Intelligent control method of semiconductor cleaning equipment

Abstract

The invention discloses an intelligent control method for semiconductor cleaning equipment, and relates to the technical field of intelligent control of semiconductor cleaning. Image data of the surface of a semiconductor are collected in real time by using an image collection module, and the surface of the semiconductor is cleaned by using an image data processing module based on the image data; sequentially calculating and outputting a preliminary distribution value Qi of the cleaning degree of the ith area, the cleaning liquid concentration NDi of the ith area, the cleaning time Ti of the ith area, the cleaning temperature Wadj, i after adjustment of the ith area and a distribution value Qadj, i of the cleaning degree after adjustment of the ith area, and carrying out intelligent control adjustment on semiconductor cleaning by utilizing a cleaning execution module, according to the method, by defining the meaning of the characters in the formula, forming the relation of circulation influence, intelligently adjusting the cleaning parameters and improving the cleaning efficiency and the cleaning quality, optimization and innovation of the prior art are achieved, and then the automation level and the intelligent level of semiconductor cleaning equipment can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent control semiconductor cleaning, and in particular to an intelligent control method for semiconductor cleaning equipment. Background Art

[0002] In the semiconductor manufacturing process, cleaning is a crucial link. Traditional cleaning methods often rely on manual experience and fixed cleaning parameters, which is not only inefficient but also difficult to ensure the consistency of cleaning quality. With the rapid development of image recognition technology, the application of image recognition technology to the intelligent control of semiconductor cleaning equipment has become a technical means. Through image recognition technology, dust and pollutants on the semiconductor surface can be accurately identified, and the degree of cleaning can be intelligently allocated according to their grayscale values, thereby realizing the automation and intelligence of the cleaning process.

[0003] However, the existing intelligent control methods of semiconductor cleaning equipment based on image recognition still have some problems and shortcomings. Specifically, some methods often use ambiguous weight factors and influence coefficients when calculating the degree of cleaning, resulting in inaccurate distribution of the degree of cleaning. At the same time, these methods often lack the relationship between systematic and cyclic influences when adjusting cleaning parameters, resulting in difficulty in achieving the best cleaning effect. In addition, the intelligent control of most cleaning equipment is a unified control of the cleaning equipment. It is worth noting that the dust and pollutant content in different areas of the semiconductor surface is inconsistent. Therefore, if unified control is performed, it is very easy to cause incomplete cleaning and excessive cleaning of some areas. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent control method for semiconductor cleaning equipment, which solves the problems raised in the above background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solutions, and the specific implementation steps are as follows: Step 1: Use the image acquisition module to collect image data of the semiconductor surface in real time; Step 2: Based on the image data and using the image data processing module, calculate and output the preliminary distribution value Q of the cleaning degree of the i-th area in sequence i , cleaning fluid concentration ND in area i i , cleaning time T for area i i , cleaning temperature W after adjustment of area i adj，i , the adjusted cleaning degree distribution value Q of the i-th area adj，i ; Step 3: Adjust the cleaning temperature W based on the i-th region adj，i and the adjusted cleaning degree distribution value Q of the i-th area adj，i, and use the cleaning execution module to perform intelligent control and adjustment of semiconductor cleaning; The image data includes the grayscale value of each pixel in each cleaning area of the semiconductor surface, as well as the cleaning temperature, cleaning degree, cleaning time, and cleaning liquid concentration of each cleaning area; The image data processing module includes a unit for preliminary allocation of cleaning degrees for different areas of the semiconductor, a unit for intelligently adjusting some cleaning parameters, and a unit for comprehensive adjustment and feedback impact; The equipment used in the image acquisition module includes a high-resolution camera, an image sensor, and an image acquisition card; The equipment used in the image data processing module includes a computer and an embedded system; The equipment used in the cleaning execution module includes cleaning equipment, temperature controller and concentration regulator; The cleaning equipment includes an ultrasonic cleaning machine and a spray cleaning machine for cleaning semiconductors; The temperature controller is used to accurately control the temperature of the cleaning equipment; The concentration regulator is used to adjust the concentration of the cleaning liquid.

[0006] Optionally, the calculation formula of the preliminary allocation unit for cleaning degrees of different areas of the semiconductor is as follows: ; in: Q i Initially assign a value to the cleaning degree of region i; N is the total number of pixels, which reflects the total number of pixels whose grayscale values are detected in the i-th area; H ij is the gray value of the j-th pixel, H ij Reflects the grayscale value of the jth pixel in the i-th area; H avg is the average gray value; W base is the basic cleaning temperature, W base Reflects the basic value of cleaning semiconductor temperature; W max is the maximum cleaning temperature, W max Reflects the maximum degree of cleaning of semiconductors at a temperature that does not damage the semiconductor surface; H max is the maximum gray value, H min is the minimum gray value, H max and H min Respectively reflect the maximum and minimum grayscale values accumulated so far when cleaning and identifying semiconductors.

[0007] Optionally, the average gray value H avg The calculation formula is as follows: H avg =(H i1 +H i2 +H i3 +......+H iN ) / N; H i1 is the gray value of the first pixel, H i2 is the gray value of the second pixel, H i3 is the gray value of the third pixel, H iN is the gray value of the Nth pixel; The gray value H of the j-th pixel ij Represents the grayscale value of the first pixel H i1 , the gray value of the second pixel H i2 , the gray value of the third pixel H i3 , the gray value of the Nth pixel H iN The grayscale value of any pixel in .

[0008] Optionally, the calculation formula of the intelligent adjustment unit for the cleaning parameters is as follows: ND i =(Q i / Q avg ) 2 ×{ND base +[(ND max -ND base ) / (Q max -Q min )]×(Q i -Q min )}; ; in: ND i is the cleaning solution concentration in the i-th region; T i is the cleaning time of the i-th area; Q avg Initially assign average values to the cleaning levels; Q max Initially assign the maximum value for the cleaning degree, Q min Initially assign a minimum value for the cleaning degree, Q max and Q min Respectively reflect the maximum and minimum levels of cleaning degree initially distributed when cleaning semiconductors up to the present time; ND base is the basic cleaning solution concentration, ND base Reflects the basic value of the cleaning solution concentration when cleaning semiconductors; ND max is the maximum cleaning solution concentration, ND max Reflects the maximum degree of cleaning of semiconductors up to now without damaging the semiconductor surface by the cleaning solution concentration; T base is the basic cleaning time, T base Reflects the basic value of cleaning time when cleaning semiconductors; T max is the maximum cleaning time, T max Reflects the maximum degree of cleaning of semiconductors up to now without damaging the semiconductor surface; ND avg is the average concentration of the cleaning solution.

[0009] Optionally, the calculation formula of the comprehensive adjustment and feedback influence unit is as follows: ; Q adj，i =Q i +[(ND i -ND avg ) / (ND max -ND base )]×[(W adj，i -W base ) / (W max -W base )]; in: W adj，i Adjust the cleaning temperature for zone i; Q adj，i Assign a value to the adjusted cleaning degree of region i; T avg is the average cleaning time.

[0010] Optionally, the cleaning temperature W after adjustment is based on the i-th region adj，i and the adjusted cleaning degree distribution value Q of the i-th area adj，i The intelligent control method is as follows: Adjust the cleaning temperature W for the i-th area adj，i If the cleaning temperature W of the i-th area is adjusted adj，i Lower than the basic cleaning temperature W base , it will lead to poor cleaning effect, the maximum cleaning temperature W should be increased max ; If the cleaning temperature W of the i-th area is adjusted adj，i Higher than the maximum cleaning temperature W max , it will cause damage to the semiconductor surface, the maximum cleaning temperature W max Lower; Adjusted cleaning degree distribution value Q for area i adj，i If the adjusted cleaning degree distribution value Q of the i-th area adj，i Below the minimum value Q of the initial allocation of cleaning degree min , it will lead to incomplete cleaning, and the cleaning time T of the i-th area should be increased. i ; If the adjusted cleaning degree distribution value Q of the i-th area adj，i Above the maximum value Q of the initial allocation of cleaning degree max , it will cause damage to the semiconductor surface and the presence of cleaning liquid residue, so the cleaning time T of the i-th area should be reduced. i .

[0011] Optionally, based on the intelligent control method, in the next round of intelligent iteration, the cleaning temperature W of the i-th area is adjusted. adj，i Replace the basic cleaning temperature W base , and the adjusted cleaning degree distribution value Q of the i-th area adj，i Replace the initial assigned value Q of the cleaning degree of the i-th area i , and perform separate introduction calculations.

[0012] Optionally, the cleaning degree is initially distributed with an average value Q avg The calculation formula is as follows: Q avg =(Q1+Q2+Q3+......+Q m ) / m; m is the total cleaning area, which reflects the total amount of area currently divided for cleaning on the semiconductor surface; Q1 is the preliminary assigned value of the cleaning degree of the first area, Q2 is the preliminary assigned value of the cleaning degree of the second area, Q3 is the preliminary assigned value of the cleaning degree of the third area, Q m Assign a preliminary value to the cleaning degree of the mth region; The average concentration of the cleaning solution ND avg The calculation formula is as follows: ND avg =(ND1+ND2+ND3+......+ND m ) / m; ND1 is the concentration of the cleaning liquid in the first area, ND2 is the concentration of the cleaning liquid in the second area, ND3 is the concentration of the cleaning liquid in the third area, ND m is the cleaning solution concentration in the mth region; The average cleaning time T avg The calculation formula is as follows: T avg =(T1+T2+T3+......+T m ) / m; T1 is the cleaning time of the first area, T2 is the cleaning time of the second area, T3 is the cleaning time of the third area, T m is the cleaning time of the mth area.

[0013] Optionally, the preliminary assigned value Q of the cleaning degree of the i-th region i represents the preliminary assigned value Q1 of the cleaning degree of the first region, the preliminary assigned value Q2 of the cleaning degree of the second region, the preliminary assigned value Q3 of the cleaning degree of the third region, the preliminary assigned value Q m Initial allocation of cleaning degree for any area; The concentration ND of the cleaning liquid in the i-th region i represents the first area cleaning liquid concentration ND1, the second area cleaning liquid concentration ND2, the third area cleaning liquid concentration ND3, the mth area cleaning liquid concentration ND m The concentration of cleaning fluid in any area; The cleaning time T of the i-th area i represents the first area cleaning time T1, the second area cleaning time T2, the third area cleaning time T3, the mth area cleaning time T m The cleaning time of any area in the system.

[0014] Compared with the prior art, the present invention has the following beneficial effects: 1. The preliminary allocation unit for the cleaning degree of different areas of the semiconductor, the intelligent adjustment unit for some cleaning parameters, and the comprehensive adjustment and feedback influence unit in the present invention all use clear character meanings and mathematical expressions, avoiding ambiguous weight factors and influence coefficients. Specifically, in the preliminary allocation unit for the cleaning degree of different areas of the semiconductor, the cleaning degree is allocated by calculating the discrete degree of the grayscale value of the semiconductor surface, ensuring the accuracy and reliability of the cleaning degree.

[0015] Second, the preliminary allocation unit for cleaning degree of different semiconductor regions, the intelligent adjustment unit for some cleaning parameters, and the comprehensive adjustment and feedback influence unit in the present invention are associated with each other through the result value and the value, forming a circular influence relationship. In the comprehensive adjustment and feedback influence unit, the cleaning temperature and the cleaning degree value are comprehensively adjusted, and the adjusted cleaning temperature W of the i-th region is adjusted. adj，i and the adjusted cleaning degree distribution value Q of the i-th area adj，i The results serve as input for the next iteration, enabling dynamic adjustment and optimization of cleaning parameters. This cyclical influence relationship helps to improve the consistency and stability of the cleaning effect.

[0016] 3. The intelligent control method of the present invention can intelligently adjust the cleaning parameters according to the degree of contamination and cleaning requirements of the semiconductor surface. Specifically, in the intelligent adjustment of the cleaning parameter unit, the average value Q of the cleaning degree is calculated. avg By setting the cleaning degree threshold parameters and accumulating them, the intelligent adjustment of cleaning liquid concentration and cleaning time is realized. This intelligent adjustment method helps to improve cleaning efficiency and cleaning quality.

[0017] 4. The intelligent control method of the present invention can significantly improve the cleaning efficiency and cleaning quality of semiconductor cleaning equipment. On the one hand, by accurately identifying dust and pollutants on the surface of the semiconductor and intelligently allocating the cleaning degree according to their grayscale values, it can ensure the pertinence and effectiveness of the cleaning process; on the other hand, by dynamically adjusting the cleaning parameters and forming a cyclic influence relationship, it can further optimize the cleaning effect and improve the consistency of the cleaning quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flow chart of the intelligent control method for semiconductor cleaning equipment; Figure 2 Schematic diagram of the structure of the image data processing module of the present invention; Figure 3 The image acquisition module in the present invention collects the grayscale value H of the i-th area and the j-th pixel point ij Schematic diagram of; Figure 4 Schematic diagram of the structure of the loop feedback of the present invention. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0020] Regarding the intelligent control method of this semiconductor cleaning equipment, it is different from the existing intelligent control methods. The existing intelligent control methods have unclear weight factors and influence coefficients, which lead to inaccurate distribution of cleaning degrees and often lack of systematic and cyclic influence relationships. In addition, there is a lack of targeted and flexible cleaning adjustments for the surface of the semiconductor. This algorithm unit optimizes and innovates the existing technology by clarifying the meaning of the characters in the formula, forming a cyclic influence relationship, intelligently adjusting the cleaning parameters, and improving the cleaning efficiency and cleaning quality, thereby helping to improve the automation and intelligence level of semiconductor cleaning equipment.

[0021] For example 1, please refer to Figures 1 to 4 This embodiment provides an intelligent control method for semiconductor cleaning equipment, and the specific implementation steps are as follows: Step 1: Use the image acquisition module to collect image data of the semiconductor surface in real time; Step 2: Based on the image data and using the image data processing module, calculate and output the preliminary distribution value Q of the cleaning degree of the i-th area in sequence i , cleaning fluid concentration ND in area i i , cleaning time T for area i i , cleaning temperature W after adjustment of area i adj，i , the adjusted cleaning degree distribution value Q of the i-th area adj，i ; Step 3: Adjust the cleaning temperature W based on the i-th area adj，i and the adjusted cleaning degree distribution value Q of the i-th area adj，i , and use the cleaning execution module to perform intelligent control and adjustment of semiconductor cleaning; The image data includes the grayscale value of each pixel in each cleaning area of the semiconductor surface, as well as the cleaning temperature, cleaning degree, cleaning time, and cleaning liquid concentration of each cleaning area; The image data processing module includes a unit for preliminary allocation of cleaning degrees for different areas of the semiconductor, a unit for intelligently adjusting some cleaning parameters, and a unit for comprehensive adjustment and feedback impact. The equipment used in the image acquisition module includes high-resolution cameras, image sensors, and image acquisition cards; The equipment used in the image data processing module includes computers and embedded systems; The equipment used in the cleaning execution module includes cleaning equipment, temperature controller and concentration regulator; Cleaning equipment includes ultrasonic cleaning machines and spray cleaning machines, which are used to clean semiconductors; The temperature controller is used to accurately control the temperature of the cleaning equipment; The concentration regulator is used to adjust the concentration of the cleaning solution.

[0022] In this embodiment, the system forms the core part of the intelligent control method of semiconductor cleaning equipment through the mutual cooperation of three algorithm units. i ,ND i With T i 、W adj，i With Q adj，i The five calculation results provide more efficient, accurate and flexible cleaning control for semiconductor cleaning. Specifically, Q iThe initial value of the cleaning degree of the i-th region is assigned. This value reflects the degree and distribution of contamination on the semiconductor surface and is an important basis for the subsequent intelligent adjustment of the cleaning solution concentration and time. i is the cleaning fluid concentration in the i-th region, T i is the cleaning time of the i-th area, the two values of the algorithm unit are in Q i On the basis of the above, the purpose of intelligently adjusting the concentration of cleaning fluid and time is achieved, thereby ensuring that the cleaning process is more efficient and accurate, while reducing unnecessary waste of cleaning fluid and cleaning time. adj，i is the adjusted cleaning temperature of the i-th zone, Q adj，i The algorithm unit comprehensively adjusts the results of the first two algorithm units and considers the relationship of the cycle influence to generate the comprehensively adjusted cleaning temperature and cleaning degree values. These adjusted values are used as Q in the next iteration. i ,ND i With T i The input is used to adjust the cleaning parameters more accurately. Through cyclic iteration and continuous optimization, the comprehensive optimization and control of the semiconductor cleaning process can be achieved, and W adj，i With Q adj，i The calculation results can also affect the feedback to Q i ,ND i With T i The calculation makes the three algorithms of this system have their own clear calculation purposes and logical relationships, as well as the mutual influence and feedback between the three algorithms. Through the comprehensive use of the three algorithms, intelligent, precise and efficient control of the semiconductor cleaning process can be achieved.

[0023] See also Figures 1 to 4 ,The calculation formula for the preliminary allocation unit of cleaning degree of different areas of semiconductor is as follows: ; in: Q i Initially assign a value to the cleaning degree of region i; N is the total number of pixels, which reflects the total number of pixels whose grayscale values are detected in the i-th area; H ij is the gray value of the j-th pixel, H ij Reflects the grayscale value of the jth pixel in the i-th area; H avg is the average gray value; W base is the basic cleaning temperature, W base Reflects the basic value of cleaning semiconductor temperature; W max is the maximum cleaning temperature, W maxReflects the maximum degree of cleaning of semiconductors at a temperature that does not damage the semiconductor surface; H max is the maximum gray value, H min is the minimum gray value, H max and H min Respectively reflect the maximum and minimum grayscale values accumulated so far when cleaning and identifying semiconductors; Average gray value H avg The calculation formula is as follows: H avg =(H i1 +H i2 +H i3 +......+H iN ) / N; H i1 is the gray value of the first pixel, H i2 is the gray value of the second pixel, H i3 is the gray value of the third pixel, H iN is the gray value of the Nth pixel; The gray value H of the j-th pixel ij Represents the grayscale value of the first pixel H i1 , the gray value of the second pixel H i2 , the gray value of the third pixel H i3 , the gray value of the Nth pixel H iN The grayscale value of any pixel in .

[0024] In this embodiment: First, in this algorithm unit, The calculation part calculates the sum of the squares of the differences between the grayscale values of all pixels in the semiconductor's i-th region and the average value, reflecting the degree of dispersion of the grayscale values on the semiconductor surface. The greater the dispersion, the more uneven the contamination of the semiconductor surface, and the more detailed cleaning is required. As a component of the calculation formula for the preliminary allocation of cleaning levels for different semiconductor regions, this calculation part can help determine the cleaning priority of different semiconductors based on the degree of dispersion of their grayscale values. “ The calculation part calculates the difference between the average grayscale value of all pixels on the semiconductor region i and the preset minimum grayscale threshold, which reflects the relative degree of contamination on the semiconductor surface. In the preliminary allocation unit of cleaning degree for different semiconductor regions, this calculation part is used to adjust the basic cleaning temperature W. base Maximum cleaning temperature W max The ratio between them can dynamically adjust the cleaning temperature according to the degree of contamination; The gray value calculation in the preliminary allocation unit of the cleaning degree of different areas of the semiconductor in this algorithm can accurately identify the contaminated area and degree of contamination on the semiconductor surface, and accumulate to the current maximum cleaning temperature W max , maximum gray value H max , minimum gray value H min , can divide the semiconductor surface into areas with different pollution levels, and intelligently allocate the cleaning degree according to the pollution degree of these areas. This targeted cleaning method can not only effectively remove pollutants, but also avoid excessive cleaning of non-polluted and lightly polluted areas, thereby protecting the integrity of the semiconductor surface. In addition, the preliminary allocation of cleaning degree is based on the precise calculation of grayscale values. Therefore, it can minimize the cleaning time and the amount of cleaning fluid used while ensuring the cleaning quality. By reasonably setting the mapping relationship between grayscale value and cleaning degree, it can achieve precise control of the cleaning degree, avoiding unnecessary cleaning operations and time waste. This optimization strategy not only improves cleaning efficiency, but also reduces production costs and environmental burden. The preliminary allocation unit for cleaning degree of different semiconductor regions also considers the influence of cleaning temperature on cleaning effect. By comprehensively considering the gray value H of the jth pixel point, ij , Maximum cleaning temperature W max , maximum gray value H max , minimum gray value H min This comprehensive consideration approach can ensure more precise temperature control during the cleaning process, thereby enhancing the cleaning effect and improving the cleanliness and quality stability of the semiconductor surface.

[0025] See also Figures 1 to 4 , the calculation formula for intelligent adjustment of some cleaning parameter units is as follows: ND i =(Q i / Q avg ) 2 ×{ND base +[(ND max -ND base ) / (Q max -Q min )]×(Q i -Q min )}; ; in: ND i is the cleaning solution concentration in the i-th region; T i is the cleaning time of the i-th area; Q avg Initially assign average values to the cleaning levels; Q max Initially assign the maximum value for the cleaning degree, Q min Initially assign a minimum value for the cleaning degree, Q max and Q min Respectively reflect the maximum and minimum levels of cleaning degree initially distributed when cleaning semiconductors up to the present time; ND base is the basic cleaning solution concentration, ND base Reflects the basic value of the cleaning solution concentration when cleaning semiconductors; ND max is the maximum cleaning solution concentration, ND max Reflects the maximum degree of cleaning of semiconductors up to now without damaging the semiconductor surface by the cleaning solution concentration; T base is the basic cleaning time, T base Reflects the basic value of cleaning time when cleaning semiconductors; T max is the maximum cleaning time, T max Reflects the maximum degree of cleaning of semiconductors up to now without damaging the semiconductor surface; ND avg is the average concentration of the cleaning solution.

[0026] In this embodiment, first, "(Q i / Q avg ) 2 The calculation part calculates the preliminary distribution value Q of the cleaning degree of the semiconductor region i i The square ratio of the average value of the preliminary cleaning degree distribution of all semiconductors reflects the cleaning requirement of the semiconductor region i relative to the overall cleaning degree. As a factor for adjusting the cleaning solution concentration, this calculation part helps determine the cleaning solution concentration required for different semiconductors to meet their specific cleaning requirements. “ "Calculation part calculates the cleaning degree and initially allocates the maximum value Q max The minimum value Q is initially assigned to the cleaning degree min The difference between the initial cleaning degree distribution value Q of the i-th area i The minimum value Q is initially assigned to the cleaning degree min The ratio of the difference between the basic cleaning solution concentration and the maximum cleaning solution concentration reflects the relative size of the cleaning degree requirement of the semiconductor region i. In the intelligent adjustment part of the cleaning parameter unit, the calculation part is used to adjust the ratio between the basic cleaning solution concentration and the maximum cleaning solution concentration, thereby dynamically adjusting the cleaning solution concentration according to the different cleaning degree requirements. The calculation part combines the relative size of the cleaning degree required and the ratio of the average value to the individual value of the cleaning solution concentration. This part calculates the adjustment factor of the cleaning time as the basis for adjusting the cleaning time. This calculation part helps determine the cleaning time required for different semiconductors to optimize the cleaning effect; This algorithm intelligently adjusts some cleaning parameter units by preliminarily assigning a value Q according to the cleaning degree of the i-th area. i Adjusting the cleaning fluid concentration ensures that the cleaning fluid is fully utilized during the cleaning process. For areas with high pollution levels, the cleaning fluid concentration can be appropriately increased to improve the cleaning effect. For areas with low pollution levels, the cleaning fluid concentration can be reduced to reduce waste. This intelligent adjustment method not only improves the utilization rate of the cleaning fluid, but also reduces production costs and environmental burden. In addition, the adjustment of the cleaning time is also based on the preliminary assigned value Q of the cleaning degree of the i-th area i For areas with heavy pollution, the cleaning time can be appropriately extended to ensure that the pollutants are completely removed. For areas with light pollution, the cleaning time can be shortened to improve efficiency. This flexible cleaning time adjustment method can be personalized according to semiconductors with different pollution levels, thereby optimizing cleaning efficiency while ensuring cleaning quality. The intelligent adjustment unit of some cleaning parameters can be intelligently adjusted according to the cleaning requirements of different semiconductors. By comprehensively considering the preliminary distribution value of the cleaning degree, the concentration of the cleaning liquid and the cleaning time factors, it can generate personalized cleaning solutions suitable for the cleaning needs of different semiconductors. This highly adaptable cleaning method not only improves the cleaning effect, but also enhances the flexibility and diversity of the cleaning method.

[0027] See also Figures 1 to 4 , the calculation formula of comprehensive adjustment and feedback influence unit is as follows: ; Q adj，i =Q i +[(ND i -ND avg ) / (ND max -ND base )]×[(W adj，i -W base ) / (W max -W base )]; in: W adj，i Adjust the cleaning temperature for zone i; Q adj，i Assign a value to the adjusted cleaning degree of region i; T avg is the average cleaning time.

[0028] In this embodiment, the algorithm unit firstly The calculation part calculates the combined effects of the cleaning time deviation from the average value, the cleaning solution concentration, and the preliminary cleaning degree distribution value, which are used to adjust the cleaning temperature. As a factor in adjusting the cleaning temperature, this calculation part reflects the comprehensive requirements of different semiconductors during the cleaning process and helps optimize the cleaning temperature to achieve the best cleaning effect. “[(ND i -ND avg ) / (ND max -ND base )]×[(W adj，i -W base ) / (W max -W base The calculation part combines the deviation of the cleaning solution concentration relative to the average value and the difference between the adjusted cleaning temperature value and the base value to calculate the adjustment amount of the cleaning degree as the basis for adjusting the cleaning degree value. The calculation part helps determine the cleaning degree that different semiconductors should achieve after cleaning to meet specific quality requirements; This algorithm unit can generate more accurate cleaning parameters by comprehensively considering factors such as cleaning temperature, cleaning solution concentration, and cleaning time. These precise cleaning parameters can ensure more accurate control of temperature, concentration, and time during the cleaning process, thereby improving cleaning accuracy and stability. This high-precision cleaning method can ensure that the cleanliness and quality stability of the semiconductor surface meet the requirements. The cyclic influence relationship formed by this algorithm unit makes the cleaning process more dynamic and flexible. In each iteration, the comprehensive adjustment and feedback influence unit will automatically adjust the cleaning parameters of the next iteration based on the cleaning results of the previous iteration. This dynamic adjustment method can continuously optimize the cleaning process, making the cleaning effect more stable and reliable. At the same time, the cyclic influence relationship can also reduce human intervention and errors, and improve the automation and intelligence level of the cleaning process. Since the parameter adjustment in the comprehensive adjustment and feedback influence unit is based on the cleaning results of the previous iteration, the dependence of the entire system on the initial conditions is reduced. This method of reducing dependence makes the system more stable and reliable, and enhances the robustness and stability of the system. Even in the face of different semiconductor cleaning requirements and changes in environmental conditions, the system can quickly adapt and adjust the cleaning parameters to ensure cleaning effect and quality stability.

[0029] See also Figures 1 to 4 , based on the adjusted cleaning temperature W of the i-th region adj，i and the adjusted cleaning degree distribution value Q of the i-th area adj，i The intelligent control method is as follows: Adjust the cleaning temperature W for the i-th areaadj，i If the cleaning temperature W of the i-th area is adjusted adj，i Lower than the basic cleaning temperature W base , it will lead to poor cleaning effect, the maximum cleaning temperature W should be increased max ; If the cleaning temperature W of the i-th area is adjusted adj，i Higher than the maximum cleaning temperature W max , it will cause damage to the semiconductor surface, the maximum cleaning temperature W max Lower; Adjusted cleaning degree distribution value Q for area i adj，i If the adjusted cleaning degree distribution value Q of the i-th area adj，i Below the minimum value Q of the initial allocation of cleaning degree min , it will lead to incomplete cleaning, and the cleaning time T of the i-th area should be increased. i ; If the adjusted cleaning degree distribution value Q of the i-th area adj，i Above the maximum value Q of the initial allocation of cleaning degree max , it will cause damage to the semiconductor surface and the presence of cleaning liquid residue, so the cleaning time T of the i-th area should be reduced. i ; Based on the intelligent control method, in the next round of intelligent iteration, the cleaning temperature W of the i-th area is adjusted. adj，i Replace the basic cleaning temperature W base , and the adjusted cleaning degree distribution value Q of the i-th area adj，i Replace the initial assigned value Q of the cleaning degree of the i-th area i , and perform separate introduction calculations.

[0030] In this embodiment, a more accurate adjusted cleaning temperature W of the i-th region can be obtained by comprehensive adjustment and adjustment of the feedback influence unit. adj，i This temperature can be used as the basic cleaning temperature W for the preliminary allocation of cleaning degrees to different areas of the semiconductor in the next iteration. base , used to adjust the cleaning temperature. This cyclic adjustment method can make the cleaning temperature more in line with actual needs and improve the cleaning effect. The comprehensive adjustment and feedback affect the adjusted cleaning degree distribution value Q of the i-th area in the unit. adj，i It can be used as the input of the intelligent adjustment unit of some cleaning parameters in the next iteration to more accurately adjust the concentration and time of the cleaning solution. At the same time, due to the adjusted cleaning degree distribution value Q of the i-th area adj，i It is adjusted based on the cleaning results of the previous iteration, so it can more accurately reflect the contamination of the semiconductor surface and thus optimize the distribution of cleaning degree; Specifically, when the cleaning temperature is too low, the basic cleaning temperature W base And set a higher maximum cleaning temperature W max , can significantly improve the cleaning efficiency. This is because the appropriate temperature can accelerate the chemical reaction between the cleaning liquid and the contaminants on the semiconductor surface, making the contaminants easier to remove. On the contrary, when the cleaning temperature is too high, lowering the basic cleaning temperature W base And set a higher maximum cleaning temperature W max It can prevent the cleaning solution from volatilizing too quickly and damaging the semiconductor surface due to excessive temperature, thereby maintaining a stable cleaning efficiency. Similarly, by increasing the cleaning solution concentration ND in the i-th region i , Extend the cleaning time T of area i i , which can ensure a more thorough cleaning process and remove more pollutants, thereby improving cleaning efficiency. When the cleaning degree value is too high, by reducing the concentration of the cleaning fluid in the i-th area ND i , Extend the cleaning time T of area i i , which can avoid damage to the semiconductor surface caused by excessive cleaning and reduce the waste of cleaning fluid, which is also a way to improve cleaning efficiency; Therefore, precise cleaning temperature control can ensure the consistency and stability of the cleaning process, thereby obtaining a uniform cleaning effect. Avoiding excessively high or low cleaning temperatures can reduce physical and chemical damage to the semiconductor surface, ensuring the quality of the semiconductor surface after cleaning. The appropriate cleaning degree value can ensure that contaminants are completely removed while avoiding unnecessary damage to the semiconductor surface. By adjusting the cleaning solution concentration and cleaning time, the cleaning degree value can be precisely controlled, thereby obtaining high-quality cleaning results. This embodiment can reduce energy consumption during the cleaning process by optimizing the cleaning temperature and cleaning time, thereby reducing production costs. By adjusting the cleaning liquid concentration and cleaning time, it can reduce cleaning liquid waste, thereby reducing cleaning liquid costs. Appropriate cleaning temperature and cleaning degree values can reduce wear and corrosion on cleaning equipment, thereby extending the service life of the equipment. By precisely controlling the cleaning parameters, it can ensure the stable operation of the cleaning equipment and reduce equipment failures and downtime caused by improper parameters. In summary, the preliminary allocation unit of cleaning degree for different areas of semiconductors, the intelligent adjustment unit of some cleaning parameters and the comprehensive adjustment and feedback influence unit each have significant beneficial effects, and the cyclical influence of the comprehensive adjustment and feedback influence unit on the preliminary allocation unit of cleaning degree for different areas of semiconductors further enhances the adaptability and cleaning effect of the entire system. These algorithm formulas together constitute the core part of the intelligent control method of semiconductor cleaning equipment, providing the semiconductor manufacturing industry with more efficient, accurate and flexible cleaning solutions, and reasonably adjusting the cleaning temperature and cleaning degree values in the semiconductor cleaning process, which can not only improve the cleaning efficiency and cleaning quality, but also save energy and costs, and increase the service life of the equipment.

[0031] For example 2, please refer to Figures 1 to 4 , the average value of the initial distribution of cleaning degree Q avg The calculation formula is as follows: Q avg =(Q1+Q2+Q3+......+Q m ) / m; m is the total cleaning area, which reflects the total amount of area currently divided for cleaning on the semiconductor surface; Q1 is the preliminary assigned value of the cleaning degree of the first area, Q2 is the preliminary assigned value of the cleaning degree of the second area, Q3 is the preliminary assigned value of the cleaning degree of the third area, Q m Assign a preliminary value to the cleaning degree of the mth region; The average concentration of the cleaning solution ND avg The calculation formula is as follows: ND avg =(ND1+ND2+ND3+......+ND m ) / m; ND1 is the concentration of the cleaning liquid in the first area, ND2 is the concentration of the cleaning liquid in the second area, ND3 is the concentration of the cleaning liquid in the third area, ND m is the cleaning solution concentration in the mth region; The average cleaning time T avg The calculation formula is as follows: T avg =(T1+T2+T3+......+T m ) / m; T1 is the cleaning time of the first area, T2 is the cleaning time of the second area, T3 is the cleaning time of the third area, T m is the cleaning time of the mth area; The initial assigned value Q for the cleaning degree of the i-th region i represents the preliminary assigned value Q1 of the cleaning degree of the first region, the preliminary assigned value Q2 of the cleaning degree of the second region, the preliminary assigned value Q3 of the cleaning degree of the third region, the preliminary assigned value Qm Initial allocation of cleaning degree for any area; The concentration of cleaning liquid in area i is ND i represents the first area cleaning liquid concentration ND1, the second area cleaning liquid concentration ND2, the third area cleaning liquid concentration ND3, the mth area cleaning liquid concentration ND m The concentration of cleaning fluid in any area; Cleaning time T for area i i represents the first area cleaning time T1, the second area cleaning time T2, the third area cleaning time T3, the mth area cleaning time T m The cleaning time of any area in the system.

[0032] In this embodiment, the average value Q of the initial distribution of cleaning degree is avg , average concentration of cleaning fluid ND avg , average cleaning time T avg The precise calculation can fully reflect the cleaning distribution, concentration, cleaning time of different areas on the semiconductor surface, as well as the average degree of cleaning distribution, concentration, and cleaning time on the semiconductor surface. Therefore, the system can perform different cleaning tasks according to different areas of the semiconductor surface to ensure that comprehensive, flexible and thorough cleaning tasks are achieved without damaging the semiconductor surface.

[0033] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent control method for semiconductor cleaning equipment, characterized in that: The specific implementation steps are as follows: Step 1: Use the image acquisition module to collect image data of the semiconductor surface in real time; Step 2: Based on the image data and using the image data processing module, calculate and output the preliminary distribution value Q of the cleaning degree of the i-th area in sequence i , cleaning fluid concentration ND in area i i , cleaning time T for area i i , cleaning temperature W after adjustment of area i adj，i , the adjusted cleaning degree distribution value Q of the i-th area adj，i ; Step 3: Adjust the cleaning temperature W based on the i-th region adj，i and the adjusted cleaning degree distribution value Q of the i-th area adj，i , and use the cleaning execution module to perform intelligent control and adjustment of semiconductor cleaning; The image data includes the grayscale value of each pixel in each cleaning area of the semiconductor surface, as well as the cleaning temperature, cleaning degree, cleaning time, and cleaning liquid concentration of each cleaning area; The image data processing module includes a unit for preliminary allocation of cleaning degrees for different areas of the semiconductor, a unit for intelligently adjusting some cleaning parameters, and a unit for comprehensive adjustment and feedback influence.

2. The intelligent control method for semiconductor cleaning equipment according to claim 1, characterized in that: The equipment used in the image acquisition module includes a high-resolution camera, an image sensor, and an image acquisition card; The equipment used by the image data processing module includes computers and embedded systems; The equipment used in the cleaning execution module includes cleaning equipment, temperature controller and concentration regulator; The cleaning equipment includes an ultrasonic cleaning machine and a spray cleaning machine for cleaning semiconductors; The temperature controller is used to accurately control the temperature of the cleaning equipment; The concentration regulator is used to adjust the concentration of the cleaning liquid.

3. The intelligent control method for semiconductor cleaning equipment according to claim 2, characterized in that: The calculation formula for the preliminary allocation unit of cleaning degree of different areas of the semiconductor is as follows: ； in: Q i Initially assign a value to the cleaning degree of region i; N is the total number of pixels, which reflects the total number of pixels whose grayscale values are detected in the i-th area; H ij is the gray value of the j-th pixel, H ij Reflects the grayscale value of the j-th pixel in the i-th area; H avg is the average gray value; W base is the basic cleaning temperature, W base Reflects the basic value of cleaning semiconductor temperature; W max is the maximum cleaning temperature, W max Reflects the maximum degree of cleaning of semiconductors at a temperature that does not damage the semiconductor surface; H max is the maximum gray value, H min is the minimum gray value, H max and H min Respectively reflect the maximum and minimum grayscale values accumulated so far when cleaning and identifying semiconductors.

4. The intelligent control method for semiconductor cleaning equipment according to claim 3, characterized in that: The average gray value H avg The calculation formula is as follows: H avg =(H i1 +H i2 +H i3 +......+H iN ) / N; H i1 is the gray value of the first pixel, H i2 is the gray value of the second pixel, H i3 is the gray value of the third pixel, H iN is the gray value of the Nth pixel; The gray value H of the j-th pixel ij Represents the grayscale value of the first pixel H i1 , the gray value of the second pixel H i2 , the gray value of the third pixel H i3 , the gray value of the Nth pixel H iN The gray value of any pixel in ; The calculation formula of the intelligent adjustment part cleaning parameter unit is as follows: ND i =(Q i / Q avg ) 2 ×{ND base +[(ND max -ND base ) / (Q max -Q min )]×(Q i -Q min )}; ； in: ND i is the cleaning solution concentration in the i-th region; T i is the cleaning time of the i-th area; Q avg Initially assign average values to the cleaning levels; Q max Initially assign the maximum value for the cleaning degree, Q min Initially assign a minimum value for the cleaning degree, Q max and Q min Respectively reflect the maximum and minimum levels of cleaning degree initially distributed when cleaning semiconductors up to the present time; ND base is the basic cleaning solution concentration, ND base Reflects the basic value of the cleaning solution concentration when cleaning semiconductors; ND max is the maximum cleaning solution concentration, ND max Reflects the maximum degree of cleaning of semiconductors up to now without damaging the semiconductor surface by the cleaning solution concentration; T base is the basic cleaning time, T base Reflects the basic value of cleaning time when cleaning semiconductors; T max is the maximum cleaning time, T max Reflects the maximum degree of cleaning of semiconductors up to now without damaging the semiconductor surface; ND avg is the average concentration of the cleaning solution.

5. The intelligent control method for semiconductor cleaning equipment according to claim 4, characterized in that: The calculation formula of the comprehensive adjustment and feedback influence unit is as follows: ； Q adj，i =Q i +[(ND i -ND avg ) / (ND max -ND base )]×[(W adj，i -W base ) / (W max -W base )]; in: W adj，i Adjust the cleaning temperature for zone i; Q adj，i Assign a value to the adjusted cleaning degree of region i; T avg is the average cleaning time.

6. The intelligent control method for semiconductor cleaning equipment according to claim 5, characterized in that: The adjusted cleaning temperature W of the i-th zone is based on adj，i and the adjusted cleaning degree distribution value Q of the i-th area adj，i The intelligent control method is as follows: Adjust the cleaning temperature W for the i-th area adj，i If the cleaning temperature W of the i-th area is adjusted adj，i Lower than the basic cleaning temperature W base , it will lead to poor cleaning effect, the maximum cleaning temperature W should be increased max ; If the cleaning temperature W of the i-th area is adjusted adj，i Higher than the maximum cleaning temperature W max , it will cause damage to the semiconductor surface, the maximum cleaning temperature W max Lower; Adjusted cleaning degree distribution value Q for area i adj，i If the adjusted cleaning degree distribution value Q of the i-th area adj，i Below the minimum value Q of the initial allocation of cleaning degree min , it will lead to incomplete cleaning, and the cleaning time T of the i-th area should be increased. i ; If the adjusted cleaning degree distribution value Q of the i-th area adj，i Above the maximum value Q of the initial allocation of cleaning degree max , it will cause damage to the semiconductor surface and the presence of cleaning liquid residue, so the cleaning time T of the i-th area should be reduced. i .

7. The intelligent control method for semiconductor cleaning equipment according to claim 6, characterized in that: Based on the intelligent control method, in the next round of intelligent iteration, the cleaning temperature W of the i-th area is adjusted. adj，i Replace the basic cleaning temperature W base , and the adjusted cleaning degree distribution value Q of the i-th area adj，i Replace the initial assigned value Q of the cleaning degree of the i-th area i , and the calculations are introduced separately.

8. The intelligent control method for semiconductor cleaning equipment according to claim 6, characterized in that: The cleaning degree is initially assigned an average value Q avg The calculation formula is as follows: Q avg =(Q1+Q2+Q3+......+Q m ) / m; m is the total cleaning area, which reflects the total amount of area currently divided for cleaning on the semiconductor surface; Q1 is the preliminary assigned value of the cleaning degree of the first area, Q2 is the preliminary assigned value of the cleaning degree of the second area, Q3 is the preliminary assigned value of the cleaning degree of the third area, Q m Assign a preliminary value to the cleaning degree of the mth region; The average concentration of the cleaning solution ND avg The calculation formula is as follows: ND avg =(ND1+ND2+ND3+......+ND m ) / m; ND1 is the concentration of the cleaning liquid in the first area, ND2 is the concentration of the cleaning liquid in the second area, ND3 is the concentration of the cleaning liquid in the third area, ND m is the cleaning solution concentration in the mth region; The average cleaning time T avg The calculation formula is as follows: T avg =(T1+T2+T3+......+T m ) / m; T1 is the cleaning time of the first area, T2 is the cleaning time of the second area, T3 is the cleaning time of the third area, T m is the cleaning time of the mth area.

9. The intelligent control method for semiconductor cleaning equipment according to claim 8, characterized in that: The preliminary assigned value Q of the cleaning degree of the i-th region i represents the preliminary assigned value Q1 of the cleaning degree of the first region, the preliminary assigned value Q2 of the cleaning degree of the second region, the preliminary assigned value Q3 of the cleaning degree of the third region, the preliminary assigned value Q m Initial allocation of cleaning degree for any area; The cleaning liquid concentration ND in the i-th region i represents the first area cleaning liquid concentration ND1, the second area cleaning liquid concentration ND2, the third area cleaning liquid concentration ND3, the mth area cleaning liquid concentration ND m The concentration of cleaning fluid in any area; The cleaning time T of the i-th area i represents the first area cleaning time T1, the second area cleaning time T2, the third area cleaning time T3, the mth area cleaning time T m The cleaning time of any area in the system.

Images