Intelligent equipment control system and method based on Internet of Things

Abstract

The invention discloses an intelligent equipment control system and method based on the Internet of Things, and relates to the technical field of equipment control. Historical data of coating through coating equipment in the past is collected through a historical operation data collection module, and the historical data is screened through a sample data preprocessing module; analyzing balance of residual data after screening processing, generating a first test sample and a second test sample through an equipment control mechanism setting module, establishing a first control reference model according to the first test sample, and establishing a second control reference model according to the second test sample; the current coating process is monitored in real time through the equipment control planning module, an early warning signal is sent when abnormity is monitored, parameter control and adjustment of the coating equipment are carried out according to the first control reference model and the second control reference model, and the film thickness uniformity of a film is guaranteed through automatic adjustment of equipment parameters; and the quality problem caused by passive change of the deposition rate of a thin film product due to adjustment of the target-substrate distance is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment control, and in particular to an intelligent equipment control system and method based on the Internet of Things. Background Art

[0002] Magnetron sputtering coating refers to a coating technology that uses the coating material as the target cathode, bombards the target material with argon ions, generates cathode sputtering, and sputters the target atoms onto the workpiece to form a deposition layer. Generally, a magnetic sputtering coating machine is used to complete the coating work. The magnetic sputtering coating machine can be used to coat various hard films, metal films, alloys, compounds, semiconductors, ceramic films, dielectric composite films and other chemical reaction films, and has a wide range of applications; During film coating, the uniformity of film thickness is an important indicator for measuring film quality. However, in the process of magnetron sputtering coating using coating equipment, the problem of film thickness uniformity deviation may occur due to improper parameter control. The target-substrate distance can be optimized during the operation of the coating equipment to effectively improve the uniformity of film thickness. The target-substrate distance refers to the distance between the target material and the substrate during magnetron sputtering coating. The existing technology fails to timely monitor the problem of improper parameter setting during the operation of the coating equipment and perform automatic adjustment planning of the target-substrate distance to effectively improve the quality of the thin film product. Secondly, the target-substrate distance adjustment will cause changes in the deposition rate. If the deposition rate does not reach the standard range, it will also affect the film quality. The existing technology cannot take into account the monitoring and adjustment of the deposition rate when performing automatic adjustment of the target-substrate distance, which is easy to cause other quality problems of the thin film product due to improper deposition rate. Summary of the invention

[0003] The purpose of the present invention is to provide an intelligent device control system and method based on the Internet of Things to solve the problems raised in the prior art.

[0004] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: an intelligent equipment control system based on the Internet of Things, comprising a historical operation data acquisition module, a sample data preprocessing module, an equipment control mechanism setting module and an equipment control planning module; The historical operation data collection module collects historical data of film coating using the film coating equipment in the past, including historical film quality demand data, film coating equipment data, and historical quality data of the film actually produced; Perform screening processing on historical data through the sample data preprocessing module, analyze the balance of the remaining data after the screening processing, and select the time to perform sample data combination processing; Generate a first test sample and a second test sample through the device control mechanism setting module, establish a first control reference model according to the first test sample, and establish a second control reference model according to the second test sample; The current coating process is monitored in real time through the equipment control planning module, and an early warning signal is sent when an abnormality is detected. The parameters of the coating equipment are controlled and adjusted according to the first control reference model and the second control reference model.

[0005] Further, the historical operation data collection module includes a demand data collection unit, a product quality parameter collection unit and an equipment parameter collection unit; The demand data acquisition unit acquires the planned demand data before coating using the coating equipment in the past, wherein the planned demand data includes the required film thickness, the required film thickness uniformity and the planned deposition rate standard range data during coating; The product quality parameter collection unit collects the quality parameters of the film produced after the previous coating work is completed, and the quality parameters refer to the film thickness and the uniformity of the film thickness; The equipment parameter acquisition unit is used to acquire equipment parameters used in the coating equipment for coating in the past, wherein the equipment parameters include the target-substrate distance and the deposition rate during coating.

[0006] Furthermore, the sample data preprocessing module includes a data balance analysis unit and a sample combination timing selection unit; The historical data are screened and processed by the data balance analysis unit: the data whose film thickness and film thickness uniformity meet the planned requirements are screened out, where meeting the planned requirements means that the film thickness is equal to the required film thickness and the film thickness uniformity of the film is not less than the film thickness uniformity of the required film, and the screened historical data are sorted out: the target-substrate distance during each previous coating, the film thickness after the corresponding coating is completed, and the film thickness uniformity are combined into a sample data set, and a total of k sample data sets are combined to analyze the balance coefficient of the sample data; The sample data balance coefficient threshold is set by the sample combination timing selection unit, and historical data is continuously collected until the balance coefficient of the analyzed sample data is less than or equal to the sample data balance coefficient threshold, and sample data sets with the same film thickness values ​​are selected for combination processing.

[0007] Further, the device control mechanism setting module includes an impact data adding and selecting unit, a test sample combining unit and a control reference model establishing unit; The influence data adding selection unit determines whether the deposition rate of the previous deposition using the deposition equipment is within the planned deposition rate standard range: if yes, choose not to add the deposition rate, the influence data, to the sample; if no, choose to add the deposition rate, the influence data, to the sample; The sample with the deposition rate data added and the sample without the deposition rate data added are divided into two test samples by the test sample combining unit: the sample without the deposition rate data added is used as the first test sample; the sample with the deposition rate data added is used as the second test sample; The control reference model establishment unit establishes a first control reference model according to the first test sample; and establishes a second control reference model according to the second test sample. For the same film thickness, there is a unique first control reference model and a unique second control reference model.

[0008] Furthermore, the equipment control planning module includes a current demand data acquisition unit, an optimal parameter prediction unit and an equipment parameter control early warning unit; Obtaining the currently required film thickness and the currently required film thickness uniformity through the currently required film data acquisition unit; Substituting the film thickness uniformity of the currently required film into the first control reference model corresponding to the currently required film thickness through the optimal parameter prediction unit to predict the optimal target-substrate distance during the current coating; The target-substrate distance during the current coating is monitored by the equipment parameter control early warning unit, and the monitored target-substrate distance is compared with the optimal target-substrate distance. If the monitored target-substrate distance is not equal to the optimal target-substrate distance, an early warning signal is sent to prompt the target-substrate distance during the current coating to be adjusted to the optimal target-substrate distance. After the target-substrate distance is adjusted to the optimal target-substrate distance, the deposition rate displayed on the coating equipment is monitored. The deposition rate data is displayed on the coating machine. It is determined whether the monitored deposition rate is within the planned deposition rate standard range during coating: if so, the deposition rate is not adjusted; if not, the optimal target-substrate distance and the uniformity of the film thickness of the currently required thin film are substituted into the second control reference model, the current optimal deposition rate is predicted, and the current deposition rate is adjusted to the current optimal deposition rate.

[0009] The intelligent device control method based on the Internet of Things includes the following steps: S100: collecting historical data of film coating using the film coating equipment in the past, including historical film quality demand data, film coating equipment data, and historical quality data of the films actually produced; S101: Filter the historical data, analyze the balance of the remaining data after the filtering process, and select the timing for sample data combination processing; S102: Generate a first test sample and a second test sample, establish a first control reference model according to the first test sample, and establish a second control reference model according to the second test sample; S103: Monitor the current coating process in real time, send an early warning signal when an abnormality is detected, and control and adjust parameters of the coating equipment according to the first control reference model and the second control reference model.

[0010] Further, the S100 includes: collecting the planned demand data before coating using the coating equipment in the past: collecting the required film thickness before coating for n times in the past, the uniformity of the required film thickness and the standard range of the deposition rate for each coating in the plan, collecting the target-substrate distance set {d1, d2, ...d n}, the deposition rate set during coating is {v1, v2, ...v n}, the quality parameters of the films produced after the previous n coating operations are collected: the thickness set of the produced films is obtained as {δ1, δ2, ...δ n}, the film thickness uniformity of the film is {σ1, σ2, ..., σ n}.

[0011] Further, the S101 includes: screening out data whose thickness and uniformity of the film meet the planned requirements, where meeting the planned requirements means that the thickness of the film is equal to the required film thickness and the uniformity of the film thickness is not less than the uniformity of the required film thickness, and obtaining the remaining data after screening, the remaining data including: the target-substrate distance set {D1, D2, ...D k}、The thickness set of the film made in the remaining data δ ’ ={δ1 ’ , δ2 ’ , ...δ k ’} and the film thickness uniformity set {σ1 ’ ,σ2 ’ , ..., σ k ’}, k represents the number of remaining data items, and the filtered historical data are sorted: the target-substrate distance during each previous coating, the film thickness after the corresponding coating, and the uniformity of the film thickness are combined into a sample data set, where a random sample data set is {D j , δ j ’ , σ j ’}, a total of k sample data sets are formed, and it is counted that there are m different film thickness values ​​in the k sample data sets, and the m film thickness value sets are {H1, H2, ...H i , ...H m}, the number of data sets corresponding to m different film thickness values ​​is counted as {A1, A2, ...A i , ...A m}, where A i Indicates that the film thickness is equal to Hi The number of sample data sets, the balance coefficient of the analyzed sample data is U: U=[∑ m i=1 (A i -(∑ m i=1 (A i )) / m) 2 ] / m; Set the sample data balance coefficient threshold to L, and compare U and L: if U≤L, select the sample data sets with the same film thickness values ​​for combined processing; if U>L, continue to collect historical data until the balance coefficient is less than or equal to L, and then perform the combined processing operation; The purpose of analyzing the balance coefficient of sample data is to maintain the balance of data volume between samples used to establish a control reference model after fitting training, so as to further improve the reference value of the model when it is used to adjust the parameters of the coating equipment. Since one film thickness corresponds to a unique first control reference model and a unique second control reference model, if the historical data collected for one of the film thicknesses is relatively small and the historical data collected for the other film thickness is relatively large, the data volume difference is large. For example, there will be a large difference between the reference data volume used for establishing the first control reference model corresponding to one of the film thicknesses and the reference data volume used for establishing the first control reference model corresponding to the other film thickness, which is not conducive to improving the reference value of the model. Therefore, the method of analyzing the balance coefficient is used to select the time to process the sample data, and the sample data is further processed when the difference between the number of samples is small, which is conducive to maintaining the balance of the reference data and improving the reference value of the model.

[0012] Further, the step S102 includes: if U≤L, selecting sample data sets with the same film thickness values ​​for combination processing: obtaining a film thickness equal to H after the combination processing; i The sample data set is {(D1 ’ ,σ1 ’’ ), (D2 ’ ,σ2 ’’ ),...(D f ’ , σ f ’’ )}, take this sample data set as the first test sample, D f ’ Indicates that the thickness of the film produced for the fth time is equal to H i The target-substrate distance is set when the film is f ’’ Indicates that the thickness of the film made for the fth time is equal to H i The uniformity of the film thickness, f=A iSince the film thicknesses in the f sample data sets are all equal, the film thickness data has been removed from the combined sample data set, and the film thickness data is retrieved from the previous sample data set that needs to be made with a thickness of H. i The standard range of deposition rate for the film before coating is [a, b], where a represents the minimum deposition rate, and b represents the maximum deposition rate. The set of deposition rates for the previous f times of coating is V = {V1, V2, ... V f}, it is determined that there are r deposition rates in the set V that are not in the range [a, b], and the set of r deposition rates is {V1 ’ , V2 ’ , ... V r ’}, add r deposition rates to the combined film thickness equal to H i The second test sample is generated from the sample data set as {(d1 ’ , B1, V1 ’ ), (d2 ’ , B2, V2 ’ ),...(d r ’ , B r , V r ’ )}, d r ’ Indicates that the deposition rate during previous coating is V r ’ The target distance set at the time, B r Indicates that the deposition rate during previous coating is V r ’ The target-substrate distance is set to d r ’ The uniformity of the film thickness obtained when the first test sample is measured is obtained. After data fitting for the first test sample, a first control reference model is established: y=ε1*x+ε2, ε1 and ε2 are fitting coefficients of the first control reference model, ε1 and ε2 are constants, x represents the independent variable representing the target-substrate distance in the first control reference model, and y represents the dependent variable representing the uniformity of the film thickness in the first control reference model; after data fitting for the second test sample, a second control reference model is established: Z=θ1*X+θ2*Y+θ3, θ1, θ2 and θ3 represent the fitting coefficients of the second control reference model, θ1, θ2 and θ3 are constants, X represents the variable representing the target-substrate distance in the second control reference model, Y represents the variable representing the uniformity of the film thickness in the second control reference model, and Z represents the variable representing the deposition rate in the second control reference model.

[0013] Further, the S103 includes: if the currently required film thickness is H i, obtain the film thickness uniformity of the currently required film as W, substitute W into the first control reference model: let y=W, predict the optimal target-substrate distance during the current coating to be (W-ε2) / ε1, monitor the target-substrate distance during the current coating to be g, if g≠(W-ε2) / ε1, send a warning signal and prompt to adjust the target-substrate distance during the current coating to (W-ε2) / ε1, monitor the deposition rate displayed on the coating equipment after the target-substrate distance is adjusted to (W-ε2) / ε1 to be M, judge whether M is within the range of [a,b]: if so, do not adjust the deposition rate; if not, let X=(W-ε2) / ε1, let Y=W, predict the current optimal deposition rate to be θ1*[(W-ε2) / ε1]+θ2*W+θ3, adjust the current deposition rate to θ1*[(W-ε2) / ε1]+θ2*W+θ3; Through big data technology, historical data of coating equipment, i.e., coating machine, is collected in the past for coating, and historical data that meets the plan requirements are screened out. After fitting the screened historical data, a control reference model is established. The first control reference model is used to predict the optimal target-substrate distance under different conditions, and helps to adjust the target-substrate distance in time when it is monitored that the target-substrate distance parameters do not meet the predicted optimal parameters during coating. The first control reference model is used to realize the automatic adjustment of the target-substrate distance during magnetron sputtering coating. Considering that the adjustment of the target-substrate distance will cause the change of the deposition rate during coating, it may cause the deposition rate to change outside the standard range. The deposition rate outside the standard range does not meet the plan requirements. If the deposition rate is not adjusted in time, it will also affect the quality of the manufactured film. After the adjustment, the deposition rate is monitored. When it is monitored that the deposition rate does not meet the plan requirements, the deposition rate is adjusted with reference to the second control reference model. When the target-substrate distance is automatically adjusted, the monitoring and adjustment of the deposition rate are taken into account, which effectively avoids other quality problems of the thin film product caused by improper deposition rate.

[0014] Compared with the prior art, the present invention has the following beneficial effects: The present invention collects historical data of previous coating operations using coating equipment through big data technology, screens out historical data that meets plan requirements, and establishes a control reference model after fitting the screened historical data. The first control reference model is used to predict the optimal target-substrate distance under different conditions, and helps to adjust the target-substrate distance in time when it is monitored that the target-substrate distance parameters do not meet the predicted optimal parameters during coating, which will lead to the film thickness uniformity failing to meet the requirements. The first control reference model is used to realize automatic adjustment of the target-substrate distance during magnetron sputtering coating, and further, considering that the adjustment of the target-substrate distance will lead to changes in the deposition rate during coating, the deposition rate is monitored after the adjustment, and when it is monitored that the deposition rate does not meet the plan requirements, the deposition rate is adjusted with reference to the second control reference model, and the monitoring and adjustment of the deposition rate are taken into account when the target-substrate distance is automatically adjusted, effectively avoiding other quality problems of thin film products caused by passive changes in the deposition rate due to the target-substrate distance adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a schematic diagram of the structure of an intelligent device control system based on the Internet of Things of the present invention; Figure 2 It is a flow chart of the intelligent device control method based on the Internet of Things of the present invention. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0017] Example 1: Figure 1 As shown, this embodiment provides an intelligent equipment control system based on the Internet of Things, the system includes: a historical operation data acquisition module, a sample data preprocessing module, an equipment control mechanism setting module and an equipment control planning module; The output end of the historical operation data acquisition module is connected to the input end of the sample data preprocessing module, the output end of the sample data preprocessing module is connected to the input end of the device control mechanism setting module, and the output end of the device control mechanism setting module is connected to the input end of the device control planning module; The historical operation data collection module collects historical data of film coating using the film coating equipment in the past, including historical film quality demand data, film coating equipment data, and historical quality data of the film actually produced; Perform screening processing on historical data through the sample data preprocessing module, analyze the balance of the remaining data after the screening processing, and select the time to perform sample data combination processing; Generate a first test sample and a second test sample through the device control mechanism setting module, establish a first control reference model according to the first test sample, and establish a second control reference model according to the second test sample; The current coating process is monitored in real time through the equipment control planning module, and an early warning signal is sent when an abnormality is detected. The parameters of the coating equipment are controlled and adjusted according to the first control reference model and the second control reference model.

[0018] The historical operation data collection module includes a demand data collection unit, a product quality parameter collection unit and an equipment parameter collection unit; The output end of the demand data acquisition unit is connected to the input ends of the product quality parameter acquisition unit and the equipment parameter acquisition unit; The demand data acquisition unit is used to acquire the planned demand data before coating using the coating equipment in the past. The planned demand data includes the required film thickness, the required film thickness uniformity and the planned deposition rate standard range data during the deposition. The data acquired by the demand data acquisition unit is the historical demand data of the film quality. Before the deposition is performed, the deposition rate range will be limited, and the limited range is the deposition rate standard range. The product quality parameter acquisition unit is used to collect the quality parameters of the film produced after the previous coating work is completed. The quality parameters refer to the film thickness and the uniformity of the film thickness. The film thickness monitor is used to measure the thickness at different positions on the film. The average value of the thickness at different positions is calculated as the film thickness. The standard deviation of the thickness at different positions is calculated as the uniformity of the film thickness. The standard deviation is the standard deviation. The data collected by the product quality parameter acquisition unit is the historical quality data of the film actually produced; The equipment parameter acquisition unit is used to acquire equipment parameters used in the coating equipment for coating in the past. The equipment parameters include the target-substrate distance and the deposition rate during coating. The data acquired by the equipment parameter acquisition unit is the coating equipment data.

[0019] The sample data preprocessing module includes a data balance analysis unit and a sample combination timing selection unit; The input end of the data balance analysis unit is connected to the output ends of the product quality parameter collection unit and the equipment parameter collection unit, and the output end of the data balance analysis unit is connected to the input end of the sample combination timing selection unit; The historical data are screened and processed by the data balance analysis unit: the data whose film thickness and film thickness uniformity meet the planned requirements are screened out, where meeting the planned requirements means that the film thickness is equal to the required film thickness and the film thickness uniformity of the film is not less than the film thickness uniformity of the required film, and the screened historical data are sorted out: the target-substrate distance during each previous coating, the film thickness after the corresponding coating is completed, and the film thickness uniformity are combined into a sample data set, and a total of k sample data sets are combined to analyze the balance coefficient of the sample data; The sample data balance coefficient threshold is set by the sample combination timing selection unit, and historical data is continuously collected until the balance coefficient of the analyzed sample data is less than or equal to the sample data balance coefficient threshold, and sample data sets with the same film thickness values ​​are selected for combination processing.

[0020] The device control mechanism setting module includes an impact data adding and selecting unit, a test sample combining unit and a control reference model establishing unit; The input end of the influence data adding selection unit is connected to the output end of the sample combination timing selection unit, the output end of the influence data adding selection unit is connected to the input end of the test sample combination unit, and the output end of the test sample combination unit is connected to the input end of the control reference model establishment unit; The influence data adding selection unit determines whether the deposition rate of the previous deposition using the deposition equipment is within the planned deposition rate standard range: if yes, choose not to add the deposition rate, the influence data, to the sample; if no, choose to add the deposition rate, the influence data, to the sample; The sample with the deposition rate data added and the sample without the deposition rate data added are divided into two test samples by the test sample combination unit: the sample without the deposition rate data added is used as the first test sample; the sample with the deposition rate data added is used as the second test sample; The control reference model establishment unit establishes a first control reference model according to the first test sample; and establishes a second control reference model according to the second test sample. For the same film thickness, there is a unique first control reference model and a unique second control reference model.

[0021] The equipment control planning module includes a current demand data acquisition unit, an optimal parameter prediction unit and an equipment parameter control early warning unit; The output ends of the current demand data acquisition unit and the control reference model establishment unit are connected to the input end of the optimal parameter prediction unit, and the output end of the optimal parameter prediction unit is connected to the input end of the equipment parameter control and early warning unit; Obtaining the currently required film thickness and the currently required film thickness uniformity through the currently required film data acquisition unit; Substituting the film thickness uniformity of the currently required film into the first control reference model corresponding to the currently required film thickness through the optimal parameter prediction unit to predict the optimal target-substrate distance during the current coating; The target-substrate distance during the current coating is monitored by the equipment parameter control early warning unit, and the monitored target-substrate distance is compared with the optimal target-substrate distance. If the monitored target-substrate distance is not equal to the optimal target-substrate distance, an early warning signal is sent to prompt the target-substrate distance during the current coating to be adjusted to the optimal target-substrate distance. After the target-substrate distance is adjusted to the optimal target-substrate distance, the deposition rate displayed on the coating equipment is monitored. The deposition rate data is displayed on the coating machine. It is determined whether the monitored deposition rate is within the planned deposition rate standard range during coating: if so, the deposition rate is not adjusted; if not, the optimal target-substrate distance and the uniformity of the film thickness of the currently required thin film are substituted into the second control reference model, the current optimal deposition rate is predicted, and the current deposition rate is adjusted to the current optimal deposition rate.

[0022] Example 2: Figure 2 As shown, this embodiment provides an intelligent device control method based on the Internet of Things, which is implemented based on the device control system in the embodiment and specifically includes the following steps: S100: Collect historical data of film coating using film coating equipment in the past, including historical film quality demand data, film coating equipment data and historical quality data of films actually produced, collect planned demand data before film coating using film coating equipment in the past: collect required film thickness before film coating for n times in the past, required film thickness uniformity and planned deposition rate standard range for each film coating, collect target-substrate distance set {d1, d2, ...d n}, the deposition rate set during coating is {v1, v2, ...v n}, the quality parameters of the films produced after the previous n coating operations are collected: the thickness set of the produced films is obtained as {δ1, δ2, ...δ n}, the film thickness uniformity of the film is {σ1, σ2, ..., σ n}; S101: Screening the historical data, analyzing the balance of the remaining data after the screening, selecting the time for combining the sample data, and screening the data whose thickness and uniformity of the film meet the planned requirements. Meeting the planned requirements means that the thickness of the film is equal to the required film thickness and the uniformity of the film thickness is not less than the uniformity of the required film thickness, and obtaining the remaining data after the screening. The remaining data includes: the target-substrate distance set {D1, D2, ...Dk}、The thickness set of the film made in the remaining data δ ’ ={δ1 ’ , δ2 ’ , ...δ k ’} and the film thickness uniformity set {σ1 ’ ,σ2 ’ , ..., σ k ’}, k represents the number of remaining data items, and the filtered historical data are sorted: the target-substrate distance during each previous coating, the film thickness after the corresponding coating, and the uniformity of the film thickness are combined into a sample data set, where a random sample data set is {D j , δ j ’ , σ j ’}, a total of k sample data sets are formed, and it is counted that there are m different film thickness values ​​in the k sample data sets, and the m film thickness value sets are {H1, H2, ...H i , ...H m}, the number of data sets corresponding to m different film thickness values ​​is counted as {A1, A2, ...A i , ...A m}, where A i Indicates that the film thickness is equal to H i The number of sample data sets, the balance coefficient of the analyzed sample data is U: U=[∑ m i=1 (A i -(∑ m i=1 (A i )) / m) 2 ] / m; Set the sample data balance coefficient threshold to L, and compare U and L: if U≤L, select the sample data sets with the same film thickness values ​​for combined processing; if U>L, continue to collect historical data until the balance coefficient is less than or equal to L, and then perform the combined processing operation; For example: Get the remaining data after screening: The set of film thickness δ in the remaining data ’ ={δ1 ’ , δ2 ’ , δ3 ’ , δ4 ’ , δ5 ’ , δ6 ’ , δ7 ’ , δ8 ’ , δ9 ’ , δ10 ’}={0.1, 0.2, 0.15, 0.16, 0.1, 0.1, 0.2, 0.1, 0.1, 0.2}, the unit is micrometer, there are 4 different film thickness values, the 4 film thickness value sets are {0.1, 0.2, 0.15, 0.16}, the number of data sets corresponding to the 4 different film thickness values ​​is {A1, A2, A3, A4}={5, 3, 1, 1}, the number of data sets is similar to the set δ ’ The number of times the same film thickness appears in the sample data is equal. The equalization coefficient of the sample data is analyzed as U, and U≈2.75 is obtained. The threshold of the equalization coefficient of the sample data is set to L=0.5, U>L, and historical data is collected until the equalization coefficient is less than or equal to 0.5, and then the sample data set is combined and processed. S102: Generate a first test sample and a second test sample, establish a first control reference model based on the first test sample, and establish a second control reference model based on the second test sample. If U≤L, select sample data sets with the same film thickness value for combination processing: obtain a film thickness equal to H after combination processing. i The sample data set is {(D1 ’ ,σ1 ’’ ), (D2 ’ ,σ2 ’’ ),...(D f ’ , σ f ’’ )}, take this sample data set as the first test sample, D f ’ Indicates that the thickness of the film produced for the fth time is equal to H i The target-substrate distance is set when the film is f ’’ Indicates that the thickness of the film made for the fth time is equal to H i The uniformity of the film thickness, f=A i Since the film thicknesses in the f sample data sets are all equal, the film thickness data has been removed from the combined sample data set, and the film thickness data is retrieved from the previous sample data set that needs to be made with a thickness of H. i The standard range of deposition rate for the film before coating is [a, b], where a represents the minimum deposition rate, and b represents the maximum deposition rate. The set of deposition rates for the previous f times of coating is V = {V1, V2, ... V f}, it is determined that there are r deposition rates in the set V that are not in the range [a, b], and the set of r deposition rates is {V1 ’ , V2 ’ , ... V r ’}, add r deposition rates to the combined film thickness equal to H i The second test sample is generated from the sample data set as {(d1 ’ , B1, V1 ’ ), (d2 ’ , B2, V2 ’ ),...(d r ’ , B r , V r ’ )}, d r ’ Indicates that the deposition rate during previous coating is V r ’ The target distance set at the time, B r Indicates that the deposition rate during previous coating is V r ’ The target-substrate distance is set to d r ’ The uniformity of the film thickness obtained when the first test sample is measured is obtained. After data fitting for the first test sample, a first control reference model is established: y=ε1*x+ε2, ε1 and ε2 are fitting coefficients of the first control reference model, ε1 and ε2 are constants, x represents the independent variable representing the target-substrate distance in the first control reference model, and y represents the dependent variable representing the uniformity of the film thickness in the first control reference model; after data fitting for the second test sample, a second control reference model is established: Z=θ1*X+θ2*Y+θ3, θ1, θ2 and θ3 represent the fitting coefficients of the second control reference model, θ1, θ2 and θ3 are constants, X represents the variable representing the target-substrate distance in the second control reference model, Y represents the variable representing the uniformity of the film thickness in the second control reference model, and Z represents the variable representing the deposition rate in the second control reference model; S103: Monitor the current coating process in real time, send an early warning signal when an abnormality is detected, and control and adjust the parameters of the coating equipment according to the first control reference model and the second control reference model. If the current required film thickness is H i, the film thickness uniformity of the currently required film is obtained as W, and W is substituted into the first control reference model: let y=W, and the optimal target-substrate distance during the current coating is predicted to be (W-ε2) / ε1. The target-substrate distance during the current coating is monitored to be g. If g≠(W-ε2) / ε1, a warning signal is sent and a prompt is given to adjust the target-substrate distance during the current coating to (W-ε2) / ε1. After the target-substrate distance is adjusted to (W-ε2) / ε1, the deposition rate displayed on the coating equipment is monitored to be M. It is determined whether M is within the range of [a,b]. If so, the deposition rate is not adjusted. If not, let X=(W-ε2) / ε1 and Y=W. The current optimal deposition rate is predicted to be θ1*[(W-ε2) / ε1]+θ2*W+θ3, and the current deposition rate is adjusted to θ1*[(W-ε2) / ε1]+θ2*W+θ3.

[0023] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. Intelligent equipment control system based on the Internet of Things, characterized by: It includes historical operation data collection module, sample data preprocessing module, equipment control mechanism setting module and equipment control planning module; The historical operation data collection module collects historical data of film coating using the film coating equipment in the past, including historical film quality demand data, film coating equipment data, and historical quality data of the film actually produced; Perform screening processing on historical data through the sample data preprocessing module, analyze the balance of the remaining data after the screening processing, and select the time to perform sample data combination processing; Generate a first test sample and a second test sample through the device control mechanism setting module, establish a first control reference model according to the first test sample, and establish a second control reference model according to the second test sample; The current coating process is monitored in real time through the equipment control planning module, and an early warning signal is sent when an abnormality is detected. The parameters of the coating equipment are controlled and adjusted according to the first control reference model and the second control reference model.

2. The intelligent device control system based on the Internet of Things according to claim 1 is characterized in that: The historical operation data acquisition module includes a demand data acquisition unit, a product quality parameter acquisition unit and an equipment parameter acquisition unit; the demand data acquisition unit is used to acquire the planned demand data before coating using the coating equipment in the past, and the planned demand data includes the required film thickness, the required film thickness uniformity and the planned deposition rate standard range data during coating; the product quality parameter acquisition unit is used to acquire the quality parameters of the film made after the coating work is completed in the past, and the quality parameters refer to the film thickness and the film thickness uniformity; the equipment parameters when coating using the coating equipment in the past are acquired through the equipment parameter acquisition unit, and the equipment parameters include the target substrate distance and the deposition rate during coating.

3. The intelligent device control system based on the Internet of Things according to claim 2 is characterized in that: The sample data preprocessing module includes a data balance analysis unit and a sample combination timing selection unit; the data balance analysis unit is used to screen the historical data: the data whose film thickness and film thickness uniformity meet the planned requirements are screened out, and meeting the planned requirements means that the film thickness is equal to the required film thickness and the film thickness uniformity is not less than the required film thickness uniformity, and the screened historical data are sorted out: the target-substrate distance during each previous coating, the film thickness formed after the corresponding coating is completed, and the film thickness uniformity are combined into a sample data set, and a total of k sample data sets are combined to analyze the balance coefficient of the sample data; the sample data balance coefficient threshold is set by the sample combination timing selection unit, and the historical data is continuously collected until the balance coefficient of the sample data obtained by analysis is less than or equal to the sample data balance coefficient threshold, and the sample data sets with the same film thickness value are selected for combination processing.

4. The intelligent device control system based on the Internet of Things according to claim 3 is characterized in that: The equipment control mechanism setting module includes an influence data adding selection unit, a test sample combination unit and a control reference model establishment unit; the influence data adding selection unit is used to judge whether the deposition rate when the coating equipment was used for coating in the past is within the planned deposition rate standard range during coating: if so, the deposition rate influence data is not added to the sample; If not, choose to add the deposition rate, an influencing data, to the sample; divide the sample with the deposition rate data and the sample without the deposition rate data into two test samples through the test sample combination unit: use the sample without the deposition rate data as the first test sample; use the sample with the deposition rate data as the second test sample; establish a first control reference model based on the first test sample through the control reference model establishment unit; and establish a second control reference model based on the second test sample.

5. The intelligent device control system based on the Internet of Things according to claim 4 is characterized in that: The equipment control planning module includes a current demand data acquisition unit, an optimal parameter prediction unit and an equipment parameter control early warning unit; Obtaining the currently required film thickness and the currently required film thickness uniformity through the currently required film data acquisition unit; Substituting the film thickness uniformity of the currently required film into the first control reference model corresponding to the currently required film thickness through the optimal parameter prediction unit to predict the optimal target-substrate distance during the current coating; The target-substrate distance during the current coating is monitored by the equipment parameter control and early warning unit, and the monitored target-substrate distance is compared with the optimal target-substrate distance. If the monitored target-substrate distance is not equal to the optimal target-substrate distance, an early warning signal is sent to prompt the target-substrate distance during the current coating to be adjusted to the optimal target-substrate distance. After the target-substrate distance is adjusted to the optimal target-substrate distance, the deposition rate displayed on the coating equipment is monitored to determine whether the monitored deposition rate is within the planned deposition rate standard range during coating: if so, no deposition rate adjustment is performed; If not, the optimal target-substrate distance and the uniformity of the thickness of the currently required thin film are substituted into the second control reference model to predict the current optimal deposition rate, and the current deposition rate is adjusted to the current optimal deposition rate.

6. An intelligent device control method based on the Internet of Things, characterized in that: The following steps are involved: S100: collecting historical data of film coating using the film coating equipment in the past, including historical film quality demand data, film coating equipment data, and historical quality data of the films actually produced; S101: Filter the historical data, analyze the balance of the remaining data after the filtering process, and select the timing for sample data combination processing; S102: Generate a first test sample and a second test sample, establish a first control reference model according to the first test sample, and establish a second control reference model according to the second test sample; S103: Monitor the current coating process in real time, send an early warning signal when an abnormality is detected, and control and adjust parameters of the coating equipment according to the first control reference model and the second control reference model.

7. The method for controlling intelligent devices based on the Internet of Things according to claim 6, characterized in that: The step S100 includes: collecting the planned demand data before coating using the coating equipment in the past: collecting the required film thickness before coating for n times in the past, the uniformity of the required film thickness and the standard range of the deposition rate for each coating in the plan, collecting the target-substrate distance set {d1, d2, ...d n }, the deposition rate set during coating is {v1, v2, ...v n }, the quality parameters of the films produced after the previous n coating operations are collected: the thickness set of the produced films is obtained as {δ1, δ2, ...δ n }, the film thickness uniformity of the film is {σ1, σ2, ..., σ n }.

8. The method for controlling intelligent devices based on the Internet of Things according to claim 7, characterized in that: The step S101 includes: screening out data whose thickness and uniformity of the film meet the planned requirements, where meeting the planned requirements means that the thickness of the film is equal to the required film thickness and the uniformity of the film thickness is not less than the uniformity of the required film thickness, and obtaining the remaining data after screening, wherein the remaining data includes: a target-substrate distance set {D1, D2, ...D k }、The thickness set of the film made in the remaining data δ ’ ={δ1 ’ , δ2 ’ , ...δ k ’ } and the film thickness uniformity set {σ1 ’ ,σ2 ’ , ..., σ k ’ }, k represents the number of remaining data items, and the filtered historical data are sorted: the target-substrate distance during each previous coating, the film thickness after the corresponding coating, and the uniformity of the film thickness are combined into a sample data set, where a random sample data set is {D j , δ j ’ , σ j ’ }, a total of k sample data sets are formed, and it is counted that there are m different film thickness values ​​in the k sample data sets, and the m film thickness value sets are {H1, H2, ...H i , ...H m }, the number of data sets corresponding to m different film thickness values ​​is counted as {A1, A2, ...A i , ...A m }, where A i Indicates that the film thickness is equal to H i The number of sample data sets, the balance coefficient of the analyzed sample data is U: U=[∑ m i=1 (A i -(∑ m i=1 (A i )) / m) 2 ] / m; Set the sample data balancing coefficient threshold to L, and compare U and L: if U≤L, select the sample data sets with the same film thickness values ​​for combined processing; if U>L, continue to collect historical data until the balancing coefficient is less than or equal to L and then perform the combined processing operation.

9. The method for controlling intelligent devices based on the Internet of Things according to claim 8, characterized in that: The step S102 includes: if U≤L, selecting sample data sets with the same film thickness values ​​for combination processing: obtaining a film thickness equal to H after the combination processing; i The sample data set is {(D1 ’ ,σ1 ’’ ), (D2 ’ ,σ2 ’’ ),...(D f ’ , σ f ’’ )}, take this sample data set as the first test sample, D f ’ Indicates that the thickness of the film made for the fth time is equal to H i The target-substrate distance is set when the film is f ’’ Indicates that the thickness of the film made for the fth time is equal to H i The uniformity of the film thickness, f=A i , retrieve the previous i The standard range of deposition rate for the film before coating is [a, b], where a represents the minimum deposition rate, and b represents the maximum deposition rate. The set of deposition rates for the previous f times of coating is V = {V1, V2, ... V f }, it is determined that there are r deposition rates in the set V that are not in the range [a, b], and the set of r deposition rates is {V1 ’ , V2 ’ , ... V r ’ }, add r deposition rates to the combined film thickness equal to H i The second test sample is generated from the sample data set as {(d1 ’ , B1, V1 ’ ), (d2 ’ , B2, V2 ’ ),...(d r ’ , B r , V r ’ )}, d r ’ Indicates that the deposition rate during previous coating is V r ’ The target distance set at the time, B r Indicates that the deposition rate during previous coating is V r ’ The target-substrate distance is set to d r ’ The uniformity of the film thickness obtained when the first test sample is measured is obtained. After data fitting for the first test sample, a first control reference model is established: y=ε1*x+ε2, ε1 and ε2 are the fitting coefficients of the first control reference model, x represents the independent variable representing the target-substrate distance in the first control reference model, and y represents the dependent variable representing the uniformity of the film thickness in the first control reference model; after data fitting for the second test sample, a second control reference model is established: Z=θ1*X+θ2*Y+θ3, θ1, θ2 and θ3 represent the fitting coefficients of the second control reference model, X represents the variable representing the target-substrate distance in the second control reference model, Y represents the variable representing the uniformity of the film thickness in the second control reference model, and Z represents the variable representing the deposition rate in the second control reference model.

10. The method for controlling intelligent devices based on the Internet of Things according to claim 9, characterized in that: The step S103 includes: if the currently required film thickness is H i , the film thickness uniformity of the currently required film is obtained as W, and W is substituted into the first control reference model: let y=W, and the optimal target-substrate distance during the current coating is predicted to be (W-ε2) / ε1. The target-substrate distance during the current coating is monitored to be g. If g≠(W-ε2) / ε1, a warning signal is sent and a prompt is given to adjust the target-substrate distance during the current coating to (W-ε2) / ε1. After the target-substrate distance is adjusted to (W-ε2) / ε1, the deposition rate displayed on the coating equipment is monitored to be M. It is determined whether M is within the range of [a,b]. If so, the deposition rate is not adjusted. If not, let X=(W-ε2) / ε1 and Y=W. The current optimal deposition rate is predicted to be θ1*[(W-ε2) / ε1]+θ2*W+θ3, and the current deposition rate is adjusted to θ1*[(W-ε2) / ε1]+θ2*W+θ3.

Images