CN101751026A - Method and device for controlling process - Google Patents
Method and device for controlling process Download PDFInfo
- Publication number
- CN101751026A CN101751026A CN200810239833A CN200810239833A CN101751026A CN 101751026 A CN101751026 A CN 101751026A CN 200810239833 A CN200810239833 A CN 200810239833A CN 200810239833 A CN200810239833 A CN 200810239833A CN 101751026 A CN101751026 A CN 101751026A
- Authority
- CN
- China
- Prior art keywords
- technology
- forecast model
- terminal time
- historical data
- data sample
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Abstract
The invention discloses a method and a device for controlling a process. The method comprises the following steps: selecting a historical data sample under a current process, wherein the historical data sample comprises spectral intensity data and to-be-predicted parameters of the current process; according to the historical data sample, constructing a prediction model which takes the spectral intensity data as an input variable and takes the to-be-predicted parameters as output variables; and based on the prediction model, controlling the process. In the method and the device, the historical data sample for the same process (like the current process to be controlled) is selected and the proper prediction model is constructed so that the defects of simple prediction in the prior art or errors caused by the single adoption of real-time spectral data are avoided.
Description
Technical field
The present invention relates to the semiconductor process techniques field, particularly relate to a kind of method and device of technology controlling and process.
Background technology
Along with the semi-conductor chip process technology is strict day by day, technology node is from 250nm to 65nm, even below the 45nm, the size of silicon chip also is increased to 300mm from 200mm, and is more and more stricter for the processing technology control of silicon chip in the case.In order to guarantee the quality of production of high novel technique lower silicon slice, a lot of integrated circuit manufacturers and etching apparatus manufacturer introduce the control technology of advanced technologies one after another, are used for the monitoring of product quality of art production process and the prediction of emphasis technological parameter.
Under perfect condition, the silicon chip of processing same kind with a kind of technological process should have identical processing speed.But in the processes process of reality, owing to the reason of system itself or the existence of other interference, cause the fluctuation of process endpoint time and processing speed or unusual, thereby cause the processing uniformity between sheet and the sheet can't reach requirement, serious also can cause useless sheet.Therefore realize the real-time monitoring of process endpoint is very important by the predict process terminal time.
In today of technology fast development, it is far from being enough that on-line sensor and real-time detection method are only arranged, engineers also wishes to dope in advance timely some important technical parameters (as the terminal time of technology, processing speed etc.) in the technological process, thereby more accurately technology is controlled, the terminal point wrong report of avoiding system interference to cause is boosted productivity.
In existing terminal time prediction scheme, simple and commonly used is the time method.In actual processing technology, if systematic comparison is stable, the fluctuation of process time and processing speed is little so, therefore (for example according to machined parameters, need the thickness of etching and the average etch rate of board), just can the predict process time, the carrying out of CONTROL PROCESS that can be relatively accurate.This prediction scheme can be useful in the not high technology of broad lines, terminal time accuracy requirement, but adopts this scheme to predict and just control and can't satisfy technological requirement in being lower than the technology of 90nm.Integrated circuit for little live width (being that live width is less) is made, and this scheme is inadvisable, and its low efficient wouldn't be said, importantly can not realize the on-line monitoring of process endpoint time in the technology.
In etching technics, another is used always and has prediction scheme now comparatively accurately is online optical emission spectroscopy.Optical emitting spectral detection system mainly is based on the spectrum that line spectrum checkout equipment article on plasma body launches and detects in real time, because etching into different material layer spectrum obvious variation can occur, when etching into the material of different layers in the etching process by monitoring, the spectral line of emission intensity level of reactant or product is judged etching terminal with this.Though optical emission spectroscopy is the spectrum that goes out of the plasma emission of detecting reactant or product in real time, and accurately judge the etching terminal time.But its accuracy relies on signal quality very much, and is lower if signal is subjected to the interference or the plasma intensity of etching system, and the accuracy of this prediction scheme will obviously reduce and occur easily the wrong report of terminal time so, influences the throughput rate of product.
In a word, press for the technical matters that those skilled in the art solve at present and be exactly: how can propose a kind of method of technology controlling and process, accurate predict process parameter is with the throughput rate of raising product with innovating.
Summary of the invention
Technical matters to be solved by this invention provides a kind of method of technology controlling and process, and accurately the predict process parameter is better carried out technology controlling and process, to improve the throughput rate of product.
In order to solve the problems of the technologies described above, the embodiment of the invention discloses a kind of process control method, comprising: choose the historical data sample under the current technology, described historical data sample comprises the spectral intensity data and the parameter to be predicted of current technology; According to described historical data sample, making up with the spectral intensity data is input variable, is the forecast model of output variable with parameter to be predicted; Based on this forecast model, this technological process is controlled.
Preferably, described parameter to be predicted can comprise terminal time, perhaps terminal time and processing speed.
Preferably, the process of described structure forecast model can comprise the standard normal step to historical data sample, and the step of setting up regression equation.Preferably, can adopt multiple linear regression method to set up regression equation.
Preferably, can control this technological process in the following manner:
Utilize the spectral intensity data of technology last time, the described forecast model of substitution, the terminal time that prediction obtains this technology carries out technology controlling and process;
Perhaps, utilize the real time spectrum intensity data of this technology, the described forecast model of substitution, the terminal time that prediction obtains this technology carries out technology controlling and process;
Perhaps, utilize the real time spectrum intensity data of this technology, the described forecast model of substitution, prediction obtains the terminal time of this technology; The terminal time of online endpoint monitoring equipment being monitored according to the terminal time of being predicted carries out feedforward control, with the terminal time of resulting optimization, carries out technology controlling and process.
Preferably, this method can also comprise optimization step: the spectral intensity data of gathering this technology enter the historical data sample storehouse, adjust described forecast model in real time.
Preferably, described current technology is etching technics.
According to another embodiment of the present invention, a kind of process control system is also disclosed, comprising:
The historical data sample storehouse is used to store the historical data sample under the selected current technology, and described historical data sample comprises the spectral intensity data and the parameter to be predicted of current technology;
Forecast model makes up module, is used for according to described historical data sample, and making up with the spectral intensity data is input variable, is the forecast model of output variable with parameter to be predicted;
The technology controlling and process module is used for based on this forecast model this technological process being controlled.
Preferably, described parameter to be predicted can comprise terminal time, perhaps terminal time and processing speed.
Preferably, described forecast model makes up module employing multiple linear regression method and sets up forecast model.
Preferably, described technology controlling and process module can be controlled this technological process in the following manner: utilize the spectral intensity data of technology last time, and the described forecast model of substitution, the terminal time that prediction obtains this technology carries out technology controlling and process; Perhaps, utilize the real time spectrum intensity data of this technology, the described forecast model of substitution, the terminal time that prediction obtains this technology carries out technology controlling and process.
Preferably, this system can also comprise: online endpoint monitoring equipment is used for the terminal time of this technology of on-line monitoring; Wherein, described technology controlling and process module can be controlled this technological process in the following manner: utilize the real time spectrum intensity data of this technology, and the described forecast model of substitution, prediction obtains the terminal time of this technology; The terminal time of online endpoint monitoring equipment being monitored according to the terminal time of being predicted carries out feedforward control, with the terminal time of resulting optimization, carries out technology controlling and process.
Preferably, this system can also comprise: the feedback adjusting module, the spectral intensity data that are used to gather this technology enter the historical data sample storehouse, adjust described forecast model in real time.
Preferably, described current technology is etching technics.
Compared with prior art, the present invention has the following advantages:
The present invention makes up suitable forecast model by choosing at the same technology historical data sample of (as, current technology to be controlled), to avoid the defective of prior art simple forecast, and the perhaps error that only relies on the real time spectrum data to be brought.
Secondly, the present invention has provided multiple different control mode, and those skilled in the art can select different control models for use according to different technological standardss and control accuracy, thereby make the present invention have universality.Wherein, online endpoint monitoring equipment is carried out feedforward control scheme, can obtain best control accuracy, reach best control effect according to the terminal time of prediction.
Moreover the present invention can also adjust forecast model in real time according to the spectral intensity data of this technology, to obtain the more accurate prediction model, improves precision of prediction.
Description of drawings
Fig. 1 is the flow chart of steps of the embodiment 1 of a kind of process control method of the present invention;
Fig. 2 is the flow chart of steps of the embodiment 2 of a kind of process control method of the present invention;
Fig. 3 is the flow chart of steps of the embodiment 3 of a kind of process control method of the present invention;
Fig. 4 is the flow chart of steps of the embodiment 4 of a kind of process control method of the present invention;
Fig. 5 is the flow chart of steps of a kind of feedforward control scheme among the embodiment 4;
Fig. 6 is the synoptic diagram that predicts the outcome of terminal time;
Fig. 7 is the synoptic diagram that predicts the outcome of etch-rate;
Fig. 8 is the structured flowchart of a kind of process control system embodiment of the present invention;
Fig. 9 is the structured flowchart of the another kind of process control system embodiment of the present invention.
Embodiment
For above-mentioned purpose of the present invention, feature and advantage can be become apparent more, the present invention is further detailed explanation below in conjunction with the drawings and specific embodiments.
One of core idea of the embodiment of the invention is by choosing suitable historical data sample, and then constructs than prior art forecast model more accurately, to obtain preferable precision of prediction, improves the production rate; Can also select different controlling schemes for use with control accuracy according to different technological standardss simultaneously.
The present invention can be applied in the process control of multiple semiconductor technology, because the control of the terminal time of etching technics is a gordian technique in the semi-conductor chip production, relatively more typical, therefore, in the instructions of this patent, be that example describes, be used for control other technologies but can not limit the present invention with this technology.
With reference to figure 1, show the embodiment 1 of a kind of process control method of the present invention, specifically can may further comprise the steps:
Because in technology controlling and process, be very important for the control of terminal time, therefore, general parameter described to be predicted can comprise terminal time, perhaps described parameter to be predicted comprises simultaneously and keeps terminal time and etch rate.
Need to prove that choosing and pre-service of historical process data is the most important condition that makes up forecast model.The rationality that historical data is chosen can have influence on the accuracy and the validity of constructed forecast model to a certain extent, provides several factors that need consideration below: the rationality of (1) data sample size, definite data volume that makes up forecast model; (2) consistance of data sample type for example, is necessary for the data of same process; (3) standard normalization of data sample.
In one embodiment of the invention, chosen following data sample (because sample data is more herein, just do not show one by one in this patent, and do not list the numerical value of concrete spectral intensity data), the form of data file is as shown in table 1, and (wherein, data type ME is meant: the main etching step in the etching technics).
Table 1
Pre-service to historical data can be finished in step 101, also can finish in step 102.For example, the process that makes up forecast model described in the step 102 can comprise the standard normal step to historical data sample, and the step of setting up regression equation (forecast model a kind of).Described standard normal step can be used for choosing suitable sample data from raw sample data, to be used in the sample data conformance with standard normal distribution that makes up forecast model.
Be the process that regression equation is set up in the example explanation below with the multiple linear regression method.
At first, the spectral intensity signal of etching technics in the sample data can be formed a matrix, as the input variable in the equation; Then with the variable that needs in the etching technics to predict (as, terminal time and etch rate) as output variable, adopt the method for multiple linear regression to set up multiple linear regression equations at last about input variable and output variable.Because the method for multiple linear regression is known mathematical method, therefore, does not repeat them here for its principle and theory.
Concrete, if establishing Y is output variable, X
1, X
2..., X
7Be 7 input variables having determined, a
1, a
2..., a
7Be regression coefficient, the form of multiple linear regression equations can be expressed as so in the present embodiment:
Y＝a
1*X
1+…+a
7*X
7
According to the historical process data in the table 1, adopt the method for multiple linear regression to calculate regression coefficient, set up under the corresponding process conditions about the regression equation of process endpoint time and etch rate, as follows: T=-0.72368X
1-0.18118X
2-0.64976X
3+ 0.90223X
4+ 0.33379X
5-0.6455X
6+ 1.3278X
7R=(B-A)/T
Wherein, T is that the process endpoint function of time, R are that etch rate function, A are thicknesses of layers and the preceding thicknesses of layers of B silicon chip erosion behind the silicon chip erosion.
Need to prove that above-mentioned regression equation only is an example, those skilled in the art can by mathematical analysis, adopt other forms of multiple linear regression equations according to practical situations and accuracy requirement, can't enumerate one by one at this.
Owing to set up more accurate forecast model in the step 102, promptly can predict the terminal time of this technology and etch rate more accurately, therefore, based on above-mentioned forecast model, just can this technological process carry out preferable control, improve the production rate.
In another preferred embodiment of the present invention, also can adopt other matches (recurrence) method to set up corresponding mathematical model and The regression equation.Simply say, so-called " match " be meant known certain function some discrete function values f1, and f2 ..., fn}, (λ 1, and λ 2 by adjusting some undetermined coefficient f in this function, ..., λ 3), make difference (least square meaning) minimum of this function and known point set.
Concrete, can with the spectral intensity data processing of etching technics in the historical data sample data point, as the input variable in the equation; Then with the variable (for example, terminal time and etch rate) that needs in the etching technics to predict as corresponding output variable, adopt the method for match (recurrence) to set up the The regression equation of input variable and output variable at last.Like this, this The regression equation can be used to predict the terminal time and the etch rate of this technology equally.
Match and homing method are from being identical in essence, and match can be divided into linearity and fit (linear regression) and non-linear fitting (non-linear regression).Wherein, linear regression (match) is meant between the match variable it is linear relationship (linear function), and common linear regression method has: linear least-squares returns (OLS), multiple linear regression (MVR) and pivot recurrence (PCR) etc.Nonlinear fitting (recurrence) is meant between the match variable it is nonlinear relationship (secondary and above funtcional relationship), and common non-linear regression has fitting of a polynomial (match of monobasic high order), nonlinear least square fitting.For example,
Multiple linear regression: Y=a
0+ a
1x
1+ ... + a
kx
k
Fitting of a polynomial: Y=a
0+ a
1T+a
2t
2+ ... + a
kt
k
The present invention can select to adopt above-mentioned feasible match (recurrence) method arbitrarily based on historical data.Certainly, in these matches (recurrence) method, may there be the difference on the effect.
The terminal time that step 204, foundation are predicted carries out technology controlling and process;
Preferably, in order to guarantee the degree of accuracy of forecast model, need to adopt a large amount of process datas that model is tested, constantly revise existing model (mainly being the coefficient of revising multiple linear regression equations) according to each assay simultaneously, thereby realize the optimization of forecast model, reach the relative error minimum of prediction.
Step 301, choose the historical data sample under the current technology, described historical data sample comprises the spectral intensity data and the parameter to be predicted of etching technics;
Step 302, the described historical data sample of foundation, making up with the spectral intensity data is input variable, is the forecast model of output variable with parameter to be predicted;
Step 303, utilize the real time spectrum intensity data of this technology, the described forecast model of substitution, prediction obtains the terminal time of this technology;
The terminal time that step 304, foundation are predicted carries out technology controlling and process;
Step 305, the spectral intensity data of gathering this technology enter the historical data sample storehouse, adjust described forecast model in real time.
Concrete, with reference to Fig. 5, can obtain more optimal terminal time by following process:
The terminal time that step 501, the terminal time of relatively being predicted and online endpoint monitoring equipment are monitored;
If the gap of step 502 between the two then adopts less in two a terminal times optimization terminal time as this technology less than 3 times standard deviation sigma, carry out technology controlling and process;
If the gap of step 503 between the two, then directly adopts the terminal time the predicted optimization terminal time as this technology more than or equal to 3 times standard deviation sigma, carry out technology controlling and process.
Wherein, standard deviation sigma is calculated by historical data sample, is used to reflect the fluctuation situation of terminal time under the normal process.When the gap between terminal time of being predicted and the online terminal time of monitoring during smaller or equal to 3 times standard deviation sigma, illustrate that then current spectral intensity data are normal, be not subjected to improper interference, thereby adopt less in two terminal times one to carry out technology controlling and process, so that guarantee not can over etching; And, illustrating that then current spectral intensity data may be subjected to excessive interference when gap during greater than 3 times standard deviation sigma, there is a strong possibility is wrong for these data, thereby directly adopt the terminal time of being predicted to carry out technology controlling and process.
Need to prove that 3 times of standard deviation sigma among Fig. 5 only are examples, those skilled in the art can select other multiples or other threshold values according to actual conditions, and the present invention does not need this to be limited.
Adopt the historical data (as shown in table 1) of many group etching technics that the forecast model among the embodiment 1 is checked in this patent, it is as shown in table 2 to obtain the result; Contrast the fitting of a polynomial homing method simultaneously and set up different forecast models, dope process endpoint time and etch rate respectively, concrete outcome such as Fig. 6, shown in Figure 7.Fig. 6 is the prediction comparison diagram of terminal time, and Fig. 7 is the prediction comparison diagram of etch rate.In Fig. 6 and Fig. 7, the curve of asterisk point is an actual value, and the curve of circle points is the predicted value of MVR, the predicted value that the curve of square dot obtains for the fitting of a polynomial homing method.
Can draw from result of calculation: the method for the multiple linear regression based on historical data of the present invention (MVR) obviously is better than the fitting of a polynomial homing method, parameter prediction degree of accuracy for etching technics is very high, the relative error of process endpoint time is 0.14%, and the relative error of etch rate is 0.19%.It is that to set up the raw sample data amount of model insufficient that there is the main cause of error in analysis, so along with the continuous increase of process data, multiple linear regression model will be constantly optimised, thereby can reach higher precision of prediction.
Table 2
In a word, experimental result shows that accuracy of predicting of the present invention meets technological standards fully, can realize robotization control and feedforward control in the etching technics, thereby guarantee accuracy and the stability of process endpoint time, avoid terminal point wrong report phenomenon, improve equipment productivity.
With reference to Fig. 8, show the embodiment of a kind of process control system of the present invention, specifically comprise:
Historical data sample storehouse 801 is used to store the historical data sample under the selected current technology, and described historical data sample comprises the spectral intensity data and the parameter to be predicted of etching technics;
Forecast model makes up module 802, is used for making up with spectral intensity data input variable according to described historical data sample, and be the forecast model 803 of output variable with parameter to be predicted;
Technology controlling and process module 804 is used for based on this forecast model this technological process being controlled.
Wherein, described parameter to be predicted can be terminal time, also can comprise terminal time and etch rate.When making up forecast model, can adopt multiple linear regression method to set up forecast model, also can adopt other feasible The regression methods to set up forecast model.Certainly, it also is feasible adopting other feasible mathematical statistics methods to set up forecast model, as long as can be by the constraint on more known discrete point set M, ask for a unknown continuous function that is defined in continuous collection S (M is contained in S), get final product thereby reach the purpose of obtaining whole rule, the application can't describe one by one at this.
Concrete, described technology controlling and process module 804 can be controlled this technological process in the following manner: utilize the spectral intensity data of technology last time, and the described forecast model 803 of substitution, the terminal time that prediction obtains this technology carries out technology controlling and process.In another embodiment, described technology controlling and process module 804 also can be utilized the real time spectrum intensity data of this technology, the described forecast model 803 of substitution, and the terminal time that prediction obtains this technology carries out technology controlling and process.
Further, in order to guarantee the accuracy of forecast model 803, system shown in Figure 8 can also comprise feedback adjusting module 805, and the spectral intensity data that are used to gather this technology enter historical data sample storehouse 801, adjust described forecast model 803 in real time.
With reference to Fig. 9, show the embodiment of the another kind of process control system of the present invention, its process control procedure is more preferably, and this embodiment specifically can may further comprise the steps:
Historical data sample storehouse 901 is used to store the historical data sample under the selected current technology, and described historical data sample comprises the spectral intensity data and the parameter to be predicted of etching technics;
Forecast model makes up module 902, is used for making up with spectral intensity data input variable according to described historical data sample, and be the forecast model 903 of output variable with parameter to be predicted;
Online endpoint monitoring equipment 904 is used for the terminal time of this technology of on-line monitoring;
Technology controlling and process module 905 is used to utilize the real time spectrum intensity data of this technology, the described forecast model 903 of substitution, and prediction obtains the terminal time of this technology; The terminal time of online endpoint monitoring equipment 904 being monitored according to the terminal time of being predicted carries out feedforward control, with the terminal time of resulting optimization, carries out technology controlling and process;
Each embodiment in this instructions all adopts the mode of going forward one by one to describe, and what each embodiment stressed all is and the difference of other embodiment that identical similar part is mutually referring to getting final product between each embodiment.In addition, because Fig. 8 and device embodiment shown in Figure 9 can correspondence be applicable among the aforesaid method embodiment that so description is comparatively simple, not detailed part can be referring to the description of this instructions front appropriate section.
More than a kind of process control method provided by the present invention and system are described in detail, used specific case herein principle of the present invention and embodiment are set forth, the explanation of above embodiment just is used for helping to understand method of the present invention and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, the part that all can change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.
Claims (14)
1. a process control method is characterized in that, comprising:
Choose the historical data sample under the current technology, described historical data sample comprises the spectral intensity data and the parameter to be predicted of current technology;
According to described historical data sample, making up with the spectral intensity data is input variable, is the forecast model of output variable with parameter to be predicted;
Based on this forecast model, this technological process is controlled.
2. the method for claim 1 is characterized in that, described parameter to be predicted comprises terminal time, perhaps terminal time and processing speed.
3. the method for claim 1 is characterized in that, the process of described structure forecast model comprises the standard normal step to historical data sample, and the step of setting up regression equation.
4. method as claimed in claim 3 is characterized in that, adopts multiple linear regression method to set up regression equation.
5. method as claimed in claim 2 is characterized in that, in the following manner this technological process is controlled:
Utilize the spectral intensity data of technology last time, the described forecast model of substitution, the terminal time that prediction obtains this technology carries out technology controlling and process;
Perhaps, utilize the real time spectrum intensity data of this technology, the described forecast model of substitution, the terminal time that prediction obtains this technology carries out technology controlling and process;
Perhaps, utilize the real time spectrum intensity data of this technology, the described forecast model of substitution, prediction obtains the terminal time of this technology; The terminal time of online endpoint monitoring equipment being monitored according to the terminal time of being predicted carries out feedforward control, with the terminal time of resulting optimization, carries out technology controlling and process.
6. method as claimed in claim 5 is characterized in that, also comprises:
The spectral intensity data of gathering this technology enter the historical data sample storehouse, adjust described forecast model in real time.
7. the method for claim 1 is characterized in that, described current technology is etching technics.
8. a process control system is characterized in that, comprising:
The historical data sample storehouse is used to store the historical data sample under the selected current technology, and described historical data sample comprises the spectral intensity data and the parameter to be predicted of current technology;
Forecast model makes up module, is used for according to described historical data sample, and making up with the spectral intensity data is input variable, is the forecast model of output variable with parameter to be predicted;
The technology controlling and process module is used for based on this forecast model this technological process being controlled.
9. system as claimed in claim 8 is characterized in that, described parameter to be predicted comprises terminal time, perhaps terminal time and processing speed.
10. system as claimed in claim 8 is characterized in that, described forecast model makes up module and adopts multiple linear regression method to set up forecast model.
11. system as claimed in claim 9 is characterized in that, described technology controlling and process module is controlled this technological process in the following manner:
Utilize the spectral intensity data of technology last time, the described forecast model of substitution, the terminal time that prediction obtains this technology carries out technology controlling and process;
Perhaps, utilize the real time spectrum intensity data of this technology, the described forecast model of substitution, the terminal time that prediction obtains this technology carries out technology controlling and process.
12. system as claimed in claim 9 is characterized in that, also comprises:
Online endpoint monitoring equipment is used for the terminal time of this technology of on-line monitoring;
Described technology controlling and process module is controlled this technological process in the following manner:
Utilize the real time spectrum intensity data of this technology, the described forecast model of substitution, prediction obtains the terminal time of this technology; The terminal time of online endpoint monitoring equipment being monitored according to the terminal time of being predicted carries out feedforward control, with the terminal time of resulting optimization, carries out technology controlling and process.
13. as claim 11 or 12 described systems, it is characterized in that, also comprise:
The feedback adjusting module, the spectral intensity data that are used to gather this technology enter the historical data sample storehouse, adjust described forecast model in real time.
14. system as claimed in claim 8 is characterized in that, described current technology is etching technics.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008102398338A CN101751026B (en) | 2008-12-12 | 2008-12-12 | Method and device for controlling process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008102398338A CN101751026B (en) | 2008-12-12 | 2008-12-12 | Method and device for controlling process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101751026A true CN101751026A (en) | 2010-06-23 |
| CN101751026B CN101751026B (en) | 2012-12-26 |
Family
ID=42478092
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2008102398338A Active CN101751026B (en) | 2008-12-12 | 2008-12-12 | Method and device for controlling process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101751026B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103605323A (en) * | 2013-08-09 | 2014-02-26 | 中国蓝星（集团）股份有限公司 | Discrete control method and device for chemical industry production |
| CN104239591A (en) * | 2013-06-24 | 2014-12-24 | 上海华虹宏力半导体制造有限公司 | Virtualization electrical characteristic optimization test (VTRIM) method |
| CN109490156A (en) * | 2018-11-29 | 2019-03-19 | 北京康仁堂药业有限公司 | The method of quantitative forecast Chinese medicinal granule mixed process terminal time |
| CN110357136A (en) * | 2019-07-17 | 2019-10-22 | 中铝视拓智能科技有限公司 | A kind of aluminium oxide calcining control method and equipment |
| CN110581208A (en) * | 2019-08-21 | 2019-12-17 | 华南理工大学 | Intelligent quality control method for LED packaging process |
| WO2020034092A1 (en) * | 2018-08-14 | 2020-02-20 | 西门子股份公司 | Process controller and method and system therefor |
| CN110850812A (en) * | 2019-11-18 | 2020-02-28 | 北京邮电大学 | Ion beam etching rate control method and device based on model |
| CN111356552A (en) * | 2018-10-26 | 2020-06-30 | 合刃科技（深圳）有限公司 | Laser processing parameter and reflection spectrum prediction model training method and device |
| CN112269963A (en) * | 2020-10-27 | 2021-01-26 | 南开大学 | Polynomial fitting nano random surface optical characteristic rapid prediction algorithm |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100568448C (en) * | 2007-01-12 | 2009-12-09 | 北京北方微电子基地设备工艺研究中心有限责任公司 | A kind of etching terminal checkout gear and method of plasma etching equipment |
| CN101226401A (en) * | 2007-01-16 | 2008-07-23 | 北京北方微电子基地设备工艺研究中心有限责任公司 | Analysis method for technique sensitiveness during plasma etching industrial process |
-
2008
- 2008-12-12 CN CN2008102398338A patent/CN101751026B/en active Active
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104239591A (en) * | 2013-06-24 | 2014-12-24 | 上海华虹宏力半导体制造有限公司 | Virtualization electrical characteristic optimization test (VTRIM) method |
| CN103605323B (en) * | 2013-08-09 | 2016-03-30 | 蓝星（北京）技术中心有限公司 | The Discrete Control Method of Chemical Manufacture and device |
| CN103605323A (en) * | 2013-08-09 | 2014-02-26 | 中国蓝星（集团）股份有限公司 | Discrete control method and device for chemical industry production |
| CN112424702B (en) * | 2018-08-14 | 2024-03-08 | 西门子股份公司 | Process controller and method and system thereof |
| CN112424702A (en) * | 2018-08-14 | 2021-02-26 | 西门子股份公司 | Process controller, method and system thereof |
| WO2020034092A1 (en) * | 2018-08-14 | 2020-02-20 | 西门子股份公司 | Process controller and method and system therefor |
| CN111356552A (en) * | 2018-10-26 | 2020-06-30 | 合刃科技（深圳）有限公司 | Laser processing parameter and reflection spectrum prediction model training method and device |
| CN109490156A (en) * | 2018-11-29 | 2019-03-19 | 北京康仁堂药业有限公司 | The method of quantitative forecast Chinese medicinal granule mixed process terminal time |
| CN109490156B (en) * | 2018-11-29 | 2021-04-02 | 北京康仁堂药业有限公司 | Method for quantitatively predicting end point time of traditional Chinese medicine formula granule mixing process |
| CN110357136A (en) * | 2019-07-17 | 2019-10-22 | 中铝视拓智能科技有限公司 | A kind of aluminium oxide calcining control method and equipment |
| CN110581208A (en) * | 2019-08-21 | 2019-12-17 | 华南理工大学 | Intelligent quality control method for LED packaging process |
| CN110581208B (en) * | 2019-08-21 | 2021-06-08 | 华南理工大学 | Intelligent quality control method for LED packaging process |
| CN110850812A (en) * | 2019-11-18 | 2020-02-28 | 北京邮电大学 | Ion beam etching rate control method and device based on model |
| CN112269963A (en) * | 2020-10-27 | 2021-01-26 | 南开大学 | Polynomial fitting nano random surface optical characteristic rapid prediction algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101751026B (en) | 2012-12-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101751026B (en) | Method and device for controlling process | |
| CN101853008B (en) | Method and system for tuning advanced process control parameters | |
| US7894927B2 (en) | Using Multi-Layer/Multi-Input/Multi-Output (MLMIMO) models for metal-gate structures | |
| US8606387B2 (en) | Adaptive and automatic determination of system parameters | |
| JP3957964B2 (en) | Lot dispatching method and lot dispatching system | |
| KR100847368B1 (en) | Method and apparatus for controlling etch selectivity | |
| US8019458B2 (en) | Creating multi-layer/multi-input/multi-output (MLMIMO) models for metal-gate structures | |
| US8108060B2 (en) | System and method for implementing a wafer acceptance test (“WAT”) advanced process control (“APC”) with novel sampling policy and architecture | |
| TW201432479A (en) | Method for searching and analyzing process parameters and computer program product thereof | |
| US7542880B2 (en) | Time weighted moving average filter | |
| US7840298B2 (en) | Method and system for advanced process control using measurement uncertainty as control input | |
| CN101207004A (en) | Method for controlling semiconductor silicon dies etching technique | |
| US20120215490A1 (en) | Method for using real-time apc information for an enhanced lot sampling engine | |
| CN102299044A (en) | System and method for data mining and feature tracking for fab-wide prediction and control | |
| US9588505B2 (en) | Near non-adaptive virtual metrology and chamber control | |
| CN101937835B (en) | Method for modifying parameters for fuse-class wafer | |
| Ruegsegger et al. | Feedforward control for reduced run-to-run variation in microelectronics manufacturing | |
| US8244500B2 (en) | Method of adjusting wafer processing sequence | |
| US20100241250A1 (en) | Feedback and feedforward control of a semiconductor process without output values from upstream processes | |
| US10811325B2 (en) | Self-healing semiconductor wafer processing | |
| US7502658B1 (en) | Methods of fabricating optimization involving process sequence analysis | |
| US8406911B2 (en) | Implementing sequential segmented interleaving algorithm for enhanced process control | |
| CN108268987B (en) | Method for estimating quality of various products | |
| Tseng et al. | An initial intercept iteratively adjusted (IIIA) controller: an enhanced double EWMA feedback control scheme | |
| KR20210071935A (en) | Search device, search program and plasma processing unit |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address | ||
| CP03 | Change of name, title or address |
Address after: 100176 Beijing economic and Technological Development Zone, Wenchang Road, No. 8, No. Patentee after: Beijing North China microelectronics equipment Co Ltd Address before: 100016, building 2, block M5, No. 1 East Jiuxianqiao Road, Beijing, Chaoyang District Patentee before: Beifang Microelectronic Base Equipment Proces Research Center Co., Ltd., Beijing |