Machining condition adjustment device and machine learning device

A machining condition adjustment device configured to adjust a laser beam machining condition fora laser beam machining device to carry out laser beam machining of a workpiece, the machining condition adjustment device comprising:

• a processor configured to produce state variables including laser beam machining condition data indicating the laser beam machining condition in the laser beam machining, workpiece data indicating information related to the workpiece, and plasma generation amount data indicating an amount of plasma generated in the laser beam machining;
• and
• a learning model storage configured to store a learned model that has learned adjustment action for the laser beam machining condition with respect to the amount of the plasma generated in the laser beam machining of the workpiece under a prescribed laser beam machining condition, wherein
• the processor is further configured to estimate the adjustment action for the laser beam machining condition with use of the learned model based on the state variables, and
• the prescribed laser beam machining condition includes a condition where the workpiece is machined by the laser beam machining and, during the laser beam machining of the workpiece, a machining surface of the workpiece is at or in a vicinity of a flow speed maximizing position for an assist gas ejected from a nozzle of the laser beam machining device to the workpiece and re-converged due to a shock wave generated from a mouth of the nozzle, wherein
• the assist gas is (1) ejected from the nozzle at a gas pressure maximizing position, and (2) then expanded and then re-converged at a position between the gas pressure maximizing position and the flow speed maximizing position in a flow direction of the assist gas.

A machine learning device configured to adjust a laser beam machining condition for a laser beam machining device to carry out laser beam machining of a workpiece, the machine learning device comprising:

• a processor configured to learn adjustment action for the laser beam machining condition with respect to an amount of plasma generated in the laser beam machining of the workpiece under a prescribed laser beam machining condition, with use of
• state variables including laser beam machining condition data indicating the laser beam machining condition in the laser beam machining, workpiece data indicating information related to the workpiece, and plasma generation amount data indicating the amount of the plasma generated in the laser beam machining, and
• determination data including plasma generation amount determination data for determination of the amount of the plasma generated in the laser beam machining by the laser beam machining device, wherein
• the plasma generation amount determination data is for determination of the amount of the plasma generated in the laser beam machining of the workpiece under the laser beam machining condition adjusted by the adjustment action for the laser beam machining condition, and
• the prescribed laser beam machining condition includes a condition where the workpiece is machined by the laser beam machining and, during the laser beam machining of the workpiece, a machining surface of the workpiece is at or in a vicinity of a flow speed maximizing position for an assist gas ejected from a nozzle of the laser beam machining device to the workpiece and re-converged due to a shock wave generated from a mouth of the nozzle, wherein
• the assist gas is (1) ejected from the nozzle at a gas pressure maximizing position, and (2) then expanded and then re-converged at a position between the gas pressure maximizing position and the flow speed maximizing position in a flow direction of the assist gas.

A machine learning device configured to adjust a laser beam machining condition for a laser beam machining device to carry out laser beam machining of a workpiece, the machine learning device comprising:

• a learning model storage configured to store a learned model that has learned adjustment action for the laser beam machining condition with respect to an amount of plasma generated in the laser beam machining of the workpiece under a prescribed laser beam machining condition;
• and
• a processor configured to estimate the adjustment action for the laser beam machining condition with use of the learned model based on state variables including laser beam machining condition data indicating the laser beam machining condition in the laser beam machining, workpiece data indicating information related to the workpiece, and plasma generation amount data indicating the amount of the plasma generated in the laser beam machining, wherein
• the prescribed laser beam machining condition includes a condition where the workpiece is machined by the laser beam machining and, during the laser beam machining of the workpiece, a machining surface of the workpiece is at or in a vicinity of a flow speed maximizing position for an assist gas ejected from a nozzle of the laser beam machining device to the workpiece and re-converged due to a shock wave generated from a mouth of the nozzle, wherein
• the assist gas is (1) ejected from the nozzle at a gas pressure maximizing position, and (2) then expanded and then re-converged at a position between the gas pressure maximizing position and the flow speed maximizing position in a flow direction of the assist gas.