Heating operation process control

Patent metadata

Field	Value
Publication number	US-2022219411-A1
Application number	US-202117505875-A
Country	US
Kind code	A1
Filing date	Oct 20, 2021
Priority date	Jan 13, 2021
Publication date	Jul 14, 2022
Grant date	—

Abstract

Heating operation control includes obtaining sensor data indicating measured temperatures within a heating vessel during a heating operation; determining sets of thermal stack parameters. Each set of candidate thermal stack parameters is descriptive of a respective configuration of a thermal stack modeled by a first machine learning model to generate one or more estimated tool temperature values. The thermal stack includes the tool and a part coupled to the tool. Heating operation control also includes determining a temperature profile for the heating operation. The temperature profile is determined, via a second machine learning model, based on the plurality of sets of thermal stack parameters and one or more process specifications of the thermal stack.

First claim

What is claimed is:

1 . A method comprising:
obtaining sensor data indicating measured temperatures within a heating vessel during a first portion of a heating operation, wherein the sensor data includes tool temperature values and interior temperature values, wherein a tool temperature value represents a temperature measurement of a portion of a tool within the heating vessel, and wherein an interior temperature value represents a temperature measurement of ambient conditions within the heating vessel; determining a plurality of sets of thermal stack parameters from a plurality of sets of candidate thermal stack parameters, wherein each set of candidate thermal stack parameters is descriptive of a respective configuration of an in-process thermal stack modeled by a first machine learning model to generate one or more estimated tool temperature values, and wherein the in-process thermal stack comprises the tool and a part coupled to the tool; and determining a temperature profile for a second portion of the heating operation, wherein the temperature profile is determined, via a second machine learning model, based on the plurality of sets of thermal stack parameters and one or more process specifications of the in-process thermal stack.

2 . The method of claim 1 , wherein the temperature profile indicates one or more target interior temperature values, one or more interior temperature change rates, one or more dwell times associated with a particular target interior temperature value, or a combination thereof.

3 . The method of claim 1 , wherein the heating operation facilitates exothermic curing of one or more materials of the part.

4 . The method of claim 1 , further comprising, for each set of candidate thermal stack parameters from the plurality of sets of candidate thermal stack parameters:
providing input to the first machine learning model, wherein the input indicates the set of candidate thermal stack parameters and a time sequence of the interior temperature values; and obtaining output from the first machine learning model, wherein the output indicates one or more estimated tool temperature values based on the input.

5 . The method of claim 4 , wherein said determining the plurality of sets of thermal stack parameters from the plurality of sets of candidate thermal stack parameters comprises selecting, as the plurality of sets of thermal stack parameters, a subset of the plurality of sets of candidate thermal stack parameters for which the one or more estimated tool temperature values most closely match the tool temperature values indicated by the sensor data.

6 . The method of claim 1 , wherein each set of candidate thermal stack parameters indicates values of one or more boundary conditions, a value of a thickness of the part, a value of a thickness of the tool, one or more temperature values associated with the tool, or some combination thereof.

7 . The method of claim 6 , wherein the values of one or more boundary conditions comprise values of a first heat transfer coefficient at a first surface of the in-process thermal stack and of a second heat transfer coefficient at a second surface of the in-process thermal stack.

8 . The method of claim 1 , further comprising determining the plurality of sets of candidate thermal stack parameters based on one or more specified start values, one or more specified step values, and one or more specified stop values.

9 . The method of claim 1 , wherein said determining the temperature profile for the second portion of the heating operation comprises:
obtaining a plurality of candidate temperature profiles; providing multiple combinations of inputs to the second machine learning model, wherein each combination of input to the second machine learning model includes a respective set of thermal stack parameters from the plurality of sets of thermal stack parameters and a respective candidate temperature profile from the plurality of candidate temperature profiles; obtaining output from the second machine learning model for each combination of input of the multiple combinations of input, wherein the output for a particular combination of input indicates whether operating the heating vessel based on parameters indicated by the particular combination of input is expected to satisfy the one or more process specifications; and selecting as the temperature profile a particular candidate temperature profile from the plurality of candidate temperature profiles associated with a combination of input that is expected to satisfy the one or more process specifications.

10 . The method of claim 9 , wherein said selecting the particular candidate temperature profile comprises:
identifying a subset of candidate temperature profiles from among the plurality of candidate temperature profiles, wherein each candidate temperature profile of the subset of candidate temperature profiles is associated with a combination of input that is expected to satisfy the one or more process specifications; and selecting the particular candidate temperature profile from among the subset of candidate temperature profiles based on a selection criterion.

11 . The method of claim 10 , wherein the selection criterion specifies selection of a candidate temperature profile based on an associated time duration of the second portion of the heating operation, an associated peak temperature of the second portion of the heating operation, or a combination thereof.

12 . The method of claim 9 , wherein said obtaining the plurality of candidate temperature profiles comprises determining the plurality of candidate temperature profiles based on one or more specified start values, one or more specified step values, and one or more specified stop values.

13 . The method of claim 1 , wherein multiple in-process thermal stacks are disposed in the heating vessel during the heating operation, and wherein the method further comprises:
obtaining sensor data for each of the multiple in-process thermal stacks; determining a plurality of sets of thermal stack parameters for each of the multiple in-process thermal stacks; and determining the temperature profile for the second portion of the heating operation based on the plurality of sets of thermal stack parameters for each of the multiple in-process thermal stacks and based on one or more process specifications for each of the multiple in-process thermal stacks.

14 . The method of claim 13 , wherein the multiple in-process thermal stacks correspond to two or more cross-sections of the part, two or more cross-sections of the tool, two or more parts, two or more tools, or some combination thereof.

15 . The method of claim 13 , further comprising selecting the second machine learning model from among a plurality of available second machine learning models based on materials of the multiple in-process thermal stacks and the one or more process specifications of the multiple in-process thermal stacks.

16 . The method of claim 1 , further comprising selecting the first machine learning model, the second machine learning model, or both, from among a plurality of available machine learning models based, at least in part, on one or more materials of the in-process thermal stack.

17 . A non-transient, computer-readable medium storing instructions executable by one or more processors to perform operations comprising:
obtaining sensor data indicating measured temperatures within a heating vessel during a first portion of a heating operation, wherein the sensor data includes tool temperature values and interior tempe

Assignees

• Boeing Co

Inventors

• Humfeld Keith D
• Butler Geoffrey A
• Nelson Karl M

Classifications

• G06N3/08

  Learning methods · CPC title

• G06N3/0442

  characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU] · CPC title

• G06N3/09

  Supervised learning · CPC title

• B29C2035/046

  dried air · CPC title

• B29C35/0288 Primary

  Controlling heating or curing of polymers during moulding, e.g. by measuring temperatures or properties of the polymer and regulating the process (controlling or regulating chemical, physical or physico- chemical processes in general B01J19/0006) · CPC title