature groups are term1 and term3, which are different original features.
[0062] Step S303: Perform the step of using a genetic algorithm to correct the lithography models to be corrected for each of the multiple lithography models to be corrected, and obtain multiple corrected lithography models.
[0063] Specifically, the calibration process for each lithography model to be calibrated is described in step S103 above, and will not be detailed here.
[0064] By using different sets of second feature terms to separately execute the step of constructing a lithography model to be calibrated based on the second feature term sets, multiple lithography models to be calibrated are obtained. Then, a step of calibrating the lithography models to be calibrated using a genetic algorithm is performed on each of these multiple calibrated lithography models, resulting in multiple calibrated lithography models. In other words, multiple different calibrated lithography models can be obtained in parallel, which can improve the comprehensiveness and comparability of model candidates and facilitate the improvement of model determination efficiency.
[0065] Furthermore, the method of this embodiment, after obtaining multiple corrected lithography models, also includes: Step S304: Select one from multiple calibration lithography models as the final calibration lithography model for this calibration.
[0066] Reference Figure 4 As shown, in one embodiment, this step includes: Step S401: Select all calibration lithography models from multiple calibration lithography models that meet the preset evaluation criteria. Specifically, the preset evaluation criteria include profile topography and offset uniformity.
[0067] Step S402: Select the calibration lithography model with the lowest root mean square error from all calibration lithography models that meet the preset standards as the calibration lithography model for the final output of this calibration.
[0068] Specifically, after selecting all calibration lithography models that meet the preset evaluation indicators and standards, the calibration lithography model with the lowest root mean square error is selected as the final calibration lithography model for this calibration. In other words, the final calibration lithography model is selected by comprehensively considering the preset evaluation indicators and the root mean square error, ensuring the rationality of the final calibration lithography model.
[0069] Furthermore, after the step of selecting one of the multiple calibration lithography models as the final output of this calibration, the following steps are also included: Step S305: Determine whether the final output calibration lithography model meets the preset requirements. If yes, end the calibration process. Specifically, if the final output calibration lithography model meets the preset requirements, a qualified calibration lithography model is obtained, and the calibration process can end. If not, proceed to step S306.
[0070] Specifically, new data can be used to verify the difference between the photoresist image predicted by the final output correction lithography model and the actual photoresist image to determine whether the final output correction lithography model meets the preset requirements.
[0071] Step S306: Use the final output corrected lithography model as the new initial lithography model and recalibrate it.
[0072] Specifically, the final output corrected lithography model is used as the new initial lithography model, and all feature terms of the final output corrected lithography model are used as the new first feature term group, and step S301 is executed again.
[0073] By determining whether the final output calibrated lithography model meets preset requirements, if it does not, the final output calibrated lithography model is used as a new initial lithography model for recalibration. This facilitates iterative model optimization, automates model optimization management, and improves the convenience of model optimization.
[0074] Based on the foregoing, it should be noted that in the solution of this embodiment: If this is the first calibration process, the initial lithography model can be created using multiple sets of modeling data, i.e., a combination of multiple sets of design graphic data and corresponding actual photoresist pattern data. The model is created according to a preset feature term template. For example, it could be a model created using the gradient method. All the feature terms contained in the preset feature term template constitute the first feature term group.
[0075] If it is not the first calibration process, the initial lithography model is the calibrated lithography model obtained by calibration, and all the feature terms of the calibrated lithography model are the first feature term group.
[0076] Reference Figure 5 As shown, to more fully illustrate the solution of this embodiment, the explanation is from a model perspective. Specifically, an initial lithography model is first obtained. Multiple lithography models to be calibrated are obtained by adding different new feature terms to the feature term composition of the initial lithography model, or by deleting different original feature terms from the feature term composition of the initial lithography model. Each lithography model to be calibrated is calibrated to obtain a calibrated lithography model, and then the final output calibrated lithography model is selected from the multiple calibrated lithography models. If the calibrated lithography model meets the preset requirements, the calibration process can end. If it does not meet the preset requirements, the final output calibrated lithography model is used as the initial lithography model for further calibration.
[0077] In one embodiment, a computer device and a computer-readable storage medium are also provided. Figure 6 This is a schematic diagram of a computer device 10 according to an embodiment of the present invention. Figure 7 This is a schematic diagram of a computer-readable storage medium 20 according to an embodiment of the present invention.
[0078] The computer device 10 may include a memory 110, a processor 120, and a computer-executable program 11 stored in the memory 110 and running on the processor 120. When the processor 120 executes the computer-executable program 11, it implements the photolithography model correction method of any of the above embodiments.
[0079] The computer-readable storage medium 20 stores a computer-executable program 11 thereon, which, when executed by a processor, implements the correction method of the photolithography model of any of the above embodiments.
[0080] This embodiment also provides a computer program product. Figure 8 This is a schematic diagram of a computer program product 30 according to an embodiment of the present invention. The computer program product 30 includes a computer executable program 11, which, when executed by a processor 120, implements the photolithography model correction method of any of the embodiments described above.
[0081] Specifically, the computer executable program 11 used to perform the operations of the present invention may be assembly instructions, instruction set architecture (ISA) instructions, computer instructions, computer-related instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages.
[0082] For the purposes of this embodiment, the computer-readable storage medium 20 can be any means capable of containing, storing, communicating, propagating, or transmitting a program for use by or in conjunction with an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, the computer-readable storage medium 20 can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.
[0083] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system.
[0084] Computer device 10 can be, for example, a server, desktop computer, laptop computer, tablet computer, or smartphone. In some examples, computer device 10 can be a cloud acquisition node. Computer device 10 can be described in the general context of computer system executable instructions (such as program modules) executed by a computer system. Typically, program modules can include routines, programs, object programs, components, logic, data structures, etc., that perform specific tasks or implement specific abstract data types. Computer device 10 can be implemented in a distributed cloud acquisition environment where tasks are performed by remote processing devices linked via a communication network. In a distributed cloud acquisition environment, program modules can reside on local or remote acquisition system storage media, including storage devices.
[0085] Computer device 10 may include a processor 120 adapted to execute stored instructions and a memory 110 that provides temporary storage space for the operation of said instructions during operation. Processor 120 may be a single-core processor, a multi-core processor, an acquisition cluster, or any other configuration. Memory 110 may include random access memory (RAM), read-only memory, flash memory, or any other suitable storage system.
[0086] The processor 120 can be connected via a system interconnect (e.g., PCI, PCI-Express, etc.) to an I / O interface (input / output interface) suitable for connecting the computer device 10 to one or more I / O devices (input / output devices). I / O devices may include, for example, a keyboard and indicating devices, where indicating devices may include a touchpad or touchscreen, etc. I / O devices may be built into the computer device 10 or may be external devices connected to the acquisition device.
[0087] The processor 120 may also be linked via a system interconnect to a display interface suitable for connecting the computer device 10 to a display device. The display device may include a display screen that is a built-in component of the computer device 10. The display device may also include an external computer monitor, television, or projector connected to the computer device 10. Furthermore, a network interface controller (NIC) may be adapted to connect the computer device 10 to a network via a system interconnect. In some embodiments, the NIC may use any suitable interface or protocol (such as an Internet Minicomputer System Interface) to transmit data. The network may be a cellular network, a radio network, a wide area network (WAN), a local area network (LAN), or the Internet, etc. Remote devices may connect to the computer device via the network.
[0088] Therefore, those skilled in the art should recognize that although numerous exemplary embodiments of the present invention have been shown and described in detail herein, many other variations or modifications conforming to the principles of the present invention can be directly determined or derived from the disclosure of the present invention without departing from the spirit and scope of the invention. Thus, the scope of the present invention should be understood and construed as covering all such other variations or modifications.
Claims
1. A method for calibrating a lithography model, comprising: obtaining an initial lithography model, wherein the initial lithography model is constructed based on a first set of features, and the initial lithography model has been preliminarily calibrated for each feature in the first set of features and determined corresponding initial parameter values and initial weight coefficient values; constructing a to-be-calibrated lithography model based on a second set of features, wherein the second set of features is obtained from the first set of features by a first type of operation or a second type of operation, the first type of operation is to add at least one new feature in the first set of features, and the second type of operation is to delete at least one original feature from the first set of features; and calibrating the to-be-calibrated lithography model by using a genetic algorithm, to obtain a calibrated lithography model in which each feature in the to-be-calibrated lithography model is determined with a final parameter value and a final weight coefficient value, wherein in an initial population generated by the genetic algorithm, a feature in the first set of features of one individual is assigned with the initial parameter value and the initial weight coefficient value. 2.The method of claim 1, wherein the step of calibrating the to-be-calibrated lithography model by using the genetic algorithm comprises: obtaining the initial population generated by the genetic algorithm, wherein the initial population includes a plurality of initial random individuals generated by assigning each feature in the to-be-calibrated lithography model with a random parameter value and setting a preset weight coefficient value; and selecting one of the initial random individuals as a replacement individual, and replacing the parameter value and the weight coefficient value of each feature in the first set of features of the replacement individual with the initial parameter value and the initial weight coefficient value of the same feature. 3.The method of claim 2, wherein the step of calibrating the to-be-calibrated lithography model by using the genetic algorithm further comprises: in a case where the second set of features is obtained from the first set of features by the first type of operation, replacing the weight coefficient value of each new feature in the replacement individual with 0 and replacing the parameter value with a middle value of a corresponding parameter interval. 4.The method of claim 1, wherein the step of obtaining the initial lithography model further comprises: using a plurality of second sets of features to respectively perform the step of constructing the to-be-calibrated lithography model based on a second set of features, to obtain a plurality of to-be-calibrated lithography models, wherein the plurality of second sets of features are obtained from the first set of features by the same type of operation; and performing the step of calibrating the to-be-calibrated lithography model by using the genetic algorithm on the plurality of to-be-calibrated lithography models respectively, to obtain a plurality of calibrated lithography models. 5.The method of claim 4, wherein after obtaining the plurality of calibrated lithography models, one of the plurality of calibrated lithography models is selected as a final output calibrated lithography model of this calibration. 6.The method of claim 5, wherein the step of selecting one of the plurality of calibrated lithography models as the final output calibrated lithography model of this calibration further comprises: determining whether the final output correction lithography model meets preset requirements, and if not, taking the final output correction lithography model as a new initial lithography model and continuing correction.
7. The correction method of claim 5, wherein the step of selecting one of the plurality of correction lithography models as the final output correction lithography model of the current correction comprises: selecting all correction lithography models that meet preset standards from the plurality of correction lithography models according to the preset evaluation index; selecting a correction lithography model with the lowest root mean square error from the correction lithography models that meet the preset standards as the final output correction lithography model of the current correction.
8. A computer device comprising a memory, a processor, and a computer executable program stored on the memory and running on the processor, and the processor implements the correction method of any one of claims 1 to 7 when executing the computer executable program.
9. A computer readable storage medium having a computer executable program stored thereon, and the computer executable program implements the correction method of any one of claims 1 to 7 when executed by a processor.
10. A computer program product comprising a computer executable program, and the computer executable program implements the correction method of any one of claims 1 to 7 when executed by a processor.