Numerical control system that detects an abnormality in an operation state

Patent metadata

Field	Value
Publication number	US-11080610-B2
Application number	US-201816141133-A
Country	US
Kind code	B2
Filing date	Sep 25, 2018
Priority date	Sep 29, 2017
Publication date	Aug 3, 2021
Grant date	Aug 3, 2021

Abstract

A numerical control system detects a state amount indicating an operation state of a machine tool, creates a characteristic amount that characterizes the state of a machining operation from the detected state amount, infers an evaluation value of the operation state of the machine tool from the characteristic amount, and detects an abnormality in the operation state of the machine tool on the basis of the inferred evaluation value. The numerical control system generates and updates a learning model by machine learning that uses the characteristic amount, and stores the learning model in correlation with a combination of conditions of the machining operation of the machine tool.

First claim

The invention claimed is:

1. A numerical control system that detects an abnormality in an operation state of a machine tool that machines a workpiece, the numerical control system comprising a processor configured to:
output a condition of a machining operation of the machine tool;
detect a state amount indicating an operation state of the machine tool;
create a characteristic amount, including data that characterizes the operation state of the machine tool, from the detected state amount;
infer an evaluation value of the operation state of the machine tool, from the state amount and the characteristic amount;
detect an abnormality in the operation state of the machine tool on the basis of the evaluation value;
generate and update a learning model by machine learning that uses the state amount and the characteristic amount; and
store at least one learning model generated by the processor in correlation with a combination of the conditions output by the processor and a use condition including an inference process including the characteristic amount or a processing ability necessary for use of the learning model, wherein
the processor computes the evaluation value of the operation state of the machine tool by selectively using at least one learning model among the stored learning models on the basis of the output condition of the machining operation and the detected state amount and characteristic amount of the inference process or the processing ability of the processor,
wherein the evaluation value includes information indicating a classification of normal/abnormal of the operation state, a location of the abnormality, and a distance between a current operation state of the machine tool and operation states of the machine tool in a normal state.

2. The numerical control system according to claim 1 , wherein
the processor generates a new learning model by altering an existing stored learning model.

3. The numerical control system according to claim 1 , wherein the processor encrypts and stores the generated or updated learning model and decrypts the encrypted learning model when the learning model is read by the processor.

4. A numerical control system that detects an abnormality in an operation state of a machine tool that machines a workpiece, the numerical control system comprising a processor configured to:
output a condition of a machining operation of the machine tool;
detect a state amount indicating an operation state of the machine tool;
create a characteristic amount, including data that characterizes the operation state of the machine tool, from the detected state amount;
infer an evaluation value of the operation state of the machine tool, from the state amount and the characteristic amount;
detect an abnormality in the operation state of the machine tool on the basis of the evaluation value; and
store at least one learning model which is correlated in advance with a combination of output conditions of the machining operation of the machine tool and a use condition including an inference process including the characteristic amount or a processing ability necessary for use of the learning model, wherein
the processor computes the evaluation value of the operation state of the machine tool by selectively using the at least one stored learning model, on the basis of the output condition of the machining operation and the detected state amount and characteristic amount of the inference process or a processing ability of the processor,
wherein the evaluation value includes information indicating a classification of normal/abnormal of the operation state, a location of the abnormality, and a distance between a current operation state of the machine tool and operation states of the machine tool in a normal state.

5. A method for detecting an abnormality in an operation state, the method comprising:
outputting a condition of a machining operation of a machine tool that machines a workpiece;
detecting a state amount indicating an operation state of the machine tool;
creating a characteristic amount that characterizes the operation state of the machine tool from the state amount;
inferring an evaluation value of the operation state of the machine tool from the state amount and the characteristic amount;
detecting an abnormality in the operation state of the machine tool on the basis of the evaluation value; and
generating and updating a learning model by machine learning that uses the state amount and the characteristic amount, wherein
in the inferring, selecting a learning model to be used, on the basis of the output condition of the machining operation and the detected state amount and characteristic amount of an inference process or processing ability of the processor, among at least one learning model correlated in advance with a combination of conditions of the machining operation of the machine tool and the inference process or the processing ability, and using the selected learning model to compute the evaluation value of the operation state of the machine tool,
wherein the evaluation value includes information indicating a classification of normal/abnormal of the operation state, a location of the abnormality, and a distance between a current operation state of the machine tool and operation states of the machine tool in a normal state.

6. A method for detecting an abnormality in an operation state, the method comprising:
outputting a condition of a machining operation of a machine tool that machines a workpiece;
detecting a state amount indicating an operation state of the machine tool;
creating a characteristic amount that characterizes the operation state of the machine tool from the state amount;
inferring an evaluation value of the operation state of the machine tool from the state amount and the characteristic amount; and
detecting an abnormality in the operation state of the machine tool on the basis of the evaluation value, wherein
in the inferring, a learning model to be used is selected, on the basis of the output condition of the machining operation and the detected state amount and characteristic amount of an inference process or processing ability of the processor, among at least one learning model correlated in advance with a combination of conditions of the machining operation of the machine tool and the inference process or the processing ability, and using the selected learning model to compute the evaluation value of the operation state of the machine tool,
wherein the evaluation value includes information indicating a classification of normal/abnormal of the operation state, a location of the abnormality, and a distance between a current operation state of the machine tool and operation states of the machine tool in a normal state.

Assignees

• Fanuc Corp

Inventors

• Iijima Kazunori
• Satou Kazuhiro
• Kamiya Yohei

Classifications

• G05B2219/34013

  Servocontroller · CPC title

• G05B19/4083

  Adapting program, configuration · CPC title

• G05B19/4063 Primary

  Monitoring general control system (G05B19/4062 takes precedence) · CPC title

• G06N5/045

  Explanation of inference; Explainable artificial intelligence [XAI]; Interpretable artificial intelligence · CPC title

• G06N5/04 Primary

  Inference or reasoning models · CPC title