Machine learning apparatus, laser apparatus and machine learning method

A machine learning apparatus for a laser apparatus that is provided with:

• a plurality of LD units including at least one or more laser diode modules;
• a power source unit that individually supplies a driving current to each LD unit;
• and a control unit that independently controls, for each LD unit, the driving current injected from the power source unit to each LD unit, and emits a laser light via a laser optical system by using a laser light from the plurality of LD units as a laser light source or as an excitation light source for laser oscillation, the machine learning apparatus being configured to learn LD unit driving condition data including output command data instructed to the power source unit,
• the machine learning apparatus comprising:
• a state amount observation unit that observes a state amount of the laser apparatus including output data from the power source unit, at least one or more optical output detection units that measure an optical output of the laser light, an optical output characteristic recording unit that records history of a measurement result of optical output characteristics of each LD unit obtained from the driving current output from the power source unit and the optical output, and a driving condition/state amount recording unit configured to record history of the LD unit driving condition data including at least a prediction result of a remaining life of each LD unit as a prediction result of characteristics of each LD unit output from the machine learning apparatus and history of the state amount of the laser apparatus;
• an operation result acquisition unit that acquires, as results of the LD unit driving condition data, a difference between the prediction result of the characteristics of each LD unit including at least the prediction result of the remaining life of each LD unit recorded in the driving condition/state amount recording unit and the measurement result of the optical output characteristics of each LD unit and/or an estimation result from the measurement result;
• a learning unit that receives an output from the state amount observation unit and an output from the operation result acquisition unit, and learns the LD unit driving condition data including at least the prediction result of the remaining life as the prediction result of the characteristics of each LD unit in association with the state amount of the laser apparatus and the results of the LD unit driving condition data;
• and
• a decision-making unit that decides, by referring to a learning result of the learning unit, an LD unit driving condition data including at least a prediction result of the remaining life as the prediction result of the characteristics of each LD unit.