A collision position estimation device allowing extraction of data related to a torque of a section related to collision of a cushion member includes a machine learning device. The machine learning device includes a state observation unit for observing axis movement data indicating information related to movement of the axis and motor torque data indicating data related to torque of a motor driving the axis as a state variable representing a current state of an environment, a label data acquisition unit for acquiring collision position data indicating the position of the axis during collision of the cushion member as label data, and a learning unit for associating and learning the information related to movement of the axis and the data related to the torque of the motor driving the axis with the position of the axis during collision of the cushion member using the state variable and the label data.

In a controller, a machine learning device, and a system that are capable of addressing change in a clamping force without use of expensive equipment, the controller includes the machine learning device that observes machining condition data indicating machining conditions for cutting, spindle torque data indicating spindle torque during the cutting, and cutting force component direction data indicating cutting force component direction information on cutting resistance against a cutting force, as state variables representing a current state of an environment, and that carries out learning or decision making with use of a learning model modelling the machining conditions for the cutting on which the cutting force that allows holding by a clamping force from a machining jig is exerted on a workpiece based on the state variables.

A relearning necessity determination method is provided for determining a necessity of relearning of a learned diagnostic model in a machine tool including a machining abnormality diagnosing unit. The machining abnormality diagnosing unit determines normal or abnormality of machining using the diagnostic model generated through machine learning. The method includes storing a cumulative cutting time or a cumulative cutting distance of a tool mounted to the machine tool as a tool usage, storing the tool usage when the machining abnormality diagnosing unit diagnoses the machining as machining abnormality, and determining the necessity of the relearning of the diagnostic model based on a frequency distribution of the tool usage stored in the storing of the tool usage.

A machine learning device of a machining condition adjustment device includes: a state observation unit that observes each of machining condition data indicative of a machining condition of each used tool and cycle time data indicative of a cycle time of one machining, as a state variable; a determination data acquisition unit that acquires determination data indicative of a result of an appropriateness determination of one machining in the case where an adjustment of the machining condition is performed; and a learning unit that performs learning by associating the machining condition data and the cycle time data with the adjustment of the machining condition using the state variable and the determination data so as to enables effective use of the allowance of a cycle time.

A calculation formula learning unit sets a coefficient relating to a time lag element in a thermal displacement estimation calculation formula by machine learning while fixing a coefficient relating to measured data except the coefficient relating to the time lag element at a predetermined value based on a difference between a thermal displacement estimated value about a machine element calculated by substituting a measured data group into the thermal displacement estimation calculation formula and a thermal displacement actual measured value about the machine element; sets the coefficient relating to the measured data except the coefficient relating to the time lag element in the thermal displacement estimation calculation formula by machine learning based on the difference while fixing the coefficient relating to the time lag element at a predetermined value; and repeats the machine learning.

A machine learning device for learning a threshold value of detecting an abnormal load in a machine tool, includes a state observation unit, and a learning unit. The state observation unit observes a state variable obtained based on at least one of information about a tool, main spindle revolution rate, and amount of coolant of the machine tool, material of a workpiece, and moving direction, cutting speed, and cut depth of the tool, and the learning unit learns the threshold value of detecting an abnormal load based on training data created from an output of the state observation unit and data related to detection of an abnormal load in the machine tool and on teacher data.

A controller includes a machine learning device for learning machining conditions when deburring is performed by controlling the robot. The machine learning device observes workpiece information indicating a shape or material of a workpiece, burr information indicating a shape or position of a burr, and machining conditions including tool information indicating a type of a tool, a feed rate of the tool and a rotational speed of the tool, as a state variable representing a current state of an environment, and acquires determination data indicating an evaluation result of the deburring. Then, using the observed state variable and the acquired determination data, the machine learning device performs learning by associating the machining conditions with the workpiece information and the burr information.

A collision position estimation device allowing extraction of data related to a torque of a section related to collision of a cushion member includes a machine learning device. The machine learning device includes a state observation unit for observing axis movement data indicating information related to movement of the axis and motor torque data indicating data related to torque of a motor driving the axis as a state variable representing a current state of an environment, a label data acquisition unit for acquiring collision position data indicating the position of the axis during collision of the cushion member as label data, and a learning unit for associating and learning the information related to movement of the axis and the data related to the torque of the motor driving the axis with the position of the axis during collision of the cushion member using the state variable and the label data.

In a controller, a machine learning device, and a system that are capable of addressing change in a clamping force without use of expensive equipment, the controller includes the machine learning device that observes machining condition data indicating machining conditions for cutting, spindle torque data indicating spindle torque during the cutting, and cutting force component direction data indicating cutting force component direction information on cutting resistance against a cutting force, as state variables representing a current state of an environment, and that carries out learning or decision making with use of a learning model modelling the machining conditions for the cutting on which the cutting force that allows holding by a clamping force from a machining jig is exerted on a workpiece based on the state variables.

A learning model is constructed for adjusting control information so that a cycle time becomes shorter while also avoiding the occurrence of overheating. A learning model construction device includes: an input data acquisition means that acquires, as input data, control information including a combination of an operation pattern of a spindle and parameters related to machining in a machine tool, and temperature information of the spindle prior to performing the machining based on the control information; a label acquisition means for acquiring temperature information of the spindle after having performed the machining based on the control information as a label; and a learning model construction means for constructing a learning model for temperature information of the spindle after having performed machining based on the control information, by performing supervised learning with a group of the input data and the label as training data.