Machine learning apparatus, servo control apparatus, servo control system, and machine learning method

A machine learning method for a machine learning apparatus to perform reinforcement learning to a servo control apparatus with respect to at least one of a backlash compensation parameter in a backlash compensating means and a backlash acceleration compensation parameter in a backlash acceleration compensating means, the servo control apparatus including the backlash compensating means for creating a backlash compensation value with respect to a position command or a position deviation and the backlash acceleration compensating means for creating a backlash acceleration compensation value with respect to a speed command, the machine learning method comprising the steps of:

• an action information outputting step of outputting action information including adjustment information on at least one of the backlash compensation parameter and the backlash acceleration compensation parameter to at least one of the backlash compensating means and the backlash acceleration compensating means;
• a state information acquiring step of acquiring, from the servo control apparatus, state information including position deviation and at least one of the backlash compensation parameter and the backlash acceleration compensation parameter, the position deviation being obtained from the position command and a fed-back position, at a time of making the servo control apparatus execute a predetermined machining program on the basis of the action information;
• a reward outputting step of outputting a reward value in the reinforcement learning on the basis of the position deviation included in the state information;
• and
• a value function updating step of updating an action-value function on the basis of the reward value, the state information, and the action information.