Method and apparatus for determining parameter of layout quality test tool, and device and medium

Abstract

A method and apparatus for determining a parameter of a layout quality test tool, and a device and a medium. The method comprises: acquiring a plurality of first benchmark values and a plurality of first candidate values of a target parameter of a layout quality test tool (201); using the tool to perform a quality test on an original layout, so as to obtain a quality index corresponding to each first benchmark value (202); if the quality index corresponding to each first benchmark value does not meet a reference condition, determining a first reference value among the plurality of first candidate values on the basis of the quality index corresponding to each first benchmark value, and determining a plurality of second benchmark values on the basis of the plurality of first benchmark values and the first reference value (203); using the tool to perform the quality test on the original layout, so as to obtain a quality index corresponding to each second benchmark value (204); and if the quality index corresponding to each second benchmark value meets the reference condition, determining a target value of the target parameter on the basis of each second benchmark value (205). Thus, the tool accuracy is improved.

Description

Parameter determination method, device, equipment and medium for layout quality detection tool

[0001] This application claims priority to Chinese patent application number 202311022025.7, filed on August 14, 2023, entitled “Parameter determination method, device, equipment and medium for layout quality detection tools”, the entire contents of which are incorporated by reference into this application. Technical Field

[0002] The embodiments of the present application relate to the field of computer technology, and in particular to a method, apparatus, device, and medium for determining parameters of a layout quality inspection tool. Background Art

[0003] With the development of computer technology, integrated circuit technology has also developed, and more and more chip layouts, mask layouts, wafer layouts, etc. have emerged. Because layouts are prone to uneven quality, it is necessary to select the target layout from multiple candidate layouts.

[0004] In related technologies, multiple candidate layouts can be input into a layout quality inspection tool, which then determines the quality indicators of each candidate layout. If a candidate layout's quality indicator exceeds a threshold, the candidate layout is determined to be the target layout. Therefore, the accuracy of the layout quality inspection tool is closely related to the accuracy of the target layout selection, and the parameters of the layout quality inspection tool determine the accuracy of the layout quality inspection tool. Therefore, how to determine the parameters of the layout quality inspection tool has become a pressing issue.

[0005] Summary of the Invention

[0006] The present application provides a method, apparatus, device and medium for determining parameters of a layout quality inspection tool, which can determine the values ​​of target parameters of the layout quality inspection tool and improve the accuracy of the layout quality inspection tool. The technical solution includes the following contents.

[0007] In one aspect, a method for determining parameters of a layout quality inspection tool is provided, the method being performed by an electronic device and comprising:

[0008] Acquire a plurality of first reference values ​​of target parameters of a layout quality inspection tool and a plurality of first candidate values ​​of the target parameters;

[0009] Performing quality inspection on the original layout using the layout quality inspection tool according to the multiple first benchmark values ​​to obtain a quality indicator corresponding to each first benchmark value;

[0010] If the quality indicators corresponding to the respective first benchmark values ​​do not meet the reference condition, determining a first reference value from the plurality of first candidate values ​​based on the quality indicators corresponding to the respective first benchmark values, and determining a plurality of second benchmark values ​​of the target parameter based on the plurality of first benchmark values ​​and the first reference value;

[0011] Performing a quality inspection on the original layout using the layout quality inspection tool according to the multiple second benchmark values ​​to obtain a quality indicator corresponding to each second benchmark value;

[0012] When the quality indicators corresponding to the respective second benchmark values ​​meet the reference conditions, the target values ​​of the target parameters are determined based on the respective second benchmark values, and a layout quality inspection tool whose target parameter value is the target value is used as a layout quality inspection tool after parameter adjustment, and the layout quality inspection tool after parameter adjustment is used to determine the quality indicators of the layout to be inspected.

[0013] In another aspect, a parameter determination device for a layout quality inspection tool is provided, the device comprising:

[0014] An acquisition module, configured to acquire a plurality of first reference values ​​of a target parameter of a layout quality inspection tool and a plurality of first candidate values ​​of the target parameter;

[0015] A quality detection module, configured to perform a quality detection on the original layout using the layout quality detection tool according to the multiple first benchmark values, and obtain a quality indicator corresponding to each first benchmark value;

[0016] a determination module, configured to, if the quality indicators corresponding to the respective first benchmark values ​​do not satisfy a reference condition, determine a first reference value from the plurality of first candidate values ​​based on the quality indicators corresponding to the respective first benchmark values, and determine a plurality of second benchmark values ​​of the target parameter based on the plurality of first benchmark values ​​and the first reference value;

[0017] The quality detection module is further configured to perform a quality detection on the original layout using the layout quality detection tool according to the plurality of second benchmark values, and obtain a quality index corresponding to each second benchmark value;

[0018] The determination module is further configured to determine, when the quality indicators corresponding to the respective second benchmark values ​​satisfy the reference conditions, the target values ​​of the target parameters based on the respective second benchmark values, and use a layout quality inspection tool whose target parameter value is the target value as a layout quality inspection tool after parameter adjustment, and the layout quality inspection tool after parameter adjustment is used to determine the quality indicators of the layout to be inspected.

[0019] On the other hand, an electronic device is provided, comprising a processor and a memory, wherein the memory stores at least one computer program, and the at least one computer program is loaded and executed by the processor so that the electronic device implements any of the above-mentioned parameter determination methods of the layout quality detection tool.

[0020] On the other hand, a non-volatile computer-readable storage medium is also provided, in which at least one computer program is stored. The at least one computer program is loaded and executed by a processor to enable an electronic device to implement any of the above-mentioned parameter determination methods of the layout quality detection tool.

[0021] On the other hand, a computer program is also provided, wherein at least one computer program is loaded and executed by a processor to enable an electronic device to implement any of the above-mentioned parameter determination methods of the layout quality detection tool.

[0022] On the other hand, a computer program product is also provided, in which at least one computer program is stored. The at least one computer program is loaded and executed by a processor to enable an electronic device to implement any of the above-mentioned parameter determination methods of the layout quality detection tool.

[0023] In the technical solution provided by the present application, the target parameters of the layout quality detection tool include various first benchmark values. The layout quality detection tool is used to perform quality detection on the original layout to obtain quality indicators corresponding to the various first benchmark values. If there is a quality indicator corresponding to the first benchmark value that does not meet the reference condition, then based on the quality indicator corresponding to each first benchmark value, a first reference value is determined from multiple first candidate values ​​of the target parameter, and based on the multiple first benchmark values ​​and the first reference value, multiple second benchmark values ​​of the target parameter are determined. By determining the second benchmark value, the accuracy of the value space of the target parameter is improved. Then, the layout quality detection tool is used to perform quality detection on the original layout to obtain quality indicators corresponding to the various second benchmark values. By improving the accuracy of the value space of the target parameter, the accuracy of the layout quality detection tool is improved, thereby improving the accuracy of the quality index. When the quality index corresponding to each second benchmark value meets the reference condition, the target value of the target parameter can be accurately determined based on each second benchmark value, thereby ensuring the accuracy of the layout quality detection tool, which is conducive to improving the accuracy of the quality index of the layout to be detected determined by the layout quality detection tool, thereby more accurately selecting a higher quality layout from the layout to be detected. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] FIG1 is a schematic diagram of an implementation environment of a parameter determination method for a layout quality detection tool provided in an embodiment of the present application;

[0025] FIG2 is a flow chart of a parameter determination method for a layout quality detection tool provided in an embodiment of the present application;

[0026] FIG3 is a schematic diagram of a Pareto front between mask error and mask complexity provided by an embodiment of the present application;

[0027] FIG4 is a schematic diagram of a flow chart of determining parameters of a layout quality detection tool for a single quality indicator provided by an embodiment of the present application;

[0028] FIG5 is a schematic diagram of a flow chart of determining parameters of a layout quality detection tool for two quality indicators provided by an embodiment of the present application;

[0029] FIG6 is a schematic diagram showing the relationship between the number of iterations and the edge distance error provided by an embodiment of the present application;

[0030] FIG7 is a schematic diagram showing the relationship between edge distance error and total complexity variation provided by an embodiment of the present application;

[0031] FIG8 is a schematic structural diagram of a parameter determination device of a layout quality detection tool provided in an embodiment of the present application;

[0032] FIG9 is a schematic structural diagram of a terminal device provided in an embodiment of the present application;

[0033] FIG10 is a schematic diagram of the structure of a server provided in an embodiment of the present application. DETAILED DESCRIPTION

[0034] In order to make the objectives, technical solutions and advantages of this application clearer, the implementation methods of this application will be further described in detail below with reference to the accompanying drawings.

[0035] FIG1 is a schematic diagram of an implementation environment for a method for determining parameters of a layout quality detection tool provided in an embodiment of the present application. As shown in FIG1 , the implementation environment includes a terminal device 101 and a server 102. The method for determining parameters of a layout quality detection tool in the embodiment of the present application can be executed by the terminal device 101, can be executed by the server 102, or can be executed jointly by the terminal device 101 and the server 102.

[0036] The terminal device 101 can be a smart phone, a game console, a desktop computer, a tablet computer, a laptop computer, a smart TV, a smart car device, an intelligent voice interaction device, a smart home appliance, etc. The server 102 can be a single server, or a server cluster consisting of multiple servers, or any one of a cloud computing platform and a virtualization center, which is not limited in the embodiments of the present application. The server 102 can be connected to the terminal device 101 via a wired network or a wireless network. The server 102 can have functions such as data processing, data storage, and data transmission and reception, which are not limited in the embodiments of the present application. The number of terminal devices 101 and servers 102 is not limited and can be one or more.

[0037] In the field of integrated circuit technology, layouts are prone to varying quality. Therefore, layout quality inspection tools are often used to determine the quality indicators of each candidate layout. If the quality indicator of a candidate layout exceeds the indicator threshold, the candidate layout is determined to be the target layout. Since the accuracy of layout quality inspection tools is closely related to the accuracy of the target layout selection, and the parameters of layout quality inspection tools determine the accuracy of layout quality inspection tools, how to determine the parameters of layout quality inspection tools has become a pressing issue.

[0038] The embodiment of the present application provides a method for determining the parameters of a layout quality detection tool. This method can be applied in the above-mentioned implementation environment to determine the values ​​of the target parameters of the layout quality detection tool and improve the accuracy of the layout quality detection tool. Taking the flowchart of the method for determining the parameters of a layout quality detection tool provided by the embodiment of the present application shown in Figure 2 as an example, for ease of description, the terminal device 101 or server 102 that executes the method for determining the parameters of the layout quality detection tool in the embodiment of the present application is referred to as an electronic device, and the method can be performed by an electronic device. As shown in Figure 2, the method includes the following steps 201 to 205.

[0039] Step 201 : Acquire a plurality of first reference values ​​of target parameters and a plurality of first candidate values ​​of target parameters of a layout quality inspection tool.

[0040] In an embodiment of the present application, a layout quality inspection tool supports quality inspection processing of a layout, where the layout includes but is not limited to a chip layout, a mask plate diagram, a wafer layout, and the like. The chip layout is also called an integrated circuit (IC) design layout. Generally, a mask plate is photolithographically processed based on the chip layout to obtain a mask layout, and a wafer is photolithographically processed based on the mask layout to obtain a wafer layout. The embodiment of the present application does not limit the layout quality inspection tool. Exemplarily, the layout quality inspection tool includes at least one of computer-aided design (CAD) software, electronic design automation (EDA) software, and the like.

[0041] There are a large number of parameters in the layout quality detection tool, some of which are configured and cannot be adjusted, while some parameters can be adjusted. The parameters that can be adjusted are called adjustable parameters. By adjusting the adjustable parameters, the values ​​of the adjustable parameters are determined so that the accuracy of the layout quality detection tool meets expectations. In an embodiment of the present application, the target parameter is the adjustable parameter in the layout quality detection tool. The embodiment of the present application does not limit the target parameter. Exemplarily, the target parameter includes a learning rate and an iteration step size. The target parameter is at least one, that is, the number of target parameters may be one or more. Since the processing methods of each target parameter are similar, the following is an introduction to the parameter adjustment method of the target parameter, i.e., the determination process of the target value of the target parameter, without considering the quantity factor.

[0042] The electronic device can obtain multiple first reference values ​​of the target parameter and multiple first candidate values ​​of the target parameter. In some embodiments, the multiple first reference values ​​can also be referred to as a plurality of first reference values, or as at least two first reference values; the multiple first candidate values ​​can also be referred to as a plurality of first candidate values, or as at least two first candidate values. The embodiments of the present application do not limit the method for obtaining any first reference value or any first candidate value. For example, the electronic device obtains each input first reference value and each first candidate value, or the electronic device randomly generates each first reference value and each first candidate value.

[0043] In one possible implementation, step 201 includes: obtaining a first value range of a target parameter, dividing the first value range into multiple first value intervals; and sampling within each of the first value intervals to obtain multiple first reference values ​​and multiple first candidate values. Sampling within the multiple first value intervals to obtain the first reference values ​​and first candidate values ​​helps improve the distribution uniformity of the first reference values ​​and first candidate values, ensuring the comprehensiveness of the first reference values ​​and first candidate values.

[0044] In an embodiment of the present application, the electronic device is configured with a layout quality detection tool, and a user can enter a first value range of a target parameter into the layout quality detection tool. Alternatively, the electronic device can randomly generate the first value range. Alternatively, the electronic device can obtain a configured first value range.

[0045] Next, the electronic device can use the Latin Hypercube Sampling (LHS) method to divide the first value range into multiple first value intervals, any two adjacent first value intervals can partially overlap or not overlap, and any two adjacent first value intervals can be of the same length or different lengths. For any first value interval, at least one value is sampled from the first value interval, and any value is a first reference value or a first candidate value. In this way, at least one value among each first reference value and each first candidate value is sampled in any first value interval.

[0046] In a possible implementation, “sampling within each first value interval to obtain multiple first reference values ​​and multiple first candidate values” includes: steps 2011 to 2013 (not shown in the figure).

[0047] Step 2011: Sampling is performed within each first value interval to obtain a plurality of first initial values ​​of the target parameter.

[0048] In the embodiment of the present application, a Latin hypercube sampling method is used to divide the first value range into multiple first value intervals, sample at least one value from any first value interval, and use each sampled value as each first initial value. In this way, multiple first initial values ​​are sampled from each first value interval.

[0049] For example, there are K target parameters (K is a positive integer). For any target parameter, the first value range of the target parameter is divided into L (L is an integer greater than 1) first value intervals of equal length and non-overlapping. For any first value interval, a value is randomly sampled from the first value interval as a first initial value. Since one target parameter corresponds to L first value intervals, L first initial values ​​can be sampled for one target parameter, and KL first initial values ​​can be sampled for K target parameters.

[0050] In addition, in addition to entering the first value range of the target parameter in the layout quality detection tool, the user can also enter other information of the target parameter. For example, the user can also enter the parameter type of the target parameter. Alternatively, the electronic device can obtain the parameter type of the configured target parameter. The parameter type of the target parameter can be a continuous parameter or a discrete parameter. The electronic device can sample at least one first initial value from any first value interval according to the parameter type of the target parameter. The first initial value sampled in this way is a value that conforms to the parameter type of the target parameter, which is conducive to ensuring the reliability of the first initial value.

[0051] It is understood that in addition to sampling multiple first initial values ​​from the first value range using the Latin hypercube sampling method, multiple first initial values ​​can also be randomly sampled from the first value range using the random sampling method. Compared to the random sampling method, sampling using the Latin hypercube sampling method can make the distribution of the first initial values ​​more uniform, ensuring the comprehensiveness of the first initial values, because the first initial values ​​are sampled from the multiple first value intervals into which the first value range is divided.

[0052] Next, a random partitioning approach can be used to partition the multiple first initial values ​​into multiple first reference values ​​and multiple first candidate values. In other words, for any first initial value, it can be randomly determined as a first reference value or a first candidate value. Alternatively, a clustering process can be used to determine individual first reference values ​​and individual first candidate values ​​from the multiple first initial values. The clustering process is described below.

[0053] Step 2012: clustering the multiple first initial values ​​to obtain multiple target clusters, where any target cluster includes at least one first initial value from the multiple first initial values.

[0054] In the embodiment of the present application, any clustering algorithm can be used to cluster the multiple first initial values ​​to obtain multiple target clusters. The embodiment of the present application does not limit the clustering algorithm. Exemplarily, the clustering algorithm includes any one of the K-means clustering algorithm, the density-based clustering algorithm, the model-based clustering algorithm, and the probability-based clustering algorithm. The following takes the K-means clustering algorithm as an example to describe the implementation method of clustering the multiple first initial values.

[0055] First, a set number M (M is a positive integer) of initial cluster centers is obtained, and the set number M is used to indicate the number of target clusters. The method for determining M is not limited in the embodiment of the present application. For example, the electronic device can obtain M input by the user, or the electronic device can randomly generate M, or the electronic device can obtain a configured M. Any initial cluster center is the cluster center of an initial cluster cluster. The method for obtaining any initial cluster center is not limited in the embodiment of the present application. For example, the electronic device can obtain M initial cluster centers input by the user, or the electronic device can randomly generate M initial cluster centers.

[0056] Then, based on the M initial clustering centers, multiple first initial values ​​are clustered into M initial clustering clusters. Optionally, for any first initial value, the distance between the first initial value and each initial clustering center is calculated. The method for determining the distance is not limited in the embodiment of the present application. For example, any distance determination method such as Euclidean distance, Manhattan distance, Chebyshev distance, etc. can be used. The minimum distance is determined from the distances between the first initial value and each initial clustering center, and the first initial value is divided into the initial clustering cluster to which the initial clustering center corresponding to the minimum distance belongs. In this way, it is possible to divide each first initial value into M initial clustering clusters.

[0057] Next, for any initial cluster, based on the first initial values ​​included in the initial cluster, the cluster center of the initial cluster is re-determined. For ease of distinction, the re-determined cluster center is referred to as the first cluster center. When the difference between the initial cluster center of the initial cluster and the first cluster center of the initial cluster is less than a first threshold (for example, the first threshold is 0 or 0.0001), it indicates that the cluster center of the initial cluster has not changed. On the contrary, when the difference between the initial cluster center of the initial cluster and the first cluster center of the initial cluster is not less than the first threshold, it indicates that the cluster center of the initial cluster has changed.

[0058] If the cluster centers of the M initial clusters do not change, it means that the algorithm converges. In this case, each initial cluster is used as the target cluster.

[0059] If the cluster center of at least one of the M initial clusters changes, it means that the algorithm has not converged. In this case, the first cluster center is the cluster center of the first cluster, and multiple first initial values ​​can be clustered into M first clusters based on the M first cluster centers. For any first initial value, the distance between the first initial value and each first cluster center is calculated, and the minimum distance is determined from the distances between the first initial value and each first cluster center, and the first initial value is divided into the first cluster to which the first cluster center corresponding to the minimum distance belongs. In this way, it is possible to divide each first initial value into M first clusters. Afterwards, for any first cluster, based on each first initial value included in the first cluster, the cluster center of the first cluster is re-determined. For the sake of distinction, the re-determined cluster center is referred to as the second cluster center. When the difference between the first cluster center of the first cluster and the second cluster center of the first cluster is less than the first threshold, it indicates that the cluster center of the first cluster has not changed. Conversely, when the difference between the first cluster center of the first cluster and the second cluster center of the first cluster is not less than the first threshold, it indicates that the cluster center of the first cluster has changed.

[0060] If the cluster centers of the M first clusters do not change, it means that the algorithm has converged. In this case, each first cluster is used as a target cluster.

[0061] If the cluster center of at least one of the M first clusters changes, it means that the algorithm has not converged. In this case, the second cluster center is the cluster center of the second cluster, and the multiple first initial values ​​can be clustered into M second clusters based on the M second cluster centers. The implementation method of clustering the multiple first initial values ​​into M second clusters is similar to the implementation method of "clustering the multiple first initial values ​​into M initial clusters" and "clustering the multiple first initial values ​​into M first clusters" described above. Through an implementation method similar to the above, the multiple first initial values ​​are continuously clustered until the algorithm converges and M target clusters are obtained.

[0062] Simply put, using the K-means clustering algorithm to cluster multiple first initial values ​​is equivalent to dividing the multiple first initial values ​​into M clusters according to the distance between each pair of first initial values ​​(for example, Euclidean distance, Manhattan distance, Chebyshev distance, etc.), so that the distance between any two first initial values ​​belonging to the same cluster is as small as possible, while the distance between any two first initial values ​​belonging to different clusters is as large as possible. When the cluster centers of the M clusters do not change, the K-means clustering algorithm converges and ultimately obtains M clusters. These M clusters are the M target clusters mentioned above.

[0063] Optionally, the K-means clustering algorithm converges, which is equivalent to minimizing the square error. Optionally, the square error is determined according to formula (1) shown below.

[0064] Among them, error represents the square error, M represents the number of cluster centers, C i represents the i-th cluster, x represents C i The first initial value in μ i Representative C i The cluster center of |C i |Characterization C i The number of first initial values ​​in .

[0065] As can be seen from formula (1), the squared error represents the square of the distance between each first initial value and the cluster center of the cluster to which it belongs. By continuously adjusting the cluster center, the squared error can be minimized, so that the K-means clustering algorithm converges and eventually obtains M target clusters.

[0066] Step 2013: For any target cluster among the multiple target clusters, select a first reference value from each first initial value included in any target cluster, and determine the first initial value in any target cluster except the first reference value as the first candidate value.

[0067] In an embodiment of the present application, at least one first initial value can be randomly selected from the various first initial values ​​included in any target cluster as the first benchmark value. Alternatively, the cluster center of the target cluster can be re-determined based on the various first initial values ​​included in any target cluster. For ease of description, the cluster center of the re-determined target cluster is referred to as the target cluster center. Calculate the distance between each first initial value in any target cluster and the target cluster center of the target cluster. Determine at least one distance that is not greater than the second threshold from the distances between each first initial value in any target cluster and the target cluster center of the target cluster. Determine the first initial value corresponding to each distance that is not greater than the second threshold as the first benchmark value. It should be noted that the number of first benchmark values ​​selected from each first initial value included in any target cluster is one or more. In other words, select at least one first benchmark value from each first initial value included in any target cluster.

[0068] The embodiment of the present application does not limit the second threshold value. Exemplarily, the second threshold value is a value determined based on experience, or the second threshold value is the minimum value of the distances between each first initial value in the target cluster and the target cluster center of the target cluster. When the second threshold value is the minimum value of the distances between each first initial value in the target cluster and the target cluster center of the target cluster, a first benchmark value can be determined based on a target cluster, and the first benchmark value is the first initial value in the target cluster that is closest to the target cluster center.

[0069] In some embodiments, the method for selecting a first benchmark value from the first initial values ​​included in any target cluster can also be: calculating the distance between each first initial value in any target cluster and the target cluster center of the target cluster, and determining at least one distance that is not less than a third threshold from the distances between each first initial value in any target cluster and the target cluster center of the target cluster. The first initial values ​​corresponding to the distances that are not less than the third threshold are determined as the first benchmark value. Exemplarily, the third threshold is a value determined empirically, or the third threshold is the maximum value of the distances between each first initial value in the target cluster and the target cluster center of the target cluster. When the third threshold is the maximum value of the distances between each first initial value in the target cluster and the target cluster center of the target cluster, a first benchmark value can be determined based on the target cluster, and the first benchmark value is the first initial value in the target cluster that is farthest from the target cluster center.

[0070] According to the above method, a first reference value can be determined from each first initial value included in each target cluster. Then, each first initial value in each target cluster other than the first reference value is determined as a first candidate value. Determining each first reference value by clustering improves the accuracy of the first reference value and can accelerate the efficiency of determining the target value.

[0071] Step 202 : Performing a quality inspection on the original layout using a layout quality inspection tool according to the plurality of first benchmark values ​​to obtain a quality index corresponding to each first benchmark value.

[0072] Since the layout quality inspection tool supports quality inspection of the layout, the layout quality inspection tool can be determined by making the target parameter include any first benchmark value, and the original layout is input into the layout quality inspection tool. The original layout is subjected to quality inspection by the layout quality inspection tool to obtain the quality index of the original layout corresponding to the first benchmark value. The original layout is a layout that meets the layout quality standards specified in the chip field, including but not limited to chip layout, mask layout, wafer layout, etc. Since the target parameter includes multiple first benchmark values, the result of the quality inspection of the original layout includes the quality index of the original layout corresponding to each first benchmark value. The quality index corresponding to the first benchmark value in the embodiment of the present application refers to the quality index of the original layout corresponding to the first benchmark value.

[0073] For any first benchmark value among multiple first benchmark values, the quality index of the original layout corresponding to the first benchmark value is used to describe the quality of the original layout determined using the layout quality detection tool when the target parameters of the layout quality detection tool include the first benchmark value.

[0074] It is understood that there are multiple indicators used to describe the quality of the original layout. That is, in the embodiments of the present application, any first benchmark value corresponds to at least one quality indicator, and any quality indicator corresponding to the first benchmark value is used to describe the quality of the original layout in a certain aspect determined by the layout quality detection tool when the target parameters of the layout quality detection tool include the first benchmark value. Any two quality indicators corresponding to any first benchmark value describe the quality of the original layout in different aspects.

[0075] For example, the mask complexity of the original layout can reflect the quality of the original layout. Generally speaking, the lower the mask complexity of the original layout, the higher the quality of the original layout. Therefore, any quality indicator corresponding to any first benchmark value can be the mask complexity corresponding to the first benchmark value, and the quality of the original layout is described by this mask complexity. The mask complexity corresponding to the first benchmark value is negatively correlated with the quality of the original layout, that is, the smaller the mask complexity corresponding to the first benchmark value, the higher the quality of the original layout.

[0076] For example, the edge distance error of the original layout can reflect the quality of the original layout. Generally speaking, the lower the edge distance error of the original layout, the higher the quality of the original layout. Therefore, any quality indicator corresponding to any first benchmark value can be the edge distance error corresponding to the first benchmark value, and the quality of the original layout is described by this edge distance error. The edge distance error corresponding to the first benchmark value is negatively correlated with the quality of the original layout, that is, the smaller the edge distance error corresponding to the first benchmark value, the higher the quality of the original layout.

[0077] In one possible implementation, step 202 includes: processing the original layout using a layout quality detection tool according to multiple first benchmark values ​​to obtain a predicted layout corresponding to each first benchmark value, and determining the quality indicator corresponding to each first benchmark value based on the predicted layout corresponding to each first benchmark value.

[0078] In an embodiment of the present application, when the target parameter includes any first benchmark value, the original layout is input into a layout quality detection tool, and the original layout is processed by the layout quality detection tool to obtain a predicted layout corresponding to the first benchmark value. Optionally, the layout quality detection tool includes a photoresist model. The photoresist model can simulate the photolithography equipment in an actual scenario, perform photolithography processing on the original layout, and obtain a predicted layout.

[0079] That is to say, if the original layout is a chip layout, the chip layout is subjected to photolithography processing through the layout quality inspection tool, that is, the layout quality inspection tool simulates the process of photolithography equipment performing photolithography processing on the mask plate based on the chip layout to obtain the mask layout, and the mask layout is the predicted layout. If the original layout is a mask layout, or the original layout is a chip layout, and the mask layout is obtained based on the chip layout through the layout quality inspection tool, the mask layout can be subjected to photolithography processing through the layout quality inspection tool, that is, the layout quality inspection tool simulates the process of photolithography equipment performing photolithography processing on the wafer based on the mask layout to obtain the wafer layout, and the wafer layout is the predicted layout.

[0080] In the embodiment of the present application, for any first reference value, the pattern of the predicted layout corresponding to any first reference value can be analyzed by the layout quality detection tool, and the analysis result is used as the quality indicator of the original layout corresponding to the first reference value. This is because the interference effect and diffraction effect in the lithography process, etc., lead to errors in the pattern of the predicted layout, so that the layout quality detection tool can analyze the quality of the predicted layout based on the pattern of the predicted layout, thereby obtaining an analysis result, which can characterize both the quality of the original layout and the quality of the predicted layout. The method of analyzing the pattern by the layout quality detection tool is not limited in the embodiment of the present application.

[0081] Alternatively, if the original layout is a chip layout and the predicted layout is a wafer layout, for any first reference value, the error between the original layout and the predicted layout corresponding to the first reference value is determined by a layout quality detection tool. Optionally, the pixels in the predicted layout correspond to the pixels in the original layout one-to-one, that is, any pixel in the predicted layout has a corresponding pixel in the original layout. The pixel difference between each pixel in the predicted layout and the corresponding pixel in the original layout is determined by a layout quality detection tool, and the sum or average of these pixel differences is determined as the error between the predicted layout and the original layout. Alternatively, the features of the predicted layout and the features of the original layout are extracted by a layout quality detection tool, and the feature distance between the features of the predicted layout and the features of the original layout is used as the error between the predicted layout and the original layout.

[0082] Next, based on the error between the original layout and the predicted layout corresponding to any first benchmark value, a quality index of the original layout corresponding to the first benchmark value (i.e., the quality index corresponding to the first benchmark value) is determined. Exemplarily, the error between the original layout and the predicted layout corresponding to any first benchmark value is used as the quality index of the original layout corresponding to the first benchmark value, or the error between the original layout and the predicted layout corresponding to any first benchmark value is weighted to obtain the quality index of the original layout corresponding to the first benchmark value.

[0083] Theoretically, if the accuracy of the layout quality detection tool is 100%, then the error between the predicted layout and the original layout is 0. If the accuracy of the layout quality detection tool is 0, then the error between the predicted layout and the original layout is 100%. Therefore, the more accurate the layout quality detection tool is, the smaller the error between the predicted layout and the original layout is, that is, the error between the predicted layout and the original layout can reflect the accuracy of the layout quality detection tool. The quality index of the original layout corresponding to each first benchmark value is calculated by the error between the original layout and the predicted layout corresponding to each first benchmark value, so that the quality index of the original layout corresponding to each first benchmark value can also reflect the accuracy of the layout quality detection tool. Therefore, based on the quality index of the original layout corresponding to each first benchmark value, the parameter values ​​of the layout quality detection tool can be adjusted in a more accurate direction to ensure the accuracy of the layout quality detection tool.

[0084] Furthermore, the layout quality detection tool can determine the quality index of the original layout directly based on the original layout without simulating the lithography equipment in the actual scenario. In other words, when the target parameter includes any first reference value, the layout quality detection tool can analyze the pattern of the original layout to obtain the quality index of the original layout corresponding to the first reference value. The method of analyzing the pattern of the original layout by the layout quality detection tool is not limited in the embodiments of the present application.

[0085] Step 203: When the quality indicators corresponding to the various first benchmark values ​​do not meet the reference conditions, a first reference value is determined from multiple first candidate values ​​based on the quality indicators corresponding to the various first benchmark values, and multiple second benchmark values ​​of the target parameter are determined based on the multiple first benchmark values ​​and the first reference value.

[0086] The fact that the quality indicators corresponding to the various first benchmark values ​​do not meet the reference conditions may mean that the quality indicators corresponding to at least one of the various first benchmark values ​​do not meet the reference conditions, that is, the fact that the quality indicators corresponding to the various first benchmark values ​​do not meet the reference conditions may mean that the quality indicators corresponding to some of the first benchmark values ​​or all of the first benchmark values ​​in the various first benchmark values ​​do not meet the reference conditions. The embodiments of the present application do not limit the fact that the quality indicators corresponding to any first benchmark value do not meet the reference conditions. Exemplarily, it is possible to determine whether the quality indicators corresponding to any first benchmark value meet the reference conditions based on the quality indicators and the reference threshold corresponding to the first benchmark value. The reference threshold is set based on experience or flexibly adjusted according to the application scenario, and the embodiments of the present application do not limit this.

[0087] Optionally, if the quality indicator is negatively correlated with the quality of the original layout, for example, the quality indicator is mask complexity, edge distance error, etc., then when the quality indicator corresponding to any first benchmark value is greater than a reference threshold, it is determined that the quality indicator corresponding to the first benchmark value does not meet the reference condition; when the quality indicator corresponding to any first benchmark value is not greater than the reference threshold, it is determined that the quality indicator corresponding to the first benchmark value meets the reference condition. If the quality indicator is positively correlated with the quality of the original layout, then when the quality indicator corresponding to any first benchmark value is less than the reference threshold, it is determined that the quality indicator corresponding to the first benchmark value does not meet the reference condition; when the quality indicator corresponding to any first benchmark value is not less than the reference threshold, it is determined that the quality indicator corresponding to the first benchmark value meets the reference condition.

[0088] Alternatively, an iterative calculation method is used to determine the quality indicator corresponding to each first benchmark value. The number of iterations used to determine the quality indicator corresponding to each first benchmark value can be obtained. If the number of iterations is less than a number threshold, it is determined that the quality indicator corresponding to each first benchmark value does not meet the reference condition; if the number of iterations is not less than the number threshold, it is determined that the quality indicator corresponding to each first benchmark value meets the reference condition. The number threshold is set based on experience or flexibly adjusted according to the application scenario, and this embodiment of the present application does not limit this.

[0089] In the embodiment of the present application, the number of first reference values ​​is M, and the number of first candidate values ​​is LM. The quality indicators corresponding to the M first reference values ​​can be recorded as (y1, ..., y m ,...,y M ). In (y1,...,y m ,...,y M ) does not meet the reference conditions, it can be based on (y1, ..., y m ,...,y M ), determine a first reference value from L M first candidate values, where M is an integer greater than 1, and L M is an integer greater than 1.

[0090] In one possible implementation, the step "determining a first reference value from a plurality of first candidate values ​​based on the quality indicators corresponding to the respective first benchmark values" in step 203 includes steps 2031 to 2032 (not shown). It should be noted that the number of first reference values ​​may be one or more. In other words, the number of first reference values ​​is at least one. That is, based on the quality indicators corresponding to the respective first benchmark values, at least one first reference value is determined from the plurality of first candidate values, and based on the plurality of first benchmark values ​​and the at least one first reference value, a plurality of second benchmark values ​​of the target parameter are determined.

[0091] Step 2031: Determine the reference indicators of each first candidate value based on the quality indicators corresponding to each first benchmark value. The reference indicators of any first candidate value are used to describe the situation where the quality indicators meet the reference conditions when the layout quality inspection tool is used to perform quality inspection on the original layout when the target parameters include any first candidate value.

[0092] In the embodiment of the present application, the quality index corresponding to each first reference value can be recorded as (y1, ..., y m ,...,y M ). Based on (y1, ..., y m ,...,y M ), determine reference indicators for the LM first candidate values. Optionally, the reference indicator for any first candidate value may be an expected improvement (EI) value, which reflects the improvement in the accuracy of the layout quality inspection tool when the target parameter includes the first candidate value.

[0093] When using a layout quality inspection tool to perform quality inspection on the original layout, a quality index of the original layout can be obtained, which reflects the accuracy of the layout quality inspection tool. The closer the quality index approaches the reference condition, the greater the accuracy of the layout quality inspection tool. Therefore, the expected improvement of any first candidate value can reflect the probability that the quality index will meet the reference condition when the layout quality inspection tool is used to perform quality inspection on the original layout, when the target parameter includes the first candidate value. The larger the expected improvement value of any first candidate value, the higher the probability.

[0094] Optionally, step 2031 includes steps S1 to S2 (not shown in the figure).

[0095] Step S1, based on each first benchmark value and the quality indicator corresponding to each first benchmark value, determine a first mapping relationship, the first mapping relationship is used to describe the mapping relationship between each first benchmark value and the statistical information of the statistical distribution when the quality indicator corresponding to each first benchmark value satisfies the statistical distribution.

[0096] In the embodiment of the present application, the quality indicators (y1, ..., y m ,...,y M ), determine the first mapping relationship. Where, (y1, ..., y m ,...,y M ) satisfies the statistical distribution, and the first mapping relationship is used to describe the mapping relationship between the M first benchmark values ​​and the statistical information of the statistical distribution.

[0097] The embodiments of the present application do not limit the statistical distribution. For example, the statistical distribution is a Gaussian distribution, a chi-square distribution, an F distribution, etc. Different statistical distributions correspond to different statistical information. The statistical information is a parameter used to determine the statistical distribution. For example, the statistical information of the Gaussian distribution includes the mean and variance.

[0098] Optionally, the M first reference values ​​and the quality indicators (y1, ..., y m ,...,y M ), determine M Gaussian Process (GP) regression models. The mth Gaussian Process regression model is used to describe the quality index y corresponding to the mth first benchmark value m When a Gaussian distribution is satisfied, the mapping relationship between the mth first benchmark value and the mean and variance of the Gaussian distribution is described by M Gaussian process regression models.

[0099] For the mth Gaussian process regression model, since the mean and variance of the Gaussian distribution can reflect y m Therefore, the mth Gaussian process regression model can reflect the mth first benchmark value and the corresponding quality index y m The mapping relationship between them. Describe y through the Gaussian process algorithm m , we get the mth Gaussian process regression model, which can be expressed as: GP(μ m (x),σ m (x)), x represents the target parameter, GP represents the quality index y corresponding to the target parameter when it is the mth first benchmark value m Satisfying Gaussian distribution, μ m (x) is the mean of the Gaussian distribution, σ m (x) is the variance of the Gaussian distribution. In some embodiments, the variance of the Gaussian distribution is also referred to as the uncertainty of the Gaussian distribution.

[0100] Optionally, the Gaussian process regression model can describe the root mean square error of the quality indicators corresponding to each first benchmark value, that is, Error(x)~GP(m(x),k(x,x′)), where Error(x) represents the root mean square error of the quality indicators corresponding to the target parameter x including each first benchmark value, m(x) is the mean function of the target parameter x, and k(x,x′) is the kernel function of the target parameter x and other parameters x′ in the layout quality inspection tool except the target parameter x. In the embodiment of the present application, an n-dimensional square exponential function is used as the kernel function, and the kernel function can be expressed as: Among them, exp represents the exponential function with the natural constant e as the base, ∑ is the summation symbol, and x iCharacterize the first benchmark value of the target parameter x, x′ i represents the i-th value of other parameters, n is the dimension of x, are hyperparameters estimated by maximum likelihood.

[0101] It is understood that the first mapping relationship can be described by M Gaussian process regression models, which are a set that satisfies a joint Gaussian distribution. In other words, under the premise that the quality indicator corresponding to each first benchmark value satisfies a Gaussian distribution, the Gaussian process regression model corresponding to each first benchmark value is determined, thereby determining the first mapping relationship.

[0102] Step S2: for any first candidate value among the first candidate values, based on any first candidate value and the first mapping relationship, determine the statistical information corresponding to any first candidate value, and determine the reference index of any first candidate value based on the statistical information corresponding to any first candidate value.

[0103] In the embodiment of the present application, the first mapping relationship can reflect the mapping relationship between each first benchmark value and the statistical information. In other words, the first mapping relationship can map any first benchmark value to the corresponding statistical information. Based on this, through the first mapping relationship, any first candidate value can be mapped to the statistical information corresponding to the first candidate value.

[0104] Then, using the statistical information corresponding to any first candidate value, a reference index for the first candidate value is calculated. The embodiment of the present application does not limit the calculation method of the reference index for the first candidate value. For example, the reference index for the first candidate value is obtained by weighting the statistical information corresponding to the first candidate value, or the reference index for the first candidate value is determined according to steps S21 to S23 mentioned below.

[0105] Optionally, the statistical information corresponding to any first candidate value includes a mean and a variance. Optionally, the first mapping relationship includes a Gaussian process regression model. As described above, the model can predict the first candidate value and obtain the mean μ corresponding to the first candidate value. m (x) and variance σ m (x) In this case, the step S2 of "determining a reference indicator of any first candidate value based on the statistical information corresponding to any first candidate value" includes steps S21 to S23 (not shown in the figure).

[0106] Step S21 : determining a reference quality indicator from the quality indicators corresponding to the first reference values.

[0107] In the embodiment of the present application, a quality indicator is randomly determined from the quality indicators corresponding to the first benchmark values ​​as a reference quality indicator, or the reference quality indicator is determined according to the implementation method A1 or implementation method A2 mentioned below.

[0108] In implementation A1, any first reference value corresponds to a quality indicator. In this case, step S21 includes: determining the minimum quality indicator from the quality indicators corresponding to the first reference values, and determining the minimum quality indicator as the reference quality indicator.

[0109] In the embodiment of the present application, for single-target minimum optimization, the quality index used to describe the original layout is one, and in this case, one first reference value corresponds to one quality index. For example, if the quality index of the original layout is the mask complexity of the original layout, then one first reference value corresponds to one mask complexity. Single-target minimum optimization refers to: when the layout quality detection tool performs a quality detection on a certain aspect of the original layout, the quality index corresponding to that aspect is obtained. In other words, single-target minimum optimization refers to optimizing a single quality index in the layout quality detection tool.

[0110] Based on this, the M first benchmark values ​​correspond to M quality indicators, which are (y1, ..., y m ,...,y M ). The minimum quality index can be determined from the quality indexes corresponding to the first reference values, and the minimum quality index is determined as the reference quality index, that is, the reference quality index is f m =min(y1, ..., y m ,...,y M ).

[0111] In implementation A2, any first reference value corresponds to multiple quality indicators. In this case, step S21 includes steps S211 to S213 (not shown in the figure).

[0112] In an embodiment of the present application, for multi-objective minimum optimization, there are multiple quality indicators used to describe the original layout. In this case, a first benchmark value corresponds to multiple quality indicators. For example, the quality indicators of the original layout include the mask complexity of the original layout and the edge distance error of the original layout, then a first benchmark value corresponds to a mask complexity and an edge distance error. Multi-objective minimum optimization refers to: optimizing the quality indicators corresponding to various aspects when the layout quality detection tool performs multiple quality tests on the original layout. In other words, multi-objective minimum optimization refers to optimizing multiple quality indicators in the layout quality detection tool.

[0113] It should be noted that there are a large number of adjustable parameters in the layout quality detection tool, and these adjustable parameters will affect the quality detection results of the layout. Generally, there are multiple quality indicators used to characterize the quality of the layout, and there is a balance between the various quality indicators. For example, the smaller the error of the mask, the greater the mask complexity, and the more difficult it is to manufacture. Therefore, the complexity of the mask and the error of the mask balance each other. The Pareto effect is formed between the mask complexity and the mask error, and the Pareto optimal solution set needs to be determined. In some embodiments, the Pareto optimal solution set is also called the Pareto front. When paying attention to a certain quality indicator, the solution on the Pareto front that is biased towards the quality indicator can be determined, thereby improving the accuracy of the quality indicator.

[0114] Please refer to Figure 3, which is a schematic diagram of the Pareto front between mask error and mask complexity, provided in an embodiment of the present application. As can be seen from Figure 3, on the Pareto front, if mask error A is greater than mask error B, then mask complexity A is less than mask complexity B. A and B are data on the Pareto front, which includes mask error and mask complexity. When focusing more on mask error, data on the Pareto front that favors mask error can be determined (i.e., the parameter values ​​enclosed by the ellipse in Figure 3).

[0115] Step S211 : for any quality indicator among the multiple quality indicators, determine the maximum value and the minimum value of any quality indicator from any quality indicator corresponding to each first reference value.

[0116] In the embodiment of the present application, assuming that there are M (M is a positive integer) quality indicators, each first reference value corresponds to M quality indicators. Taking any first reference value as the i-th first reference value and any quality indicator as the m-th quality indicator as an example, the m-th quality indicator corresponding to the i-th first reference value can be recorded as y m,i The maximum value of the mth quality indicator can be determined from the mth quality indicators corresponding to the first reference values. and the minimum value of the mth quality index

[0117] Step S212: For any quality indicator corresponding to any first benchmark value, determine the priority indicator of any quality indicator corresponding to any first benchmark value based on the first selected quality indicator, the weight of the first selected quality indicator, the maximum value and the minimum value of the first selected quality indicator. The first selected quality indicator is at least one quality indicator among the multiple quality indicators corresponding to any first benchmark value, and the first selected quality indicator includes any quality indicator corresponding to any first benchmark value.

[0118] In an embodiment of the present application, the electronic device can obtain the weight of any quality indicator, and the weight is used to describe the importance of the quality indicator. That is, the greater the weight of any quality indicator, the more important the quality indicator is, and the more it can reflect the quality of the original layout. The embodiment of the present application does not limit the method for obtaining the weight of any quality indicator. For example, the electronic device can obtain the weight of any input quality indicator, or the electronic device can determine the weight of any quality indicator based on the number of quality indicators. For example, if there are two quality indicators, the weight of each quality indicator is 0.5.

[0119] Optionally, the quality indicators are sorted to obtain sorted quality indicators. For example, the quality indicators can be sorted manually or randomly. When calculating the priority indicator of any quality indicator corresponding to any first benchmark value, the quality indicator and the quality indicators ranked after the quality indicator can be used for calculation. The quality indicator used to calculate the priority indicator of any quality indicator corresponding to any first benchmark value is called the first selected quality indicator. That is to say, for the first selected quality indicator in step S212 (that is, at least one quality indicator among the multiple quality indicators corresponding to any first benchmark value), when any quality indicator corresponding to any first benchmark value is the last quality indicator after sorting, the first selected quality indicator (that is, at least one quality indicator among the multiple quality indicators corresponding to any first benchmark value) includes any quality indicator corresponding to any first benchmark value; when any quality indicator corresponding to any first benchmark value is not the last quality indicator after sorting, the first selected quality indicator (that is, at least one quality indicator among the multiple quality indicators corresponding to any first benchmark value) includes any quality indicator corresponding to any first benchmark value and each quality indicator corresponding to any first benchmark value that is sorted after any quality indicator.

[0120] Assuming that each first benchmark value corresponds to M quality indicators, taking any first benchmark value as the i-th first benchmark value and any quality indicator as the m-th quality indicator as an example, at least one quality indicator among the multiple quality indicators corresponding to the i-th first benchmark value includes the m-th to M-th quality indicators corresponding to the i-th first benchmark value.

[0121] In an embodiment of the present application, the electronic device may determine the priority index of the mth quality index corresponding to the i-th first reference value according to formula (2) shown below.

[0122] Among them, d iThe priority index of the mth quality index corresponding to the i-th first benchmark value. M represents the number of quality indicators. ω m Characterize the weight of the mth quality indicator, y m,i Characterizes the mth quality index corresponding to the i-th first benchmark value, Characterizes the maximum value of the mth quality index, Characterizes the minimum value of the mth quality indicator.

[0123] In formula (2), when determining the priority indicator of the mth quality indicator corresponding to the i-th first benchmark value, the m-th to M-th quality indicators corresponding to the i-th first benchmark value need to be used. Therefore, it can be seen that formula (2) determines the priority indicator of the mth quality indicator corresponding to the i-th first benchmark value under the premise of considering the checks and balances between the M quality indicators, thereby improving the accuracy of the quality indicator priority. When determining the reference quality indicator based on the priority indicator of the quality indicators in step S213, the reference quality indicator approaches the Pareto optimal solution of the M quality indicators, thereby improving the accuracy of the reference quality indicator.

[0124] In some embodiments, after ranking the quality indicators, when calculating the priority indicator of any quality indicator corresponding to any first benchmark value, the quality indicator and the quality indicators ranked before the quality indicator can also be used. That is, for the first selected quality indicator in step S212 (i.e., at least one quality indicator among the multiple quality indicators corresponding to any first benchmark value), when any quality indicator corresponding to any first benchmark value is the first quality indicator after ranking, the first selected quality indicator (i.e., at least one quality indicator among the multiple quality indicators corresponding to any first benchmark value) includes any quality indicator corresponding to any first benchmark value; when any quality indicator corresponding to any first benchmark value is not the first quality indicator after ranking, the first selected quality indicator (i.e., at least one quality indicator among the multiple quality indicators corresponding to any first benchmark value) includes any quality indicator corresponding to any first benchmark value and each quality indicator ranked before any quality indicator corresponding to any first benchmark value.

[0125] Assuming that each first benchmark value corresponds to M quality indicators, taking any first benchmark value as the i-th first benchmark value and any quality indicator as the m-th quality indicator as an example, at least one quality indicator among the multiple quality indicators corresponding to the i-th first benchmark value includes the 1st to m-th quality indicators corresponding to the i-th first benchmark value.

[0126] Step S213 : determining a reference quality indicator from the quality indicators corresponding to the first reference values ​​based on the priority indicators of the quality indicators corresponding to the first reference values.

[0127] In the embodiment of the present application, the priority indicators of the quality indicators corresponding to the first reference values ​​can be determined in accordance with step S212. Then, the minimum priority indicator is selected from these priority indicators, and the quality indicator corresponding to the minimum priority indicator is determined as the reference quality indicator. In other words, the reference quality indicator f m is the smallest d i Alternatively, a priority indicator not greater than the priority threshold is selected from these priority indicators, and the quality indicator corresponding to the priority indicator not greater than the priority threshold is determined as the reference quality indicator. In this way, the reference quality indicator is determined from each quality indicator corresponding to each first benchmark value.

[0128] For example, each quality indicator corresponding to each first reference value is used as a training set, and the reference quality indicator f determined by implementation method A2 is m , is the evaluation result corresponding to the most balanced parameter of the Pareto front of the training set. By calculating the Pareto front in the training set and calculating the weighted Euclidean distance d on the Pareto front i , through d i Determine the reference quality index f m , which can improve f m The accuracy of the reference index of the first candidate value is improved.

[0129] Step S22: Determine a normal distribution function of any first candidate value based on the reference quality indicator, the mean value, and the variance corresponding to any first candidate value.

[0130] In the embodiment of the present application, the difference between the reference quality indicator and the mean corresponding to any first candidate value can be determined; the quotient of the difference and the variance corresponding to any first candidate value can be used as the normal distribution function of any first candidate value. In other words, the difference between the reference quality indicator and the mean corresponding to any first candidate value can be divided by the variance corresponding to any first candidate value to obtain the normal distribution function of any first candidate value. Optionally, the normal distribution function z of any first candidate value m It can be expressed as: m =(f m -μ m (x)) / σ m (x), where f m is the reference quality indicator, μ m (x) is the mean value corresponding to any first candidate value, σm (x) is the variance corresponding to any first candidate value.

[0131] Step S23: Determine a reference index for any first candidate value based on the variance corresponding to any first candidate value, the normal distribution function corresponding to any first candidate value, the cumulative density function and the probability density function of the normal distribution function corresponding to any first candidate value.

[0132] In an embodiment of the present application, the normal distribution function corresponding to any first candidate value and the cumulative density function of the normal distribution function can be multiplied to obtain a multiplication result. The multiplication result is added to the probability density function of the normal distribution function corresponding to any first candidate value to obtain an addition result. The addition result is multiplied by the variance corresponding to any first candidate value to obtain a reference index for any first candidate value. Optionally, the reference index EI of any first candidate value is m (x) can be expressed as: Among them, σ m (x) represents the variance corresponding to any first candidate value, z m Characterize the normal distribution function of any first candidate value, Φ(z m ) characterizes z m The cumulative density function of Characterization z m The probability density function of .

[0133] Reference index EI of any first candidate value m (x) consists of two terms. The first term is σ m (x)z m Φ(z m ), which tends to find a better result than the current optimal value, representing a local search during the detection process. The second term is This term tends to search for regions with greater uncertainty within the space of the first candidate value, representing a global search during the exploration process. Determining the reference metric for any first candidate value through the first and second terms enables both global and local searches during the exploration process, improving the accuracy of the reference metric for the first candidate value.

[0134] Determining the reference indicator for the first candidate value by considering statistical distribution and statistical information, etc., is conducive to ensuring the accuracy of the reference indicator for the first candidate value. In addition, in the process of determining the reference indicator for the first candidate value, different situations in which the quality indicator corresponding to the first benchmark value is one or more are considered. In different situations, the reference indicator for the first candidate value is determined in different ways. The reliability of the method for determining the reference indicator for the first candidate value is higher, thereby further ensuring the accuracy of the reference indicator for the first candidate value, and thus improving the accuracy of the target value finally determined.

[0135] Step 2032: Determine a first reference value from multiple first candidate values ​​based on the reference indicators of the respective first candidate values.

[0136] In an embodiment of the present application, the reference index of each first candidate value can be determined in accordance with step 2031. For any first candidate value, it can be determined whether the first candidate value can be used as the first reference value based on the reference index of the first candidate value, thereby determining the first reference value from multiple first candidate values. Optionally, the first reference value is determined in accordance with implementation method B1 or implementation method B2 shown below. Exemplarily, the number of first reference values ​​is at least one, that is, based on the reference index of each first candidate value, at least one first reference value is determined from multiple first candidate values.

[0137] In implementation B1, any first candidate value corresponds to a reference indicator. In this case, step 2032 includes: based on the reference indicators of the first candidate values, selecting a first candidate value whose reference indicator is not less than a set indicator threshold from the multiple first candidate values ​​as the first reference value.

[0138] For single-objective minimum optimization, there is one quality indicator. According to the content of the first mapping relationship in step S1, it can be concluded that there is one first mapping relationship. According to the content of the reference indicator of the first candidate value in step S2, one first candidate value corresponds to one reference indicator.

[0139] For any first candidate value, if the reference index of the first candidate value is not less than the set index threshold, the first candidate value is used as the first reference value; if the reference index of the first candidate value is less than the set index threshold, the first candidate value is not the first reference value. In this way, the first reference value can be determined.

[0140] The embodiment of the present application does not limit the method for determining the threshold value of the set indicator. For example, the threshold value of the set indicator is a value set according to experience. Alternatively, the reference indicators of each first candidate value are sorted from large to small or from small to large to obtain the sorted reference indicators, and the set number of reference indicators after sorting are used as the set indicator threshold value. In this case, it is equivalent to sorting multiple first candidate values ​​from large to small according to the reference indicators, and selecting the first set number of first candidate values ​​as the first reference values; or it is equivalent to sorting multiple first candidate values ​​from small to large according to the reference indicators, and selecting the set number of first candidate values ​​and the first candidate values ​​located after the set number of first candidate values ​​as the first reference values.

[0141] In implementation B2, any first candidate value corresponds to multiple reference indicators. In this case, step 2032 includes steps 20321 to 20324 (not shown in the figure).

[0142] For multi-objective minimum optimization, there are multiple quality indicators. Based on the first mapping relationship in step S1, it can be concluded that each quality indicator corresponds to a first mapping relationship. Based on the reference indicator for the first candidate value in step S2, it can be seen that a reference indicator corresponding to a first candidate value can be determined through a first mapping relationship. Based on this, a first candidate value corresponds to multiple reference indicators.

[0143] Step 20321: For any reference indicator among the multiple reference indicators, determine the maximum value and the minimum value of any reference indicator from any reference indicator corresponding to each first candidate value.

[0144] In the embodiment of the present application, assuming that there are M (M is a positive integer) reference indicators, each first candidate value corresponds to M reference indicators. Taking any first candidate value as the i-th first candidate value and any reference indicator as the m-th reference indicator as an example, the m-th reference indicator corresponding to the i-th first candidate value can be recorded as EI m,i The maximum value of the mth reference index can be determined from the mth reference indexes corresponding to the first candidate values. and the minimum value of the mth reference index

[0145] Step 20322. For any reference indicator corresponding to any first candidate value, determine the priority indicator of any reference indicator corresponding to any first candidate value based on the selected reference indicator, the maximum value and the minimum value of the selected reference indicator. The selected reference indicator is at least one reference indicator among the multiple reference indicators corresponding to any first candidate value. The selected reference indicator includes any reference indicator corresponding to any first candidate value.

[0146] Optionally, the reference indicators are sorted to obtain sorted reference indicators. For example, the reference indicators can be sorted manually or randomly. When calculating the priority index of any reference indicator corresponding to any first candidate value, the reference indicator and the reference indicators ranked after the reference indicator can be used for calculation. The reference indicator used to calculate the priority indicator of any reference indicator corresponding to any first candidate value is called the selected reference indicator. That is, for the selected reference indicator in step 20322 (that is, at least one reference indicator among the multiple reference indicators corresponding to any first candidate value), when any reference indicator corresponding to any first candidate value is the last reference indicator after sorting, the selected reference indicator (that is, at least one reference indicator among the multiple reference indicators corresponding to any first candidate value) includes any reference indicator corresponding to any first candidate value; when any reference indicator corresponding to any first candidate value is not the last reference indicator after sorting, the selected reference indicator (that is, at least one reference indicator among the multiple reference indicators corresponding to any first candidate value) includes any reference indicator corresponding to any first candidate value and each reference indicator corresponding to any first candidate value that is sorted after any reference indicator.

[0147] Assuming that each first candidate value corresponds to M reference indicators, taking any first candidate value is the i-th first candidate value and any reference indicator is the m-th reference indicator as an example, at least one reference indicator among the multiple reference indicators corresponding to the i-th first candidate value includes the m-th to M-th reference indicators corresponding to the i-th first candidate value.

[0148] In an embodiment of the present application, the electronic device can obtain the weight of any reference indicator, and the weight is used to describe the importance of the reference indicator. In other words, the greater the weight of any reference indicator, the more important the reference indicator. In an embodiment of the present application, the method for obtaining the weight of any reference indicator is not limited. For example, the electronic device can obtain the weight of any input reference indicator, or the electronic device can determine the weight of any reference indicator based on the number of reference indicators. For example, if there are two reference indicators, the weight of each reference indicator is 0.5.

[0149] Optionally, the electronic device may determine the priority index of the mth reference index corresponding to the i-th first candidate value according to formula (3) shown below.

[0150] Among them, d ′ iThe priority index of the mth reference index corresponding to the i-th first candidate value. M represents the number of reference indicators. ω M-m Characterizes the weight of the Mm-th reference index, EI m,i Represents the mth reference index corresponding to the i-th first candidate value, Represents the maximum value of the mth reference index, Represents the minimum value of the mth reference indicator.

[0151] In some embodiments, after sorting the reference indicators, when calculating the priority indicator of any reference indicator corresponding to any first candidate value, the reference indicator and the reference indicator ranked before the reference indicator can be used for calculation. That is, for the selected reference indicator in step 20322 (i.e., at least one reference indicator among the multiple reference indicators corresponding to any first candidate value), when any reference indicator corresponding to any first candidate value is the first reference indicator after sorting, the selected reference indicator (i.e., at least one reference indicator among the multiple reference indicators corresponding to any first candidate value) includes any reference indicator corresponding to any first candidate value; when any reference indicator corresponding to any first candidate value is not the first reference indicator after sorting, the selected reference indicator (i.e., at least one reference indicator among the multiple reference indicators corresponding to any first candidate value) includes any reference indicator corresponding to any first candidate value and each reference indicator ranked before any reference indicator corresponding to any first candidate value.

[0152] Assuming that each first candidate value corresponds to M reference indicators, taking any first candidate value is the i-th first candidate value and any reference indicator is the m-th reference indicator as an example, at least one reference indicator among the multiple reference indicators corresponding to the i-th first candidate value includes the 1st to m-th reference indicators corresponding to the i-th first candidate value.

[0153] Step 20323: Based on the priority indicators of the reference indicators corresponding to the first candidate values, determine the target indicator from the reference indicators corresponding to the first candidate values.

[0154] In an embodiment of the present application, the priority indicators of each reference indicator corresponding to each first candidate value can be determined in the manner of step 20322. Then, these priority indicators are sorted, and the reference indicators corresponding to each priority indicator located before the specified position after sorting are determined as target indicators. Alternatively, these priority indicators are sorted, and each priority indicator after sorting is randomly or evenly divided into multiple segments of indicators, and each segment of indicator includes at least one priority indicator after sorting. For any segment of indicator, each priority indicator included in the segment of indicator is sampled, and the reference indicator corresponding to the sampled priority indicator is determined as the target indicator. For example, the minimum priority indicator is sampled from each priority indicator included in a segment of indicator, and the reference indicator corresponding to the minimum priority indicator is determined as the target indicator. In this way, the target indicator is sampled from each segment of indicator. The number of target indicators is one or more, in other words, the number of target indicators is at least one, that is, based on the priority indicators of each reference indicator corresponding to each first candidate value, at least one target indicator is determined from each reference indicator corresponding to each first candidate value.

[0155] Step 20324: determine the first candidate value corresponding to the target indicator as the first reference value.

[0156] That is, for any target indicator, the first candidate value corresponding to the target indicator is determined as a first reference value. Since there is at least one target indicator, at least one first reference value can be determined in this way.

[0157] For example, the reference indicators corresponding to the first candidate values ​​are used as the training set, and the target indicator determined by implementation method B2 is the evaluation result corresponding to the most balanced parameter of the Pareto front of the training set. By calculating the Pareto front in the training set and calculating the weighted Euclidean distance d on the Pareto front ′ i , through d ′ i Determining the target indicator can improve the accuracy of the target indicator, thereby improving the accuracy of the first reference value.

[0158] Next, based on the multiple first benchmark values ​​and the first reference values, multiple second benchmark values ​​are determined. For example, one first benchmark value can be one second benchmark value, and one first reference value can be one second benchmark value. That is, each first benchmark value and each first reference value can be used as a second benchmark value. Alternatively, any first benchmark value can be adjusted based on the first reference value to obtain an adjusted benchmark value corresponding to the first benchmark value, and the adjusted benchmark values ​​corresponding to each first benchmark value can be used as the multiple second benchmark values.

[0159] Exemplarily, the number of first reference values ​​is at least one, and any first benchmark value is adjusted based on the first reference value to obtain the adjusted benchmark value corresponding to any first benchmark value. The method can be: taking the average value of at least one first reference value as the first numerical value, adjusting any first benchmark value based on the first numerical value, and obtaining the adjusted benchmark value corresponding to any first benchmark value. For example, the average value of the first numerical value and any first benchmark value is taken as the adjusted benchmark value corresponding to any first benchmark value. For another example, the first numerical value and any first benchmark value are weighted averaged to obtain the adjusted benchmark value corresponding to any first benchmark value. In the process of weighted averaging the first numerical value and any first benchmark value, the weight corresponding to the first numerical value and the weight corresponding to any first benchmark value can be set according to experience, or flexibly adjusted according to the application scenario, and the embodiments of the present application do not limit this.

[0160] In the process of determining the first reference value, different situations are taken into consideration, namely, whether the first candidate value corresponds to one or more reference indicators. In different situations, the first reference value is determined in different ways. The reliability of the method for determining the first reference value is higher, which is conducive to ensuring the accuracy of the first reference value, and thereby improving the accuracy of the target value finally determined.

[0161] Step 204 : Performing a quality inspection on the original layout using a layout quality inspection tool according to the plurality of second benchmark values ​​to obtain a quality index corresponding to each second benchmark value.

[0162] In an embodiment of the present application, by making the target parameter include any second benchmark value, a layout quality detection tool is determined, and the original layout is input into the layout quality detection tool. The original layout is subjected to quality detection processing by the layout quality detection tool to obtain a quality index of the original layout corresponding to the second benchmark value. There is at least one quality index, that is, any second benchmark value corresponds to at least one quality index, and the result of the quality detection of the original layout includes the quality index corresponding to each second benchmark value. The quality index corresponding to the second benchmark value in the embodiment of the present application refers to the quality index of the original layout corresponding to the second benchmark value.

[0163] It is understandable that the implementation of step 204 is similar to the implementation of step 202. Please refer to the description of step 202 and no further details will be given here.

[0164] Step 205: When the quality indicators corresponding to the respective second benchmark values ​​meet the reference conditions, the target values ​​of the target parameters are determined based on the respective second benchmark values, and the layout quality inspection tool with the target parameter value as the target value is used as the layout quality inspection tool after parameter adjustment. The layout quality inspection tool after parameter adjustment is used to determine the quality indicators of the layout to be inspected.

[0165] The fact that the quality indicators corresponding to the respective second benchmark values ​​satisfy the reference conditions means that the quality indicators corresponding to all the second benchmark values ​​satisfy the reference conditions. The embodiments of the present application do not limit the quality indicators corresponding to any one of the second benchmark values ​​to satisfying the reference conditions. Exemplarily, based on the quality indicators corresponding to any one of the second benchmark values ​​and the reference threshold, it is determined whether the quality indicators corresponding to the second benchmark value satisfy the reference conditions. Alternatively, the number of iterations when determining the quality indicators corresponding to the respective second benchmark values ​​is obtained, and based on the number of iterations and the number threshold, it is determined whether the quality indicators corresponding to the respective second benchmark values ​​satisfy the reference conditions.

[0166] It can be understood that the content of the quality indicator corresponding to any second benchmark value meeting the reference condition is similar to the content of the quality indicator corresponding to any first benchmark value meeting the reference condition, which can be seen in the description of step 203 and will not be repeated here.

[0167] If the quality indicators corresponding to the respective second benchmark values ​​meet the reference conditions, the respective second benchmark values ​​can be determined as the target values ​​of the target parameters. Alternatively, for single-target minimum optimization, one second benchmark value corresponds to one quality indicator, and according to the implementation principle of implementation method B1, the respective second benchmark values ​​whose quality indicators are greater than or less than the set threshold value can be selected from the plurality of second benchmark values, and the selected second benchmark values ​​can be used as the respective target values. For example, a second benchmark value corresponds to a quality indicator for characterizing the edge distance error of the original layout, and the second benchmark value with the smallest quality indicator is selected from the plurality of second benchmark values ​​as the target value. The implementation principle of the aforementioned content can be found in implementation method B1 and will not be repeated here.

[0168] For multi-objective minimum optimization, a second benchmark value corresponds to multiple quality indicators. According to the implementation principle of implementation method B2, for any of the multiple quality indicators, the maximum and minimum values ​​of any quality indicator can be determined from any quality indicator corresponding to each second benchmark value. For any quality indicator corresponding to any second benchmark value, a priority indicator for the quality indicator corresponding to any second benchmark value is determined based on a second selected quality indicator and the maximum and minimum values ​​of the second selected quality indicator. The second selected quality indicator is at least one quality indicator from the multiple quality indicators corresponding to any second benchmark value, and the second selected quality indicator includes any quality indicator corresponding to any second benchmark value. Based on the priority indicators of each quality indicator corresponding to each second benchmark value, a target quality indicator is determined from each quality indicator corresponding to each second benchmark value, and the second benchmark value corresponding to the target quality indicator is determined as the target value. The number of target quality indicators is one or more, in other words, the number of target quality indicators is at least one. That is, based on the priority indicators of each quality indicator corresponding to each second benchmark value, at least one target quality indicator is determined from each quality indicator corresponding to each second benchmark value. The target value determined in this way is a value in the Pareto optimal solution set of each second benchmark value. The implementation principle of the above content can be seen in implementation method B2, which will not be repeated here.

[0169] Alternatively, for multi-objective minimum optimization, a second benchmark value corresponds to multiple quality indicators. By annotating the original layout, the indicator ranges corresponding to the quality indicators of the original layout are determined. For example, through manual annotation, the mask complexity of the original layout is determined to be between 0.2 and 0.6, and the edge error is between 0.1 and 0.3. If the quality indicators corresponding to a second benchmark value are within the corresponding indicator ranges, the second benchmark value is determined to be the target value. In this way, at least one target value for the target parameter can be determined.

[0170] In a possible implementation, the method of the embodiment of the present application further includes steps 206 to 208 (not shown in the figure), which are performed after step 204.

[0171] Step 206: When the quality indicators corresponding to the respective second benchmark values ​​do not meet the reference conditions, a second reference value is determined from a plurality of second candidate values ​​based on the quality indicators corresponding to the respective second benchmark values; and a plurality of third benchmark values ​​are determined based on the plurality of second benchmark values ​​and the second reference values, where the plurality of second candidate values ​​include the first candidate values ​​among the plurality of first candidate values ​​excluding the first reference value.

[0172] In the embodiment of the present application, each first candidate value other than the first reference value among the plurality of first candidate values ​​can be used as each second candidate value. If there is a quality indicator corresponding to the second reference value that does not meet the reference condition, according to an implementation principle similar to step 203, a second reference value is determined from each second candidate value, and each third reference value is determined based on each second reference value and the second reference value. This part can be found in the description of step 203 and will not be repeated here.

[0173] The number of second reference values ​​is one or more, in other words, the number of second reference values ​​is at least one, that is, based on the quality indicators corresponding to each second reference value, at least one second reference value is determined from multiple second candidate values, and based on the multiple second reference values ​​and at least one second reference value, multiple third reference values ​​are determined, and the multiple second candidate values ​​include the first candidate values ​​among the multiple first candidate values ​​except for at least one first reference value.

[0174] Optionally, in step 206, "based on the quality indicators corresponding to each second benchmark value, determine a second reference value from multiple second candidate values, and based on the multiple second benchmark values ​​and the second reference values, determine multiple third benchmark values" includes steps 2061 to 2065 (not shown in the figure).

[0175] Step 2061: Adjust the first value range of the target parameter to obtain the second value range of the target parameter.

[0176] In the embodiment of the present application, the first value range can be flexibly selected to be expanded, narrowed, or not adjusted according to actual conditions. When the first value range is expanded or narrowed, the expanded or narrowed first value range becomes the second value range.

[0177] Optionally, step 2061 includes: determining a first quality indicator whose quality indicator is not greater than a first indicator threshold value from the quality indicators corresponding to each first benchmark value; determining a second quality indicator whose quality indicator is not greater than a second indicator threshold value from the quality indicators corresponding to each second benchmark value; when the first quality indicator is less than the second quality indicator, narrowing the first value range to obtain a second value range; when the first quality indicator is greater than the second quality indicator, expanding the first value range to obtain a second value range.

[0178] The embodiments of the present application do not limit the first indicator threshold and the second indicator threshold. For example, the first indicator threshold and the second indicator threshold are numerical values ​​set based on experience, or the first indicator threshold is the minimum quality indicator among the quality indicators corresponding to the first benchmark values, and the second indicator threshold is the minimum quality indicator among the quality indicators corresponding to the second benchmark values.

[0179] For any quality indicator corresponding to the first benchmark value, if the quality indicator is less than or equal to the first indicator threshold, the quality indicator is determined as the first quality indicator. In this way, at least one first quality indicator is determined from the quality indicators corresponding to the first benchmark values. Similarly, at least one second quality indicator can be determined from the quality indicators corresponding to the second benchmark values.

[0180] It is understood that if the first indicator threshold is the minimum quality indicator among the quality indicators corresponding to the first benchmark values, then the first quality indicator is the minimum quality indicator among the quality indicators corresponding to the first benchmark values. Similarly, the second quality indicator may also be the minimum quality indicator among the quality indicators corresponding to the second benchmark values.

[0181] Next, based on each first quality indicator and each second quality indicator, it is determined whether to scale the first value range. Optionally, if each first quality indicator is smaller than each second quality indicator, the first value range is reduced; if each first quality indicator is larger than each second quality indicator, the first value range is expanded.

[0182] Exemplarily, the number of first quality indicators may be one or more, the number of second quality indicators may also be one or more, the first quality indicator being less than the second quality indicator may mean that the average value of the first quality indicator is less than the average value of the second quality indicator, and the first quality indicator being greater than the second quality indicator may mean that the average value of the first quality indicator is greater than the average value of the second quality indicator.

[0183] For example, for single-objective optimization or multi-objective optimization, the minimum quality index can be determined from the quality indexes corresponding to the first benchmark values ​​to obtain the first quality index. The minimum quality index can be determined from the quality indexes corresponding to the second benchmark values ​​to obtain the second quality index. If the first quality index is smaller than the second quality index, the first value range is halved with the first benchmark value corresponding to the first quality index as the center, or the second benchmark value corresponding to the second quality index as the center. If the first quality index is larger than the second quality index, the first value range is doubled with the first benchmark value corresponding to the first quality index as the center, or the second benchmark value corresponding to the second quality index as the center.

[0184] Optionally, for multi-objective optimization, whether to scale the first value range can also be determined in the following manner. In step S213, a reference quality indicator is determined from the quality indicators corresponding to the first benchmark values, and the reference quality indicator can be used as the first quality indicator. In addition, according to the implementation principle of step S213, the second quality indicator can be determined from the quality indicators corresponding to the second benchmark values. If the first quality indicator is smaller than the second quality indicator, the first value range is reduced by half with the first benchmark value corresponding to the first quality indicator as the center, or with the second benchmark value corresponding to the second quality indicator as the center. If the first quality indicator is larger than the second quality indicator, the first value range is doubled with the first benchmark value corresponding to the first quality indicator as the center, or with the second benchmark value corresponding to the second quality indicator as the center.

[0185] It is understandable that the embodiment of the present application determines the target value in an iterative manner. Therefore, it is possible to determine whether to scale the value range every certain number of iterations. That is, in N (N is a positive integer, for example, N is 3) iterations, the previous iteration corresponds to the first quality indicator in the embodiment of the present application, and the next iteration corresponds to the second quality indicator in the embodiment of the present application. If the first quality indicator is less than the second quality indicator for each of the two adjacent times, the value range is narrowed. If the first quality indicator is greater than the second quality indicator for two adjacent times, the value range is enlarged.

[0186] It should be noted that the length of the value range after reduction is not less than the first length, and the length of the value range after enlargement is not greater than the second length. In other words, it is necessary to constrain the value range before and after scaling to be reasonable to prevent extreme cases. In other words, the length of the second value range is not less than the first length, and the length of the second value range is not greater than the second length. The first length is the minimum length of a reasonable value range, and the second length is the maximum length of a reasonable value range. The first length and the second length can be set based on experience, and can also be flexibly adjusted according to the application scenario. The embodiments of the present application do not limit this.

[0187] In some embodiments, narrowing the value range is also referred to as narrowing the grid, and enlarging the value range is also referred to as enlarging the grid. By scaling the grid, the value space of the target parameter is optimized. When the second candidate value and the third benchmark value are determined based on the optimized value space later (see steps 2062-2065), the accuracy of the second candidate value and the third benchmark value can be improved, and the efficiency of finding the Pareto optimal solution can be improved, that is, the efficiency of determining the target value can be improved. In addition, it is also possible to avoid blindly increasing the number of values, achieve a balance between computational complexity and computational efficiency, and thus determine the target value with a lower computational complexity and higher computational efficiency.

[0188] Step 2062: Divide the second value range into multiple second value intervals.

[0189] In the embodiment of the present application, the implementation principle of step 2062 is similar to the implementation principle of "dividing the first value range into multiple first value intervals", which can be seen in the description of step 201 and will not be repeated here.

[0190] Step 2063: Sampling is performed in each second value interval to obtain a plurality of first supplementary values ​​and a plurality of second supplementary values.

[0191] In the embodiment of the present application, the method for determining the first supplementary value is similar to the method for determining the first candidate value in step 201, and the method for determining the second supplementary value is similar to the method for determining the first reference value in step 201. Therefore, the implementation principle of step 2063 is similar to the implementation principle of "sampling within each first value interval to obtain multiple first reference values ​​and multiple first candidate values". Please refer to the description of step 201 and will not be repeated here.

[0192] Step 2064: Determine the first candidate values ​​other than the first reference value from the multiple first supplementary values ​​and the multiple first candidate values ​​as multiple second candidate values, and determine the second reference value from the multiple second candidate values ​​based on the quality indicators corresponding to each second reference value.

[0193] In the embodiment of the present application, the plurality of second candidate values ​​include each first supplementary value and first candidate values ​​other than the first reference value among the plurality of first candidate values. The second reference value can be determined from the plurality of second candidate values ​​according to an implementation principle similar to step 203. The number of the first reference value and the number of the second reference value are both at least one, that is, the plurality of first supplementary values ​​and the first candidate values ​​other than the at least one first reference value among the plurality of first candidate values ​​are determined as the plurality of second candidate values, and at least one second reference value is determined from the plurality of second candidate values ​​based on the quality indicators corresponding to the respective second reference values.

[0194] Step 2065: Determine the plurality of second supplementary values, the plurality of second base values, and the second reference value as a plurality of third base values. In other words, the plurality of third base values ​​includes each second supplementary value, each second base value, and the second reference value.

[0195] The number of the second reference values ​​is at least one, that is, the multiple second supplementary values, the multiple second reference values ​​and each second reference value are determined as multiple third reference values.

[0196] Step 207 : Using a layout quality inspection tool to perform quality inspection on the original layout according to the multiple third benchmark values, and obtaining quality indicators corresponding to the respective third benchmark values.

[0197] In an embodiment of the present application, a layout quality detection tool is determined by including any third benchmark value in the target parameter, and the original layout is input into the layout quality detection tool. The original layout is subjected to quality detection processing by the layout quality detection tool to obtain a quality indicator of the original layout corresponding to the third benchmark value. There is at least one quality indicator, that is, any third benchmark value corresponds to at least one quality indicator, and the result of the quality detection of the original layout includes the quality indicators corresponding to each third benchmark value. The quality indicator corresponding to the third benchmark value in the embodiment of the present application refers to the quality indicator of the original layout corresponding to the third benchmark value.

[0198] It is understandable that the implementation of step 207 is similar to the implementation of step 202. Please refer to the description of step 202 and no further details will be given here.

[0199] Step 208 : When the quality indicators corresponding to the respective third reference values ​​satisfy the reference conditions, target values ​​of the target parameters are determined based on the respective third reference values.

[0200] The embodiments of the present application do not limit the quality indicator corresponding to any third benchmark value to satisfying the reference condition. For example, whether the quality indicator corresponding to any third benchmark value satisfies the reference condition can be determined based on the quality indicator corresponding to the third benchmark value and a reference threshold. Alternatively, the number of iterations used to determine the quality indicator corresponding to each third benchmark value is obtained, and based on the number of iterations and the number threshold, whether the quality indicator corresponding to each third benchmark value satisfies the reference condition is determined.

[0201] It can be understood that the content of the quality indicator corresponding to any third benchmark value meeting the reference condition is similar to the content of the quality indicator corresponding to any first benchmark value meeting the reference condition, which can be seen in the description of step 203 and will not be repeated here.

[0202] If the quality indicators corresponding to the various third benchmark values ​​meet the reference conditions, the target value can be determined from the multiple third benchmark values. The method of determining the target value from the multiple third benchmark values ​​can be seen in the description of step 205. The implementation principles of the two are similar and will not be repeated here.

[0203] In the embodiment of the present application, determining the target value of the target parameter is equivalent to completing the parameter adjustment of the layout quality detection tool, and using the layout quality detection tool with the target parameter value as the layout quality detection tool after parameter adjustment, and the accuracy of the layout quality detection tool is relatively high. Subsequently, any chip layout, any mask layout, or any wafer layout can be used as the layout to be tested. According to the target value, the layout quality detection tool after parameter adjustment is used to perform quality inspection on the layout to be tested, and the quality index of the layout to be tested is obtained.

[0204] Optionally, a layout quality detection tool with adjusted parameters is used to simulate the photolithography equipment in an actual scenario, and the layout to be detected is subjected to photolithography processing to obtain a photolithography layout. The quality index of the layout to be detected is obtained by performing pattern analysis on the photolithography layout. For example, when the layout to be detected is a chip layout, the photolithography layout can be at least one of a mask layout and a wafer layout. When the layout to be detected is a mask layout, the photolithography layout can be a wafer layout. If the layout to be detected is a chip layout and the photolithography layout is a wafer layout, in this case, the layout quality detection tool can also be used to determine the quality index of the layout to be detected based on the layout to be detected and the photolithography layout.

[0205] Alternatively, the pattern of the layout to be inspected can be analyzed by a layout quality inspection tool according to the target value to obtain the quality index of the layout to be inspected.

[0206] It should be noted that the method for determining the quality index of the layout to be inspected is similar to the method for determining the quality index of the original layout, which can be seen in the description of step 202 and will not be repeated here.

[0207] Next, based on the relationship between the quality index of the layout to be detected and the target threshold, it is determined whether the layout to be detected is a reference layout. Optionally, if the quality index is negatively correlated with the quality of the layout to be detected, then when the quality index of the layout to be detected is not greater than the target threshold, the layout to be detected is determined to be a reference layout; when the quality index of the layout to be detected is greater than the target threshold, the layout to be detected is determined not to be a reference layout. Similarly, if the quality index is positively correlated with the quality of the layout to be detected, then when the quality index of the layout to be detected is not less than the target threshold, the layout to be detected is determined to be a reference layout; when the quality index of the layout to be detected is less than the target threshold, the layout to be detected is determined not to be a reference layout.

[0208] In the case where it is determined that the layout to be tested is a reference layout. If the reference layout is a chip layout, the photolithography equipment in the actual scene can be called first, and the photolithography process is performed on the mask plate based on the chip layout to obtain the mask plate after the photolithography process. Then, the photolithography equipment in the actual scene is called again to perform photolithography on the wafer based on the mask plate after the photolithography process to obtain the wafer after the photolithography process. If the reference layout is a mask layout, the photolithography equipment in the actual scene can be called, and the photolithography process is performed on the wafer based on the mask plate after the photolithography process corresponding to the mask layout to obtain the wafer after the photolithography process. Of course, when applied, the reference layout can also be a layout on the wafer after the photolithography process, that is, a wafer layout. In some embodiments, the wafer after the photolithography process is also called an integrated circuit wafer, which is a product in the chip manufacturing process.

[0209] When the reference layout is a chip layout or a mask layout, by screening out high-quality reference layouts, high-quality integrated circuit wafers can be manufactured based on the reference layouts, reducing the defective rate of the integrated circuit wafers, thereby reducing the production cost of the integrated circuit wafers. When the reference layout is a wafer layout, by screening out high-quality reference layouts, high-quality integrated circuit wafers can be screened out, reducing the defective rate of the integrated circuit wafers.

[0210] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.) and signals involved in this application are all authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant regions. For example, the original map and the first value range involved in this application were obtained with full authorization.

[0211] In the above method, the target parameters of the layout quality inspection tool include various first benchmark values. The layout quality inspection tool is used to perform quality inspection on the original layout to obtain quality indicators corresponding to each first benchmark value. If there is a quality indicator corresponding to the first benchmark value that does not meet the reference condition, then based on the quality indicator corresponding to each first benchmark value, a first reference value is determined from multiple first candidate values ​​of the target parameter. Based on the multiple first benchmark values ​​and the first reference value, multiple second benchmark values ​​of the target parameter are determined. By determining the second benchmark value, the accuracy of the value space of the target parameter is improved. Next, the layout quality inspection tool is used to perform quality inspection on the original layout to obtain quality indicators corresponding to each second benchmark value. By improving the accuracy of the value space of the target parameter, the accuracy of the layout quality detection tool is improved, thereby improving the accuracy of the quality index. When the quality index corresponding to each second benchmark value meets the reference condition, the target value of the target parameter can be accurately determined based on each second benchmark value, thereby ensuring the accuracy of the layout quality detection tool, which is conducive to improving the accuracy of the quality index of the layout to be detected determined by the layout quality detection tool, thereby more accurately selecting a higher quality layout from the layout to be detected.

[0212] Furthermore, sampling within multiple first value intervals to obtain first benchmark values ​​and first candidate values ​​improves the uniformity of their distribution and ensures their comprehensiveness. Determining each first benchmark value through clustering improves the accuracy of the first benchmark values ​​and accelerates the efficiency of determining the target value.

[0213] The quality index of the original layout corresponding to each first benchmark value is calculated through the error between the original layout and the predicted layout corresponding to each first benchmark value, so that the quality index of the original layout corresponding to each first benchmark value can reflect the accuracy of the layout quality detection tool, thereby adjusting the parameter values ​​of the layout quality detection tool in a more accurate direction based on the quality index of the original layout corresponding to each first benchmark value, thereby ensuring the accuracy of the layout quality detection tool.

[0214] Determining the reference indicator for the first candidate value by considering statistical distribution and statistical information, etc., is conducive to ensuring the accuracy of the reference indicator for the first candidate value. In addition, in the process of determining the reference indicator for the first candidate value, different situations in which the quality indicator corresponding to the first benchmark value is one or more are considered. In different situations, the reference indicator for the first candidate value is determined in different ways. The reliability of the method for determining the reference indicator for the first candidate value is higher, thereby further ensuring the accuracy of the reference indicator for the first candidate value, and thus improving the accuracy of the target value finally determined.

[0215] In the process of determining the first reference value, different situations are taken into consideration, namely, whether the first candidate value corresponds to one or more reference indicators. In different situations, the first reference value is determined in different ways. The reliability of the method for determining the first reference value is higher, which is conducive to ensuring the accuracy of the first reference value, and thereby improving the accuracy of the target value finally determined.

[0216] By scaling the target parameter's range, the target parameter's value space is optimized. When determining the second candidate value and the third benchmark value based on this optimized value space, the accuracy of these two values ​​can be improved, thereby increasing the efficiency of determining the target value. Furthermore, this avoids blindly increasing the number of values, achieving a balance between computational effort and efficiency, thereby determining the target value with low computational effort and high efficiency.

[0217] The above describes the method provided by the embodiment of the present application from the perspective of method steps. The following describes it in conjunction with scenarios. The embodiment of the present application provides two scenarios, namely, the scenario of determining the parameters of the layout quality detection tool for a single quality indicator and the scenario of determining the parameters of the layout quality detection tool for multiple quality indicators. These two scenarios are described below.

[0218] First, we will introduce a scenario where parameters for a layout quality inspection tool are determined based on a single quality indicator. Taking the quality indicator being error as an example, see Figure 4 , which is a schematic diagram of a process for determining parameters for a layout quality inspection tool based on a single quality indicator, as provided in an embodiment of the present application. The process includes steps 1 to 4 as shown below.

[0219] Step 1: Determine the benchmark value and candidate values ​​through clustering algorithm.

[0220] In an embodiment of the present application, multiple initial values ​​can be obtained, and clusters can be obtained by clustering the multiple initial values. For any cluster, the initial value with the smallest distance from the cluster center of the cluster is selected from the cluster, and the selected initial value is used as the reference value, and the unselected initial value is used as the candidate value. In this way, the multiple initial values ​​are divided into multiple reference values ​​and multiple candidate values. The implementation method of step 1 can be found in the description of step 201 and will not be repeated here.

[0221] Step 2: By making the target parameters of the layout quality inspection tool include multiple benchmark values, the layout quality inspection tool is used to perform quality inspection on the original layout to obtain the error of the original layout corresponding to each benchmark value.

[0222] In an embodiment of the present application, the layout quality detection tool is determined by setting the target parameter of the layout quality detection tool to any benchmark value. The original layout is input into the layout quality detection tool, and the layout quality detection tool outputs a predicted layout corresponding to the benchmark value. Next, the predicted layout corresponding to the benchmark value is input into the layout quality detection tool, and the layout quality detection tool determines the error of the original layout corresponding to the benchmark value. In this way, the error corresponding to each benchmark value can be determined. The implementation method of step 2 can be seen in the description of step 202 and will not be repeated here.

[0223] If the errors corresponding to the respective benchmark values ​​are less than or equal to the expected value, a target value is determined based on the respective benchmark values. The method for determining the target value can be found in the description of step 205 and will not be repeated here. If the errors corresponding to the respective benchmark values ​​are greater than the expected value, step 3 is executed.

[0224] Step 3: Determine the Gaussian process regression model based on the errors corresponding to each benchmark value.

[0225] In the embodiment of the present application, the errors corresponding to the respective benchmark values ​​satisfy a Gaussian distribution. In this case, a Gaussian process regression model is used to describe the mapping relationship between the benchmark values ​​and the mean and variance of the Gaussian distribution. The implementation principle of step 3 can be found in the description of step S1 and will not be repeated here.

[0226] Step 4: Based on the Gaussian process regression model and multiple candidate values, determine the reference index of each candidate value, and select a benchmark value from the multiple candidate values ​​based on the reference index of each candidate value.

[0227] In the embodiment of the present application, since the Gaussian process regression model describes the mapping relationship between the reference value and the mean of the Gaussian distribution and the variance of the Gaussian distribution, the Gaussian process regression model can be used to map any candidate value to the corresponding mean and variance. Then, based on the mean and variance corresponding to any candidate value, the reference index of the candidate value is determined, and then whether the candidate value is a reference value is determined based on the reference index of the candidate value. In this way, the selection of a reference value from multiple candidate values ​​is achieved. The implementation principle of step 4 can be found in the description of step S2, implementation method A1, and implementation method B1, and will not be repeated here.

[0228] Next, consider the benchmark values ​​selected in step 4 and the benchmark values ​​determined in step 1 as a set of benchmark values. Step 2 is then repeated based on these benchmark values ​​to determine the errors corresponding to each benchmark value. If the errors corresponding to each benchmark value are less than or equal to the expected value, the target value is determined based on each benchmark value. If the errors corresponding to each benchmark value are greater than the expected value, steps 3, 4, and 2 are repeated until the target value is determined.

[0229] Next, we will introduce a scenario where parameters for a layout quality inspection tool are determined based on multiple quality indicators. For example, if the multiple quality indicators are two quality indicators, and the two quality indicators are error and complexity, please refer to Figure 5. Figure 5 is a schematic diagram of a process for determining parameters for a layout quality inspection tool based on two quality indicators, provided in an embodiment of the present application. The process includes steps 1 to 4 as shown below.

[0230] Step 1: Determine the benchmark value and candidate values ​​through clustering algorithm.

[0231] In an embodiment of the present application, multiple initial values ​​can be obtained, and clusters can be obtained by clustering the multiple initial values. For any cluster, the initial value with the smallest distance from the cluster center of the cluster is selected from the cluster, and the selected initial value is used as the reference value, and the unselected initial value is used as the candidate value. In this way, the multiple initial values ​​are divided into multiple reference values ​​and multiple candidate values. The implementation method of step 1 can be found in the description of step 201 and will not be repeated here.

[0232] Step 2: By making the target parameters of the layout quality inspection tool include multiple benchmark values, the layout quality inspection tool is used to perform quality inspection on the original layout to obtain the error and complexity of the original layout corresponding to each benchmark value.

[0233] In an embodiment of the present application, by making the target parameter of the layout quality detection tool any benchmark value, the layout quality detection tool is determined. The original layout is input into the layout quality detection tool, and the predicted layout corresponding to the benchmark value is output by the layout quality detection tool. Then, the predicted layout corresponding to the benchmark value is input into the layout quality detection tool. On the one hand, the error of the original layout corresponding to the benchmark value is determined by the layout quality detection tool. On the other hand, the complexity of the original layout corresponding to the benchmark value is determined by the layout quality detection tool. In this way, the error and complexity corresponding to each benchmark value can be determined. The implementation method of step 2 can be seen in the description of step 202, which will not be repeated here.

[0234] If the error corresponding to each benchmark value is less than or equal to the expected value corresponding to the error, and the complexity corresponding to each benchmark value is less than or equal to the expected value corresponding to the complexity, then a target value is determined based on each benchmark value. The method for determining the target value can be found in the description of step 205 and will not be repeated here. If the error corresponding to any benchmark value is greater than the expected value corresponding to the error, and / or the complexity corresponding to any benchmark value is greater than the expected value corresponding to the complexity, then step 3 is executed.

[0235] Step 3: Determine the Gaussian process regression model corresponding to the error based on the error corresponding to each benchmark value, and determine the Gaussian process regression model corresponding to the complexity based on the complexity corresponding to each benchmark value.

[0236] In the embodiment of the present application, the errors corresponding to the respective benchmark values ​​satisfy the Gaussian distribution. In this case, the Gaussian process regression model corresponding to the error is used to describe the mapping relationship between the benchmark value and the mean of the Gaussian distribution satisfied by the error, and the variance of the Gaussian distribution. Similarly, the complexity corresponding to the respective benchmark values ​​also satisfies the Gaussian distribution. In this case, the Gaussian process regression model corresponding to the complexity is used to describe the mapping relationship between the benchmark value and the mean of the Gaussian distribution satisfied by the complexity, and the variance of the Gaussian distribution. The implementation principle of step 3 can be found in the description of step S1 and will not be repeated here.

[0237] Step 4: Based on the Gaussian process regression model and multiple candidate values, determine the reference index of each candidate value, and select a benchmark value from the multiple candidate values ​​based on the reference index of each candidate value.

[0238] In the embodiment of the present application, since the Gaussian process regression model corresponding to the error describes the mapping relationship between the reference value and the mean and variance of the Gaussian distribution satisfied by the error, the Gaussian process regression model corresponding to the error can be used to map any candidate value to the mean and variance corresponding to the error. Based on the same principle, the Gaussian process regression model corresponding to the complexity is used to map any candidate value to the mean and variance corresponding to the complexity.

[0239] Next, based on the mean and variance corresponding to the error of any candidate value and the mean and variance corresponding to the complexity of the candidate value, the reference index corresponding to the error of the candidate value and the reference index corresponding to the complexity of the candidate value are determined, thereby determining whether the candidate value is a benchmark value based on the reference index corresponding to the error of the candidate value and the reference index corresponding to the complexity of the candidate value. In this way, a benchmark value is selected from multiple candidate values. The implementation principle of step 4 can be found in the description of step S2, implementation method A2, and implementation method B2, and will not be repeated here.

[0240] Next, consider the benchmark values ​​selected in step 4 and the benchmark values ​​determined in step 1 as a set of benchmark values. Re-execute step 2 based on these benchmark values ​​to obtain the error and complexity corresponding to each benchmark value. If the error corresponding to each benchmark value is less than or equal to the expected value corresponding to the error, and the complexity corresponding to each benchmark value is less than or equal to the expected value corresponding to the complexity, then determine the target value based on each benchmark value; otherwise, repeat steps 3, 4, and 2 until the target value is determined.

[0241] In an embodiment of the present application, according to the flowchart shown in Figure 4, the parameters of the layout quality detection tool are determined for the edge distance error. By counting the relationship between the number of iterations and the edge distance error, the relationship lines corresponding to the present application in Figure 6 are obtained. In addition, four related algorithms are used to determine the parameters of the layout quality detection tool for the edge distance error. By counting the relationship between the number of iterations and the edge distance error corresponding to various related algorithms, the relationship lines corresponding to related algorithms 1 to 4 in Figure 6 are obtained. Related algorithm 1 is a random algorithm, related algorithm 2 is a particle swarm optimization (PSO) algorithm, related algorithm 3 is a social learning particle swarm optimization (SLPSO) algorithm, and related algorithm 4 is a random walking (RW) algorithm.

[0242] As can be seen from Figure 6, after 20 iterations, the edge distance error of the relationship line corresponding to the present application is the smallest, and the convergence is faster than that of the related algorithms. This shows that the method of the embodiment of the present application can determine the value of the target parameter with high accuracy with a low amount of calculation, thereby improving the accuracy of the layout quality detection tool.

[0243] In addition, the parameters of the layout quality detection tool can be determined for edge distance error and complexity according to the flowchart shown in Figure 5. By statistically analyzing the relationship between the edge distance error and the total variation of the relevant complexity, the data distribution corresponding to the present application in Figure 7 is obtained. In addition, the parameters of the layout quality detection tool can be determined for edge distance error and complexity using the related algorithm 1. By statistically analyzing the relationship between the edge distance error and the total variation of the relevant complexity, the data distribution corresponding to the related algorithm 1 in Figure 7 is obtained.

[0244] As can be seen in Figure 7, in multi-objective optimization, the method of the embodiment of the present application can achieve a smaller edge distance error compared to the related algorithms. Each data point shown in Figure 7 is a Pareto optimal solution. The data distribution of the present application corresponds to 8 Pareto optimal solutions, while the data distribution of the related algorithm 1 corresponds to 5 Pareto optimal solutions. This shows that the method of the embodiment of the present application has more Pareto optimal solutions, making the layout quality detection tool more accurate.

[0245] FIG8 is a schematic diagram showing the structure of a parameter determination device for a layout quality inspection tool provided in an embodiment of the present application. As shown in FIG8 , the device includes:

[0246] An acquisition module 801 is configured to acquire a plurality of first reference values ​​of a target parameter and a plurality of first candidate values ​​of the target parameter of a layout quality inspection tool;

[0247] A quality detection module 802 is configured to perform a quality detection on the original layout using a layout quality detection tool according to the plurality of first reference values, and obtain a quality index corresponding to each first reference value;

[0248] a determination module 803 configured to determine, if the quality indicators corresponding to the respective first benchmark values ​​do not satisfy the reference condition, a first reference value from a plurality of first candidate values ​​based on the quality indicators corresponding to the respective first benchmark values, and determine a plurality of second benchmark values ​​of the target parameter based on the plurality of first benchmark values ​​and the first reference value;

[0249] The quality detection module 802 is further configured to perform a quality detection on the original layout using a layout quality detection tool according to the plurality of second reference values, and obtain a quality index corresponding to each second reference value;

[0250] The determination module 803 is also used to determine the target value of the target parameter based on each second benchmark value when the quality index corresponding to each second benchmark value meets the reference condition, and use the layout quality detection tool with the target parameter value as the target value as the layout quality detection tool after parameter adjustment, and the layout quality detection tool with the adjusted parameters is used to determine the quality index of the layout to be detected.

[0251] In a possible implementation, the acquisition module 801 is used to obtain a first value range of the target parameter; divide the first value range into multiple first value intervals; and perform sampling in each first value interval to obtain multiple first reference values ​​and multiple first candidate values.

[0252] In one possible implementation, the acquisition module 801 is used to sample within each first value interval to obtain multiple first initial values ​​of the target parameter; cluster the multiple first initial values ​​to obtain multiple target cluster clusters, and any target cluster cluster includes at least one first initial value among the multiple first initial values; for any target cluster cluster among the multiple target cluster clusters, select a first benchmark value from the various first initial values ​​included in any target cluster cluster, and determine the first initial value of any target cluster cluster other than the first benchmark value as the first candidate value.

[0253] In one possible implementation, the quality detection module 802 is used to process the original layout using a layout quality detection tool according to multiple first benchmark values ​​to obtain a predicted layout corresponding to each first benchmark value, and determine the quality indicator corresponding to each first benchmark value based on the predicted layout corresponding to each first benchmark value.

[0254] In one possible implementation, the quality detection module 802 is used to determine, for any first benchmark value among the first benchmark values, an error between the original layout and a predicted layout corresponding to any first benchmark value; and determine a quality indicator corresponding to any first benchmark value based on the error between the original layout and the predicted layout corresponding to any first benchmark value.

[0255] In one possible implementation, the determination module 803 is used to determine the reference indicators of each first candidate value based on the quality indicators corresponding to each first benchmark value. The reference indicators of any first candidate value are used to describe the situation where the quality indicators meet the reference conditions when the layout quality inspection tool is used to perform quality inspection on the original layout when the target parameters include any first candidate value; based on the reference indicators of each first candidate value, the first reference value is determined from multiple first candidate values.

[0256] In one possible implementation, determination module 803 is used to determine a first mapping relationship based on each first benchmark value and the quality indicator corresponding to each first benchmark value, where the first mapping relationship is used to describe the mapping relationship between each first benchmark value and the statistical information of the statistical distribution when the quality indicator corresponding to each first benchmark value satisfies the statistical distribution; for any first candidate value among the first candidate values, based on any first candidate value and the first mapping relationship, determine the statistical information corresponding to any first candidate value, and determine the reference indicator of any first candidate value based on the statistical information corresponding to any first candidate value.

[0257] In one possible implementation, the statistical information corresponding to any first candidate value includes a mean and a variance; a determination module 803 is used to determine a reference quality indicator from the quality indicators corresponding to each first benchmark value; based on the reference quality indicator, the mean and variance corresponding to any first candidate value, a normal distribution function of any first candidate value is determined; based on the variance corresponding to any first candidate value, the normal distribution function corresponding to any first candidate value, and the cumulative density function and probability density function of the normal distribution function corresponding to any first candidate value, a reference indicator of any first candidate value is determined.

[0258] In a possible implementation, any first reference value corresponds to a quality indicator; the determination module 803 is configured to determine a minimum quality indicator from the quality indicators corresponding to the first reference values, and determine the minimum quality indicator as a reference quality indicator.

[0259] In one possible implementation, any first benchmark value corresponds to multiple quality indicators; the determination module 803 is used to determine, for any quality indicator among the multiple quality indicators, the maximum value and the minimum value of any quality indicator from any quality indicator corresponding to each first benchmark value; for any quality indicator corresponding to any first benchmark value, based on the first selected quality indicator, the weight of the first selected quality indicator, the maximum value and the minimum value of the first selected quality indicator, determine the priority indicator of any quality indicator corresponding to any first benchmark value, the first selected quality indicator is at least one quality indicator among the multiple quality indicators corresponding to any first benchmark value, and the first selected quality indicator includes any quality indicator corresponding to any first benchmark value; based on the priority indicator of each quality indicator corresponding to each first benchmark value, determine the reference quality indicator from each quality indicator corresponding to each first benchmark value.

[0260] In one possible implementation, the determination module 803 is configured to determine a difference between a reference quality indicator and a mean corresponding to any first candidate value; and use a quotient of the difference and a variance corresponding to any first candidate value as a normal distribution function of any first candidate value.

[0261] In one possible implementation, the determination module 803 is used to multiply the normal distribution function corresponding to any first candidate value and the cumulative density function of the normal distribution function corresponding to any first candidate value to obtain a multiplication result; add the multiplication result and the probability density function of the normal distribution function corresponding to any first candidate value to obtain an addition result; and multiply the addition result and the variance corresponding to any first candidate value to obtain a reference index of any first candidate value.

[0262] In one possible implementation, any first candidate value corresponds to a reference indicator; the determination module 803 is used to select a first candidate value whose reference indicator is not less than a set indicator threshold from multiple first candidate values ​​based on the reference indicators of each first candidate value, as the first reference value.

[0263] In one possible implementation, any first candidate value corresponds to multiple reference indicators; a determination module 803 is used to determine, for any reference indicator among the multiple reference indicators, the maximum value and the minimum value of any reference indicator from any reference indicator corresponding to each first candidate value; for any reference indicator corresponding to any first candidate value, based on the selected reference indicator and the maximum value and the minimum value of the selected reference indicator, determine the priority indicator of any reference indicator corresponding to any first candidate value, the selected reference indicator is at least one reference indicator among the multiple reference indicators corresponding to any first candidate value, and the selected reference indicator includes any reference indicator corresponding to any first candidate value; based on the priority indicator of each reference indicator corresponding to each first candidate value, determine the target indicator from each reference indicator corresponding to each first candidate value; determine the first candidate value corresponding to the target indicator as the first reference value.

[0264] In one possible implementation, the determination module 803 is further configured to, if the quality indicators corresponding to the respective second reference values ​​do not satisfy the reference condition, determine a second reference value from a plurality of second candidate values ​​based on the quality indicators corresponding to the respective second reference values, and determine a plurality of third reference values ​​based on the plurality of second reference values ​​and the second reference values, where the plurality of second candidate values ​​include first candidate values ​​other than the first reference value among the plurality of first candidate values;

[0265] The quality detection module 802 is further configured to perform a quality detection on the original layout using a layout quality detection tool according to the plurality of third reference values, and obtain a quality index corresponding to each third reference value;

[0266] The determination module 803 is further configured to determine a target value of the target parameter based on each third reference value when the quality indicator corresponding to each third reference value satisfies a reference condition.

[0267] In one possible implementation, the determination module 803 is used to adjust the first value range of the target parameter to obtain the second value range of the target parameter; divide the second value range into multiple second value intervals; perform sampling in each second value interval to obtain multiple first supplementary values ​​and multiple second supplementary values; determine the first candidate values ​​other than the first reference value among the multiple first supplementary values ​​and the multiple first candidate values ​​as multiple second candidate values; determine the second reference value from the multiple second candidate values ​​based on the quality indicators corresponding to each second benchmark value; and determine the multiple second supplementary values, the multiple second benchmark values, and the second reference value as multiple third benchmark values.

[0268] In one possible implementation, the determination module 803 is used to determine a first quality indicator whose quality indicator is not greater than a first indicator threshold value from the quality indicators corresponding to each first benchmark value; determine a second quality indicator whose quality indicator is not greater than a second indicator threshold value from the quality indicators corresponding to each second benchmark value; when the first quality indicator is less than the second quality indicator, narrow the first value range to obtain the second value range; when the first quality indicator is greater than the second quality indicator, expand the first value range to obtain the second value range.

[0269] In a possible implementation, the length of the second value range is not less than the first length, and the length of the second value range is not greater than the second length.

[0270] In one possible implementation, the determination module 803 is used to adjust any first benchmark value among multiple first benchmark values ​​based on the first reference value to obtain an adjusted benchmark value corresponding to any first benchmark value; and use the adjusted benchmark values ​​corresponding to each first benchmark value as multiple second benchmark values.

[0271] In one possible implementation, any second benchmark value corresponds to a quality indicator, and the determination module 803 is used to select each second benchmark value whose quality indicator is greater than or less than a set threshold from multiple second benchmark values, and use the selected second benchmark values ​​as each target value.

[0272] In one possible implementation, any second benchmark value corresponds to multiple quality indicators, and the determination module 803 is used to determine, for any quality indicator among the multiple quality indicators, the maximum value and the minimum value of any quality indicator from any quality indicator corresponding to each second benchmark value; for any quality indicator corresponding to any second benchmark value, based on the second selected quality indicator and the maximum value and the minimum value of the second selected quality indicator, determine the priority indicator of any quality indicator corresponding to any second benchmark value, the second selected quality indicator is at least one quality indicator among the multiple quality indicators corresponding to any second benchmark value, and the second selected quality indicator includes any quality indicator corresponding to any second benchmark value; based on the priority indicator of each quality indicator corresponding to each second benchmark value, determine the target quality indicator from each quality indicator corresponding to each second benchmark value; and determine the second benchmark value corresponding to the target quality indicator as the target value.

[0273] In the above-mentioned device, the target parameters of the layout quality inspection tool include various first benchmark values. The layout quality inspection tool is used to perform quality inspection on the original layout to obtain quality indicators corresponding to each first benchmark value. If there is a quality indicator corresponding to a first benchmark value that does not meet the reference condition, a first reference value is determined from multiple first candidate values ​​of the target parameter based on the quality indicator corresponding to each first benchmark value. Based on the multiple first benchmark values ​​and the first reference value, multiple second benchmark values ​​of the target parameter are determined. By determining the second benchmark values, the accuracy of the value space of the target parameter is improved. Next, the layout quality inspection tool is used to perform quality inspection on the original layout to obtain quality indicators corresponding to each second benchmark value. By improving the accuracy of the value space of the target parameter, the accuracy of the layout quality detection tool is improved, thereby improving the accuracy of the quality index. When the quality index corresponding to each second benchmark value meets the reference condition, the target value of the target parameter can be accurately determined based on each second benchmark value, thereby ensuring the accuracy of the layout quality detection tool, which is conducive to improving the accuracy of the quality index of the layout to be detected determined by the layout quality detection tool, thereby more accurately selecting a higher quality layout from the layout to be detected.

[0274] It should be understood that the device provided in FIG8 is merely an example of the division of the functional modules described above when implementing its functions. In actual applications, the functions described above can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiments are based on the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0275] FIG9 shows a block diagram of a terminal device 900 provided by an exemplary embodiment of the present application. The terminal device 900 includes a processor 901 and a memory 902 .

[0276] The processor 901 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 901 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), or PLA (Programmable Logic Array). The processor 901 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 901 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 901 may also include an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.

[0277] The memory 902 may include one or more computer-readable storage media, which may be non-transitory. The memory 902 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices and flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 902 is used to store at least one computer program, which is executed by the processor 901 to enable the terminal device 900 to implement the parameter determination method of the layout quality detection tool provided in the method embodiment of the present application.

[0278] In some embodiments, the terminal device 900 may optionally further include: a display screen 905 .

[0279] The display screen 905 is used to display a user interface (UI). This UI may include graphics, text, icons, videos, or any combination thereof. When the display screen 905 is a touch screen, it is also capable of collecting touch signals on or above the surface of the display screen 905. These touch signals can be input as control signals to the processor 901 for processing. In this case, the display screen 905 can also be used to provide virtual buttons and / or a virtual keyboard, also known as soft buttons and / or a soft keyboard. In some embodiments, there can be one display screen 905, located on the front panel of the terminal device 900. In other embodiments, there can be at least two display screens 905, located on different surfaces of the terminal device 900 or in a foldable design. In other embodiments, the display screen 905 can be a flexible display, located on a curved or foldable surface of the terminal device 900. Furthermore, the display screen 905 can be configured as a non-rectangular irregular shape, i.e., a special-shaped screen. The display screen 905 can be made of materials such as LCD (Liquid Crystal Display) and OLED (Organic Light-Emitting Diode). For example, the original layout, the predicted layout, the layout to be tested, etc. can be displayed using the display screen 905.

[0280] Those skilled in the art will understand that the structure shown in FIG9 does not constitute a limitation on the terminal device 900 , and may include more or fewer components than shown, or combine certain components, or adopt a different component arrangement.

[0281] FIG10 is a schematic diagram of the structure of the server provided in an embodiment of the present application. The server 1000 may have relatively large differences due to different configurations or performances, and may include one or more processors 1001 and one or more memories 1002, wherein the one or more memories 1002 store at least one computer program, and the at least one computer program is loaded and executed by the one or more processors 1001 to implement the parameter determination method of the layout quality detection tool provided in the above-mentioned various method embodiments. Exemplarily, the processor 1001 is a CPU. Of course, the server 1000 may also have components such as a wired or wireless network interface, a keyboard, and an input and output interface for input and output. The server 1000 may also include other components for implementing device functions, which will not be described in detail here.

[0282] In an exemplary embodiment, a non-volatile computer-readable storage medium is also provided, in which at least one computer program is stored. The at least one computer program is loaded and executed by a processor to enable an electronic device to implement any of the above-mentioned parameter determination methods of the layout quality detection tool.

[0283] Optionally, the above-mentioned non-volatile computer-readable storage medium can be a read-only memory (ROM), a random access memory (RAM), a compact disc (CD-ROM), a magnetic tape, a floppy disk, an optical data storage device, etc.

[0284] In an exemplary embodiment, a computer program is further provided. The computer program is at least one, and the at least one computer program is loaded and executed by a processor to enable an electronic device to implement any of the above-mentioned parameter determination methods of the layout quality detection tool.

[0285] In an exemplary embodiment, a computer program product is further provided, in which at least one computer program is stored. The at least one computer program is loaded and executed by a processor to enable an electronic device to implement any of the above-mentioned parameter determination methods of the layout quality detection tool.

[0286] It should be understood that the term "plurality" used herein refers to two or more. "And / or" describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can mean: A exists alone, A and B exist simultaneously, or B exists alone. The character " / " generally indicates an "or" relationship between the associated objects.

[0287] The above description is merely an exemplary embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for determining parameters of a layout quality inspection tool, wherein: The method is performed by an electronic device, and includes: Acquire a plurality of first reference values ​​of a target parameter of a layout quality inspection tool and a plurality of first candidate values ​​of the target parameter; Performing quality inspection on the original layout using the layout quality inspection tool according to the multiple first benchmark values ​​to obtain quality indicators corresponding to each first benchmark value; In the case where the quality indicators corresponding to the respective first benchmark values ​​do not meet the reference condition, determining a first reference value from the plurality of first candidate values ​​based on the quality indicators corresponding to the respective first benchmark values, and determining a plurality of second benchmark values ​​of the target parameter based on the plurality of first benchmark values ​​and the first reference value; Performing quality inspection on the original layout using the layout quality inspection tool according to the multiple second benchmark values ​​to obtain quality indicators corresponding to each second benchmark value; In the case that the quality indicators corresponding to the respective second benchmark values ​​satisfy the reference conditions, the target value of the target parameter is determined based on the respective second benchmark values, and a layout quality inspection tool whose target parameter value is the target value is used as a layout quality inspection tool after parameter adjustment, and the layout quality inspection tool after parameter adjustment is used to determine the quality indicators of the layout to be inspected.

2. The method according to claim 1, wherein: The step of obtaining a plurality of first reference values ​​of a target parameter of a layout quality inspection tool and a plurality of first candidate values ​​of the target parameter comprises: Acquire a first value range of the target parameter; Dividing the first value range into a plurality of first value intervals; Sampling is performed in each first value interval to obtain the multiple first reference values ​​and the multiple first candidate values.

3. The method according to claim 2, wherein: The sampling in each first value interval to obtain the plurality of first reference values ​​and the plurality of first candidate values ​​includes: Sampling is performed in each first value interval to obtain a plurality of first initial values ​​of the target parameter; Performing clustering processing on the multiple first initial values ​​to obtain multiple target clusters, any one of which includes at least one first initial value among the multiple first initial values; For any one of the multiple target clusters, a first reference value is selected from each first initial value included in the any one of the target clusters, and the first initial value in the any one of the target clusters except the first reference value is determined as a first candidate value.

4. The method according to any one of claims 1 to 3, wherein: The method of performing quality inspection on the original layout using the layout quality inspection tool according to the multiple first benchmark values ​​to obtain quality indicators corresponding to each first benchmark value includes: The original layout is processed using the layout quality detection tool according to the multiple first benchmark values ​​to obtain predicted layouts corresponding to the respective first benchmark values, and the quality indicators corresponding to the respective first benchmark values ​​are determined based on the predicted layouts corresponding to the respective first benchmark values.

5. The method according to claim 4, wherein: The determining, based on the predicted layout corresponding to each of the first reference values, the quality indicator corresponding to each of the first reference values ​​comprises: For any first reference value among the first reference values, determining an error between the original layout and a predicted layout corresponding to the first reference value; Based on the error between the original layout and the predicted layout corresponding to any one of the first reference values, a quality indicator corresponding to any one of the first reference values ​​is determined.

6. The method according to any one of claims 1 to 5, wherein: The determining a first reference value from the plurality of first candidate values ​​based on the quality indicators corresponding to the respective first benchmark values ​​includes: Based on the quality indicators corresponding to the first reference values, reference indicators of the first candidate values ​​are determined, and the reference indicator of any first candidate value is used to describe the case where the target parameter includes any first candidate value. When the layout quality inspection tool is used to perform quality inspection on the original layout, the quality index meets the reference condition; based on the reference index of each first candidate value, a first reference value is determined from the multiple first candidate values.

7. The method according to claim 6, wherein: The determining, based on the quality indicators corresponding to the first reference values, reference indicators for the first candidate values ​​includes: Determine a first mapping relationship based on the first benchmark values ​​and the quality indicators corresponding to the first benchmark values, wherein the first mapping relationship is used to describe a mapping relationship between the first benchmark values ​​and the statistical information of the statistical distribution when the quality indicators corresponding to the first benchmark values ​​satisfy the statistical distribution; For any first candidate value among the first candidate values, based on any first candidate value and the first mapping relationship, determine the statistical information corresponding to any first candidate value, and based on the statistical information corresponding to any first candidate value, determine the reference indicator of any first candidate value.

8. The method according to claim 7, wherein: The statistical information corresponding to any of the first candidate values ​​includes a mean and a variance; and determining a reference indicator of any of the first candidate values ​​based on the statistical information corresponding to any of the first candidate values ​​includes: Determine a reference quality indicator from the quality indicators corresponding to the first reference values; Determine a normal distribution function of any first candidate value based on the reference quality indicator and the mean and variance corresponding to any first candidate value; Based on the variance corresponding to any one of the first candidate values, the normal distribution function corresponding to any one of the first candidate values, the cumulative density function and the probability density function of the normal distribution function corresponding to any one of the first candidate values, a reference indicator of any one of the first candidate values ​​is determined.

9. The method according to claim 8, wherein: Any first benchmark value corresponds to a quality indicator; The determining of the reference quality indicator from the quality indicators corresponding to the first reference values ​​includes: A minimum quality indicator is determined from the quality indicators corresponding to the first reference values, and the minimum quality indicator is determined as the reference quality indicator.

10. The method according to claim 8, wherein: Any first benchmark value corresponds to multiple quality indicators; The determining of the reference quality indicator from the quality indicators corresponding to the first reference values ​​includes: For any one of the multiple quality indicators, determine a maximum value and a minimum value of the any one of the quality indicators from the any one of the quality indicators corresponding to the respective first reference values; For any quality indicator corresponding to any first benchmark value, determine a priority indicator of any quality indicator corresponding to any first benchmark value based on a first selected quality indicator, a weight of the first selected quality indicator, and a maximum value and a minimum value of the first selected quality indicator, wherein the first selected quality indicator is at least one quality indicator among a plurality of quality indicators corresponding to any first benchmark value, and the first selected quality indicator includes any quality indicator corresponding to any first benchmark value; Based on the priority indicators of the quality indicators corresponding to the first reference values, a reference quality indicator is determined from the quality indicators corresponding to the first reference values.

11. The method according to any one of claims 8 to 10, wherein: The determining, based on the reference quality indicator and the mean and variance corresponding to any of the first candidate values, a normal distribution function of any of the first candidate values ​​comprises: Determine a difference between the reference quality indicator and a mean value corresponding to any one of the first candidate values; The quotient of the difference and the variance corresponding to any one of the first candidate values ​​is used as the normal distribution function of the any one of the first candidate values.

12. The method according to any one of claims 8 to 11, wherein: The determining a reference index of any first candidate value based on the variance corresponding to any first candidate value, the normal distribution function corresponding to any first candidate value, the cumulative density function and the probability density function of the normal distribution function corresponding to any first candidate value includes: Multiplying the normal distribution function corresponding to any one of the first candidate values ​​and the cumulative density function of the normal distribution function corresponding to any one of the first candidate values ​​to obtain a multiplication result; Add the multiplication result to the probability density function of the normal distribution function corresponding to any one of the first candidate values, Get the addition result; The addition result is multiplied by the variance corresponding to any one of the first candidate values ​​to obtain a reference index for any one of the first candidate values.

13. The method according to any one of claims 6 to 12, wherein: Any first candidate value corresponds to a reference indicator, and determining a first reference value from the plurality of first candidate values ​​based on the reference indicators of the first candidate values ​​includes: Based on the reference indicators of the first candidate values, a first candidate value whose reference indicator is not less than a set indicator threshold is selected from the multiple first candidate values ​​as the first reference value.

14. The method according to any one of claims 6 to 12, wherein: Any first candidate value corresponds to a plurality of reference indicators; and determining a first reference value from the plurality of first candidate values ​​based on the reference indicators of the respective first candidate values ​​comprises: For any one of the multiple reference indicators, determine a maximum value and a minimum value of the any one of the reference indicators from the any one of the reference indicators corresponding to the respective first candidate values; For any one of the reference indicators corresponding to any one of the first candidate values, determining a priority indicator of the reference indicator corresponding to any one of the first candidate values ​​based on a selected reference indicator, a maximum value and a minimum value of the selected reference indicator, wherein the selected reference indicator is at least one reference indicator among a plurality of reference indicators corresponding to any one of the first candidate values, and the selected reference indicator includes the reference indicator corresponding to any one of the first candidate values; Based on the priority indicators of the reference indicators corresponding to the first candidate values, a target indicator is determined from the reference indicators corresponding to the first candidate values; and the first candidate value corresponding to the target indicator is determined as the first reference value.

15. The method according to any one of claims 1 to 14, wherein: The method further comprises: In the case that the quality indicators corresponding to the respective second reference values ​​do not satisfy the reference condition, determining a second reference value from a plurality of second candidate values ​​based on the quality indicators corresponding to the respective second reference values, and determining a plurality of third reference values ​​based on the plurality of second reference values ​​and the second reference value, wherein the plurality of second candidate values ​​include first candidate values ​​other than the first reference value among the plurality of first candidate values; Performing quality inspection on the original layout using the layout quality inspection tool according to the plurality of third benchmark values ​​to obtain quality indicators corresponding to each third benchmark value; In a case where the quality indicators corresponding to the respective third reference values ​​satisfy the reference condition, a target value of the target parameter is determined based on the respective third reference values.

16. The method according to claim 15, wherein: The determining a second reference value from a plurality of second candidate values ​​based on the quality indicators corresponding to the respective second reference values, and determining a plurality of third reference values ​​based on the plurality of second reference values ​​and the second reference value, comprises: Adjusting the first value range of the target parameter to obtain a second value range of the target parameter; Dividing the second value range into a plurality of second value intervals; Sampling is performed in each second value interval to obtain a plurality of first supplementary values ​​and a plurality of second supplementary values; Determine the first candidate values ​​other than the first reference value among the plurality of first supplementary values ​​and the plurality of first candidate values ​​as the plurality of second candidate values; Determining a second reference value from the plurality of second candidate values ​​based on the quality indicators corresponding to the respective second benchmark values; The plurality of second supplementary values, the plurality of second benchmark values ​​and the second reference value are determined as the plurality of third benchmark values.

17. The method according to claim 16, wherein: The adjusting the first value range of the target parameter to obtain the second value range of the target parameter includes: Determine, from the quality indicators corresponding to the first reference values, a first quality indicator whose quality indicator is not greater than a first indicator threshold; Determine the second quality index whose quality index is not greater than the second index threshold value from the quality index corresponding to each of the second reference values. Quality indicators; When the first quality indicator is less than the second quality indicator, narrowing the first value range to obtain the second value range; When the first quality indicator is greater than the second quality indicator, the first value range is expanded to obtain the second value range.

18. The method according to claim 17, wherein: The length of the second value range is not less than the first length, and the length of the second value range is not greater than the second length.

19. The method according to any one of claims 1 to 18, wherein: The determining, based on the plurality of first benchmark values ​​and the first reference value, a plurality of second benchmark values ​​of the target parameter comprises: For any first reference value among the multiple first reference values, adjusting the any first reference value based on the first reference value to obtain an adjusted reference value corresponding to the any first reference value; The adjusted reference values ​​corresponding to the first reference values ​​are used as the multiple second reference values.

20. The method according to any one of claims 1 to 19, wherein: Any second reference value corresponds to a quality indicator, and determining the target value of the target parameter based on each of the second reference values ​​includes: Second reference values ​​whose quality indicators are greater than or less than a set threshold are selected from a plurality of second reference values, and the selected second reference values ​​are used as target values.

21. The method according to any one of claims 1 to 19, wherein: Any second reference value corresponds to a plurality of quality indicators, and determining the target value of the target parameter based on each of the second reference values ​​includes: For any one of the multiple quality indicators, determine a maximum value and a minimum value of the any one of the quality indicators from the any one of the quality indicators corresponding to the respective second reference values; For any quality indicator corresponding to any second benchmark value, determine a priority indicator of any quality indicator corresponding to any second benchmark value based on a second selected quality indicator, a maximum value and a minimum value of the second selected quality indicator, wherein the second selected quality indicator is at least one quality indicator among a plurality of quality indicators corresponding to any second benchmark value, and the second selected quality indicator includes the any quality indicator corresponding to any second benchmark value; Based on the priority indicators of the quality indicators corresponding to the second reference values, a target quality indicator is determined from the quality indicators corresponding to the second reference values; and the second reference value corresponding to the target quality indicator is determined as the target value.

22. A parameter determination device for a layout quality inspection tool, wherein: The device comprises: An acquisition module, used to acquire a plurality of first reference values ​​of a target parameter of a layout quality inspection tool and a plurality of first candidate values ​​of the target parameter; A quality detection module, configured to perform quality detection on the original layout using the layout quality detection tool according to the multiple first reference values, and obtain a quality index corresponding to each first reference value; a determination module, configured to determine, when the quality indicators corresponding to the respective first benchmark values ​​do not satisfy a reference condition, a first reference value from the plurality of first candidate values ​​based on the quality indicators corresponding to the respective first benchmark values, and determine a plurality of second benchmark values ​​of the target parameter based on the plurality of first benchmark values ​​and the first reference value; The quality detection module is further used to perform quality detection on the original layout using the layout quality detection tool according to the multiple second benchmark values ​​to obtain quality indicators corresponding to each second benchmark value; The determination module is further used to determine the target value of the target parameter based on the respective second benchmark values ​​when the quality indicators corresponding to the respective second benchmark values ​​satisfy the reference conditions, and use the layout quality detection tool with the target parameter value as the layout quality detection tool after parameter adjustment, and the layout quality detection tool with adjusted parameters is used to determine the quality indicators of the layout to be detected.

23. The device according to claim 22, wherein: The acquisition module is used to acquire a first value range of the target parameter; divide the first value range into a plurality of first value intervals; and perform sampling in each first value interval to obtain the plurality of first reference values ​​and the plurality of first candidate values.

24. The device according to claim 23, wherein: The acquisition module is used to sample within each first value interval to obtain multiple first initial values ​​of the target parameter; cluster the multiple first initial values ​​to obtain multiple target cluster clusters, any target cluster cluster includes at least one first initial value among the multiple first initial values; for any target cluster cluster among the multiple target cluster clusters, select a first benchmark value from the various first initial values ​​included in any target cluster cluster, and determine the first initial value of any target cluster cluster except the first benchmark value as the first candidate value.

25. The device according to any one of claims 22 to 24, wherein: The quality detection module is used to process the original layout using the layout quality detection tool according to the multiple first benchmark values ​​to obtain a predicted layout corresponding to each first benchmark value, and determine the quality indicator corresponding to each first benchmark value based on the predicted layout corresponding to each first benchmark value.

26. The device according to claim 25, wherein The quality detection module is used to determine, for any first reference value among the first reference values, an error between the original layout and a predicted layout corresponding to any first reference value; Based on the error between the original layout and the predicted layout corresponding to any one of the first reference values, a quality indicator corresponding to any one of the first reference values ​​is determined.

27. The device according to any one of claims 22 to 26, wherein: The determination module is used to determine the reference index of each first candidate value based on the quality index corresponding to each first reference value, and the reference index of any first candidate value is used to describe the situation that the quality index satisfies the reference condition when the layout quality detection tool is used to perform quality detection on the original layout when the target parameter includes any first candidate value; Based on the reference indicators of the first candidate values, a first reference value is determined from the multiple first candidate values.

28. The device according to claim 27, wherein The determination module is used to determine a first mapping relationship based on the first benchmark values ​​and the quality indicators corresponding to the first benchmark values, wherein the first mapping relationship is used to describe a mapping relationship between the first benchmark values ​​and the statistical information of the statistical distribution when the quality indicators corresponding to the first benchmark values ​​satisfy the statistical distribution; For any first candidate value among the first candidate values, based on any first candidate value and the first mapping relationship, determine the statistical information corresponding to any first candidate value, and based on the statistical information corresponding to any first candidate value, determine the reference indicator of any first candidate value.

29. The device according to claim 28, wherein The statistical information corresponding to any one of the first candidate values ​​includes a mean and a variance; the determination module is used to determine a reference quality indicator from the quality indicators corresponding to the first benchmark values; based on the reference quality indicator, the mean and the variance corresponding to any one of the first candidate values, determine the normal distribution function of any one of the first candidate values; based on the variance corresponding to any one of the first candidate values, the normal distribution function corresponding to any one of the first candidate values, and the cumulative density function and the probability density function of the normal distribution function corresponding to any one of the first candidate values, determine the reference indicator of any one of the first candidate values.

30. The device according to claim 29, wherein: Any first reference value corresponds to a quality indicator; the determination module is used to determine the minimum quality indicator from the quality indicators corresponding to the first reference values, and determine the minimum quality indicator as the reference quality indicator.

31. The device according to claim 29, wherein: Any first reference value corresponds to a plurality of quality indicators; the determination module is used to determine, for any one of the plurality of quality indicators, a maximum value and a minimum value of the any one of the quality indicators from the any one of the quality indicators corresponding to each of the first reference values; For any quality indicator corresponding to any first benchmark value, determine a priority indicator of any quality indicator corresponding to any first benchmark value based on a first selected quality indicator, a weight of the first selected quality indicator, and a maximum value and a minimum value of the first selected quality indicator, wherein the first selected quality indicator is at least one quality indicator among a plurality of quality indicators corresponding to any first benchmark value, and the first selected quality indicator includes any quality indicator corresponding to any first benchmark value; Based on the priority indicators of the quality indicators corresponding to the first reference values, a reference quality indicator is determined from the quality indicators corresponding to the first reference values.

32. The device according to any one of claims 29 to 31, wherein: The determination module is used to determine the difference between the reference quality indicator and the mean value corresponding to any one of the first candidate values; The quotient of the difference and the variance corresponding to any one of the first candidate values ​​is used as the normal distribution function of the any one of the first candidate values.

33. The device according to any one of claims 29 to 32, wherein: The determination module is used to multiply the normal distribution function corresponding to any one of the first candidate values ​​and the cumulative density function of the normal distribution function corresponding to any one of the first candidate values ​​to obtain a multiplication result; add the multiplication result and the probability density function of the normal distribution function corresponding to any one of the first candidate values ​​to obtain an addition result; multiply the addition result and the variance corresponding to any one of the first candidate values ​​to obtain a reference index of any one of the first candidate values.

34. The device according to any one of claims 27 to 33, wherein: Any first candidate value corresponds to a reference indicator, and the determination module is used to select a first candidate value whose reference indicator is not less than a set indicator threshold from the multiple first candidate values ​​based on the reference indicators of the various first candidate values ​​as the first reference value.

35. The device according to any one of claims 27 to 33, wherein: Any first candidate value corresponds to a plurality of reference indicators; the determination module is used to determine, for any one of the plurality of reference indicators, a maximum value and a minimum value of the any one of the reference indicators from the any one of the reference indicators corresponding to each of the first candidate values; For any reference indicator corresponding to any first candidate value, determine the priority indicator of any reference indicator corresponding to any first candidate value based on the selected reference indicator, the maximum value and the minimum value of the selected reference indicator, the selected reference indicator is at least one reference indicator among the multiple reference indicators corresponding to any first candidate value, and the selected reference indicator includes the any reference indicator corresponding to any first candidate value; based on the priority indicators of the reference indicators corresponding to the first candidate values, determine the target indicator from the reference indicators corresponding to the first candidate values; determine the first candidate value corresponding to the target indicator as the first reference value.

36. The device according to any one of claims 22 to 35, wherein: The determination module is further configured to determine, if the quality indicators corresponding to the respective second reference values ​​do not satisfy the reference condition, a second reference value from a plurality of second candidate values ​​based on the quality indicators corresponding to the respective second reference values, and determine a plurality of third reference values ​​based on the plurality of second reference values ​​and the second reference value, wherein the plurality of second candidate values ​​include first candidate values ​​other than the first reference value among the plurality of first candidate values; The quality detection module is further used to perform quality detection on the original layout using the layout quality detection tool according to the multiple third benchmark values ​​to obtain quality indicators corresponding to each third benchmark value; The determination module is further configured to determine a target value of the target parameter based on each of the third reference values ​​when the quality indicators corresponding to each of the third reference values ​​meet the reference condition.

37. The device according to claim 36, wherein The determination module is used to adjust the first value range of the target parameter to obtain a second value range of the target parameter; and divide the second value range into a plurality of second value intervals; Sampling is performed in each second value interval to obtain a plurality of first supplementary values ​​and a plurality of second supplementary values; and a first candidate value other than the first reference value among the plurality of first supplementary values ​​and the plurality of first candidate values ​​is determined as the plurality of second candidate values; Determining a second reference value from the plurality of second candidate values ​​based on the quality indicators corresponding to the respective second benchmark values; The plurality of second supplementary values, the plurality of second benchmark values ​​and the second reference value are determined as the plurality of third benchmark values.

38. The device according to claim 37, wherein The determination module is used to determine a first quality indicator whose quality indicator is not greater than a first indicator threshold value from the quality indicators corresponding to the various first benchmark values; determine a second quality indicator whose quality indicator is not greater than a second indicator threshold value from the quality indicators corresponding to the various second benchmark values; when the first quality indicator is less than the second quality indicator, narrow the first value range to obtain the second value range; when the first quality indicator is greater than the second quality indicator, expand the first value range to obtain the second value range.

39. The device according to claim 38, wherein The length of the second value range is not less than the first length, and the length of the second value range is not greater than the second length.

40. The device according to any one of claims 22 to 39, wherein: The determining module is configured to adjust any first reference value among the plurality of first reference values ​​based on the first reference value to obtain an adjusted reference value corresponding to the any first reference value; The adjusted reference values ​​corresponding to the first reference values ​​are used as the multiple second reference values.

41. The device according to any one of claims 22 to 40, wherein: Any second reference value corresponds to a quality indicator. The determination module is used to select second reference values ​​whose quality indicators are greater than or less than a set threshold from multiple second reference values, and use the selected second reference values ​​as target values.

42. The device according to any one of claims 22 to 40, wherein: Any second reference value corresponds to a plurality of quality indicators, and the determination module is used to determine, for any one of the plurality of quality indicators, a maximum value and a minimum value of the any one of the quality indicators from the any one of the quality indicators corresponding to each of the second reference values; For any quality indicator corresponding to any second benchmark value, determine the priority indicator of any quality indicator corresponding to any second benchmark value based on the second selected quality indicator, the maximum value and the minimum value of the second selected quality indicator, the second selected quality indicator is at least one quality indicator among the multiple quality indicators corresponding to any second benchmark value, and the second selected quality indicator includes the any quality indicator corresponding to any second benchmark value; based on the priority indicators of the quality indicators corresponding to the respective second benchmark values, determine the target quality indicator from the quality indicators corresponding to the respective second benchmark values; determine the second benchmark value corresponding to the target quality indicator as the target value.

43. An electronic device, wherein: The electronic device includes a processor and a memory, wherein at least one computer program is stored in the memory, and the at least one computer program is loaded and executed by the processor so that the electronic device implements the parameter determination method of the layout quality detection tool as described in any one of claims 1 to 21.

44. A non-volatile computer-readable storage medium, wherein: At least one computer program is stored in the non-volatile computer-readable storage medium, and the at least one computer program is loaded and executed by the processor to enable the electronic device to implement the parameter determination method of the layout quality detection tool as described in any one of claims 1 to 21.

45. A computer program product, wherein: The computer program product stores at least one computer program, and the at least one computer program is loaded and executed by a processor so that the electronic device implements the parameter determination method of the layout quality detection tool as described in any one of claims 1 to 21.

Images