Numerical controller and machine learning device

A numerical controller that predicts an abnormality which will occur in a machine to be controlled, the numerical controller comprising:

• a data logger device that acquires information, relating to operation of the machine, as time-series data;
• a machine learning device that predicts occurrence of the abnormality in the machine based on the time-series data acquired by the data logger device;
• and
• an abnormality prediction notification unit that makes notification of prediction results, given by the machine learning device, on the occurrence of the abnormality in the machine, wherein
• the machine learning device includes
• a database that stores results of machine learning of trends in the operation of the machine on occasions of occurrence of abnormalities,
• a state observation unit that acquires the time-series data, acquired by the data logger device, as input data, and
• a prediction unit that predicts the occurrence of the abnormality in the machine based on the input data acquired by the state observation unit and the results of the machine learning that are stored in the database, wherein
• in the database, feature vectors representing the trends in the operation of the machine on the occasions of the occurrence of the abnormalities in the machine and abnormality information relating to the abnormalities are stored in association with each other, and results of clustering of the feature vectors are stored and,
• the prediction unit includes
• a similar cluster determination unit that finds approximate expressions for the time-series data included in the input data, that generates a feature vector based on the found approximate expressions, and that determines which cluster stored in the database is an approximation to the generated feature vector, and
• an abnormality ranking determination unit that produces ranking information on abnormalities associated with feature vectors belonging to the cluster determined as the approximation to the feature vector by the similar cluster determination unit and that outputs the produced ranking information as prediction results on the occurrence of the abnormality in the machine.

A machine learning device that conducts machine learning of trends in operation of a machine on occasions of occurrence of abnormalities in the machine, based on time-series data representing information relating to the operation of the machine and acquired by a data logger device provided in a numerical controller which controls the machine, the machine learning device comprising:

• a state observation unit that acquires the time-series data, acquired by the data logger device, as input data;
• an abnormality observation unit that acquires abnormality information relating to the abnormalities which occur in the machine;
• a learning unit that conducts the machine learning of the trends in the operation of the machine on the occasions of the occurrence of the abnormalities, based on the input data acquired by the state observation unit and the abnormality information acquired by the abnormality observation unit;
• and
• a database that stores results of the machine learning in the learning unit, wherein
• the learning unit includes
• an analysis unit that finds approximate expressions for the time-series data included in the input data, that generates a feature vector based on the found approximate expressions, and that stores the generated feature vector, associated with the abnormality information, in the database, and
• a clustering unit that carries out clustering of the feature vector generated by the analysis unit and that stores results of the clustering in the database.

A machine learning device that predicts an abnormality which will occur in a machine controlled by a numerical controller including a data logger device, based on time-series data acquired by the data logger device and representing information relating to operation of the machine, the machine learning device comprising:

• a database that stores results of machine learning of trends in the operation of the machine on occasions of occurrence of abnormalities;
• a state observation unit that acquires the time-series data, acquired by the data logger device, as input data;
• and
• a prediction unit that predicts occurrence of the abnormality in the machine based on the input data acquired by the state observation unit and the results of the machine learning that are stored in the database, wherein
• in the database, feature vectors representing the trends in the operation of the machine on the occasions of the occurrence of the abnormalities in the machine and abnormality information relating to the abnormalities are stored in association with each other, and results of clustering of the feature vectors are stored, and
• the prediction unit includes
• a similar cluster determination unit that finds approximate expressions for the time-series data included in the input data, that generates a feature vector based on the found approximate expressions, and that determines which cluster stored in the database is an approximation to the generated feature vector, and
• an abnormality ranking determination unit that produces ranking information on abnormalities associated with feature vectors belonging to the cluster determined as the approximation to the feature vector by the similar cluster determination unit and that outputs the produced ranking information as prediction results on the occurrence of the abnormality in the machine.