Systems and methods for operating a motor according to parameters provided by an autotuning component if available are described. A controller can be coupled to a drive which operates a motor for executing a task that can be related to a drilling operation for oil and gas. The controller stores initial parameters and checks for new parameters provided by the autotuning component which are stored on the drive after the autotuning component autotunes the motor. If there are new parameters, they are given priority over the initial parameters.

Learning related to a device having a driving unit is performed more easily. An information processing device includes: a storage unit that stores a machining program for operating a motor of a machine tool, a robot, or an industrial machine; and a generation unit that generates a learning program for performing learning based on operating characteristics of the motor by extracting a partial machining program including a characteristic element from the machining program stored in the storage unit.

A machine learning device includes: a training data acquisition unit configured to acquire multiple pieces of training data each including insulation resistances of a servomotor at the beginning and the end of a certain period and time-series data indicating conditions of the servomotor in the certain period; and a learning model generating unit configured to perform a supervised learning using the training data to thereby generate a learning model.

Provided is a machine learning device configured to perform machine learning related to optimization of a compensation value of a compensation generation unit with respect to a servo control device configured to control a servo motor configured to drive an axis of a machine tool, a robot, or an industrial machine, and that includes at least one feedback loop, a compensation generation unit configured to generate a compensation value to be applied to the feedback loop, and an abnormality detection unit configured to detect an abnormal operation of the servo motor, wherein, during a machine learning operation, when the abnormality detection unit detects an abnormality, the compensation from the compensation generation unit is stopped and the machine learning device continues optimization of the compensation value generated by the compensation generation unit.

A control circuit includes: a controller; a controlled system; and a filter for smoothing a return signal. The controller acts on the controlled system vis--vis a control signal and the return signal acts on the controller. The controller and the filter are simultaneously adjustable by an adjustment. In an embodiment, the adjustment is made on the basis of a method that includes: measuring or estimating an output signal as a measurement or estimate; transferring, using the measurement or estimate, the output signal into the return signal; determining a power density spectrum of the return signal; and limiting a control signal of the controller such that a power of the control signal does not exceed a predefined limiting value.

A machining parameter adjustment system is established in a controller. The controller is connected to a machine. The controller is configured to receive a machining program and analyze at least one process included in the machining program. Also, the controller sets a tuning program corresponding to the process and inserts the tuning program before or after the code of the process, so as to generate an integration program and upload the integration program to the machine.

A motor control apparatus including a controller that controls a servo motor or a spindle motor and includes a switching determining part that determines a switching condition of the controller based on axis position information on a motor related to control of the motor control apparatus, a machine learning part that adjusts one or more parameters for the controller by machine learning for each switching condition, and a parameter holding part that holds the parameter adjusted by the machine learning part for each switching condition. The switching determining part, when determining the switching condition after adjustment of the parameter, uses the adjusted parameter corresponding to the switching condition in the controller. The apparatus enables changing, and automatic adjustment, of a parameter or controller to be used depending on a switching condition of the parameter related to axis position information or a switching condition of the controller using the parameter.

A machine learning device performs machine learning with respect to a servo motor controller that converts a three-phase current to a two-phase current of the d- and q-phase. The machine learning device includes: a state information acquisition unit configured to acquire, from the servo motor controller, state information including velocity or a velocity command, reactive current, and an effective current command and effective current or a voltage command; an action information output unit configured to output action information including a reactive current command to the servo motor controller; a reward output unit configured to output a value of a reward of reinforcement learning based on the voltage command or the effective current command and the effective current; and a value function updating unit configured to update a value function on the basis of the output value of the reward, the state information, and the action information.

A machining parameter adjustment system is established in a controller. The controller is connected to a machine. The controller is configured to receive a machining program and analyze at least one process included in the machining program. Also, the controller sets a tuning program corresponding to the process and inserts the tuning program before or after the code of the process, so as to generate an integration program and upload the integration program to the machine.

Training and/or using a recurrent neural network model for visual servoing of an end effector of a robot. In visual servoing, the model can be utilized to generate, at each of a plurality of time steps, an action prediction that represents a prediction of how the end effector should be moved to cause the end effector to move toward a target object. The model can be viewpoint invariant in that it can be utilized across a variety of robots having vision components at a variety of viewpoints and/or can be utilized for a single robot even when a viewpoint, of a vision component of the robot, is drastically altered. Moreover, the model can be trained based on a large quantity of simulated data that is based on simulator(s) performing simulated episode(s) in view of the model. One or more portions of the model can be further trained based on a relatively smaller quantity of real training data.