Coating Thickness Control Method, Device and Storage Medium

Abstract

The present application discloses a method, device and storage medium for controlling coating thickness, which relates to the field of intelligent control technology. The method includes: obtaining target clustering information of target parameters, where the target clustering information includes first clustering information of film thickness data and second clustering information of process operation parameters; periodically collecting real-time process operation parameters during the coating process; and controlling the coating thickness according to the Bayesian algorithm, the target clustering information and the periodically collected real-time process operation parameters. The problem that the existing coating control method has deficiencies in coating quality is solved, and the effect of periodically collecting real-time process operation parameters, thereby controlling the coating thickness in real time and improving the coating quality is achieved.

Description

Technical Field

[0001] The present invention relates to a method, device, and storage medium for controlling coating thickness, belonging to the field of intelligent control technology. Background Art

[0002] With the rapid development of the photovoltaic industry, the quality of photovoltaic cells has become a key factor affecting the vitality of enterprises. During the production process of the cells, they need to go through the PECVD (Plasma Enhanced Chemical Vapor Deposition) coating process, which aims to deposit a film on the cells and has many parameters. Any change in each parameter will affect the film thickness. Therefore, it is necessary to promptly detect parameter abnormalities, estimate whether the film thickness can reach the desired value, and make parameter adjustments in a timely manner when the desired value cannot be reached.

[0003] In the existing solutions, in order to predict the PECVD coating thickness, the commonly used method is to use the coating time to predict the film thickness. If the coating time is insufficient, it is considered that the film thickness has not reached the desired value and the coating time needs to be further extended. However, the above method only focuses on the length of the coating time, and the coating time is only one of the many factors affecting the coating quality during the coating process, without considering the influence of other process parameters of the coating on the film formation quality. The film formation quality under this method may have significant deficiencies. Summary of the Invention

[0004] The purpose of the present invention is to provide a method, device, and storage medium for controlling coating thickness to solve the problems existing in the prior art.

[0005] To achieve the above purpose, the present invention provides the following technical solutions:

[0006] According to the first aspect, an embodiment of the present invention provides a method for controlling coating thickness, the method comprising:

[0007] Obtaining target clustering information of target parameters, where the target clustering information includes first clustering information of film thickness data and second clustering information of process operation parameters;

[0008] Periodically collecting real-time process operation parameters during the coating process;

[0009] Controlling the coating thickness according to the Bayesian algorithm, the target clustering information, and the periodically collected real-time process operation parameters.

[0010] Optionally, the obtaining target clustering information of target parameters includes:

[0011] Obtain historical data, where the historical data includes historical film thickness data during a historical time period and historical process operation parameters during the film coating process;

[0012] Cluster the film thickness data according to the historical film thickness data to obtain the first clustering information;

[0013] Cluster the process operation parameters according to the historical process operation parameters to obtain the second clustering information.

[0014] Optionally, the controlling the film coating thickness according to the Bayesian algorithm, the target clustering information, and the periodically collected real-time process operation parameters includes:

[0015] Calculate the probability that the film thickness is the target thickness according to the Bayesian algorithm, the target clustering information, and the periodically collected real-time process operation parameters;

[0016] If the calculated probability is lower than a preset threshold, query the target process operation parameters from the database according to the target thickness, and adjust the process operation parameters during the film coating process to the target process operation parameters.

[0017] Optionally, the method further includes:

[0018] If the calculated probability reaches the preset threshold, continue the film coating according to the current process operation parameters.

[0019] Optionally, the calculating the probability that the film thickness is the target thickness according to the Bayesian algorithm, the target clustering information, and the periodically collected real-time process operation parameters includes:

[0020] Determine the thickness clustering category to which the target thickness belongs according to the first clustering information and the target thickness;

[0021] Determine the parameter clustering category to which the current belongs according to the second clustering information and the periodically collected real-time process operation parameters;

[0022] Obtain a first probability, where the first probability is the ratio of the number of times the film thickness belongs to the thickness clustering category to the total number of times;

[0023] Obtain a second probability, where the second probability is the ratio of the number of times the process operation parameters belong to the parameter clustering category to the total number of times;

[0024] Obtain a third probability, where the third probability is the ratio of the number of times the process operation parameters belong to the parameter clustering category when the film thickness belongs to the thickness clustering category to the number of times the film thickness belongs to the thickness clustering category;

[0025] Calculate the probability that the film thickness is the target thickness according to the Bayesian algorithm, the first probability, the second probability, and the third probability.

[0026] Optionally, determining the thickness clustering category to which the target thickness belongs according to the first clustering information and the target thickness includes:

[0027] Calculate the distances between the target thickness and each film thickness clustering center in the first clustering information;

[0028] Determine the film thickness clustering center corresponding to the distance with the smallest value among the calculated distances as the thickness clustering category to which the target thickness belongs.

[0029] Optionally, determining the parameter clustering category to which the current belongs according to the second clustering information and the periodically collected real-time process operation parameters includes:

[0030] For each parameter in the process operation parameters, calculate the mean values of the periodically collected parameters, and determine the process operation parameters composed of the calculated mean values as the collected operation parameters;

[0031] Calculate the distances between the collected operation parameters and each parameter clustering center in the second clustering information;

[0032] Determine the parameter clustering center corresponding to the distance with the smallest value among the calculated distances as the parameter clustering category to which the current belongs.

[0033] Optionally, calculating the probability that the film thickness is the target thickness according to the Bayesian algorithm, the first probability, the second probability, and the third probability includes:

[0034] Let the first probability be , the second probability be , and the third probability be ; then the calculated probability is:

[0035]

[0036] where a and b are non-zero positive numbers.

[0037] In a second aspect, a coating thickness control device is provided. The device includes a memory and a processor. At least one program instruction is stored in the memory, and the processor loads and executes the at least one program instruction to implement the method as described in the first aspect.

[0038] In a third aspect, a computer storage medium is provided, in which at least one program instruction is stored, and the at least one program instruction is loaded and executed by a processor to implement the method as described in the first aspect.

[0039] By obtaining the target clustering information of the target parameters, the target clustering information includes the first clustering information of the film thickness data and the second clustering information of the process operation parameters; periodically collecting the real-time process operation parameters during the film coating process; and controlling the film coating thickness according to the Bayesian algorithm, the target clustering information, and the periodically collected real-time process operation parameters. The problem that the existing film coating control method has deficiencies in the obtained film coating quality is solved, and the effect that the real-time process operation parameters can be periodically collected, and then the film coating thickness can be real-time controlled to improve the film coating quality is achieved.

[0040] At the same time, by discretizing and clustering the film thickness data and the process operation parameters, and calculating the probability by statistical means, the present application achieves the effect of reducing the calculation amount.

[0041] The above description is only an overview of the technical solution of the present invention. In order to be able to understand the technical means of the present invention more clearly and implement it according to the content of the specification, the following describes the preferred embodiments of the present invention in detail in conjunction with the accompanying drawings. Description of the Drawings

[0042] Figure 1 It is a method flow chart of the film coating thickness control method provided by an embodiment of the present invention. Detailed Embodiments

[0043] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0044] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present invention. In addition, the terms "first", "second", and "third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0045] In the description of the present invention, it should be noted that unless otherwise clearly specified and defined, the terms "installation", "connection", and "coupling" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, and it can be the communication inside two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0046] In addition, the technical features involved in different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0047] Please refer to Figure 1 , which shows a flowchart of a method for controlling the coating thickness provided by an embodiment of the present application. As Figure 1 shown, the method includes:

[0048] Step 101, obtaining the target clustering information of the target parameters, where the target clustering information includes the first clustering information of the film thickness data and the second clustering information of the process operation parameters;

[0049] Among them, the process operation parameters may include at least one of the equipment operating temperature, pressure, radio frequency power, and gas ratio.

[0050] Optionally, this step includes:

[0051] First, obtaining historical data, where the historical data includes the historical film thickness data of the coating within a historical time period and the historical process operation parameters during the coating process;

[0052] Among them, the historical time period can be one month in history, half a year in history, or one year in history, and the present application does not limit this.

[0053] Second, clustering the film thickness data according to the historical film thickness data to obtain the first clustering information;

[0054] In actual implementation, clustering can be performed by the K-means method. And since the allowable error of the PECVD coating thickness is within , therefore, the number of film thickness types K is taken to be between 16 and 20. This value is an empirical value and can be fine-tuned according to the clustering effect and the actual situation of the workshop data. The distance formula used in the clustering algorithm is the Euclidean distance, that is:

[0055]

[0056] The film thickness has only one dimension, namely thickness. Therefore, using the thickness data as the input and simultaneously inputting the number of classification categories, which is 16, the film thickness data is divided into 16 categories, and the clustering centers of each category are denoted as . The dimension of each class center is 1.

[0057] In actual implementation, the specific clustering steps include:

[0058] (1) Initialize the clustering centers: Randomly assign 16 clustering centers , and calculate the distances from all the film thickness data to the initial 16 clustering centers. Among them, the distance from any film thickness to the given initial clustering center is denoted as ;

[0059] (2) Clustering: Classify all points according to the minimum distance to all clustering centers, that is: the minimum value in min{ } is used as the class to which the point belongs;

[0060] (3) Update the clustering centers: For all classes, calculate the mean value of all film thicknesses within the class, and use this mean film thickness as the new clustering center. Repeat the above steps until the specified number of iterations, which is 10,000, is reached. After that, the final clustering result can be obtained.

[0061] Third, cluster the process operation parameters according to the historical process operation parameters to obtain the second clustering information.

[0062] Assume that the process operation parameters include K, then the historical process operation parameters can be expressed as , and in this step, each piece of data is used as a K-dimensional point as the input for clustering. Moreover, in actual implementation, the number of clustering categories for the process operation parameters can be determined according to the actual situation of the workshop. For example, if it is set to categories, then the historical process operation parameters can be divided into γ categories. Among them, the parameter clustering centers of each classification are respectively denoted as , and each clustering center C is a K-dimensional number.

[0063] In actual implementation, the clustering method in this step is similar to the clustering method for the film thickness above, and will not be elaborated here.

[0064] Step 102, periodically collect the real-time process operation parameters during the coating process;

[0065] Optionally, the real-time process operation parameters during the coating process can be collected every t seconds. Among them, t is a preset time, and this embodiment does not limit it.

[0066] Step 103: Control the coating thickness according to the Bayesian algorithm, the target clustering information, and the periodically collected real-time process operation parameters.

[0067] Optionally, this step includes:

[0068] First, calculate the probability that the film thickness is the target thickness according to the Bayesian algorithm, the target clustering information, and the periodically collected real-time process operation parameters.

[0069] In actual implementation, this step includes:

[0070] (1) Determine the thickness clustering category to which the target thickness belongs according to the first clustering information and the target thickness.

[0071] A. Calculate the distances between the target thickness and each film thickness clustering center in the first clustering information.

[0072] Calculate the distance from the target thickness h to the clustering center Denote them as in turn. Among them, the target thickness is the film thickness of the expected coating, and this thickness can be a preset value or a system default value, and there is no limitation on this.

[0073] B. Determine the film thickness clustering center corresponding to the smallest distance among the calculated distances as the thickness clustering category to which the target thickness belongs.

[0074] Take the minimum value of the set as the category to which the target film thickness belongs .

[0075] (2) Determine the parameter clustering category to which it currently belongs according to the second clustering information and the periodically collected real-time process operation parameters.

[0076] A. For each parameter in the process operation parameters, calculate the mean value of each periodically collected parameter, and determine the process operation parameters composed of the calculated mean values as the collected operation parameters.

[0077] Each periodically collected parameter is denoted as .

[0078] In actual implementation, the entire coating process can be evenly divided into N intervals according to time. The value of N needs to be reasonably selected according to the actual production situation of the factory. At the beginning of any interval n, all parameters in the above parameter set are collected, and the average value of the parameters in the previous (n - 1) intervals and the collected values in the nth interval is calculated, and this average value is used as the value of all parameters in this interval.

[0079] That is:

[0080]

[0081]

[0082]

[0083] From this, the acquisition operation parameters can be calculated.

[0084] B. Calculate the distances between the acquisition operation parameters and the parameter cluster centers in the second cluster information;

[0085] For the acquisition operation parameters , calculate the distances to all the cluster centers in sequence , and record the distances to each cluster center in sequence as .

[0086] C. Determine the parameter cluster category to which the current belongs as the parameter cluster center corresponding to the distance with the smallest value among the calculated distances.

[0087] Take the minimum value in the set as the category to which the current condition belongs .

[0088] (3). Obtain the first probability, where the first probability is the ratio of the number of times the film thickness belongs to the thickness cluster category to the total number of times;

[0089] As described in step (1), after determining the thickness cluster category to which the target thickness belongs, the number of all thicknesses belonging to the thickness cluster category can be counted, calculate the ratio of the counted number to the total number, and take the calculated ratio as the first probability.

[0090] (4). Obtain the second probability, where the second probability is the ratio of the number of times the process operation parameters belong to the parameter cluster category to the total number of times;

[0091] As described in step (2), after determining the parameter cluster category to which the acquisition operation parameters belong, the number of process operation parameters belonging to this parameter cluster category can be counted, calculate the ratio of the counted number to the total number, and take the calculated ratio as the second probability.

[0092] (5). Obtain the third probability, where the third probability is the ratio of the number of times the process operation parameters belong to the parameter cluster category when the film thickness belongs to the thickness cluster category to the number of times the film thickness belongs to the thickness cluster category;

[0093] (6) Calculate the probability that the film thickness is the target thickness based on the Bayesian algorithm, the first probability, the second probability, and the third probability.

[0094] Let the first probability be , the second probability be , and the third probability be ; then the calculated probability is:

[0095]

[0096] where a and b are non-zero positive numbers.

[0097] In a possible embodiment, to ensure the accuracy rate of the calculated probability, a and b can be integers approaching 0. For example, a = b = 0.001.

[0098] By setting non-zero positive numbers, the problem of probability being 0 is avoided.

[0099] Second, if the calculated probability is lower than the preset threshold, query the target process operation parameters from the database according to the target thickness, and adjust the process operation parameters during the film coating process to the target process operation parameters.

[0100] The preset threshold can be an empirical value. In actual implementation, the preset threshold can be 0.9.

[0101] After calculating the probability through the above steps, the magnitude relationship between the calculated probability and the preset threshold can be detected. If it is lower than the preset threshold, it means that the probability of the film thickness reaching the target probability is very low. At this time, to ensure that the film thickness meets the requirements, query the target process operation parameters from the database according to the target thickness, and then adjust the current operation parameters to the target process operation parameters. After that, continue to execute step 102 and step 103, which will not be elaborated here.

[0102] Among them, the database stores the corresponding relationship between different thicknesses and process operation parameters, and this corresponding relationship is a pre-set relationship, which is not limited here.

[0103] Third, if the calculated probability reaches the preset threshold, continue film coating according to the current process operation parameters.

[0104] If the calculated probability reaches the preset threshold, it means that continuing film coating according to the current situation can meet the requirements. At this time, film coating can be continued according to the current process operation parameters.

[0105] In summary, by obtaining the target clustering information of the target parameters, the target clustering information includes the first clustering information of the film thickness data and the second clustering information of the process operation parameters; periodically collecting the real-time process operation parameters during the film coating process; and controlling the film coating thickness according to the Bayesian algorithm, the target clustering information, and the periodically collected real-time process operation parameters. The problem of insufficient film coating quality in the existing film coating control method is solved, and the effect of periodically collecting real-time process operation parameters, thereby controlling the film coating thickness in real time and improving the film coating quality is achieved.

[0106] The present application also provides a film coating thickness control device, the device includes a memory and a processor, at least one program instruction is stored in the memory, and the processor realizes the method as described above by loading and executing the at least one program instruction.

[0107] The present application also provides a computer storage medium, at least one program instruction is stored in the computer storage medium, and the at least one program instruction is loaded and executed by a processor to realize the method as described above.

[0108] The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification.

[0109] The above-described embodiments only represent several implementation manners of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the invention patent should be subject to the appended claims.

Claims

1. A method for controlling coating thickness, characterized in that, The method includes: Obtaining target clustering information of target parameters, where the target clustering information includes first clustering information of film thickness data and second clustering information of process operation parameters; Periodically collecting real-time process operation parameters during the film coating process; Controlling the film thickness according to the Bayesian algorithm, the target clustering information, and the periodically collected real-time process operation parameters, including: Calculating the probability that the film thickness is the target thickness according to the Bayesian algorithm, the target clustering information, and the periodically collected real-time process operation parameters, including: Determining the thickness clustering category to which the target thickness belongs according to the first clustering information and the target thickness; Determining the current parameter clustering category according to the second clustering information and the periodically collected real-time process operation parameters; Obtaining a first probability, where the first probability is the ratio of the number of times the film thickness belongs to the thickness clustering category to the total number of times; Obtaining a second probability, where the second probability is the ratio of the number of times the process operation parameters belong to the parameter clustering category to the total number of times; Obtaining a third probability, where the third probability is the ratio of the number of times the process operation parameters belong to the parameter clustering category when the film thickness belongs to the thickness clustering category to the number of times the film thickness belongs to the thickness clustering category; Calculating the probability that the film thickness is the target thickness according to the Bayesian algorithm, the first probability, the second probability, and the third probability; If the calculated probability is lower than a preset threshold, query the target process operation parameters from the database according to the target thickness, and adjust the process operation parameters during the film coating process to the target process operation parameters.

2. The method according to claim 1, wherein The obtaining of the target clustering information of the target parameters includes: Obtaining historical data, where the historical data includes historical film thickness data during a historical time period and historical process operation parameters during the film coating process; Clustering the film thickness data according to the historical film thickness data to obtain the first clustering information; Clustering the process operation parameters according to the historical process operation parameters to obtain the second clustering information.

3. The method according to claim 1, wherein The method further includes: If the calculated probability reaches the preset threshold, continue film coating according to the current process operation parameters.

4. The method according to claim 1, wherein The determining of the thickness clustering category to which the target thickness belongs according to the first clustering information and the target thickness includes: Calculating the distances between the target thickness and each film thickness clustering center in the first clustering information; Determining the film thickness clustering center corresponding to the smallest value among the calculated distances as the thickness clustering category to which the target thickness belongs.

5. The method according to claim 1, wherein The determining of the current parameter clustering category according to the second clustering information and the periodically collected real-time process operation parameters includes: For each parameter in the process operation parameters, calculating the mean values of the periodically collected parameters, and determining the process operation parameters composed of the calculated mean values as the collected operation parameters; Calculating the distances between the collected operation parameters and each parameter clustering center in the second clustering information; Determining the parameter clustering center corresponding to the smallest value among the calculated distances as the current parameter clustering category.

6. The method according to claim 1, wherein Calculating the probability that the film thickness is the target thickness according to the Bayesian algorithm, the first probability, the second probability, and the third probability includes: Let the first probability be , the second probability be , and the third probability be ; then the calculated probability is: ， Where a and b are non-zero positive numbers.

7. A coating thickness control device, characterized in that, The device includes a memory and a processor. At least one program instruction is stored in the memory, and the processor loads and executes the at least one program instruction to implement the method according to any one of claims 1 to 6.

8. A computer storage medium, characterized in that, At least one program instruction is stored in the computer storage medium, and the at least one program instruction is loaded and executed by the processor to implement the method according to any one of claims 1 to 6.

Images