An adjustment method and device of an integrated circuit, a storage medium and a terminal device

Abstract

The application discloses an integrated circuit adjustment method and device, a storage medium and a terminal device. The method comprises the following steps: acquiring feature information corresponding to an integrated circuit, wherein the feature information comprises a position parameter corresponding to a circuit layout of the integrated circuit and photoelectric characteristic information corresponding to the integrated circuit; determining a target position parameter corresponding to the integrated circuit based on the feature information, and determining an adjusted integrated circuit based on the target position parameter. In this way, the position parameter is automatically adjusted based on the photoelectric characteristic, the circuit layout is automatically adjusted, and therefore, the design efficiency of the circuit layout can be improved, and the production efficiency of the integrated circuit can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of integrated circuits, and particularly relates to an adjustment method and device of an integrated circuit, a storage medium and a terminal device. BACKGROUND

[0002] Electronic Design Automation (EDA) is widely applied in the field of circuit design, and through the electronic design automation, the design of the circuit is more efficient. However, in the field of TFT circuit design, the process of TFT layout design generally needs professional personnel to operate software for drawing, which makes the TFT circuit design highly dependent on manpower, thereby causing low production efficiency. SUMMARY

[0003] The present application aims at solving the problems in the prior art, and provides an adjustment method and device of an integrated circuit, a storage medium and a terminal device.

[0004] In order to solve the above technical problems, the first aspect of the present application provides an adjustment method of an integrated circuit, which comprises the following steps:

[0005] obtaining feature information corresponding to the integrated circuit, wherein the feature information comprises a position parameter corresponding to a circuit layout of the integrated circuit and photoelectric characteristic information corresponding to the integrated circuit;

[0006] determining a target position parameter corresponding to the integrated circuit based on the feature information, and determining an adjusted integrated circuit based on the target position parameter.

[0007] The method for obtaining the photoelectric characteristic, wherein the determination of the target position parameter corresponding to the integrated circuit based on the feature information specifically comprises:

[0008] obtaining a reference position parameter corresponding to the position parameter, wherein the dimension of the reference position parameter is lower than that of the position parameter;

[0009] determining an optimized position parameter corresponding to the reference position parameter based on the photoelectric characteristic and the reference position parameter, wherein the dimension of the optimized position parameter is equal to that of the reference position parameter;

[0010] determining the target position parameter corresponding to the position parameter based on the optimized position parameter.

[0011] The method for obtaining the photoelectric characteristic, wherein the determination of the optimized position parameter corresponding to the reference position parameter based on the photoelectric characteristic and the reference position parameter specifically comprises:

[0012] acquire a target function corresponding to the integrated circuit, and determine a target value corresponding to the optoelectronic feature based on the target function;

[0013] based on the target value, use a Bayesian optimizer to optimize the reference position parameter to obtain an optimized position parameter.

[0014] The method for acquiring the optoelectronic feature, wherein the determining the target position parameter corresponding to the position parameter based on the optimized position parameter specifically comprises:

[0015] inputting the optimized position parameter into a trained third network model, and outputting a target position parameter corresponding to the position parameter through the third network model.

[0016] The method for acquiring the optoelectronic feature, wherein the detection network model comprises a cascaded first network model and a second network model; the input of the first network model is a first high-dimensional position parameter, and the output is a first low-dimensional position parameter; the input of the third network model is a second low-dimensional position parameter, and the output is a second high-dimensional position parameter, and the dimension of the first high-dimensional position parameter is equal to the dimension of the second high-dimensional position parameter, and the dimension of the first low-dimensional position parameter is equal to the dimension of the second low-dimensional position parameter.

[0017] The method for acquiring the optoelectronic feature, wherein the training process of the third network model specifically comprises:

[0018] inputting the first position parameter corresponding to each integrated circuit in the training sample into a trained first network model, and outputting a second position parameter through the first network model, wherein the first network model and the second network model are jointly trained;

[0019] inputting the second position parameter into a preset network model, and outputting a third position parameter through the preset network model;

[0020] based on the first position parameter and the third position parameter, training the preset network model to obtain a third network model.

[0021] The method for acquiring the optoelectronic feature, wherein after the target position parameter corresponding to the integrated circuit is determined based on the feature information, and the adjusted integrated circuit is determined based on the target position parameter, the method further comprises:

[0022] acquire a target optoelectronic feature corresponding to the target position parameter, and save the circuit parameters formed by the target position parameter and the target optoelectronic feature.

[0023] The method for obtaining the optoelectrical characteristic, wherein after the target optoelectrical characteristic corresponding to the target position parameter is obtained and the circuit parameter composed of the target position parameter and the target optoelectrical characteristic is saved, the method further comprises:

[0024] selecting a target circuit parameter corresponding to the expected optoelectrical characteristic from the pre-stored circuit parameters based on the preset expected optoelectrical characteristic;

[0025] determining the integrated circuit corresponding to the expected optoelectrical characteristic based on the target position parameter in the target circuit parameter.

[0026] The method for obtaining the optoelectrical characteristic, wherein the selecting the target circuit parameter corresponding to the expected optoelectrical characteristic from the pre-stored circuit parameters based on the preset expected optoelectrical characteristic specifically comprises:

[0027] obtaining a weight coefficient set corresponding to the expected optoelectrical characteristic, wherein the weight coefficient set comprises a weight coefficient corresponding to each optoelectrical parameter in the expected optoelectrical characteristic;

[0028] selecting the target circuit parameter from the circuit parameters based on the expected optoelectrical characteristic and the weight coefficient set.

[0029] The method for obtaining the optoelectrical characteristic, wherein the pre-stored circuit parameters are stored in the form of KD tree.

[0030] The method for obtaining the optoelectrical characteristic, wherein the selecting the target circuit parameter from the circuit parameters based on the expected optoelectrical characteristic and the weight coefficient set specifically comprises:

[0031] performing KD tree search on the circuit parameters based on the expected optoelectrical characteristic to obtain the target circuit parameter corresponding to the expected optoelectrical characteristic, wherein the search range of the node optoelectrical characteristic in the backtracking process is determined based on the node optoelectrical characteristic, the expected optoelectrical characteristic and the weight coefficient set.

[0032] The method for obtaining the optoelectrical characteristic, wherein the integrated circuit is TFT circuit.

[0033] The second aspect of the embodiments of the present application provides an adjusting device of integrated circuit, the adjusting device comprises:

[0034] a first obtaining module, configured to obtain characteristic information corresponding to the integrated circuit, wherein the characteristic information comprises a position parameter corresponding to a circuit layout of the integrated circuit and optoelectrical characteristic information corresponding to the integrated circuit;

[0035] a first determining module, configured to determine a target position parameter corresponding to the integrated circuit based on the characteristic information, and determine an adjusted integrated circuit based on the target position parameter.

[0036] The third aspect of the embodiments of the present application provides a computer readable storage medium, which stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps in the adjustment method of the integrated circuit according to any one of the above.

[0037] The fourth aspect of the embodiments of the present application provides a terminal device, which comprises a processor, a memory and a communication bus; the memory stores a computer readable program which can be executed by the processor;

[0038] The communication bus realizes the connection and communication between the processor and the memory.

[0039] The processor realizes the steps in the adjustment method of the integrated circuit according to any one of the above when executing the computer readable program.

[0040] Advantages: compared with the prior art, the present application provides an adjustment method, device, storage medium and terminal device of an integrated circuit, the method comprises obtaining feature information corresponding to the integrated circuit, the feature information comprises a position parameter corresponding to a circuit layout of the integrated circuit and photoelectric characteristic information corresponding to the integrated circuit; determining a target position parameter corresponding to the integrated circuit based on the feature information, and determining an adjusted integrated circuit based on the target position parameter, so as to automatically adjust the position parameter based on the photoelectric characteristic, so as to realize automatic adjustment of the circuit layout, thereby improving the design efficiency of the circuit layout, and improving the production efficiency of the integrated circuit. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0042] Figure 1 An example diagram of the prior method of the adjustment method of the integrated circuit provided by the present application.

[0043] Figure 2 A flowchart of the adjustment method of the integrated circuit provided by the present application.

[0044] Figure 3 An example diagram of TFT circuit.

[0045] Figure 4 An example diagram of a sub-pixel structure in TFT circuit.

[0046] Figure 5An example diagram of a position parameter of a sub-pixel structure in a TFT circuit.

[0047] Figure 6 An example diagram of a position parameter adjustment process of an integrated circuit adjustment method provided by the present application.

[0048] Figure 7 An example diagram of a plurality of reference position parameter acquisition processes of an integrated circuit adjustment method provided by the present application.

[0049] Figure 8 An example diagram of a position parameter adjustment process of an integrated circuit adjustment method provided by the present application.

[0050] Figure 9 An example diagram of a target optoelectronic feature acquisition process in an integrated circuit adjustment method provided by the present application.

[0051] Figure 10 An example diagram of a target optoelectronic feature acquisition process in an integrated circuit adjustment method provided by the present application.

[0052] Figure 11 A structural schematic diagram of an integrated circuit adjustment device provided by the present application.

[0053] Figure 12 A structural schematic diagram of a terminal device provided by the present application. DETAILED DESCRIPTION

[0054] The present application provides an integrated circuit adjustment method, device, storage medium and terminal device, in order to make the purpose, technical scheme and effect of the present application more clear and explicit, the present application is further explained in detail below with reference to the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application, and are not used to limit the present application.

[0055] Those skilled in the art can understand that, unless specifically stated otherwise, the singular form "a", "an", "said" and "the" used herein can also include the plural form. It should be further understood that the use of the word "comprise" in the specification of the present application means that the features, integers, steps, operations, elements and / or components exist, but do not exclude the existence or addition of one or more other features, integers, steps, operations, elements, components and / or their combinations. It should be understood that when we say that an element is "connected" or "coupled" to another element, it can be directly connected or coupled to the other element, or there can be intermediate elements. In addition, the "connection" or "coupling" used herein can include wireless connection or wireless coupling. The phrase "and / or" used herein includes all or any single unit and all combinations of the associated listed items.

[0056] Those skilled in the art can understand that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It should also be understood that the terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein. It should be understood that the sequence of the steps and the size in the embodiments do not mean the order of execution, and the execution order of the processes is determined by its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the application.

[0057] The inventors have found that electronic design automation (EDA) is widely used in the field of circuit design, and through electronic design automation, the design of the circuit is more efficient. However, as shown in the prior art, in the field of TFT circuit design, professional personnel generally need to operate software to draw in the process of TFT layout design, which makes the TFT circuit design rely on manpower, thereby causing low production efficiency. Figure 1

[0058] To solve the above problems, in the embodiments of the present application, the feature information corresponding to the integrated circuit is obtained, the target position parameter corresponding to the integrated circuit is determined based on the feature information, and the adjusted integrated circuit is determined based on the target position parameter. In this way, the position parameter is automatically adjusted based on the photoelectric characteristics to automatically adjust the circuit layout, thereby improving the design efficiency of the circuit layout, thereby improving the production efficiency of the integrated circuit.

[0059] For example, the embodiments of the present application can be applied to the scene of designing the circuit layout of the TFT circuit through the terminal device. In this scene, the terminal device can determine the TFT circuit to be designed, and obtain the circuit layout corresponding to the TFT circuit. After obtaining the circuit layout, the position parameter corresponding to the integrated circuit is determined based on the circuit layout. Based on the trained detection network model and the position parameter, the photoelectric characteristics corresponding to the integrated circuit are determined to obtain the feature information corresponding to the integrated circuit, the adjustment parameter corresponding to the position parameter is determined based on the feature information. The target position parameter corresponding to the integrated circuit is determined based on the feature information, and the adjusted integrated circuit is determined based on the target position parameter.

[0060] ​It can be understood that, in the above application scenarios, although the actions of the embodiments of the present application are described as being performed entirely by the terminal device, these actions can also be partially performed by the terminal device and partially performed by the server connected to the terminal device. For example, after the terminal device obtains the circuit layout corresponding to the integrated circuit, the terminal device inputs the circuit layout to the server, so that the server obtains the circuit layout. The server can determine the position parameters corresponding to the integrated circuit based on the input circuit layout of the terminal device; determine the optoelectronic characteristics corresponding to the integrated circuit based on the trained detection network model and the position parameters, to obtain the feature information corresponding to the integrated circuit, determine the target position parameters corresponding to the integrated circuit based on the feature information, and determine the adjusted integrated circuit based on the target position parameters. Thus, the present application is not limited in terms of execution subject, as long as the actions disclosed in the embodiments of the present application are performed.

[0061] It should be noted that the above application scenarios are only shown for the purpose of facilitating understanding of the present application, and the embodiments of the present application are not limited in this respect. On the contrary, the embodiments of the present application can be applied to any applicable scenario.

[0062] The application content will be further described by describing the embodiments in combination with the accompanying drawings.

[0063] The present embodiment provides a method for adjusting an integrated circuit, as shown in Figure 2 and Figure 8 The method comprises:

[0064] S10, obtaining feature information corresponding to an integrated circuit.

[0065] Specifically, the integrated circuit can be a chip integrated circuit, or an integrated circuit of a display panel, or an integrated circuit of a pixel unit. In one implementation manner of the present embodiment, the integrated circuit is a TFT circuit, and the TFT circuit is used to manufacture a TFT backplane. For example, as shown in Figure 3 The TFT circuit can include a GOA (Gate Driver on Array, gate driver on array) and a plurality of pixels, each of the plurality of pixels including three sub-pixel units, i.e., an R pixel unit, a G pixel unit, and a B pixel unit, wherein, as shown in Figure 4 The sub-pixel unit can include circuit elements and connection lines, the circuit elements can include capacitors, TFT transistors, light-emitting diodes, etc., and the connection lines can include scan lines and data lines, etc.

[0066] The feature information includes position parameters corresponding to the circuit layout of the integrated circuit and photoelectric characteristics corresponding to the integrated circuit. The position parameters are a parameterized representation of the circuit layout, by which the various components included in the circuit layout, the relative positional relationship between the various components, and the positional information of the various components in the circuit layout can be determined. In other words, the position parameters are a parameter vector for representing the circuit layout, and the circuit layout can be drawn based on the position parameters. The photoelectric characteristics are used to reflect the photoelectric properties of the driving circuit corresponding to the circuit layout, and the photoelectric characteristics can include aperture ratio, charge rate, maximum voltage, and the like.

[0067] The circuit layout is used to map the circuit design of the integrated circuit to the physical description level, so that the integrated circuit can be mapped to the wafer for production. The circuit layout contains the types of components in the integrated circuit, the sizes of the components, the relative positions between the components, and the connection relationships between the components, and other related physical information. The circuit layout of the integrated circuit can be automatically generated in advance, or manually designed by a layout designer. In one implementation manner of the embodiment, the circuit layout is automatically generated, and the generation process can be: determining the integrated circuit to be designed, obtaining a plurality of components corresponding to the integrated circuit and the relative positional relationship between the components, and generating the circuit layout of the integrated circuit based on the relative positional relationship.

[0068] In one implementation manner of the embodiment, the position parameters can be pre-obtained and stored locally in the terminal device, or transmitted to the terminal device by an external device, or obtained through the cloud, or determined by the terminal device based on the circuit layout corresponding to the integrated circuit. In one implementation manner of the embodiment, the feature information is determined by the terminal device based on the circuit layout corresponding to the integrated circuit, and accordingly, the obtaining the feature information corresponding to the integrated circuit specifically includes:

[0069] A10, obtaining the circuit layout corresponding to the integrated circuit.

[0070] A20, identifying the area information of the components in the circuit layout.

[0071] A30, determining the position parameters of the circuit layout based on the obtained area information;

[0072] A40, determining the photoelectric characteristics corresponding to the integrated circuit based on the position information, to obtain the feature information corresponding to the integrated circuit.

[0073] Specifically, the component is a component used to form the circuit layout, wherein the component can include a metal layer, a TFT transistor, a capacitor, an ITO film, and the like. The region information is used to locate the component, and through the region information, a location region of the component in the circuit layout and a component type of the component can be determined. It can be understood that the region information includes location information of the component, size information of the component, and category information of the component, the location information is used to reflect a position of the component in the circuit layout, the size information is used to reflect a size of the component, and the category information is used to reflect a component category of the component. For example, the size information is a height and a width of an identification region corresponding to the component, and the location information is a distance between a center of the identification region corresponding to the component and an image center of the panel image. Of course, in actual application, the location information can also be determined in other manners, for example, a distance between a center point of the identification region corresponding to the component and a top-left corner of the panel image, a distance between a top-left corner of the identification region corresponding to the component and a top-left corner of the panel image, and the like. The size information can also be determined in other manners, for example, a perimeter of the identification region corresponding to the component, an area of the identification region corresponding to the component, and the like.

[0074] Based on this, in one implementation manner of the embodiment, the region information can be in the form of a circuit parameter, and the circuit parameter includes three data items, namely, the location information, the size information, and the category information. For example, the region information is {(100, 100), (20, 30), capacitor}, (100, 100) indicates coordinate information of a location point in the component in a coordinate system corresponding to the circuit layout, 20 in (20, 30) can indicate a width of a rectangular region corresponding to the component, 30 can indicate a width of the rectangular region corresponding to the component, and capacitor indicates that the component type of the component is a capacitor. In addition, the rectangular region refers to a minimum rectangle containing the component. Of course, it is worth noting that the rectangular region here is only an example, and it can also be a circle, an ellipse, a triangle, a regular pentagon, and the like, and when the shape is different, the form of the size information is also different, which is not described one by one here, and only one example is described, for example, when the component corresponds to a circular region, the region information can be circular coordinates and a circular radius.

[0075] In one implementation manner of the embodiment, the region information of the component in the circuit layout specifically includes:

[0076] inputting the circuit layout into the trained recognition network model;

[0077] determining, by the recognition network model, the region information of the component in the circuit layout.

[0078] Specifically, the recognition network model can be a pre-trained network model for recognizing the region information of each component in the circuit layout. It can be understood that the recognition network model is a pre-trained network model, the input of the recognition network model is the circuit layout, and the output of the recognition network model is the region information. When the circuit layout includes one component, the region information is one; when the circuit layout includes multiple components, the region information is multiple, and the multiple region information corresponds to the multiple components one by one. For example, the circuit layout includes component A and component B, and the region information output by the recognition network model includes region information a and region information b, where the region information a corresponds to the component A and is used to locate the component A, and the region information b corresponds to the component B and is used to locate the component B.

[0079] In one implementation manner of the embodiment, the determining the position parameter of the circuit layout based on the obtained region information specifically includes:

[0080] For each component in the circuit layout, a reference component corresponding to the component is obtained;

[0081] Based on the region information of the component, a sub-position parameter corresponding to the component is determined with reference to the reference component;

[0082] Based on all the obtained sub-position parameters, the position parameter of the circuit layout is determined.

[0083] Specifically, the reference component is a reference of the component, which is used to determine the relative position relationship between the component and the reference component, such as the distance between the component and the reference component, the distance between the edge of the component and the edge of the reference component, the inclination angle of the component relative to the reference component, and the like. The reference component can be determined based on the obtained region information corresponding to each component, or can be determined based on a pre-set component position relationship list, wherein the component position relationship list can store the position relationship between all components included in the circuit layout, for example, the circuit version includes component A and component B, and the component A and the component B are adjacent in the component position relationship list.

[0084] In one implementation manner of the embodiment, the reference component is determined based on the obtained region information corresponding to each component, and the obtaining process thereof can specifically be: after obtaining the region information corresponding to each component, candidate components located around the component are selected based on the region information in all components, and all the obtained candidate components are taken as the reference component of the component. In this way, the reference component corresponding to the component can be automatically determined according to the obtained region information, so that the speed of parameterization of the circuit layout can be improved.

[0085] In one implementation of this embodiment, the reference component is determined based on a pre-set list of component position relationships. The determination process specifically includes: for each component, selecting candidate components associated with that component from the list of component position relationships, and using all selected candidate components as the reference component corresponding to that component. By determining the reference component through a pre-set list of component position relationships, the problem of incorrect reference components caused by errors in regional information can be avoided, thereby improving the accuracy of the reference component.

[0086] The sub-positional parameters reflect the component's own size information and its positional information relative to the reference component. These parameters determine the component's size and distance from the reference component. Therefore, the sub-positional parameters reflect several attributes of the component, including its size and distance from the reference component. Correspondingly, each sub-positional parameter includes several positional parameter items, each corresponding to one of the attributes. The value of each positional parameter item is the attribute value of its corresponding attribute. These attributes are determined based on circuit layout design rules, such as: linewidth rule: minimum width of polygons in the layout; maximum (minimum) size limit: width or length of polygons; spacing rule: minimum distance between polygons; enclosure rule: minimum size of overlap between two layers of lines; overlap rule: minimum size of overlap between two layers; minimum area rule: minimizing the layout area while meeting the above basic requirements.

[0087] Based on this, after obtaining the reference component corresponding to the component, the position parameter items included in the sub-position parameters corresponding to the component are determined according to the rules of the circuit layout. Then, the parameter values ​​corresponding to each position parameter item are calculated sequentially based on the region information to obtain the sub-position parameters corresponding to the component. For example, if the integrated circuit is a TFT sub-pixel circuit, and the capacitor in this TFT sub-pixel circuit is a rectangle with a chamfered corner, and the reference components corresponding to the capacitor are the TFT and the bottom edge of the substrate, then the sub-position parameters corresponding to the capacitor may include the rectangle length, matrix width, distance to the bottom edge of the substrate, distance to the TFT, and the size of the chamfered corner region, etc.

[0088] For example: Figure 5 As shown, a capacitor in the TFT sub-pixel circuit has 12 sub-position parameters, namely the distances A, B, C, D, E, F, G, H, I, J, K and L in the figure. The sub-position parameters corresponding to the capacitor can be expressed as (A, B, C, D, E, F, G, H, I, J, K, L).

[0089] In one implementation form of the embodiment, the sub-position parameters are sub-position vectors; and the determining the position parameter of the circuit layout based on the obtained all sub-position parameters specifically includes:

[0090] splicing the sub-position parameters to obtain the position parameter of the circuit layout.

[0091] Specifically, the position parameter is spliced from the sub-position parameters, and the dimension of the position parameter is equal to the sum of the dimensions of the sub-position parameters. For example, the sub-position parameters corresponding to the circuit layout include a sub-position parameter A and a sub-position parameter B, the position parameter A is (A1, A2), and the sub-position parameter B is (B1, B2), and then the position parameter is (A1, A2, B1, B2). In addition, in order to determine the corresponding component of each position parameter item in the spliced position parameter, each position parameter item in the position parameter can be configured with the component category of the corresponding component, wherein the component category can be used as the subscript of each position parameter item in the sub-position parameter, or as the suffix of each position parameter item in the sub-position parameter, or as the prefix of each position parameter item in the sub-position parameter, and the like.

[0092] In one implementation form of the embodiment, the photoelectric characteristic is determined based on a trained detection network model. Correspondingly, the determining the photoelectric characteristic corresponding to the integrated circuit based on the position information to obtain the feature information corresponding to the integrated circuit specifically includes:

[0093] inputting the position information into the detection network model, and outputting the photoelectric characteristic corresponding to the integrated circuit through the detection network model.

[0094] Specifically, the detection network model is pre-trained and used to determine the photoelectric characteristic corresponding to the integrated circuit. It can be understood that the detection network model is used to convert the position parameter into the photoelectric characteristic. Correspondingly, the input item of the detection network model is the position parameter, and the output item of the detection network model is the photoelectric characteristic, wherein the photoelectric characteristic can include the aperture ratio, the charging rate, the RC, the LCS, the voltage deviation Bestvcom, the feedthrough, the charging time, and the like.

[0095] In one implementation form of the embodiment, the detection network model includes a first full connection module, a transformation module, and a second full connection module. The determining the photoelectric characteristic corresponding to the integrated circuit based on the trained detection network model and the position parameter specifically includes:

[0096] inputting the position parameter into the first full connection module, and outputting a first feature vector through the first full connection module;

[0097] inputting the first feature vector into a transformation module, and outputting a second feature vector through the transformation module, wherein a dimension of the first feature vector is equal to a dimension of the second feature vector;

[0098] inputting the second feature vector into a second full connection module, and outputting the photoelectric characteristic corresponding to the integrated circuit through the second full connection module.

[0099] Specifically, the first full connection module is configured to perform linear transformation on the position parameter to reduce the dimension of the position parameter. The first feature vector is a low-dimensional vector obtained by performing linear transformation on the position parameter through the first full connection module. Correspondingly, a vector dimension of the first feature vector is less than a vector dimension of the position parameter. The vector dimension of the position parameter can be 50-500, and the vector dimension of the first feature vector can be 5-500. In a specific implementation, the vector dimension of the position parameter can be 50-100, and the vector dimension of the first feature vector can be 5-50. For example, the vector dimension of the position parameter is 100, and the vector dimension of the first feature vector is 50.

[0100] The transformation module is configured to convert the first characteristic vector into the second characteristic vector, and a vector dimension of the second characteristic vector is equal to a vector dimension of the first characteristic vector. For example, the vector dimension of the first characteristic vector is 50, and the vector dimension of the second characteristic vector is also 50. In an implementation of the embodiment, the transformation module can use a sigmoid function, a tanh function, or the like. The output item of the second full connection layer is the photoelectric characteristic, and the dimension of the output item of the second full connection layer can be determined according to the characteristic items included in the photoelectric characteristic obtained according to actual application requirements. For example, the photoelectric characteristic includes an opening rate, a charging rate, an RC, an LCS, a voltage deviation Bestvcom, a feedthrough, and a charging time. Therefore, the dimension of the output item of the second full connection layer is 7.

[0101] In an implementation of the embodiment, the detection network model can include two cascaded network models, namely a first network model and a second network model. The first network model is configured to reduce the dimension of the position parameter, and the second network model is configured to determine the photoelectric characteristic corresponding to the position parameter. Correspondingly, determining the photoelectric characteristic corresponding to the integrated circuit based on the trained detection network model and the position parameter specifically includes:

[0102] inputting the position parameter into the first network model, and outputting a candidate position parameter through the first network model;

[0103] The candidate position parameter is input into the second network model, and photoelectric characteristics corresponding to the integrated circuit are output by the second network model.

[0104] Specifically, the candidate position parameter is an output item of the first network model, and an input item of the second network model is the candidate position parameter, and an output item is photoelectric characteristics. The first network model is used for dimension reduction of the position parameter to obtain the candidate position parameter after dimension reduction. Correspondingly, the vector dimension of the candidate position parameter is smaller than the vector dimension of the position parameter. The value range of the vector dimension of the position parameter can be 50-500, and the value range of the vector dimension of the candidate position parameter can be 5-500. In a specific implementation manner, the value range of the vector dimension of the position parameter can be 50-100, and the value range of the vector dimension of the candidate position parameter can be 5-50. For example, the vector dimension of the position parameter is 100, and the vector dimension of the candidate position parameter is 50.

[0105] In an implementation manner of the embodiment, the first network model can include a third full connection module and a nonlinear transformation module, and the second network model can include a fourth full connection module. The third full connection module is used for linear transformation of the position parameter to reduce the dimension of the position parameter. The third full connection module is connected with the nonlinear transformation module. An output item of the third full connection module is an input item of the nonlinear transformation module. An output item of the nonlinear transformation module is the candidate position parameter. The vector dimension of the candidate position parameter is equal to the vector dimension of the output item of the third full connection module. For example, the vector dimension of the output item of the third full connection module is 50, and the vector dimension of the candidate position parameter is 50. In an implementation manner of the embodiment, the nonlinear transformation module can adopt a sigmoid function and a tanh function.

[0106] In an implementation manner of the embodiment, the second network model includes a fourth full connection module. The dimension of an output item of the fourth full connection module can be determined according to the characteristic items included in the photoelectric characteristics obtained according to actual application requirements. For example, the photoelectric characteristics include an aperture ratio, a charging rate, an RC, an LCS, a voltage deviation Bestvcom, a feedthrough, and a charging time. The dimension of the output item of the second full connection layer is 7.

[0107] In an implementation manner of the embodiment, the training process of the detection network model specifically includes:

[0108] obtaining a training sample set;

[0109] inputting a training position parameter in the training sample set into a preset network model, and outputting a predicted photoelectric characteristic corresponding to the training position parameter by the preset network model;

[0110] Based on the predicted photoelectric characteristic and the target photoelectric characteristic, the preset network model is trained to obtain the detection network model.

[0111] Specifically, the preset network model can be pre-set for generating the detection network model based on the training sample set; and the detection network model is obtained by training the preset network model based on the training sample set, and is used for determining the photoelectric characteristic corresponding to the position parameter. It can be understood that when the preset network model is trained based on the training sample set, the detection network model can be obtained, wherein the model structure of the detection network model is the same as that of the preset network model, and the difference between the detection network model and the preset network model is that the model parameters configured by the detection network model are trained model parameters, and the model parameters configured by the preset network model are initial model parameters.

[0112] In one implementation manner of the embodiment, the training sample set includes a plurality of training position parameters and target photoelectric characteristics corresponding to the training position parameters respectively, wherein each of the plurality of training position parameters corresponds to a circuit layout, and the target photoelectric characteristic corresponding to the training position parameter is the photoelectric characteristic of the integrated circuit corresponding to the circuit layout. The target photoelectric characteristic is used as the label value of the corresponding training position parameter, and after the predicted photoelectric characteristic corresponding to the training position parameter is determined by the preset network model, the target photoelectric characteristic is used as the standard to determine the loss value corresponding to the predicted photoelectric characteristic, so as to perform backward training on the preset network model based on the loss value, so as to optimize the model parameters of the preset network model.

[0113] In one implementation manner of the embodiment, the training process of the detection network model specifically includes:

[0114] Obtaining a training integrated circuit set;

[0115] For each training integrated circuit, determining a training position parameter corresponding to the circuit layout of the training integrated circuit, and determining a target photoelectric characteristic corresponding to the training integrated circuit by a circuit simulator;

[0116] Based on the training position parameter and the target photoelectric characteristic corresponding to each training integrated circuit, a training sample set is determined.

[0117] Specifically, the training integrated circuit set can include a plurality of training integrated circuits, each of which is pre-designed and tested. Each of the plurality of training integrated circuits corresponds to a training position parameter, wherein the determination process of the training position parameter can refer to the above-mentioned determination process of the position parameter, which will not be repeated here. In addition, the target optoelectronic characteristic is determined based on a circuit simulator, and the determination process can be: for each training integrated circuit, obtaining the circuit layout corresponding to the training integrated circuit, and outputting the target optoelectronic characteristic corresponding to the training integrated circuit through the circuit simulator.

[0118] Further, after obtaining the training position parameter and the target optoelectronic characteristic corresponding to the training integrated circuit, the training position parameter and the target optoelectronic characteristic are taken as a circuit parameter. A plurality of training integrated circuits can determine a plurality of groups of circuit parameters, each group of circuit parameters including a training position parameter and a target optoelectronic characteristic. Thus, after obtaining a plurality of groups of circuit parameters corresponding to a plurality of training integrated circuits, the set of a plurality of groups of circuit parameters corresponding to a plurality of training integrated circuits can be taken as a training sample set of a preset network model.

[0119] S20, determining a target position parameter corresponding to the integrated circuit based on the feature information, and determining an adjusted integrated circuit based on the target position parameter.

[0120] Specifically, the target position parameter is an adjusted position parameter corresponding to the position parameter, and the vector dimension of the target position parameter is the same as that of the position parameter. For example, if the vector dimension of the position parameter is 100, the vector dimension of the target position parameter is 100. In addition, after the target position parameter is determined, the circuit layout can be determined based on the target position parameter, and the integrated circuit can be processed based on the circuit layout.

[0121] In one implementation of the embodiment, the determination of the target position parameter corresponding to the integrated circuit based on the feature information specifically includes:

[0122] Obtaining a reference position parameter corresponding to the position parameter;

[0123] Based on the optoelectronic characteristic and the reference position parameter, determining an optimized position parameter corresponding to the reference position parameter;

[0124] Based on the optimized position parameter, determining a target position parameter corresponding to the position parameter.

[0125] Specifically, the vector dimension of the reference position parameter is less than that of the position parameter, and the reference position parameter is obtained by dimension reduction transformation on the position parameter. In the embodiment, the reference position parameter can be determined by the detection network model described above. When the detection network model includes a first network model and a second network model, the reference position parameter is an output item of the first network model. When the detection network model includes a first full connection module, a transformation module, and a second full connection module, the reference position parameter is an output item of the transformation module.

[0126] In one implementation of the embodiment, the determining of the optimization position parameter corresponding to the reference position parameter based on the photoelectric characteristic and the reference position parameter specifically includes:

[0127] obtaining a target function corresponding to the integrated circuit, and determining a target value corresponding to the photoelectric characteristic based on the target function;

[0128] based on the target value, using a Bayesian optimizer to optimize the reference position parameter to obtain an optimization position parameter.

[0129] Specifically, the vector dimension of the optimization position parameter is the same as that of the reference position parameter, and the optimization position parameter is a position parameter that maximizes the target function. It can be understood that the optimization position parameter can maximize the probability that the error based on the target function is reduced. The target function is determined based on the photoelectric characteristic restriction condition of the integrated circuit, and the determination process and the form of the target function are the same as those of the target function described below. In addition, the target value is determined by inputting the photoelectric characteristic into the target function, and the target function is a target function with the photoelectric characteristic as the independent variable. After obtaining the target value, the Bayesian optimizer calculates the optimization position parameter that maximizes the target function value based on the target value and the reference position parameter. In the embodiment, the Bayesian optimizer is used for Gaussian regression optimization. In other implementations, other optimizers can also be used, such as genetic algorithm, annealing algorithm, etc.

[0130] In one implementation of the embodiment, the determining of the optimization position parameter corresponding to the reference position parameter based on the photoelectric characteristic and the reference position parameter specifically includes:

[0131] inputting the optimization position parameter into a third network model trained, and outputting a target position parameter corresponding to the position parameter by the third network model.

[0132] Specifically, the third network model is pre-trained for dimension increasing of the optimization position parameter to obtain a target position parameter, wherein a vector dimension of the target position parameter is greater than a vector dimension of the optimization position parameter, and the vector dimension of the target position parameter is equal to a vector dimension of the position parameter. It can be understood that a dimension of an output item of the third network model is the same as a dimension of an input item of the detection model, when the detection network model comprises a first full connection module, a transformation module and a second full connection module, the dimension of the input item of the third network model is the same as a dimension of an output item of the transformation module; when the detection network model comprises a first network model and a second network model, the dimension of the input item of the third network model is the same as a dimension of an output item of the first network model.

[0133] For example, the detection network model comprises a first network model and a second network model in cascade; an input item of the first network model is a first high-dimensional position parameter, and an output item of the first network model is a first low-dimensional position parameter; an input item of the third network model is a second low-dimensional position parameter, and an output item of the third network model is a second high-dimensional position parameter, and a dimension of the first high-dimensional position parameter is equal to a dimension of the second high-dimensional position parameter, and a dimension of the first low-dimensional position parameter is equal to a dimension of the second low-dimensional position parameter.

[0134] In one implementation form of the embodiment, the training process of the third network model specifically comprises:

[0135] inputting the first position parameter corresponding to each integrated circuit in the training sample into the trained first network model, and outputting a second position parameter through the first network model, wherein the first network model and the second network model are jointly trained;

[0136] inputting the second position parameter into a preset network model, and outputting a third position parameter through the preset network model;

[0137] training the preset network model based on the first position parameter and the third position parameter to obtain a third network model.

[0138] Specifically, the first network model is included in the detection network model, when the detection network model comprises a first network model and a second network model, the first network model is the first network model in the detection network model, when the detection network model comprises a first full connection module, a transformation module and a second full connection module, the detection network model can be divided into the first network model and the second network model, wherein the first network model comprises the first full connection module and the transformation module, and the second network model comprises the second full connection module.

[0139] Further, the input of the third network model is determined based on the first network model, and the input of the third network model is the output of the first network model, so that when the third network model is trained, the training sample corresponding to the third network model can be determined based on the trained first network model, the first network model and the third network model are jointly trained to train the third network model, and the model coefficients of the first network model remain unchanged in the process of training the third network model.

[0140] In one implementation of the embodiment, the third network model is obtained by jointly training the first network model and the third network model. In the training process, the first position parameter is the input of the first network model, the second position parameter is the output of the first network model and the input of the third network model, and the third position parameter is the output of the third network model, wherein the first position parameter is the target value corresponding to the third position parameter, so that the loss term can be determined based on the first position parameter and the third position parameter, so as to train the third network model based on the loss term. In addition, the first network model is trained, and the first network model is trained based on the second network model in the detection network model, in other words, the first network model is obtained by training the detection network model.

[0141] In one embodiment, as shown in Figure 7 After the position parameter is adjusted to obtain the target position parameter, the target optoelectronic characteristic corresponding to the target position parameter can be obtained, the target optoelectronic characteristic corresponding to the target position parameter is obtained, and the circuit parameter composed of the target position parameter and the target optoelectronic characteristic is saved. In addition, after the target optoelectronic characteristic is obtained, the above adjustment process can be continued to be executed until the target optoelectronic characteristic reaches the preset condition, or the number of cycles reaches the preset requirement. Thus, a plurality of circuit parameters can be obtained, each of the plurality of circuit parameters including a position parameter and an optoelectronic characteristic, wherein the position parameter corresponds to a circuit layout, and the optoelectronic characteristic is the optoelectronic characteristic corresponding to the circuit layout.

[0142] In summary, the embodiment provides an integrated circuit adjustment method, which includes obtaining feature information corresponding to the integrated circuit, determining an adjustment parameter corresponding to the position parameter based on the feature information, determining a target position parameter corresponding to the integrated circuit based on the feature information, and determining an adjusted integrated circuit based on the target position parameter. Thus, the position parameter is automatically adjusted based on the optoelectronic characteristic to automatically adjust the circuit layout, so that the design efficiency of the circuit layout can be improved, and the production efficiency of the integrated circuit can be improved.

[0143] In one embodiment of the present embodiment, as shown in Figure 6 The integrated circuit adjustment method specifically includes:

[0144] acquire feature information corresponding to the integrated circuit, wherein the feature information comprises position parameters corresponding to a circuit layout of the integrated circuit and optoelectronic characteristics corresponding to the integrated circuit;

[0145] determine adjustment parameters corresponding to the position parameters based on the feature information;

[0146] adjust the position parameters based on the adjustment parameters to obtain target position parameters, and determine an adjusted integrated circuit based on the target position parameters.

[0147] Specifically, the acquisition process of the feature information can refer to the description of the above embodiments, which will not be described here. Here, the process of determining the target position parameters corresponding to the position parameters based on the optoelectronic characteristics and the position parameters, and determining the adjusted integrated circuit based on the target position parameters will be described.

[0148] The adjustment parameters are used to adjust the position parameters. Through the adjustment parameters, the position parameters can be adjusted to obtain target position parameters corresponding to the position parameters. The vector dimension of the adjustment parameters is the same as that of the position parameters. For each position parameter item in the position parameters, there is an adjustment parameter item in the adjustment parameters corresponding to the position parameter item. The adjustment parameter item is used to adjust the position parameter item corresponding to it. The corresponding between the adjustment parameter item and the position parameter item means that the position sequence number of the adjustment parameter item in the adjustment parameters is the same as that of the position parameter item in the position parameters. For example, the position parameters are 50-dimensional vectors, and the adjustment parameters are 50-dimensional vectors. The 25th adjustment parameter item in the adjustment parameters corresponds to the 25th position parameter item in the position parameters. In addition, the vector dimension of the target position parameters is the same as that of the position parameters. Each target position parameter item in the target position parameters corresponds to each position parameter item in the position parameters one by one. Each target position parameter item is calculated from the position parameter item corresponding to it and the adjustment parameter item corresponding to the position parameter item.

[0149] In one implementation manner of the embodiment, the determination of the adjustment parameters corresponding to the position parameters based on the optoelectronic characteristics and the position parameters specifically comprises:

[0150] acquiring error values corresponding to the optoelectronic characteristics;

[0151] determining the adjustment parameters corresponding to the position parameters based on the error values and the position parameters

[0152] Specifically, the error value is used to reflect an error of the optoelectronic characteristic and an optoelectronic characteristic limit condition corresponding to the integrated circuit, which is used to limit a minimum optoelectronic characteristic requirement that the integrated circuit reaches. The optoelectronic characteristic includes a plurality of optoelectronic parameters, and the optoelectronic characteristic limit condition includes a lower threshold value corresponding to part of the optoelectronic parameters. For example, the optoelectronic characteristic includes an aperture ratio, a charge rate and a peak voltage, and the optoelectronic characteristic limit condition can include a lower threshold value of the aperture ratio, a lower threshold value of the charge rate and a lower threshold value of the peak voltage; or the optoelectronic characteristic limit condition can include a lower threshold value of the charge rate and a lower threshold value of the peak voltage; or the optoelectronic characteristic limit condition can include a lower threshold value of the peak voltage. Of course, the optoelectronic characteristic limit condition can not include a lower threshold value of any optoelectronic parameter, and the optoelectronic characteristic limit condition is that the greater the optoelectronic parameter, the better the reference position parameter corresponding to the reference optoelectronic characteristic, and vice versa.

[0153] In one implementation of the embodiment, the obtaining of the error value corresponding to the optoelectronic characteristic specifically includes:

[0154] obtaining an optoelectronic characteristic limit condition corresponding to the integrated circuit, and determining a target function corresponding to the integrated circuit based on the optoelectronic characteristic limit condition;

[0155] determining the error value corresponding to the optoelectronic characteristic based on the target function.

[0156] Specifically, the target function can be pre-set, and the target function corresponds to the optoelectronic characteristic limit condition. After the optoelectronic characteristic limit condition corresponding to the integrated circuit is determined, a corresponding target function can be selected according to the optoelectronic characteristic limit condition. It can be understood that a target function set is pre-set, the target function set includes a plurality of target functions, and each target function in the plurality of target functions corresponds to an optoelectronic characteristic limit condition. After the optoelectronic characteristic limit condition corresponding to the integrated circuit is determined, a target function can be selected in the target function set based on the optoelectronic characteristic limit condition, and the target function is used as a target function for determining the error value corresponding to the optoelectronic characteristic.

[0157] The target functions in the target function set can be established according to actual application. Here, an application scenario is described. In the application scenario, the optoelectronic characteristic includes three optoelectronic parameters of an aperture ratio a, a charge rate b and a peak voltage c, and the correspondence between the target function and the optoelectronic characteristic limit condition can be:

[0158] When the photoelectric characteristic restriction condition is open ratio a>a0, charge ratio b>b0, peak voltage c>c0, wherein a0 is a lower threshold of open ratio, b0 is a lower threshold of charge ratio, and c0 is a lower threshold of peak voltage, the objective function f(a, b, c) can be:

[0159] f(a, b, c) = sigmoid(a-a0) * sigmoid(b-b0) * sigmoid(c-c0)

[0160] Wherein a represents open ratio, b represents charge ratio, c represents peak voltage, and sigmoid represents sigmoid function.

[0161] When the photoelectric characteristic restriction condition is charge ratio b>b0, peak voltage c>c0, wherein b0 is a lower threshold of charge ratio, and c0 is a lower threshold of peak voltage, the objective function f(a, b, c) can be:

[0162] f(a, b, c) = (a-a0) * sigmoid(b-b0) * sigmoid(c-c0)

[0163] Wherein a represents open ratio, b represents charge ratio, c represents peak voltage, and sigmoid represents sigmoid function.

[0164] When the photoelectric characteristic restriction condition is open ratio a, charge ratio b and peak voltage c are not set with lower threshold, the objective function f(a, b, c) can be:

[0165] f(a, b, c) = ka*(a-a0) + kb*(b-b0) + kc*(c-c0);

[0166] Wherein ka, kb, kc are weight coefficients, a0 is a lower threshold of open ratio, b0 is a lower threshold of charge ratio, and c0 is a lower threshold of peak voltage, wherein ka, kb, kc, a0, b0 and c0 can be set according to actual demand.

[0167] Further, after obtaining the target function, since the target function is a function with the optoelectronic characteristic as the independent variable, the target function can be converted into a function with the position parameter as the independent variable. In this way, since the optoelectronic characteristic is determined based on the position parameter, the position parameter and the optoelectronic characteristic have a candidate function, based on which each optoelectronic parameter in the optoelectronic characteristic can be represented by the position parameter, and then each optoelectronic characteristic represented by the position parameter is substituted into the target function to obtain the converted target function, and the converted target function is taken as the target function corresponding to the optoelectronic characteristic. After obtaining the target function with the position parameter as the independent variable, the position parameter is input into the target function, and the error value can be obtained. Of course, in actual application, the optoelectronic characteristic can also be directly input into the target function with the optoelectronic characteristic as the independent variable to obtain the error value.

[0168] In one implementation form of the embodiment, the determining, based on the error value and the position parameter, of the adjustment parameter corresponding to the position parameter specifically comprises:

[0169] For each position parameter term in the position parameter, a gradient value corresponding to the position parameter term is determined based on the error value and the target function, and an adjustment parameter term corresponding to the position parameter term is determined based on the gradient value.

[0170] All the adjustment parameter terms obtained constitute the adjustment parameter corresponding to the position parameter.

[0171] Specifically, the gradient value is used to determine the adjustment parameter term corresponding to the position parameter term. In one implementation form of the embodiment, the calculation formula of the gradient value can be: dY / dXi=f(X1, X2, X3, …Xi+△Xi, …Xn)-Y0, i=1, 2, 3…n, where dY / dXi represents the gradient value corresponding to the position parameter term Xi, Y=f(X1, X2, X3, …Xi, …Xn) is the target function with the position parameter term as the independent variable; Xi is the i-th position parameter term in the position parameter, △Xi is the change amount corresponding to the position parameter term Xi, n represents the number of position parameter terms, and Y0 represents the error value. The change amounts corresponding to the position parameter terms can be the same, or the change amounts corresponding to part of the position parameter terms are the same and the change amounts corresponding to part of the position parameter terms are different; or the change amounts corresponding to all the position parameter terms are different, wherein the change amount can be pre-set according to actual application.

[0172] The calculation formula of the adjustment parameter term can be:

[0173] Xi=X1+gamma_i*(dY / dXi)*△Xi, i=1, 2, 3…n

[0174] Wherein, △Xi is the change of the position parameter item Xi, and gamma_i is the weight coefficient corresponding to the position parameter item Xi.

[0175] In one implementation, the adjusting the position parameter based on the adjustment parameter to obtain a target position parameter specifically includes:

[0176] For each position parameter item in the position parameter, determining an adjustment parameter item corresponding to the position parameter item in the adjustment parameter;

[0177] Determining the element sum of the position parameter item and the adjustment parameter item, and taking the element sum as a target position parameter item corresponding to the position parameter item;

[0178] Taking the position parameter composed of all target position parameter items as a target position parameter.

[0179] Specifically, the vector dimension of the adjustment parameter is the same as that of the position parameter, and the adjustment parameter item in the adjustment parameter corresponds to the position parameter item in the position parameter one by one, so that for each position parameter item, its corresponding adjustment parameter item can be determined, and after the adjustment parameter item is determined, the element sum of the position parameter item and the adjustment parameter item is calculated to obtain an adjusted position parameter item, i.e. a target position parameter item corresponding to the position parameter item. Based on this, after obtaining the target position parameter items corresponding to all position parameter items, the position parameter composed of all target position parameter items can be taken as a target position parameter.

[0180] In one embodiment, after obtaining a plurality of circuit parameters, the circuit parameters can be saved to form a circuit parameter set. After obtaining the desired optoelectronic characteristic, the target circuit parameter can be selected based on the desired optoelectronic characteristic in the circuit parameter set, and the position parameter corresponding to the target circuit parameter in the circuit layout is taken as the circuit layout corresponding to the desired optoelectronic characteristic, so as to improve the acquisition speed of the circuit layout.

[0181] Based on this, after obtaining the circuit parameter set, as shown in Figure 9 and Figure 10 The method can further include:

[0182] Based on the preset desired optoelectronic characteristic, selecting a target circuit parameter corresponding to the desired optoelectronic characteristic from the pre-stored plurality of circuit parameters;

[0183] Based on the target position parameter in the target circuit parameter, determining the integrated circuit corresponding to the desired optoelectronic characteristic.

[0184] Specifically, each of the plurality of circuit parameters comprises a position parameter and an optoelectronic characteristic, wherein the position parameter corresponds to a circuit layout, and the optoelectronic characteristic is an optoelectronic characteristic corresponding to the circuit layout. It can be understood that the position parameter and the optoelectronic characteristic correspond to the same circuit layout, the position parameter is parameterized from the circuit layout, and the optoelectronic characteristic is determined based on the position parameter or determined by a circuit simulator.

[0185] The target circuit parameter is included in the plurality of circuit parameters, and the target optoelectronic characteristic in the plurality of circuit parameters has the highest matching degree with the expected optoelectronic characteristic, wherein the matching degree is used to reflect the similarity between the target optoelectronic characteristic and the expected optoelectronic characteristic, and the higher the matching degree, the higher the similarity between the target optoelectronic characteristic and the expected optoelectronic characteristic, and vice versa, the lower the matching degree, the lower the similarity between the target optoelectronic characteristic and the expected optoelectronic characteristic.

[0186] The matching degree can be the Euclidean distance between the optoelectronic characteristic and the expected optoelectronic characteristic, or each optoelectronic parameter in the optoelectronic characteristic can be configured with a weight coefficient. When determining the matching degree, each optoelectronic parameter can be updated based on the weight coefficient of each optoelectronic parameter (for example, the product of the optoelectronic parameter and the weight coefficient is used as the optoelectronic parameter), and based on the Euclidean distance between the updated optoelectronic characteristic and the expected optoelectronic characteristic, the matching degree between the optoelectronic characteristic and the expected optoelectronic characteristic can be obtained.

[0187] In one implementation of the embodiment, the target circuit parameter corresponding to the expected optoelectronic characteristic is selected from the plurality of pre-stored circuit parameters based on the pre-set expected optoelectronic characteristic, and specifically includes:

[0188] The weight coefficient set corresponding to the expected optoelectronic characteristic is obtained, wherein the weight coefficient set comprises a weight coefficient corresponding to each optoelectronic parameter in the expected optoelectronic characteristic;

[0189] The target circuit parameter is selected from the plurality of circuit parameters based on the expected optoelectronic characteristic and the weight coefficient set.

[0190] Specifically, the weight coefficient set comprises a plurality of weight coefficients, and the plurality of weight coefficients correspond one-to-one to a plurality of optoelectronic parameters included in the optoelectronic characteristic. Each weight coefficient is used to reflect the importance of the corresponding optoelectronic parameter. The higher the weight coefficient, the higher the importance of the optoelectronic parameter, and vice versa. The lower the weight coefficient, the lower the importance of the optoelectronic parameter. The weight coefficient set can be pre-set, and the weight coefficient set corresponding to different expected optoelectronic characteristics is different.

[0191] In one implementation form of the embodiment, after the weight coefficient set is obtained, the matching degrees of the photoelectric characteristics in each of the circuit parameters to the expected photoelectric characteristics are calculated respectively when the target photoelectric characteristic corresponding to the expected photoelectric characteristic is determined based on the weight coefficient set, and the circuit parameter corresponding to the photoelectric characteristic with the highest matching degree is selected as the target circuit parameter.

[0192] In one implementation form of the embodiment, the circuit parameters are stored in the form of a KD tree, and when the target circuit parameter is determined based on the weight coefficient set after the weight coefficient set is obtained, KD tree search can be used to determine the target circuit parameter, which can improve the speed of obtaining the target circuit parameter. Correspondingly, the selecting the target circuit parameter from the circuit parameters based on the expected photoelectric characteristic and the weight coefficient set specifically includes:

[0193] KD tree search is performed on the circuit parameters based on the expected photoelectric characteristic to obtain the target circuit parameter corresponding to the expected photoelectric characteristic, wherein the search range of the node photoelectric characteristic in the backtracking process of the KD tree search is determined based on the node photoelectric characteristic, the expected photoelectric characteristic and the weight coefficient set.

[0194] Specifically, after the expected photoelectric characteristic is obtained, KD tree search is used to search the circuit parameters to obtain the candidate circuit parameter. In the KD tree search process, the value of the split dimension of the reference photoelectric characteristic in the circuit parameter of the split node can be compared with the expected photoelectric characteristic, and if the expected photoelectric characteristic is less than the reference photoelectric characteristic, the left subtree branch is entered; if the expected photoelectric characteristic is equal to or greater than the reference photoelectric characteristic, the right subtree branch is entered, and the process is repeated until the leaf node is reached, and the circuit parameter corresponding to the leaf node is taken as the candidate circuit parameter. After the candidate circuit parameter is determined, the leaf node in the same subspace as the candidate circuit parameter is searched, wherein in the search of the leaf node in the same subspace as the candidate circuit parameter, the candidate circuit parameter is taken as the query circuit parameter, the photoelectric characteristic in the candidate circuit parameter is taken as the node photoelectric characteristic, and the search radius is determined based on the node photoelectric characteristic, the expected photoelectric characteristic and the weight coefficient set. For example, the process of determining the search radius can be as follows: first, the photoelectric parameters in the expected photoelectric characteristic are multiplied by the corresponding weight coefficients to update the photoelectric parameters, then the Euclidean distance between the expected photoelectric characteristic and the node photoelectric characteristic is calculated, and finally the calculated Euclidean distance is taken as the search radius.

[0195] Further, after searching the leaf nodes in the same subspace as the to-be-queried point, the search path is traced back, and it is determined whether there is a data point with a distance from the expected optoelectronic feature less than the search radius in other child node spaces of the nodes on the search path. If there is, the other child node spaces need to be jumped to for searching, and the other child nodes are added to the search path. The determination of whether there is a data point with a distance from the expected optoelectronic feature less than the search radius in other child node spaces of the nodes on the search path is continued until the search path is empty, so as to obtain the target circuit parameter corresponding to the expected optoelectronic characteristic.

[0196] Based on the above integrated circuit adjustment method, the embodiment provides an integrated circuit adjustment device, as shown in the figure, which comprises: Figure 11

[0197] A first acquisition module 100 is configured to acquire feature information corresponding to an integrated circuit, wherein the feature information comprises position parameters corresponding to a circuit layout of the integrated circuit and optoelectronic characteristic information corresponding to the integrated circuit.

[0198] A first determination module 200 is configured to determine a target position parameter corresponding to the integrated circuit based on the feature information, and determine an adjusted integrated circuit based on the target position parameter.

[0199] In an implementation manner, the second determination module specifically comprises:

[0200] A first acquisition unit is configured to acquire a reference position parameter corresponding to the position parameter, wherein the dimension of the reference position parameter is lower than the dimension of the position parameter.

[0201] A first determination unit is configured to determine an optimized position parameter corresponding to the reference position parameter based on the optoelectronic characteristic and the reference position parameter, wherein the dimension of the optimized position parameter is equal to the dimension of the reference position parameter.

[0202] The target position parameter corresponding to the position parameter is determined based on the optimized position parameter.

[0203] In an implementation manner, the first determination unit specifically comprises:

[0204] A first acquisition subunit is configured to acquire a target function corresponding to the integrated circuit, and determine a target value corresponding to the optoelectronic feature based on the target function.

[0205] An optimization subunit is configured to optimize the reference position parameter by using a Bayesian optimizer based on the target value, so as to obtain the optimized position parameter.

[0206] ​In an implementation manner, the optimization subunit is specifically configured to: input the optimized position parameter into a trained third network model, and output a target position parameter corresponding to the position parameter through the third network model.

[0207] In an implementation manner, the detection network model comprises a first network model and a second network model in cascade; the first network model has a first high-dimensional position parameter as an input item and a first low-dimensional position parameter as an output item; the third network model has a second low-dimensional position parameter as an input item and a second high-dimensional position parameter as an output item, and the first high-dimensional position parameter has the same dimension as the second high-dimensional position parameter, and the first low-dimensional position parameter has the same dimension as the second low-dimensional position parameter.

[0208] In an implementation manner, the training process of the third network model specifically comprises:

[0209] inputting the first position parameter corresponding to each integrated circuit in the training sample into a trained first network model, and outputting a second position parameter through the first network model, wherein the first network model and the second network model are jointly trained;

[0210] inputting the second position parameter into a preset network model, and outputting a third position parameter through the preset network model;

[0211] training the preset network model based on the first position parameter and the third position parameter to obtain a third network model.

[0212] In an implementation manner, the adjustment device of the integrated circuit comprises:

[0213] The second acquisition module is configured to acquire a target photoelectric characteristic corresponding to the target position parameter, and save a circuit parameter formed by the target position parameter and the target photoelectric characteristic.

[0214] In an implementation manner, the adjustment device of the integrated circuit comprises:

[0215] The second determination module is configured to select a target circuit parameter corresponding to an expected photoelectric characteristic from a plurality of pre-stored circuit parameters based on the expected photoelectric characteristic;

[0216] The third determination module is configured to determine an integrated circuit corresponding to the expected photoelectric characteristic based on a target position parameter in the target circuit parameter.

[0217] In an implementation manner, the second determination module specifically comprises:

[0218] The second acquisition unit is used to acquire the set of weight coefficients corresponding to the desired photoelectric characteristics, wherein the set of weight coefficients includes the weight coefficients corresponding to each photoelectric parameter in the desired photoelectric characteristics.

[0219] The selection unit is used to select target circuit parameters from the plurality of circuit parameters based on the desired photoelectric characteristics and the set of weighting coefficients.

[0220] In one implementation, the pre-stored circuit parameters are stored in the form of a KD tree.

[0221] In one implementation, the selection unit is specifically used to: perform a KD-tree search on several circuit parameters based on the desired photoelectric characteristics to obtain the target circuit parameters corresponding to the desired photoelectric characteristics, wherein the search range of the node photoelectric characteristics in the backtracking process of the KD-tree search is determined based on the node photoelectric characteristics, the desired photoelectric characteristics, and the set of weight coefficients.

[0222] In one implementation, the integrated circuit is a TFT circuit.

[0223] Based on the above-described integrated circuit adjustment method, this embodiment provides a computer-readable storage medium storing one or more programs that can be executed by one or more processors to implement the steps in the integrated circuit adjustment method described in the above embodiment.

[0224] Based on the aforementioned integrated circuit adjustment method, this application also provides a terminal device, such as... Figure 12 As shown, it includes at least one processor 20; a display screen 21; and a memory 22, and may also include a communications interface 23 and a bus 24. The processor 20, display screen 21, memory 22, and communications interface 23 can communicate with each other via the bus 24. The display screen 21 is configured to display a preset user guide interface in the initial setup mode. The communications interface 23 can transmit information. The processor 20 can invoke logical instructions in the memory 22 to execute the methods described in the above embodiments.

[0225] Furthermore, the logical instructions in the aforementioned memory 22 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium.

[0226] The memory 22, as a computer readable storage medium, can be configured to store software programs, computer executable programs, such as program instructions or modules corresponding to the method in the embodiments of the present disclosure. The processor 20 executes the functions of the application and data processing by running the software programs, instructions or modules stored in the memory 22, that is, implements the method in the above embodiments.

[0227] The memory 22 can include a program storage area and a data storage area, wherein the program storage area can store an operating system and at least one application required by a function; the data storage area can store data created according to the use of the terminal device, etc. In addition, the memory 22 can include a high-speed random access memory, and can also include a non-volatile memory. For example, a variety of media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc., can also be a transitory storage medium.

[0228] In addition, the specific process of the memory medium and the plurality of instructions in the terminal device loaded and executed by the processor has been described in detail in the above method, and will not be repeated here.

[0229] Finally, it should be pointed out that: the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for adjusting an integrated circuit, characterized in that, The method includes: Obtain the feature information corresponding to the integrated circuit, wherein the feature information includes the position parameters corresponding to the circuit layout of the integrated circuit and the photoelectric characteristic information corresponding to the integrated circuit; Based on the feature information, the target position parameters corresponding to the integrated circuit are determined, and based on the target position parameters, the adjusted integrated circuit is determined; The acquisition of the feature information corresponding to the integrated circuit includes: Obtain the circuit layout and determine the location parameters corresponding to the integrated circuit based on the circuit layout; Based on the trained detection network model and the position parameters, the photoelectric characteristics of the integrated circuit are determined to obtain the feature information of the integrated circuit. The specific steps of determining the target location parameters corresponding to the integrated circuit based on the feature information include: Obtain the reference position parameter corresponding to the position parameter, wherein the dimension of the reference position parameter is lower than the dimension of the position parameter; Based on the photoelectric properties and the reference position parameter, an optimized position parameter corresponding to the reference position parameter is determined, wherein the dimension of the optimized position parameter is equal to the dimension of the reference position parameter; Based on the optimized position parameters, the target position parameters corresponding to the position parameters are determined.

2. The method for adjusting an integrated circuit according to claim 1, characterized in that, The step of determining the optimized position parameter corresponding to the reference position parameter based on the photoelectric properties and the reference position parameter specifically includes: Obtain the target function corresponding to the integrated circuit, and determine the target value corresponding to the photoelectric feature based on the target function; Based on the target value, a Bayesian optimizer is used to optimize the reference position parameter to obtain the optimized position parameter.

3. The method for adjusting an integrated circuit according to claim 1, characterized in that, The step of determining the target position parameter corresponding to the optimized position parameter specifically includes: The optimized position parameters are input into a trained third network model, and the target position parameters corresponding to the optimized position parameters are output by the third network model.

4. The method for adjusting an integrated circuit according to claim 3, characterized in that, The detection network model includes a cascaded first network model and a second network model; the input of the first network model is a first high-dimensional position parameter, and the output is a first low-dimensional position parameter; the input of the third network model is a second low-dimensional position parameter, and the output is a second high-dimensional position parameter, wherein the dimension of the first high-dimensional position parameter is equal to the dimension of the second high-dimensional position parameter, and the dimension of the first low-dimensional position parameter is equal to the dimension of the second low-dimensional position parameter.

5. The method for adjusting an integrated circuit according to claim 4, characterized in that, The training process of the third network model specifically includes: The first position parameter corresponding to each integrated circuit in the training samples is input into the trained first network model, and the second position parameter is output through the first network model. The first network model and the second network model are jointly trained. The second position parameter is input into a preset network model, and the third position parameter is output through the preset network model. Based on the first position parameter and the third position parameter, the preset network model is trained to obtain the third network model.

6. The method for adjusting an integrated circuit according to any one of claims 1-5, characterized in that, After determining the target location parameters corresponding to the integrated circuit based on the feature information, and determining the adjusted integrated circuit based on the target location parameters, the method further includes: Obtain the target photoelectric characteristics corresponding to the target position parameters, and save the circuit parameters formed by the target position parameters and the target photoelectric characteristics.

7. The method for adjusting an integrated circuit according to claim 6, characterized in that, After acquiring the target photoelectric characteristics corresponding to the target position parameters and saving the circuit parameters formed by the target position parameters and the target photoelectric characteristics, the method further includes: Based on the preset desired photoelectric characteristics, the target circuit parameters corresponding to the desired photoelectric characteristics are selected from a number of pre-stored circuit parameters. Based on the target position parameter in the target circuit parameters, determine the integrated circuit corresponding to the desired photoelectric characteristics.

8. The method for adjusting an integrated circuit according to claim 7, characterized in that, Based on the preset desired photoelectric characteristics, the selection of target circuit parameters corresponding to the desired photoelectric characteristics from a number of pre-stored circuit parameters specifically includes: Obtain the set of weight coefficients corresponding to the desired photoelectric characteristics, wherein the set of weight coefficients includes the weight coefficients corresponding to each photoelectric parameter in the desired photoelectric characteristics; Based on the desired photoelectric properties and the set of weighting coefficients, target circuit parameters are selected from the plurality of circuit parameters.

9. The method for adjusting an integrated circuit according to claim 8, characterized in that, The pre-stored circuit parameters are stored in the form of a KD tree.

10. The method for adjusting an integrated circuit according to claim 9, characterized in that, The step of selecting target circuit parameters from the plurality of circuit parameters based on the desired photoelectric characteristics and the weighting coefficient set specifically includes: KD-tree search is performed on several circuit parameters based on the desired photoelectric characteristics to obtain the target circuit parameters corresponding to the desired photoelectric characteristics. The search range of the node photoelectric characteristics during the backtracking process of the KD-tree search is determined based on the node photoelectric characteristics, the desired photoelectric characteristics, and the set of weight coefficients.

11. The method for adjusting an integrated circuit according to any one of claims 1-5, characterized in that, The integrated circuit is a TFT circuit.

12. An adjustment apparatus for an integrated circuit, used to implement the adjustment method for an integrated circuit as described in any one of claims 1-11, characterized in that, The adjustment device includes: The first acquisition module is used to acquire feature information corresponding to the integrated circuit, wherein the feature information includes position parameters corresponding to the circuit layout of the integrated circuit and photoelectric characteristic information corresponding to the integrated circuit. The first determining module is used to determine the target position parameters corresponding to the integrated circuit based on the feature information, and to determine the adjusted integrated circuit based on the target position parameters.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores one or more programs, which can be executed by one or more processors to implement the steps in the adjustment method of the integrated circuit as described in any one of claims 1-11.

14. A terminal device, characterized in that, include: Processor, memory, and communication bus; the memory stores a computer-readable program that can be executed by the processor; The communication bus enables communication between the processor and the memory; When the processor executes the computer-readable program, it implements the steps in the adjustment method of the integrated circuit as described in any one of claims 1-11.

Images