Failure prediction device and machine learning device
AssigneeFanuc Corporation
InventorYuu Oota
Life of this patent
Estimating the term…
Abstract
A failure prediction device is provided with a machine learning device configured to learn the state of a brake of a motor with respect to data on the brake. The machine learning device observes brake operating state data indicative of an operating state of the brake when the brake is in a normal state, as state variables representative of a current environmental state, and uses the observed state variables to learn a distribution of the state variables with the brake in the normal state.
Claims15
A machine learning device configured to learn a state of a brake of a motor based on data on the brake, the machine learning device comprising:
- a state observation unit configured to observe brake operating state data indicative of an operating state of the brake when the brake is in a normal state, as state variables representative of a current environmental state;
- a learning unit configured to
- perform a cluster analysis based on the state variables to learn a distribution of the state variables with the brake in the normal state, and
- record a cluster learned by the cluster analysis as a learned model configured to predict a failure state of the brake of the motor,
- wherein an operation of a machine having the motor is controlled to be stopped in response to a prediction of the failure state of the brake based on the learned model, and
- wherein the brake operating state data includes at least one of
- 1) data on a gravity load torque,
- 2) data on a mechanical friction torque, or
- 3) data on a brake reaction time.
A machine learning device that has learned a state of a brake of a motor based on data on the brake, the machine learning device comprising:
- a state observation unit configured to observe brake operating state data indicative of an operating state of the brake, as state variables representative of a current environmental state;
- a learning unit that has learned a distribution of data indicative of the operating state of the brake when the brake is in a normal state by
- performing a cluster analysis based on the state variables to learn the distribution of data, and
- recording a cluster learned by the cluster analysis as a learned model;
- and
- an estimation result output unit configured to estimate and output a failure state of the brake, based on the state variables observed by the state observation unit and the learned model,
- wherein an operation of a machine having the motor is controlled to be stopped in response to the failure state of the brake output by the estimation result output unit, and
- wherein the brake operating state data includes at least one of
- 1) data on a gravity load torque,
- 2) data on a mechanical friction torque, or
- 3) data on a brake reaction time.
A machine learning device configured to learn a state of a brake of a motor based on data on the brake, the machine learning device comprising:
- a state observation unit configured to observe brake operating state data indicative of an operating state of the brake, as state variables representative of a current environmental state;
- a label data acquisition unit configured to acquire label data indicative of a failure state of the brake;
- and
- a learning unit configured to learn the operating state of the brake and the failure state of the brake in association with each other, using the state variables and the label data, wherein the learning unit comprises:
- an error calculation unit configured to calculate an error between (i) a correlation model configured to derive the failure state of the brake from the state variables, based on the state variables and the label data, and (ii) a correlation feature identified from teacher data prepared in advance;
- and
- a model update part configured to update the correlation model so as to reduce the error,
- wherein an operation of a machine having the motor is controlled to be stopped in response to a prediction of the failure state of the brake based on the correlation model.
A machine learning device that has learned a state of a brake of a motor based on data on the brake, the machine learning device comprising:
- a state observation unit configured to observe brake operating state data indicative of an operating state of the brake, as state variables representative of a current environmental state;
- a learning unit that has learned the operating state of the brake and a failure state of the brake in association with each other, wherein the learning unit comprises:
- an error calculation unit configured to calculate an error between (i) a correlation model configured to derive the failure state of the brake from the state variables, based on the state variables and the label data, and (ii) a correlation feature identified from teacher data prepared in advance;
- and
- a model update part configured to update the correlation model so as to reduce the error;
- and
- an estimation result output unit configured to estimate and output the failure state of the brake, based on the state variables observed by the state observation unit and the correlation model learned by the learning unit,
- wherein an operation of a machine having the motor is controlled to be stopped in response to the failure state of the brake output by the estimation result output unit.
Dependencies are inferred from the claim text.