IPC분류정보
국가/구분 |
United States(US) Patent
등록
|
국제특허분류(IPC7판) |
|
출원번호 |
US-0003829
(2004-12-06)
|
등록번호 |
US-7505879
(2009-03-17)
|
우선권정보 |
JP-2002-163869(2002-06-05); JP-2002-168653(2002-06-10) |
발명자
/ 주소 |
- Tomoyasu,Masayuki
- Oh,Hin
- Tanaka,Hideki
|
출원인 / 주소 |
|
대리인 / 주소 |
Finnegan, Henderson, Farabow, Garrett & Dunner, LLP
|
인용정보 |
피인용 횟수 :
35 인용 특허 :
13 |
초록
▼
According to the present invention, multivariate analysis model expressions are generated for a plasma processing apparatus 100A and a plasma processing apparatus 100B by executing a multivariate analysis of detection data provided by a plurality of sensors included in each plasma processing apparat
According to the present invention, multivariate analysis model expressions are generated for a plasma processing apparatus 100A and a plasma processing apparatus 100B by executing a multivariate analysis of detection data provided by a plurality of sensors included in each plasma processing apparatus when the plasma processing apparatuses 100A and 100B operate based upon first setting data. Then, when the plasma processing apparatus 100A operates based upon new second setting data, detection data provided by the plurality of sensors in the plasma processing apparatus 100A are used to generate a corresponding multivariate analysis model expression, and by using the new multivariate analysis model expression corresponding to the plasma processing apparatus 100A generated based upon the second setting data and to the plasma processing apparatus 100B, a multivariate analysis model expression corresponding to the new second setting data is generated four the plasma processing apparatus 100B.
대표청구항
▼
What is claimed is: 1. A method for generating a multivariate analysis model expression for a processing apparatus, the method comprising: a first step in which a recurrent matrix of detection data provided by a plurality of sensors in a processing apparatus operating with first setting data is asc
What is claimed is: 1. A method for generating a multivariate analysis model expression for a processing apparatus, the method comprising: a first step in which a recurrent matrix of detection data provided by a plurality of sensors in a processing apparatus operating with first setting data is ascertained through multivariate analysis for each processing apparatus among a plurality of processing apparatuses that operate based upon the first setting data, the plurality of processing apparatuses including a processing apparatus designated as a reference processing apparatus and a first processing apparatus that is not the reference processing apparatus; a second step in which a recurrent matrix of detection data provided by the plurality of sensors in the reference processing apparatus is ascertained through multivariate analysis when the reference processing apparatus operates based upon new second setting data with the second setting data; a third step in which a recurrent matrix of detection data with the second setting data in the first processing apparatus is ascertained based upon the recurrent matrix determined through the first step for the first processing apparatus, the recurrent matrix determined through the first step for the reference processing apparatus and the recurrent matrix determined through the second step for the reference processing apparatus, wherein the third step does not use detection data obtained from the first processing apparatus operating with the second setting data; a fourth step in which a multivariate analysis model expression to be used to evaluate an apparatus state of the first processing apparatus or to predict processing results in the first processing apparatus is generated based upon the recurrent matrix ascertained through the third step; and a fifth step of storing the multivariate analysis model expression in a computer-readable storage medium. 2. A method for generating a multivariate analysis model expression for a processing apparatus according to claim 1, wherein: in the third step, the third recurrent matrix between the second setting data and the detection data in the first processing apparatus is determined based upon a proportional relationship between the recurrent matrix of the detection data with the second setting data in the first processing apparatus relative to the recurrent matrix ascertained through the first step for the first processing apparatus and the recurrent matrix ascertained through the second step for the reference processing apparatus relative to the recurrent matrix ascertained through the first step for the reference processing apparatus. 3. A method for generating a multivariate analysis model expression for a processing apparatus according to claim 1, wherein: the multivariate analysis is executed by adopting a method of partial least squares. 4. A method for generating a multivariate analysis model expression for a processing apparatus according to claim 1, wherein: the processing apparatuses are plasma processing apparatuses. 5. A method for generating a multivariate analysis model expression for a processing apparatus according to claim 1, wherein: the processing apparatuses are plasma processing apparatuses; and the setting data are constituted with a plurality of control parameters with which the state of the plasma can be controlled and the detection data are constituted with at least one parameter or two or more parameters selected from a group of parameters including a plurality of plasma state parameters indicating the plasma state, a plurality of apparatus state parameters related to the apparatus state and parameters that reflect the processing results. 6. A method for generating a multivariate analysis model expression for a processing apparatus according to claim 1, wherein: the processing apparatuses are plasma processing apparatuses; and the setting data are constituted with a plurality of control parameters with which the state of the plasma can be controlled and the detection data are constituted with at least one parameter or two or more parameters selected from a group of parameters including a plurality of plasma state parameters indicating the plasma state, a plurality of apparatus state parameters related to the apparatus state and parameters that reflect the processing results; and the multivariate analysis model expression is a correlational expression indicating a correlation of detection data calculated based upon the recurrent matrix determined through the third step for the first processing apparatus with the second setting data. 7. A control device of a processing apparatus, included in the processing apparatus which processes a workpiece to control the processing apparatus based upon specific setting data, comprising: a transmission/reception section capable of data exchange through a connection with a network with which the processing apparatus and a host apparatus at least are connected, wherein: the control device transmits the first setting data and detection data provided by a plurality of sensors in the processing apparatus operating based upon the first setting data to the host apparatus via the network by engaging the transmission/reception section in operation, receives a recurrent matrix between the first setting data and the detection data ascertained by the host apparatus through a multivariate analysis from the host apparatus via the network by engaging the transmission/reception section in operation; the control device transmits new second setting data to the host apparatus via the network by engaging the transmission/reception section and receives a recurrent matrix between the second setting data and detection data corresponding to the second setting data, which is ascertained through multivariate analysis by the host apparatus based upon the transmitted data via the network by engaging the transmission/reception section in operation; and the control device generates a multivariate analysis model expression based upon the recurrent matrix with regard to the second setting data received from the host apparatus, evaluates the apparatus state of the processing apparatus or predicts the processing results to be achieved in the processing apparatus based upon the multivariate analysis model expression and controls the processing apparatus based upon the results of the evaluation or the prediction; wherein the recurrent matrix with regard to the second setting data in the processing apparatus is calculated by the host apparatus based upon the recurrent matrix with regard to the first setting data in the processing apparatus determined by the host apparatus through multivariate analysis, a recurrent matrix of detection data provided by a plurality of sensors in a reference processing apparatus operating based upon the first setting data with the first setting data, which is determined by the host apparatus through multivariate analysis, and a recurrent matrix of detection data provided by the plurality of sensors in the reference processing apparatus operating based upon the second setting data with the second setting data, which is determined by the host apparatus through multivariate analysis, wherein calculating the recurrent matrix with regard to the second setting data in the processing apparatus does not use detection data obtained from the processing apparatus operating with the second setting data. 8. A control device of a processing apparatus according to claim 7, wherein: the multivariate analysis is executed by adopting a method of partial least squares. 9. A control device of a processing apparatus according to claim 7, wherein: the processing apparatus is a plasma processing apparatus. 10. A control device of a processing apparatus according to claim 7, wherein: the processing apparatus is a plasma processing apparatus; and the setting data are constituted with a plurality of control parameters with which the state of the plasma can be controlled and the detection data is constituted with at least one parameter or two or more parameters selected from a group of parameters including a plurality of plasma state parameters indicating the plasma state, a plurality of apparatus state parameters related to the apparatus state and parameters that reflect the processing results. 11. A control system for processing apparatuses that process workpieces, each having a control device that controls the processing apparatus based upon specific setting data, comprising: a plurality of processing apparatuses each connected to a network via a transmission/reception section; and a host apparatus connected to the network, wherein: upon receiving first setting data and detection data provided by a plurality of sensors in each processing apparatus among the plurality of processing apparatuses operating based upon the first setting data from the plurality of processing apparatuses via the network, the host apparatus ascertains a recurrent matrix between the first setting data and the detection data through multivariate analysis for each processing apparatus and transmits the recurrent matrix to its corresponding processing apparatus via the network; upon receiving new second setting data and detection data provided by the plurality of sensors in one of the processing apparatuses designated as a reference processing apparatus operating based upon the second setting data from the reference processing apparatus via the network, the host apparatus ascertains a recurrent matrix between the second setting data and the detection data having been received through multivariate analysis and transmits the recurrent matrix thus ascertained to the reference processing apparatus via the network; upon receiving the second setting data from an other processing apparatus that is not the reference processing apparatus via the network, the host apparatus ascertains a recurrent matrix between the received second setting data and detection data corresponding to the second setting data based upon the recurrent matrix with regard to the first setting data in the other processing apparatus ascertained through the multivariate analysis, the recurrent matrix with regard to the first setting data in the reference processing apparatus ascertained through the multivariate analysis and the recurrent matrix with regard to the second setting data in the reference processing apparatus having been ascertained through the multivariate analysis and transmits the recurrent matrix thus ascertained to the other processing apparatus via the network, wherein ascertaining the recurrent matrix between the received second setting data and detection data corresponding to the second setting data based upon the recurrent matrix with regard to the first setting data in the other processing apparatus does not use the detection data obtained from the other processing apparatus operating with the second setting data; and wherein the other processing apparatus generates a multivariate analysis model expression based upon the recurrent matrix with regard to the second setting data received from the host apparatus, the apparatus state of the processing apparatus is evaluated or the processing results to be achieved in the processing apparatus are predicted based upon the multivariate analysis model expression, and the processing apparatus is controlled based upon the evaluation results or the prediction results. 12. A control system for processing apparatuses according to claim 11, wherein: the multivariate analysis is executed by adopting a method of partial least squares. 13. A control system for processing apparatuses according to claim 11, wherein: the processing apparatuses are plasma processing apparatuses. 14. A control system for processing apparatuses according to claim 11, wherein: the processing apparatuses are plasma processing apparatuses; and the setting data are constituted with a plurality of control parameters with which the state of the plasma can be controlled and the detection data may be constituted with at least one parameter or two or more parameters selected from a group of parameters including a plurality of plasma state parameters indicating the plasma state, a plurality of apparatus state parameters related to the apparatus state and parameters that reflect the processing results.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.