Productivity prediction method and device, storage medium and electronic equipment

Abstract

The invention discloses a productivity prediction method and device, a storage medium and electronic equipment. The method comprises the following steps: preliminarily classifying wafer products to be classified; on the basis of product characteristics, fuzzy clustering is carried out on the wafer products in each preliminary classification, and a dynamic classification graph corresponding to the wafer products in each preliminary classification is obtained; obtaining a final classification result of the to-be-classified wafer product based on the dynamic classification graph; and predicting productivity based on the final classification result. By adopting the scheme, the future productivity bottleneck of the factory can be accurately predicted.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor technology, and in particular to a production capacity prediction method and device, a storage medium, and an electronic device. Background Art

[0002] With the continuous development of chip technology, the demand for wafer manufacturing capacity is gradually increasing. The classification of wafer products is becoming more and more complex. For semiconductor wafer foundries that produce wafers on a large scale, factory capacity forecasting becomes more difficult as the product combination becomes more complex.

[0003] Therefore, how to reasonably classify and group wafer products to simplify complex product requirements so as to accurately predict the factory's future production capacity bottleneck has become an urgent problem to be solved. Summary of the invention

[0004] The problem to be solved by the present invention is: how to accurately predict the future production capacity bottleneck of a factory.

[0005] To solve the above problem, an embodiment of the present invention provides a capacity prediction method, which includes:

[0006] Carry out preliminary classification of wafer products to be classified;

[0007] Based on product characteristics, fuzzy clustering is performed on the wafer products in each preliminary classification to obtain a dynamic classification diagram corresponding to the wafer products in each preliminary classification;

[0008] Based on the dynamic classification diagram, obtaining a final classification result of the wafer product to be classified;

[0009] Forecast production capacity based on the final classification results.

[0010] Optionally, the product characteristics include at least two of the following:

[0011] Number of mask layers;

[0012] Number of metal layers;

[0013] Number of ion implantation layers;

[0014] Chemical mechanical polishing times.

[0015] Optionally, based on the product characteristics, fuzzy clustering is performed on the wafer products in each preliminary classification to obtain a dynamic classification diagram corresponding to the wafer products in each preliminary classification, including:

[0016] According to product characteristics, construct a characteristic index matrix corresponding to each wafer product in the preliminary classification;

[0017] Perform data standardization on each characteristic index matrix to obtain a standardized characteristic matrix;

[0018] Based on each of the standardized characteristic matrices, corresponding fuzzy similarity matrices are constructed respectively;

[0019] Using the fuzzy transfer package method, a fuzzy equivalent matrix corresponding to each of the fuzzy similarity matrices is obtained;

[0020] Based on the fuzzy equivalence matrix, a dynamic classification diagram corresponding to the wafer products in each preliminary classification is obtained.

[0021] Optionally, performing data standardization processing on each characteristic indicator matrix to obtain a standardized characteristic matrix includes:

[0022] The maximum value method is used to perform data standardization on each characteristic index matrix to obtain a standardized characteristic matrix.

[0023] Optionally, constructing corresponding fuzzy similarity matrices based on each of the standardized characteristic matrices includes:

[0024] Based on each of the standardized characteristic matrices, a corresponding fuzzy similarity matrix is ​​constructed using the maximum-minimum method.

[0025] Optionally, obtaining a dynamic classification diagram corresponding to wafer products in each preliminary classification based on the fuzzy equivalence matrix includes:

[0026] Sort the elements in the fuzzy equivalence matrix from large to small and use them as the value points of the classification threshold in sequence;

[0027] The fuzzy equivalence matrix is ​​respectively intercepted by using the classification threshold at each value point to obtain a dynamic classification diagram corresponding to the wafer product in each preliminary classification.

[0028] Optionally, the preliminary classification of the wafer products to be classified includes:

[0029] The wafer products to be classified are preliminarily classified according to one or more of the technology node, size, and application platform.

[0030] Optionally, obtaining a final classification result of the wafer product to be classified based on the dynamic classification diagram includes:

[0031] In combination with actual production capacity bottlenecks and demands, the classification results in the dynamic classification diagram are dynamically adjusted.

[0032] The embodiment of the present invention further provides a production capacity prediction device, the device comprising:

[0033] A preliminary classification unit, suitable for preliminary classification of wafer products to be classified;

[0034] The fuzzy clustering unit is suitable for performing fuzzy clustering on the wafer products in each preliminary classification based on product characteristics, and obtaining a dynamic classification diagram corresponding to the wafer products in each preliminary classification;

[0035] A classification adjustment unit, adapted to obtain a final classification result of the wafer product to be classified based on the dynamic classification diagram;

[0036] The capacity prediction unit is adapted to predict the capacity based on the final classification result.

[0037] Optionally, the fuzzy clustering unit is adapted to perform fuzzy clustering on the wafer products in each preliminary classification based on at least two of the following product characteristics:

[0038] Number of mask layers;

[0039] Number of metal layers;

[0040] Number of ion implantation layers;

[0041] Chemical mechanical polishing times.

[0042] Optionally, the fuzzy clustering unit includes:

[0043] The characteristic index matrix construction subunit is adapted to construct the characteristic index matrix corresponding to the wafer product in each preliminary classification according to the product characteristics;

[0044] The standardization subunit is suitable for performing data standardization processing on each characteristic index matrix to obtain a standardized characteristic matrix;

[0045] A fuzzy similarity matrix construction subunit, adapted to construct corresponding fuzzy similarity matrices based on each of the standardized characteristic matrices;

[0046] A fuzzy equivalent matrix construction subunit is adapted to obtain the fuzzy equivalent matrix corresponding to each of the fuzzy similarity matrices by using the fuzzy transfer package method;

[0047] The dynamic clustering subunit is suitable for obtaining a dynamic classification diagram corresponding to the wafer products in each preliminary classification based on the fuzzy equivalence matrix.

[0048] Optionally, the standardization subunit is adapted to use a maximum value method to perform data standardization processing on each characteristic index matrix to obtain a standardized characteristic matrix.

[0049] Optionally, the fuzzy similarity matrix construction subunit is adapted to construct a corresponding fuzzy similarity matrix based on each of the standardized characteristic matrices using a maximum-minimum method.

[0050] Optionally, the dynamic clustering subunit is suitable for sorting the elements in the fuzzy equivalence matrix from large to small, and taking them as the value points of the classification threshold in turn. The classification threshold under each value point is used to intercept the fuzzy equivalence matrix respectively to obtain a dynamic classification map corresponding to the wafer products in each preliminary classification.

[0051] Optionally, the preliminary classification unit is suitable for preliminarily classifying the wafer products to be classified according to one or more of technology nodes, sizes, and application platforms.

[0052] Optionally, the classification adjustment unit is adapted to dynamically adjust the classification results in the dynamic classification diagram in combination with actual capacity bottlenecks of production equipment and requirements of production equipment.

[0053] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. The computer program is executed by a processor to implement the steps of any of the above methods.

[0054] An embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor executes the steps of any of the above methods when running the computer program.

[0055] Compared with the prior art, the technical solution of the embodiment of the present invention has the following advantages:

[0056] By applying the solution of the present invention, after the wafer products are preliminarily classified, the wafer products in each preliminary classification are fuzzy clustered to obtain a dynamic classification diagram corresponding to the wafer products in each preliminary classification. Subsequently, the final classification results of the wafer products to be classified can be obtained based on the dynamic classification diagram, and the production capacity can be predicted based on the final classification results. Compared with predicting the production capacity based only on the preliminary classification structure, the wafer products can be further combined based on the product characteristics, so that the final classification results are more in line with the product characteristics, thereby simplifying the wafer product combination, and the prediction of the production capacity bottleneck is more accurate, ensuring the maximum output of the factory production capacity. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 is a flow chart of a capacity prediction method in an embodiment of the present invention;

[0058] Figure 2 is a schematic diagram of a dynamic classification diagram in an embodiment of the present invention;

[0059] Figure 3 It is a structural schematic diagram of a production capacity prediction device in an embodiment of the present invention. DETAILED DESCRIPTION

[0060] At present, there are many key points of wafer products: technology node, size, application platform, number of illumination layers, etc. Based on the above factors, products can be initially classified. Since there is no longer a strict boundary between products, further classification makes the operation more difficult, so the production capacity is currently only predicted based on the preliminary classification results.

[0061] However, in the preliminary classification results, the combination of wafer products is very complex, and it is very difficult for factories to predict production capacity bottlenecks under different wafer product combinations.

[0062] To address this problem, the present invention provides a capacity prediction method. After preliminary classification of wafer products, fuzzy clustering is performed on the wafer products in each preliminary classification based on product characteristics to obtain a dynamic classification graph corresponding to the wafer products in each preliminary classification. Subsequently, the final classification results of the wafer products to be classified are obtained based on the dynamic classification graph, thereby simplifying the product combination in the final classification structure, thereby more accurately predicting the future capacity bottleneck of the factory.

[0063] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0064] Reference Figure 1 The embodiment of the present invention provides a capacity prediction method, which may include the following steps:

[0065] Step 11, preliminarily classify the wafer products to be classified.

[0066] In a specific implementation, the wafer products to be classified can be preliminarily classified based on one or more of the technology node, size, and application platform. For example, wafer products with the same technology node, the same size, and corresponding to several application platforms can be preliminarily classified into one category. Among them, the technology node, size, and application platform of the wafer product are information that can be known in advance.

[0067] In another embodiment, the wafer products to be classified can also be preliminarily classified based on the number of mask layers. For example, wafer products with the same technology node, the same size, corresponding to several application platforms and with mask layers within a certain range can be preliminarily classified into one category. Among them, the number of mask layers of the wafer product is also information that can be known in advance.

[0068] Step 12, based on product characteristics, fuzzy clustering is performed on the wafer products in each preliminary classification to obtain a dynamic classification diagram corresponding to the wafer products in each preliminary classification.

[0069] In a specific implementation, the product characteristics, i.e., the process characteristics of the wafer product itself, include but are not limited to at least two of the following: the number of mask layers, the number of metal layers, the number of ion implantation layers, and the number of chemical mechanical polishing times. For example, the wafer products can be fuzzily clustered based on only any two of the number of mask layers, the number of metal layers, the number of ion implantation layers, and the number of chemical mechanical polishing times, or based on any three of the number of mask layers, the number of metal layers, the number of ion implantation layers, and the number of chemical mechanical polishing times.

[0070] In a preferred embodiment, the wafer products can be fuzzily clustered based on the number of mask layers, the number of metal layers, the number of ion implantation layers, and the number of chemical mechanical polishing times, thereby making the final classification result of the wafer products more accurate.

[0071] In a specific implementation, the number of mask layers, metal layers, ion implantation layers, and chemical mechanical polishing times of a wafer product can be known in advance. For example, referring to Table 1, the number of mask layers, metal layers, ion implantation layers, and chemical mechanical polishing times of wafer products X1, X2, X3, X4, and X5 can be obtained as follows:

[0072] Table 1

[0073]

[0074] In one embodiment of the present invention, when fuzzy clustering is performed on wafer products in each preliminary classification based on product characteristics, a characteristic indicator matrix corresponding to the wafer products in each preliminary classification can be constructed according to the product characteristics, and then each characteristic indicator matrix can be subjected to data standardization to obtain a standardized characteristic matrix. Subsequently, based on each of the standardized characteristic matrices, corresponding fuzzy similarity matrices are constructed respectively, and then the fuzzy transfer package method is used to obtain a fuzzy equivalence matrix corresponding to each of the fuzzy similarity matrices. Finally, based on the fuzzy equivalence matrix, a dynamic classification diagram corresponding to the wafer products in each preliminary classification is obtained.

[0075] According to product characteristics, when constructing the characteristic index matrix corresponding to the wafer products in each preliminary classification, it is assumed that the same preliminary classification includes n wafer products, and the wafer product set X in the preliminary classification is {x1, x2, ..., x n Each wafer product has m characteristic indicators, namely x i It can be represented by an m-dimensional characteristic index vector: x i ={x i1 , x i2 , ..., x im}, then all characteristic indices of n objects are constructed into a characteristic index matrix X * , denoted as X * =(x ij )n*m .

[0076] In a specific implementation, a variety of methods can be used to perform data standardization processing on each characteristic indicator matrix to obtain a standardized characteristic matrix.

[0077] In one embodiment of the present invention, the maximum value method can be used to perform data standardization on the characteristic index matrix to obtain a standardized characteristic matrix. * The jth column takes the maximum value M j =max{x 1j , x 2j , ..., x nj}, and use the formula x′ ij =x ij / M j , compress the characteristic matrix data to the interval [0, -1], and obtain the standardized characteristic matrix X′. ij is the element in the i-th row and j-th column of the standardized feature matrix X′.

[0078] Using the maximum value method to obtain the standardized characteristic matrix can simplify the computational complexity of data standardization processing and improve the efficiency of fuzzy clustering.

[0079] In a specific implementation, based on each of the standardized characteristic matrices, a variety of methods can be used to construct a corresponding fuzzy similarity matrix.

[0080] In one embodiment of the present invention, based on each of the standardized characteristic matrices, a corresponding fuzzy similarity matrix R may be constructed using the maximum-minimum method.

[0081] Specifically, the element r in the i-th row and j-th column of the fuzzy similarity matrix R is ij The value of is equal to the sum of the minimum values ​​of the elements corresponding to the i-th row and the v-th row divided by the sum of the maximum values ​​of the elements corresponding to the i-th row and the j-th row, as shown in formula (1).

[0082]

[0083] Among them, k represents the k-th row.

[0084] In the specific implementation, the fuzzy transfer package method is used to perform square self-synthesis on the fuzzy similarity matrix to obtain the corresponding fuzzy equivalent matrix t(R), specifically referring to formula (2):

[0085]

[0086] In a specific implementation, after obtaining the fuzzy equivalence matrix t(R), the elements in the fuzzy equivalence matrix can be sorted from large to small and used as the value points of the classification threshold λ in turn, and then the classification threshold λ under each value point is used to intercept the fuzzy equivalence matrix respectively to obtain the dynamic classification map corresponding to the wafer products in each preliminary classification. Wherein, λ∈[0,1].

[0087] Taking wafer products x1 to x5 in Table 1 as all wafer products in the same initial classification as an example, according to the product characteristics of wafer products x1 to x5, the characteristic index matrix X is constructed * as follows:

[0088]

[0089] Based on the characteristic index matrix X * , the standardized characteristic matrix X′ is obtained as follows:

[0090]

[0091] Based on the standardized characteristic matrix X′, the corresponding fuzzy similarity matrix t(R) is obtained as follows:

[0092]

[0093] Based on the fuzzy similarity matrix t(R), the fuzzy equivalence matrix is ​​as follows:

[0094]

[0095] By arranging the elements in the fuzzy similarity matrix t(R) from large to small, we can obtain the following set of value points of the classification threshold λ: λ = {1, 0.7, 0.63, 0.62, 0.53}.

[0096] If λ=1, we can get:

[0097]

[0098] At this time, wafer products x1 to x5 are divided into 5 categories, namely: {x1}, {x2}, {x3}, {x4}, {x5}.

[0099] If λ = 0.7, we can get:

[0100]

[0101] At this time, wafer products x1 to x5 are divided into 4 categories, namely {x1}, {x2, x4}, {x3}, {x5}.

[0102] If λ = 0.63, we can get:

[0103]

[0104] At this time, wafer products x1 to x5 are divided into three categories, namely: {x1, x2, x4}, {x3}, {x5}.

[0105] If λ = 0.62, we can get:

[0106]

[0107] At this time, wafer products x1 to x5 are divided into 2 categories, namely: {x1, x2, x3, x4}, {x5}.

[0108] If λ = 0.53, we can get:

[0109]

[0110] At this time, wafer products x1 to x5 are classified into 1 category, namely {x1, x2, x3, x4, x5}.

[0111] Based on the above fuzzy equivalence matrix, the dynamic classification diagram is as follows: Figure 2 shown.

[0112] Step 13, based on the dynamic classification diagram, obtaining the final classification result of the wafer product to be classified.

[0113] In specific implementation, the value conditions of the classification threshold λ can be set according to actual needs. It can be understood that the higher the value of the classification threshold λ, the more accurate the classification of wafer products, but the higher the amount of calculation required for subsequent capacity prediction. For example, referring to Figure 2 , the classification threshold λ can be set to be greater than 85%. Figure 2 , λ should be equal to 1, that is, wafer products x1 to x5 are divided into 5 categories.

[0114] In one embodiment of the present invention, after obtaining the dynamic classification diagram, the classification results in the dynamic classification diagram can be dynamically adjusted in combination with the actual capacity bottleneck and demand production of the production equipment, so that the classification of wafer products is optimal. For example, it can be determined whether each type of product in the dynamic classification diagram has passed through a key process that affects the actual capacity bottleneck during the actual manufacturing process, and then the product classification can be further adjusted based on the judgment result. In addition, the product classification can also be adjusted according to some special production needs. For example, products that are required to be produced as the same type of products should be treated as a type of product for subsequent capacity forecasting.

[0115] Taking a certain platform product as an example, the existing technology is used for classification, and the classification accuracy is about 97.1%. The classification method in the embodiment of the present invention is used, and the confidence level greater than 95% is used as the value condition of the classification threshold λ. The classification result obtained is 98.3%, and the classification accuracy is improved.

[0116] Step 14: predicting production capacity based on the final classification result.

[0117] In specific implementations, after obtaining the final classification results, wafer products in the same category have basically the same product characteristics and are usually manufactured using the same process. At this time, wafer products in the same category can be dispatched to the same equipment, thereby maximizing the factory's production capacity.

[0118] By adopting the solution of the present invention, wafer products are dynamically clustered based on product characteristics, so that products can be grouped according to actual factory needs to simplify product combinations, accurately predict factory production capacity, and ensure maximum factory production capacity.

[0119] In order to enable those skilled in the art to better understand and implement the present invention, the device, storage medium and electronic device corresponding to the above method are described in detail below.

[0120] Reference Figure 3 The embodiment of the present invention provides a production capacity prediction device 30, which may include: a preliminary classification unit 31, a fuzzy clustering unit 32, a classification adjustment unit 33 and a production capacity prediction unit 34. Among them:

[0121] The preliminary classification unit 31 is adapted to perform preliminary classification of wafer products to be classified;

[0122] The fuzzy clustering unit 32 is adapted to perform fuzzy clustering on the wafer products in each preliminary classification based on product characteristics, and obtain a dynamic classification diagram corresponding to the wafer products in each preliminary classification;

[0123] The classification adjustment unit 33 is adapted to obtain a final classification result of the wafer product to be classified based on the dynamic classification diagram;

[0124] The production capacity prediction unit 34 is adapted to predict the production capacity based on the final classification result.

[0125] In one embodiment of the present invention, the fuzzy clustering unit 32 is adapted to perform fuzzy clustering on the wafer products in each preliminary classification based on at least two of the following product characteristics: number of mask layers; number of metal layers; number of ion implantation layers; number of chemical mechanical polishing times.

[0126] In one embodiment of the present invention, the fuzzy clustering unit 32 includes: a characteristic index matrix construction subunit 321, a standardization subunit 322, a fuzzy similarity matrix construction subunit 323, a fuzzy equivalent matrix construction subunit 324 and a dynamic clustering subunit 325. Among them:

[0127] The characteristic index matrix construction subunit 321 is adapted to construct a characteristic index matrix corresponding to each wafer product in the preliminary classification according to product characteristics;

[0128] The standardization subunit 322 is adapted to perform data standardization processing on each characteristic index matrix to obtain a standardized characteristic matrix;

[0129] The fuzzy similarity matrix construction subunit 323 is adapted to construct corresponding fuzzy similarity matrices based on each of the standardized characteristic matrices;

[0130] The fuzzy equivalent matrix construction subunit 324 is adapted to obtain the fuzzy equivalent matrix corresponding to each of the fuzzy similarity matrices by using the fuzzy transfer package method;

[0131] The dynamic clustering subunit 325 is adapted to obtain a dynamic classification diagram corresponding to the wafer products in each preliminary classification based on the fuzzy equivalence matrix.

[0132] In one embodiment of the present invention, the standardization subunit 322 is adapted to use a maximum value method to perform data standardization processing on each characteristic index matrix to obtain a standardized characteristic matrix.

[0133] In one embodiment of the present invention, the fuzzy similarity matrix construction subunit 323 is adapted to construct a corresponding fuzzy similarity matrix based on each of the standardized characteristic matrices using a maximum-minimum method.

[0134] In one embodiment of the present invention, the dynamic clustering subunit 325 is suitable for sorting the elements in the fuzzy equivalence matrix from large to small, and sequentially using them as the value points of the classification threshold. The classification threshold under each value point is used to intercept the fuzzy equivalence matrix respectively to obtain a dynamic classification diagram corresponding to the wafer products in each preliminary classification.

[0135] In a specific implementation, the preliminary classification unit 31 is suitable for preliminarily classifying the wafer products to be classified according to one or more of the technology node, size, and application platform.

[0136] In a specific implementation, the classification adjustment unit 33 is suitable for dynamically adjusting the classification results in the dynamic classification diagram in combination with the actual capacity bottleneck and production demand of the production equipment.

[0137] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the computer instructions are executed, the steps of the production capacity forecasting method in any one of the above embodiments are executed, which will not be repeated here.

[0138] In a specific implementation, the computer-readable storage medium may include: ROM, RAM, magnetic disk or optical disk, etc.

[0139] An embodiment of the present invention further provides an electronic device, wherein the terminal may include a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor runs the computer program, the steps of production capacity prediction in any one of the above embodiments are executed, which will not be repeated here.

[0140] Regarding the various modules / units included in the various devices and products described in the above embodiments, they can be software modules / units, or hardware modules / units, or they can be partially software modules / units and partially hardware modules / units. For example, for various devices and products applied to or integrated in a chip, the various modules / units included therein can all be implemented in the form of hardware such as circuits, or at least some of the modules / units can be implemented in the form of software programs, which run on a processor integrated inside the chip, and the remaining (if any) modules / units can be implemented in the form of hardware such as circuits; for various devices and products applied to or integrated in a chip module, the various modules / units included therein can all be implemented in the form of hardware such as circuits, and different modules / units can be located in the same component (such as a chip, circuit module, etc.) or different components of the chip module, or at least some of the modules / units can be implemented in the form of hardware such as circuits. The element can be implemented in the form of a software program, which runs on a processor integrated inside the chip module, and the remaining (if any) modules / units can be implemented in the form of hardware such as circuits; for various devices and products applied to or integrated in the terminal, the various modules / units contained therein can be implemented in the form of hardware such as circuits, and different modules / units can be located in the same component (for example, chip, circuit module, etc.) or in different components in the terminal, or, at least some modules / units can be implemented in the form of a software program, which runs on a processor integrated inside the terminal, and the remaining (if any) modules / units can be implemented in the form of hardware such as circuits.

[0141] Although the present invention is disclosed as above, the present invention is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the scope defined by the claims.

Claims

1. A capacity forecasting method, characterized in that: include: Carry out preliminary classification of wafer products to be classified; Based on product characteristics, fuzzy clustering is performed on the wafer products in each preliminary classification to obtain a dynamic classification diagram corresponding to the wafer products in each preliminary classification; Based on the dynamic classification diagram, obtaining a final classification result of the wafer product to be classified; Forecast production capacity based on the final classification results.

2. The capacity prediction method according to claim 1, characterized in that: The product characteristics include at least two of the following: Number of mask layers; Number of metal layers; Number of ion implantation layers; Chemical mechanical polishing times.

3. The capacity prediction method according to claim 1, characterized in that: Based on the product characteristics, fuzzy clustering is performed on the wafer products in each preliminary classification to obtain a dynamic classification diagram corresponding to the wafer products in each preliminary classification, including: According to product characteristics, construct a characteristic index matrix corresponding to each wafer product in the preliminary classification; Perform data standardization on each characteristic index matrix to obtain a standardized characteristic matrix; Based on each of the standardized characteristic matrices, corresponding fuzzy similarity matrices are constructed respectively; Using the fuzzy transfer package method, a fuzzy equivalent matrix corresponding to each of the fuzzy similarity matrices is obtained; Based on the fuzzy equivalence matrix, a dynamic classification diagram corresponding to the wafer products in each preliminary classification is obtained.

4. The capacity prediction method according to claim 3, characterized in that: The data standardization process is performed on each characteristic index matrix to obtain a standardized characteristic matrix, including: The maximum value method is used to perform data standardization on each characteristic index matrix to obtain a standardized characteristic matrix.

5. The capacity prediction method according to claim 3, characterized in that: The step of constructing corresponding fuzzy similarity matrices based on the standardized characteristic matrices comprises: Based on each of the standardized characteristic matrices, a corresponding fuzzy similarity matrix is ​​constructed using the maximum-minimum method.

6. The capacity prediction method according to claim 3, characterized in that: Based on the fuzzy equivalent matrix, a dynamic classification diagram corresponding to the wafer products in each preliminary classification is obtained, including: Sort the elements in the fuzzy equivalence matrix from large to small and use them as the value points of the classification threshold in sequence; The fuzzy equivalence matrix is ​​respectively intercepted by using the classification threshold at each value point to obtain a dynamic classification diagram corresponding to the wafer product in each preliminary classification.

7. The capacity prediction method according to any one of claims 1 to 6, characterized in that: The preliminary classification of the wafer products to be classified includes: The wafer products to be classified are preliminarily classified according to one or more of the technology node, size, and application platform.

8. The capacity prediction method according to any one of claims 1 to 6, characterized in that: The obtaining of the final classification result of the wafer product to be classified based on the dynamic classification diagram includes: In combination with the actual capacity bottleneck and production demand of the production equipment, the classification results in the dynamic classification diagram are dynamically adjusted.

9. A production capacity prediction device, characterized in that: include: A preliminary classification unit, suitable for preliminary classification of wafer products to be classified; The fuzzy clustering unit is suitable for performing fuzzy clustering on the wafer products in each preliminary classification based on product characteristics, and obtaining a dynamic classification diagram corresponding to the wafer products in each preliminary classification; A classification adjustment unit, adapted to obtain a final classification result of the wafer product to be classified based on the dynamic classification diagram; The capacity prediction unit is adapted to predict the capacity based on the final classification result.

10. The production capacity prediction device according to claim 9, characterized in that: The fuzzy clustering unit is adapted to perform fuzzy clustering on the wafer products in each preliminary classification based on at least two of the following product characteristics: Number of mask layers; Number of metal layers; Number of ion implantation layers; Chemical mechanical polishing times.

11. The production capacity prediction device according to claim 9, characterized in that: The fuzzy clustering unit comprises: The characteristic index matrix construction subunit is adapted to construct the characteristic index matrix corresponding to the wafer product in each preliminary classification according to the product characteristics; The standardization subunit is suitable for performing data standardization processing on each characteristic index matrix to obtain a standardized characteristic matrix; A fuzzy similarity matrix construction subunit, adapted to construct corresponding fuzzy similarity matrices based on each of the standardized characteristic matrices; A fuzzy equivalent matrix construction subunit is adapted to obtain the fuzzy equivalent matrix corresponding to each of the fuzzy similarity matrices by using the fuzzy transfer package method; The dynamic clustering subunit is suitable for obtaining a dynamic classification diagram corresponding to the wafer products in each preliminary classification based on the fuzzy equivalence matrix.

12. The production capacity prediction device according to claim 11, characterized in that: The standardization subunit is suitable for using the maximum value method to perform data standardization processing on each characteristic index matrix to obtain a standardized characteristic matrix.

13. The production capacity prediction device according to claim 11, characterized in that: The fuzzy similarity matrix construction subunit is suitable for constructing a corresponding fuzzy similarity matrix based on each of the standardized characteristic matrices using the maximum-minimum method.

14. The production capacity prediction device according to claim 11, characterized in that: The dynamic clustering subunit is suitable for sorting the elements in the fuzzy equivalence matrix from large to small, and taking them as the value points of the classification threshold in turn. The classification threshold under each value point is used to intercept the fuzzy equivalence matrix respectively to obtain the dynamic classification diagram corresponding to the wafer products in each preliminary classification.

15. The production capacity prediction device according to any one of claims 9 to 14, characterized in that: The preliminary classification unit is suitable for preliminarily classifying the wafer products to be classified according to one or more of the technology node, size, and application platform.

16. The production capacity prediction device according to any one of claims 9 to 14, characterized in that: The classification adjustment unit is suitable for dynamically adjusting the classification results in the dynamic classification diagram in combination with the actual production capacity bottleneck and production demand of the production equipment.

17. A computer-readable storage medium having a computer program stored thereon, characterized in that: The computer program is executed by a processor to implement the steps of the method according to any one of claims 1 to 8.

18. An electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, wherein: When the processor runs the computer program, the steps of the method according to any one of claims 1 to 8 are performed.

Images