Robot controlling device, robot apparatus, robot control method, program for executing robot control method, and recording medium on which program is recorded

Abstract

Disclosed is a technique that reduces the amount of calculation necessary for time optimal control. An interpolation function calculating partcalculates an interpolation function that passes through a plurality of interpolated teach points for interpolation between respective teach points. Further, a differential coefficient calculating partcalculates each differential coefficient obtained by differentiating each vector component included each interpolated teach point by a variable number of the interpolation function. Further, an estimated value calculating partcalculates an estimated velocity of each joint in each interpolated teach point on the basis of each pass velocity and each differential coefficient. Further, a time optimal controlling partchanges the respective pass velocities and minimizes an objective function obtained by integrating reciprocal numbers of the respective pass velocities by a variable number under a constraint condition including a condition that the ratio of the estimated velocity of each joint and a predetermined allowable velocity is within a predetermined constraint range. A position command generating partgenerates a position command by substituting, whenever a pass velocity function is time-integrated by a value of a predetermined time interval, a value of the integration in the variable number in the interpolation function.