Intelligent processing monitoring system of integrated circuit chip

Abstract

The invention relates to the technical field of integrated circuit chip processing monitoring, in particular to an intelligent processing monitoring system of an integrated circuit chip. The method comprises the following steps: acquiring image data of an integrated circuit chip in a target area and processing data acquired by a sensor on a processing production line, calculating a defect characteristic index of each sub-image, segmenting an integrated circuit chip defect part in the sub-image based on the defect characteristic index, extracting the characteristic of the integrated circuit chip defect part, and determining the defect part of the integrated circuit chip according to the characteristic of the integrated circuit chip defect part. Identifying and analyzing the features of the defect part, generating an appearance monitoring analysis result, analyzing the processing data, generating a non-appearance monitoring analysis result, and marking the integrated circuit chip in the target area based on the appearance monitoring analysis result and the non-appearance monitoring analysis result. The surface defect of the integrated circuit chip can be quickly identified, the abnormal change of the processing technology parameters can be found in time, and the processing monitoring accuracy and the processing monitoring efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of integrated circuit chip processing monitoring, and particularly relates to an intelligent processing monitoring system for integrated circuit chips. Background Art

[0002] The chip processing process involves hundreds of complex processes, including multiple links such as wafer preparation, lithography, etching, thin film deposition, doping, packaging and testing. Each link has extremely strict requirements on process parameters. Any minor parameter deviation may lead to a decline in chip performance, a reduction in yield, or even the scrapping of an entire batch of chips, resulting in huge economic losses.

[0003] Currently, traditional chip processing monitoring mainly relies on manual operation. Operators need to regularly inspect the production line and record the operating parameters of various equipment, such as temperature, pressure, speed, etc. This method not only has low efficiency but is also easily affected by human factors, such as operator fatigue and negligence, resulting in inaccurate and untimely data recording and an inability to promptly detect abnormal changes in process parameters. Further, there is currently no means in the prior art to monitor the chip processing quality using the image data of integrated circuit chips, resulting in low monitoring efficiency.

[0004] Therefore, there is an urgent need for an intelligent processing monitoring system for integrated circuit chips to solve the above problems. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent processing monitoring system for integrated circuit chips, aiming to solve the technical problems that traditional chip processing monitoring mainly relies on manual operation, resulting in inaccurate and untimely monitoring, and an inability to promptly detect abnormal changes in process parameters and chip surface defects.

[0006] The purpose of the present invention can be achieved through the following technical solutions: An intelligent processing monitoring system for integrated circuit chips, the system includes: A monitoring data acquisition unit, configured to acquire the image data of the integrated circuit chips in the target area and the processing data collected by sensors on the processing production line, and send the image data of the integrated circuit chips to the appearance analysis unit and the processing data to the non-appearance analysis unit; An appearance analysis unit, configured to receive the image data of the integrated circuit chips, divide the image of the integrated circuit chips into M sub-images, calculate the defect feature index of each sub-image, segment the defect parts of the integrated circuit chips in the sub-images based on the defect feature index, extract the defect part features of the integrated circuit chips, perform identification and analysis on the defect part features, generate an appearance monitoring analysis result, and send the appearance monitoring analysis result to the monitoring and early warning unit; The non - appearance analysis unit is used to receive the processing data, analyze the processing data, generate the non - appearance monitoring and analysis results, and send the non - appearance monitoring and analysis results to the monitoring and warning unit; The monitoring and warning unit is used to receive the appearance monitoring and analysis results and the non - appearance monitoring and analysis results to mark the integrated circuit chips in the target area.

[0007] Further, dividing the integrated circuit chip image into M sub - images specifically includes the following process: Step 1, initialize the population: Set the initial population number m, the fitness threshold , the maximum number of iterations , the search range , randomly generate an n - dimensional vector with each component between 0 and 1 , with as the initial value, based on the regression equation iterate to obtain N vectors , where, ; Step 2, divide the initial population into NP group and C group: Calculate the fitness values of each particle in the initial population and sort them. Select the first m particles as the NP - group particles, and the remaining m particles as the C - group particles; and record the individual extreme value of the NP group and the global extreme value of the NP group, record the global extreme value of the C group; Step 3, determine whether the maximum number of iterations is reached. If the maximum number of iterations is reached, execute Step 5. Otherwise, execute Steps 3 - 1 to 3 - 4; Step 4, cross the NP group and the C group: Select the G particles with the best fitness values from the C group to replace the G particles with the worst fitness values in the NP group, and return to Step 3 to continue the iteration until the number of iterations is reached; Step 5, the iteration ends, output the target optimization parameter M, and divide the integrated circuit chip image into M sub - images based on the target optimization parameter M.

[0008] Further, executing Steps 3 - 1 to 3 - 4 specifically includes the following process: Step 3 - 1, update the particles: Update the velocity and position of the NP - group particles with the following formula ; ; where, the learning factors and are both 0.01, is the number of iterations; is the inertia weight, is the The historical best position experienced by a particle is the best position experienced by all particles in the population and is a random number uniformly distributed between 0 and 1 is the particle velocity at the th iteration is the particle position at the th iteration is the particle position at the th iteration is the particle velocity at the th iteration; ; where is the particle position at the th iteration is the particle position at the th iteration has a value of 3.57; Step Three Two, update the individual extreme value of the NP group and the global extreme value of the NP group , evaluate the fitness value of the particles in the NP group after update. If the individual fitness value of a particle is better than , then replace with the current particle position; compare and . If is greater than , then replace with the current ; Step Three Three, update the global extreme value of the C group : Evaluate the fitness value of the particles in the C group. If the fitness value is better than , then replace with the current particle position; Step Three Four, calculate the population fitness variance : ; ; where is the total number of particles in the NP group is the individual fitness value of the i-th particle is the overall average fitness value of the NP group, and max represents the maximum value operation; When is less than the preset threshold, it is determined that premature convergence has occurred, and Step Four is executed. When is greater than or equal to the preset threshold, then return to Step Three.

[0009] Further, calculating the defect feature index of each sub-image specifically includes the following process: Calculate the contrast significant value between the sub-image and the preset defect-free sub-image : ; Wherein, is the number of superpixel regions in the sub-image, is the number of pixels in each superpixel region, is the superpixel region and the superpixel region of the preset defect-free sub-image the Euclidean distance between them, is the superpixel region and the superpixel region of the preset defect-free sub-image the centroid distance, is the median of the distances between each pair of superpixel centroids; Calculate the color distribution feature value of the sub-image : ; Wherein, take the maximum value of and 0, and are the R, B, and G values of the jth pixel point in the superpixel region in the RGB color space respectively; Calculate the defect feature index of each sub-image : ; Wherein, , , are weight coefficients respectively, is the preset contrast significant threshold, is the preset color distribution feature threshold, is the defect coefficient of the sub-image in the hyperspectral space.

[0010] Further, obtaining the defect coefficient of the sub-image in the hyperspectral space specifically includes the following process: Obtain the characteristic wavelength reflectivity map of the sub-image in the hyperspectral space, add the preset band to the characteristic wavelength reflectivity map, and extract three characteristic vectors of the characteristic wavelength reflectivity map in the preset band , , , and the wavelengths corresponding to the vectors are respectively , , , calculate the defect coefficient of the sub-image in the hyperspectral space ; Among them, the ratios of the reflectance differences of the three eigenvectors in the axis projection direction are equal.

[0011] Furthermore, based on the defect feature index, the defective parts of the integrated circuit chips in the sub-images are segmented, and the extraction of the features of the defective parts of the integrated circuit chips specifically includes the following process: The sub-image is segmented into several parts, and it is judged whether each part's exceeds a preset threshold. If so, this part is taken as the defective part of the integrated circuit chip, and feature extraction is performed on the defective part of the integrated circuit chip to obtain the features of the defective part of the integrated circuit chip : ; Among them, represents the number of channels in the image, represents the th channel color value at the th pixel point, represents the number of pixel points, represents the color mean value of the

[0012] Furthermore, the defective part features are identified and analyzed to generate the appearance monitoring analysis result, which specifically includes the following process: The defective part features are compared and learned with the defective part data set to obtain the learned representation of the node. It is judged whether the learned representation of the node has a representation of component missing. If so, it is judged that the integrated circuit chip has a component missing; It is judged whether the learned representation of the node has a representation of component position soldering error. If so, it is judged that the integrated circuit chip has a component position soldering error; It is judged whether the learned representation of the node has a representation of component redundancy. If so, it is judged that the integrated circuit chip has a component redundancy; The judgment result is recorded as the appearance monitoring analysis result.

[0013] Furthermore, the processing data is analyzed to generate the non-appearance monitoring analysis result, which specifically includes the following process: Based on the processing data, the equipment operation data of the integrated circuit chip processing line is obtained. Among them, the equipment includes a wafer cleaning machine, a lithography machine, an etching machine, a chemical mechanical polishing machine, an electroplating machine, and a wafer bonding machine. The equipment operation data is compared and analyzed with the equipment operation standard data in the processing indicators to judge whether the equipment operation data is the same as the equipment operation standard data in the processing indicators. If so, the processing line equipment parameter coefficient JG is 1. If not, the processing line equipment parameter coefficient JG is 0; Obtain the temperature characterization value WD during the processing period based on the processing data; Obtain the wafer cutting speed, lithography speed, etching speed, and metal filling speed during the processing period based on the processing data. Calculate the difference between the wafer cutting speed and the preset wafer cutting speed threshold, the difference between the lithography speed and the preset lithography speed threshold, the difference between the etching speed and the preset etching speed threshold, and the difference between the metal filling speed and the preset metal filling speed threshold. After taking the absolute value of all differences, sum them up to obtain the sum value SD; Obtain the internal pressure of the processing equipment, the gas pipeline pressure for etching or cleaning, and the liquid pressure in the liquid processing process during the processing period. Among them, the liquid processing process includes the coating of photoresist and the spraying of developer. Determine whether all pressures are within the preset range. If so, the pressure parameter YL is 1; if not, the pressure parameter YL is 0; Substitute the processing line equipment parameter coefficient JG, temperature characterization value WD, SD, and pressure parameter YL into the processing qualification index calculation formula to calculate the processing qualification index LMS. The calculation formula is as follows: ; Among them, The value of is 2.72, , Are weight coefficients respectively, and their values are 0.6 and 0.4 respectively; Determine whether the processing qualification index exceeds the preset processing qualification index threshold. If so, determine that the processing of the integrated circuit chip is qualified; if not, determine that the processing of the integrated circuit chip is unqualified. Record the judgment result as the generated non-appearance monitoring analysis result.

[0014] Furthermore, obtaining the temperature characterization value WD during the processing period includes: dividing the processing period into several sub-periods, detecting the temperature during the wafer manufacturing process, the temperature during the lithography and etching processes, and the temperature during the ion implantation and metal filling processes in each sub-period. Establish a rectangular coordinate system with the execution time of each sub-period as the X-axis and the temperature value as the Y-axis. Connect the temperatures in each manufacturing process to generate the temperature curve of the wafer manufacturing process, the temperature curve of the lithography and etching processes, and the temperature curve of the ion implantation and metal filling processes. Calculate the total length of the curves located above the preset temperature curve for all temperature curves, and record the total curve length as the temperature characterization value WD during the processing period.

[0015] Compared with the existing solutions, the beneficial effects achieved by the present invention: The present invention obtains the integrated circuit chip image data within the target area and the processing data collected by sensors on the processing production line, divides the integrated circuit chip image into M sub-images, calculates the defect feature index of each sub-image, segments the defect parts of the integrated circuit chip in the sub-image based on the defect feature index, extracts the defect part features of the integrated circuit chip, identifies and analyzes the defect part features to generate an appearance monitoring analysis result, analyzes the processing data to generate a non-appearance monitoring analysis result, and marks the integrated circuit chips within the target area based on the appearance monitoring analysis result and the non-appearance monitoring analysis result, which can quickly identify the surface defects of the integrated circuit chips and timely detect abnormal changes in the processing process parameters, improving the accuracy and efficiency of processing monitoring.

[0016] Furthermore, the intelligent processing monitoring system for the integrated circuit chip realizes the full-range monitoring and quality control of the chip manufacturing process through functions such as comprehensive and accurate data acquisition and integration, refined appearance defect analysis and identification, in-depth non-appearance parameter analysis and evaluation, comprehensive early warning and accurate marking, etc., effectively improving the production efficiency, reducing the cost, and enhancing the quality and reliability of the chips. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0018] Figure 1 It is a system block diagram of an intelligent processing monitoring system for an integrated circuit chip according to an embodiment of the present invention; Figure 2 It is a working flowchart of a first intelligent processing monitoring system for an integrated circuit chip according to an embodiment of the present invention; Figure 3 It is a working flowchart of a second intelligent processing monitoring system for an integrated circuit chip according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0020] In addition, the described features, structures, or characteristics may be combined in any suitable manner in one or more example embodiments. In the following description, numerous specific details are provided to give a thorough understanding of example embodiments of the present disclosure. However, those skilled in the art will recognize that the technical solutions of the present disclosure may be practiced without one or more of the specific details, or other methods, components, steps, etc. may be employed. In other cases, well-known structures, methods, implementations, or operations are not shown or described in detail to avoid obscuring aspects of the present disclosure.

[0021] This embodiment provides an intelligent processing monitoring system for integrated circuit chips. Figure 1 It is a system block diagram of an intelligent processing monitoring system for integrated circuit chips according to an embodiment of the present invention, as Figure 1 shown. The system includes: A monitoring data acquisition unit, configured to acquire integrated circuit chip image data within a target area and processing data collected by sensors on a processing production line, and send the integrated circuit chip image data to an appearance analysis unit and the processing data to a non-appearance analysis unit; An appearance analysis unit, configured to receive the integrated circuit chip image data, divide the integrated circuit chip image into M sub-images, calculate the defect feature index of each sub-image, segment the defect parts of the integrated circuit chip in the sub-images based on the defect feature index, extract the defect part features of the integrated circuit chip, perform identification and analysis on the defect part features, generate an appearance monitoring analysis result, and send the appearance monitoring analysis result to a monitoring and warning unit; A non-appearance analysis unit, configured to receive the processing data, analyze the processing data, generate a non-appearance monitoring analysis result, and send the non-appearance monitoring analysis result to the monitoring and warning unit; A monitoring and warning unit, configured to receive the appearance monitoring analysis result and the non-appearance monitoring analysis result to mark the integrated circuit chips within the target area.

[0022] It should be noted that the monitoring and warning unit marks the unqualified integrated circuit chips and issues a warning signal.

[0023] In summary, the present invention obtains the image data of the integrated circuit chip in the target area and the processing data collected by the sensors on the processing production line, divides the integrated circuit chip image into M sub-images, calculates the defect feature index of each sub-image, segments the defect parts of the integrated circuit chip in the sub-image based on the defect feature index, extracts the features of the defect parts of the integrated circuit chip, identifies and analyzes the features of the defect parts to generate an appearance monitoring analysis result, analyzes the processing data to generate a non-appearance monitoring analysis result, and marks the integrated circuit chips in the target area based on the appearance monitoring analysis result and the non-appearance monitoring analysis result, which can quickly identify the surface defects of the integrated circuit chips and timely detect the abnormal changes of the processing process parameters, improving the accuracy and efficiency of processing monitoring.

[0024] In some embodiments, dividing the integrated circuit chip image into M sub-images specifically includes the following process: Step 1, initialize the population: Set the initial population number m, fitness threshold , maximum number of iterations , search range , randomly generate an n-dimensional vector with each component between 0 and 1 , with as the initial value, and based on the regression equation iterate to obtain N vectors , where ; Step 2, divide the initial population into NP group and C group: Calculate the fitness value of each particle in the initial population and sort them, select the first m particles as the NP group particles, and the remaining m particles as the C group particles; and record the individual extreme value of the NP group and the global extreme value of the NP group, and record the global extreme value of the C group; Step 3, determine whether the maximum number of iterations is reached. If the maximum number of iterations is reached, execute Step 5; otherwise, execute Steps 3-1 to 3-4; Step 3-1, update the particles: Update the velocity and position of the NP group particles with the following formula ; ; where the learning factors and are both 0.01, is the number of iterations; is the inertia weight, is the historical best position experienced by the -th particle, is the best position experienced by all particles in the group, and is a random number uniformly distributed between 0 and 1, is the particle velocity at the -th iteration, is the particle position at the -th iteration, is the particle position at the -th iteration, is the particle velocity at the -th iteration; Update the position of the C group with the following formula: ; where, is the particle position at the -th iteration, is the particle position at the -th iteration, has a value of 3.57; Step Three Two, update the individual extreme value of the NP group and the global extreme value of the NP group , evaluate the fitness value of the particles in the NP group after update. If the individual fitness value of the particle is better than , then replace with the current particle position; compare and . If is greater than , then replace with the current ; Step Three Three, update the global extreme value of the C group : Evaluate the fitness value of the particles in the C group. If the fitness value is better than , then replace with the current particle position; Step Three Four, calculate the population fitness variance : ; ; where, is the total number of particles in the NP group, is the individual fitness value of the i-th particle, is the overall average fitness value of the NP group, and max represents the operation of taking the maximum value; When is less than the preset threshold, it is determined that the premature state is entered, and Step Four is executed. When is greater than or equal to the preset threshold, then return to Step Three.

[0025] Step 4: Cross the NP group and the C group: Select the G particles with the best fitness values from the C group to replace the G particles with the worst fitness values in the NP group, and return to Step 3 to continue the iteration until the number of iterations is reached. ; Step 5: End the iteration, output the target optimization parameter M, and divide the integrated circuit chip image into M sub-images based on the target optimization parameter M.

[0026] It should be noted that dividing the integrated circuit chip image into M sub-images according to the target optimization parameter M can achieve refined processing of the chip image. The range of each sub-image is relatively small, which enables the system to focus more on the feature analysis of local areas. For example, during the chip manufacturing process, some minor defects such as pinholes and scratches may not be easily detected in the entire image. However, when the image is divided into sub-images, these defects will be more prominent in the local sub-images, thereby improving the accuracy of defect detection. There may be some background noise or irrelevant areas in the chip image, and these interference factors will affect the accuracy of defect detection. By dividing the image into sub-images, the interference factors can be confined within a smaller range, reducing their impact on defect detection. For example, some stains or light and shadow changes at the edge of the chip may interfere with the overall defect detection, but in the sub-image, these interference factors will only affect some sub-images, while other sub-images can perform defect detection more accurately.

[0027] In some embodiments, calculating the defect feature index of each sub-image specifically includes the following process: Calculate the contrast significance value between the sub-image and the preset defect-free sub-image : ; Among them, is the number of superpixel regions in the sub-image, is the number of pixels in each superpixel region, is the superpixel region and the superpixel region of the preset defect-free sub-image the Euclidean distance between them, is the superpixel region and the superpixel region of the preset defect-free sub-image the centroid distance, is the median of the distances between each pair of superpixel centroids; Calculate the color distribution feature value of the sub-image : ; Among them, take the maximum value of and 0, and They are the R, B, and G values of the j-th pixel in the superpixel region in the RGB color space respectively. The R, B, and G values of the j-th pixel in Calculate the defect feature index of each sub-image : ; Among them, , , are weight coefficients respectively, is the preset contrast significant threshold, is the preset color distribution feature threshold, is the defect coefficient of the sub-image in the hyperspectral space.

[0028] It should be noted that obtaining the defect coefficient of the sub-image in the hyperspectral space specifically includes the following process: Obtain the characteristic wavelength reflectivity map of the sub-image in the hyperspectral space, add the preset band to the characteristic wavelength reflectivity map, and extract three characteristic vectors of the characteristic wavelength reflectivity map in the preset band , , , and the wavelengths corresponding to the vectors are respectively , , , calculate the defect coefficient of the sub-image in the hyperspectral space ; among them, the ratio of the reflectivity difference in the projection direction of the three characteristic vectors on the axis is equal.

[0029] In some embodiments, based on the defect feature index, segment the defective part of the integrated circuit chip in the sub-image, and extract the features of the defective part of the integrated circuit chip, which specifically includes the following process: Divide the sub-image into several parts, and judge whether the of each part exceeds the preset threshold. If so, regard this part as the defective part of the integrated circuit chip, and perform feature extraction on the defective part of the integrated circuit chip to obtain the features of the defective part of the integrated circuit chip : ; Among them, represents the number of channels in the image, represents the th pixel point on the th channel's color value, represents the number of pixel points, represents the th channel's color mean.

[0030] In some embodiments, the identification and analysis of the characteristics of the defective part to generate the appearance monitoring analysis result specifically include the following processes: Perform contrastive learning on the characteristics of the defective part and the defective part dataset to obtain the learned representation of the node: Data preparation: Collect a large number of image data of integrated circuit chips containing various defects, and annotate them to clarify the characteristics of the defective parts and the corresponding defect types in each image, such as component missing, component position soldering error, component redundancy, etc.

[0031] Construct a defective part dataset, which should cover various possible defect situations to ensure the diversity and representativeness of the data, so as to improve the accuracy of subsequent contrastive learning.

[0032] Feature extraction Extract the characteristics of the defective part from the image of the integrated circuit chip to be analyzed. These characteristics can include visual characteristics such as the shape, size, color, and texture of the defect, or advanced characteristics extracted through image processing techniques, such as edge contours, corner points, etc.

[0033] Adopt a suitable feature extraction algorithm, such as the convolutional layer and pooling layer in the convolutional neural network (CNN), to convert the defective part image into a feature vector that can be processed by a computer.

[0034] Contrastive learning Perform contrastive learning on the extracted characteristics of the defective part and the characteristics in the defective part dataset. Similarity measurement methods, such as Euclidean distance, cosine similarity, etc., can be used to calculate the similarity between the feature to be analyzed and each feature in the dataset.

[0035] Through contrastive learning, find the feature in the dataset that is most similar to the feature to be analyzed, and obtain the corresponding node learned representation. The node learned representation contains the feature information and category information of similar defects in the dataset.

[0036] Judge whether the learned representation of the node has a component missing representation. If so, judge that the integrated circuit chip has a component missing; judge whether the learned representation of the node has a component position soldering error representation. If so, judge that the integrated circuit chip has a component position soldering error; judge whether the learned representation of the node has a component redundancy representation. If so, judge that the integrated circuit chip has a component redundancy; Record the judgment result as the appearance monitoring analysis result.

[0037] In some embodiments, Figure 2 is the working flowchart of the first intelligent processing monitoring system for integrated circuit chips in the embodiments of the present invention, as Figure 2As shown in the figure, analyzing the processed data to generate the non-appearance monitoring analysis results specifically includes the following processes: Step S201: Obtain the equipment operation data of the integrated circuit chip processing line based on the processed data. Among them, the equipment includes a wafer cleaning machine, a lithography machine, an etching machine, a chemical mechanical polishing machine, an electroplating machine, and a wafer bonding machine. Compare and analyze the equipment operation data with the equipment operation standard data in the processing indicators to determine whether the equipment operation data is the same as the equipment operation standard data in the processing indicators. If so, the processing line equipment parameter coefficient JG is 1; if not, the processing line equipment parameter coefficient JG is 0. Step S202: Obtain the temperature characterization value WD during the processing period based on the processed data. Step S203: Obtain the wafer cutting speed, lithography speed, etching speed, and metal filling speed during the processing period based on the processed data. Calculate the difference between the wafer cutting speed and the preset wafer cutting speed threshold, the difference between the lithography speed and the preset lithography speed threshold, the difference between the etching speed and the preset etching speed threshold, and the difference between the metal filling speed and the preset metal filling speed threshold respectively. After taking the absolute value of all differences and summing them, obtain the sum value SD. Step S204: Obtain the internal pressure of the processing equipment, the gas pipeline pressure for etching or cleaning, and the liquid pressure in the liquid processing process during the processing period based on the processed data. The liquid processing process includes the coating of photoresist and the spraying of developer. Determine whether all pressures are within the preset range. If so, the pressure parameter YL is 1; if not, the pressure parameter YL is 0. Step S205: Substitute the processing line equipment parameter coefficient JG, the temperature characterization value WD, SD, and the pressure parameter YL into the processing qualification index calculation formula to calculate the processing qualification index LMS. The calculation formula is as follows: ; Among them, The value of is 2.72, , Are weight coefficients respectively, and their values are 0.6 and 0.4 respectively; Step S206: Determine whether the processing qualification index exceeds the preset processing qualification index threshold. If so, determine that the processing of the integrated circuit chip is qualified; if not, determine that the processing of the integrated circuit chip is unqualified. Record the judgment result as the generated non-appearance monitoring analysis result.

[0038] In some embodiments, Figure 3 Is the working flow chart of the second intelligent processing monitoring system for integrated circuit chips in the embodiments of the present invention. As Figure 3 Shown, obtaining the temperature characterization value WD during the processing period includes the following processes: Step S301: Divide the processing period into several sub-periods, and detect the temperatures during the wafer manufacturing process, the temperatures during the photolithography and etching processes, and the temperatures during the ion implantation and metal filling processes in each sub-period. Step S302: Establish a rectangular coordinate system with the execution time of each sub-period as the X-axis and the temperature value as the Y-axis, connect the temperatures in each manufacturing process, and generate the temperature curve of the wafer manufacturing process, the temperature curve of the photolithography and etching processes, and the temperature curve of the ion implantation and metal filling processes. Step S303: Calculate the total length of the curves that are above the preset temperature curve among all the temperature curves, and record the total curve length as the temperature characterization value WD during the processing period.

[0039] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0040] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Skilled professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0041] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0042] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only for some logical function divisions, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0043] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0044] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, and all should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An intelligent processing monitoring system for an integrated circuit chip, characterized in that, The system includes: A monitoring data acquisition unit, configured to acquire integrated circuit chip image data within a target area and processing data collected by sensors on a processing production line, and send the integrated circuit chip image data to an appearance analysis unit and the processing data to a non-appearance analysis unit; An appearance analysis unit, configured to receive the integrated circuit chip image data, divide the integrated circuit chip image into M sub-images, calculate the defect feature index of each sub-image, segment the defective parts of the integrated circuit chip in the sub-image based on the defect feature index, extract the defective part features of the integrated circuit chip, perform identification and analysis on the defective part features, generate an appearance monitoring analysis result, and send the appearance monitoring analysis result to a monitoring and warning unit; A non-appearance analysis unit, configured to receive the processing data, analyze the processing data, generate a non-appearance monitoring analysis result, and send the non-appearance monitoring analysis result to the monitoring and warning unit; A monitoring and warning unit, configured to receive the appearance monitoring analysis result and the non-appearance monitoring analysis result to mark the integrated circuit chips within the target area.

2. The intelligent processing monitoring system for an integrated circuit chip according to claim 1, characterized in that, Dividing the integrated circuit chip image into M sub-images specifically includes the following process: Step 1, Initialize the population: Set the initial population size m, fitness threshold , maximum number of iterations , search range , randomly generate an n-dimensional vector with each component between 0 and 1 , with as the initial value, based on the regression equation iterate to obtain N vectors , where ; Step 2: Divide the initial population into NP group and C group: Calculate the fitness values of each particle in the initial population and sort them. Select the first m particles as the particles in the NP group, and the remaining m particles as the particles in the C group; and record the individual extreme value of the NP group and the global extreme value of the NP group , record the global extreme value of the C group ; Step 3, determine whether the maximum number of iterations is reached. If the maximum number of iterations is reached, execute Step 5; otherwise, execute Steps 3-1 to 3-4; Step 4: Cross the NP group and the C group: Select the G particles with the best fitness values from the C group to replace the G particles with the worst fitness values in the NP group, and return to Step 3 to continue the iteration until the number of iterations is reached ; Step 5, the iteration ends, output the target optimization parameter M, and divide the integrated circuit chip image into M sub-images based on the target optimization parameter M.

3. The intelligent processing monitoring system for an integrated circuit chip according to claim 2, wherein, Executing Steps 3-1 to 3-4 specifically includes the following process: Step 3-1, update particles: update the velocity and position of the particles in the NP group according to the following formula ； ； Among them, the learning factor and are both 0.01, is the number of iterations; is the inertia weight, is the historical best position experienced by the th particle, is the best position experienced by all particles in the swarm, and are random numbers uniformly distributed between 0 and 1, is the particle velocity at the th iteration, is the particle position at the th iteration, is the particle position at the th iteration, is the particle velocity at the th iteration; Update the position of the C group according to the following formula: ； Among them, is the particle position at the -th iteration, is the particle position at the -th iteration, has a value of 3.57; Step 32, update the individual extreme value of the NP group and the global extreme value of the NP group , evaluate the particle fitness value after the NP group is updated. If the particle individual fitness value is better than , then is replaced with the current particle position; compare and . If is greater than , then is replaced with the current ; Step 3: Update the global extreme value of group C : Evaluate the fitness value of the particles in group C. If the fitness value is better than , then is replaced with the position of the current particle; Steps three and four, calculate the variance of the population fitness : ； ； Among them, is the total number of particles in the NP group, is the individual fitness value of the i-th particle, is the overall average fitness value of the NP group, and max represents the operation of taking the maximum value; When is less than the preset threshold, it is determined that the premature state is entered, and step four is executed. When is greater than or equal to the preset threshold, then return to step three.

4. An intelligent processing monitoring system for an integrated circuit chip according to claim 1, wherein, Calculating the defect feature index of each sub-image specifically includes the following process: Calculate the contrast significant value between the sub-image and the preset defect-free sub-image : ； Among them, is the number of superpixel regions in the sub-image, is the number of pixels in each superpixel region, is the superpixel region and the superpixel region of the preset defect-free sub-image the Euclidean distance between them, is the superpixel region and the superpixel region of the preset defect-free sub-image the centroid distance, is the median of the distances between the centroids of each pair of superpixels; Calculate the eigenvalue of the color distribution of the sub-image : ； Among them, Take The maximum value of And Are respectively the R, B, and G values of the j-th pixel point in the superpixel region in the RGB color space ; Calculate the defect feature index of each sub-image : ； Among them, , , are weight coefficients respectively, is a preset significant contrast threshold, is a preset color distribution feature threshold, is the defect coefficient of the sub-image in the hyperspectral space.

5. The intelligent processing monitoring system for an integrated circuit chip according to claim 4, characterized in that, Obtaining the defect coefficient of the sub-image in the hyperspectral space specifically includes the following process: Obtain the characteristic wavelength reflectance map of the sub-image in the hyperspectral space, add the preset band to the characteristic wavelength reflectance map, and extract three characteristic vectors of the characteristic wavelength reflectance map in the preset band , , . The wavelengths corresponding to the vectors are respectively , , . Calculate the defect coefficient of the sub-image in the hyperspectral space ; among them, the ratios of the reflectance differences of the three characteristic vectors in the projection direction of the axis are equal.

6. The intelligent processing monitoring system for an integrated circuit chip according to claim 1, wherein, Segmenting the defective parts of the integrated circuit chip in the sub-image based on the defect feature index and extracting the defective part features of the integrated circuit chip specifically Includes the following process: Divide the sub-image into several parts and judge each part's whether it exceeds a preset threshold. If so, regard this part as the defective part of the integrated circuit chip, and perform feature extraction on the defective part of the integrated circuit chip to obtain the features of the defective part of the integrated circuit chip : ； Among them, represents the number of channels in the image, represents the th pixel point, and represents the color value of the th channel, represents the number of pixel points, represents the average color value of the th channel.

7. An intelligent processing monitoring system for an integrated circuit chip according to claim 1, characterized in that, Performing identification and analysis on the defective part features to generate an appearance monitoring analysis result specifically includes the following process: Perform contrast learning on the defective part features and the defective part data set to obtain the learned representation of the node, and determine whether the learned representation of the node has a representation of missing components. If so, determine that the integrated circuit chip has missing components; Determine whether the learned representation of the node has a representation of incorrect component position welding. If so, determine that the integrated circuit chip has incorrect component position welding; Determine whether the learned representation of the node has a representation of redundant components. If so, determine that the integrated circuit chip has redundant components; Record the judgment result as the appearance monitoring analysis result.

8. An intelligent processing monitoring system for an integrated circuit chip according to claim 1, characterized in that, Analyzing the processing data to generate a non-appearance monitoring analysis result specifically includes the following process: Obtain the equipment operation data of the integrated circuit chip processing line based on the processed data. Among them, the equipment includes a wafer cleaning machine, a lithography machine, an etching machine, a chemical mechanical polishing machine, an electroplating machine, and a wafer bonding machine. Compare and analyze the equipment operation data with the equipment operation standard data in the processing indicators to determine whether the equipment operation data is the same as the equipment operation standard data in the processing indicators. If so, the processing line equipment parameter coefficient JG is 1; if not, the processing line equipment parameter coefficient JG is 0. Obtain the temperature characterization value WD during the processing period based on the processed data. Obtain the wafer cutting speed, lithography speed, etching speed, and metal filling speed during the processing period based on the processed data. Calculate the differences between the wafer cutting speed and the preset wafer cutting speed threshold, the lithography speed and the preset lithography speed threshold, the etching speed and the preset etching speed threshold, and the metal filling speed and the preset metal filling speed threshold respectively. After taking the absolute values of all the differences and summing them up, obtain the sum value SD. Obtain the internal pressure of the processing equipment, the gas pipeline pressure for etching or cleaning, and the liquid pressure in the liquid processing process during the processing period based on the processed data. Among them, the liquid processing process includes the coating of photoresist and the spraying of developer. Determine whether all the pressures are within the preset range. If so, the pressure parameter YL is 1; if not, the pressure parameter YL is 0. Substitute the processing line equipment parameter coefficient JG, the temperature characterization value WD, SD, and the pressure parameter YL into the processing qualification index calculation formula to calculate the processing qualification index LMS. The calculation formula is as follows: ； Among them, The value of which is 2.72, , are weight coefficients respectively, and their values are 0.6 and 0.4 respectively; Determine whether the processing qualification index exceeds the preset processing qualification index threshold. If so, determine that the processing of the integrated circuit chip is qualified; if not, determine that the processing of the integrated circuit chip is unqualified. Record the judgment result as the generated non-appearance monitoring analysis result.

9. An intelligent processing monitoring system for an integrated circuit chip according to claim 8, characterized in that, Obtaining the temperature characterization value WD during the processing period includes: dividing the processing period into several sub-periods, detecting the temperature during the wafer manufacturing process, the temperature during the lithography and etching processes, and the temperature during the ion implantation and metal filling processes in each sub-period. Establish a rectangular coordinate system with the execution time of each sub-period as the X-axis and the temperature value as the Y-axis. Connect the temperatures in each manufacturing process to generate the temperature curve of the wafer manufacturing process, the temperature curve of the lithography and etching processes, and the temperature curve of the ion implantation and metal filling processes. Calculate the total length of the curves where all the temperature curves are above the preset temperature curve, and record the total curve length as the temperature characterization value WD during the processing period.

Images