Machine learning correction parameter adjustment apparatus and method for use with a motor drive control system

A correction parameter adjustment method, comprising:

• machine learning, performed by a machine learning apparatus, for learning a correction parameter used in correction of a command value that controls a motor in a motor drive system including a plurality of kinds of correction functions, the machine learning including:
• observing, by state observation circuitry, as a state variable, each of a feature extracted from drive data of the motor drive system for each of a plurality of correction functions on the basis of the kind of any of the correction functions and the correction parameter;
• and
• leaning, by learning circuitry, the correction parameter for each of the correction functions according to a training data set created on the basis of the state variable;
• selecting, by correction function selection circuitry, one of the plurality of kinds of correction functions;
• acquiring, by drive data acquisition circuitry, the drive data;
• extracting, by feature extraction circuitry, the feature on the basis of the correction function selected and the drive data acquired;
• inputting, by adjustment requirement input circuitry, an adjustment requirement of the correction parameter;
• and
• determining, by correction parameter determination circuitry, the correction parameter satisfying the adjustment requirement on the basis of a learning result of the learning circuitry.

A correction parameter adjustment method, comprising:

• machine learning, performed by a machine learning apparatus, for learning a correction parameter used in correction of a command value that controls a motor in a motor drive system including a plurality of kinds of correction functions, the machine learning including:
• observing, by state observation circuitry, as a state variable, each of a feature extracted from drive data of the motor drive system for each of a plurality of correction functions on the basis of the kind of any of the correction functions and the correction parameter;
• and
• leaning, by learning circuitry, the correction parameter for each of the correction functions according to a training data set created on the basis of the state variable;
• selecting, by correction function selection circuitry, one of the plurality of kinds of correction functions;
• acquiring, by drive data acquisition circuitry, the drive data;
• selecting, by algorithm selection circuitry, an algorithm used for extracting a feature from the drive data for each of the correction functions selected;
• and
• using, by feature calculation circuitry, the selected algorithm to extract the feature from the drive data acquired.

A correction parameter adjustment method, comprising:

• machine learning, performed by a machine learning apparatus, for learning a correction parameter used in correction of a command value that controls a motor in a motor drive system including a plurality of kinds of correction functions, the machine learning including:
• observing, by state observation circuitry, as a state variable, each of a feature extracted from drive data of the motor drive system for each of a plurality of correction functions on the basis of the kind of any of the correction functions and the correction parameter;
• and
• learning, by learning circuitry, the correction parameter for each of the correction functions according to a training data set created on the basis of the state variable;
• and
• selecting, by correction function selection circuitry, the plurality of kinds of correction functions,
• wherein the correction function selection circuitry selects in turn a predetermined combination of correction functions.