# US 11,907,235 B2 · Plasma processing apparatus including predictive control

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 11,907,235 B2
- Document ID: US11907235B2
- Status: Granted
- Kind: B2
- Application number: 17/241,826
- Filed: 2021-04-27
- Published: 2021-08-26
- Granted: 2024-02-20
- Earliest claimed: 2018-03-14
- Assignees: Hitachi High-Tech Corporation
- Applicants: Hitachi High-Tech Corporation
- Inventors: Yutaka Okuyama; Takeshi Ohmori; Masaru Kurihara; Hyakka Nakada
- Web page: https://patentgrep.com/patent/US11907235B2

## CPC

- G06F 16/2462: {Approximate or statistical queries}
- G05B 13/04: involving the use of models or simulators
- G05B 13/042: {in which a parameter or coefficient is automatically adjusted to optimise the performance}
- G05B 13/048: {using a predictor}
- G06F 3/04883: for inputting data by handwriting, e.g. gesture or text
- G06F 16/90335: {Query processing}
- G06F 16/9535: Search customisation based on user profiles and personalisation
- G06N 3/04: Architecture, e.g. interconnection topology
- G06N 3/045: Combinations of networks
- G06N 3/0499: Feedforward networks
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- G06N 5/04: Inference or reasoning models
- G06N 20/00: Machine learning
- G06N 20/10: using kernel methods, e.g. support vector machines [SVM]
- H01J 37/3211: {Antennas, e.g. particular shapes of coils}
- H10P 72/0612: {Production flow monitoring, e.g. for increasing throughput}
- G05B 13/02: electric
- H01J 2237/327: Arrangements for generating the plasma
- H01J 2237/334: Etching
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- H01J: ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS (spark-gaps H01T; electric arc lamps H05B31/00; particle accelerators H05H)
- H10P: GENERIC PROCESSES OR APPARATUS FOR THE MANUFACTURE OR TREATMENT OF DEVICES COVERED BY CLASS H10

## Abstract

A model learning unit learns a prediction model on the basis of learning data, a target setting unit sets a target output parameter value by interpolating between a goal output parameter value and an output parameter value which is the closest to the goal output parameter value in output parameter values in the learning data, a processing condition search unit estimates input parameter values which corresponds to the goal output parameter value and the target output parameter value, a model learning unit updates the prediction model by using a set of the estimated input parameter value and an output parameter value which is a result of processing that a processing device performs as additional learning data.

## Claims

### Claim 1

A plasma processing apparatus which performs plasma processing on a sample using a plasma, comprising:
a processing chamber;
a plasma generation device which generates the plasma in the processing chamber; and
a control device which searches for an input parameter value of each control parameter of the apparatus which corresponds to a goal output parameter value of the sample and controls the plasma generation device to perform the plasma processing on the sample which is installed in the processing chamber on the basis of the searched-out input parameter value,
wherein the control device
learns a prediction model on the basis of learning data which is sets of predetermined input parameter values and output parameter values which are results of plasma processing performed on the sample on the basis of the predetermined input parameter values,
sets a target output parameter value by interpolating values between the goal output parameter value and an output parameter value which is the closest to the goal output parameter value in the output parameter values in the learning data,
estimates input parameter values which correspond to the goal output parameter value and the target output parameter value by using the prediction model,
in a case where it is decided that the output parameter value which is the result of the plasma processing performed on the sample on the basis of the input parameter value which is estimated by using the prediction model and corresponds to the goal output parameter value converges to the goal output parameter value, controls the plasma generation device to perform the plasma processing on the sample on the basis of the input parameter value with which convergence thereof is decided, and
in a case where it is decided that the output parameter value which is the result of the plasms processing performed on the sample on the basis of the input parameter value which is estimated by using the prediction model and corresponds to the goal output parameter value does not converge to the goal output parameter value, updates the prediction model by adding a set of the estimated input parameter value and the output parameter value which is the result of the plasma processing performed on the sample on the basis of the input parameter value to the learning data and resets the target output parameter value and thereby continues searching-for of an input parameter value which corresponds to the goal output parameter value.

### Claim 2 (depends on 1)

The plasma processing apparatus according to claim 1,
wherein the control device
sets a plurality of initial input parameter values in such a manner that each of input parameters used in the prediction model includes three level values on the basis of input parameter value settable ranges and
the prediction model is learned by using a set of the initial input parameter value and an initial output parameter value which is a result of the plasma processing performed on the sample on the basis of the initial input parameter value as initial learning data.

### Claim 3 (depends on 1)

The plasma processing apparatus according to claim 1, wherein the goal output parameter value is given as a dimension of a processed shape of the sample obtained after plasma-processed.