Исследование алгоритмов адаптации управляю- щего воздействия для приводов робототехнических систем с помощью симулятора межкоординатных возмущений

Abstract

In this paper we review algorithms of technologic trajectory correction for coordinate actuators of robot effectors. The target of controlling these effectors is essentially the motion of controlled object along the given trajectory and also (if necessary) the adaptation of an programmed traectory to the actual one under the influence of perturbations from the object or from the technological process. The object of our research is the motion actuator with the feedback of the position of an effector. The adjacent coordinate actuator is implemented as a simulation of the intercoordinate perturbations. Control is based on a fourth form of invariance. The first algorithm in this article includes the synthesis of the additional component and its inclusion into the already devised control action. This additional component is a torque error function of an uncalculated perturbation. The second algorithm is a correction of derivative signal levels performed during a stage of the control action synthesis while the technological process executes. Correction of the control action in the control component is performed on a software level and does not involve changes in the main actuator circuit, i.e. the inner properties and therefore the dynamic characteristics of an actuator are not influenced. In this paper we provide an analysis of the control system precision in the mode of constant velocity motion, while processing the harmonic input signal for channels of the position, velocity and acceleration signal adaptation. The test actuator in constant velocity motion exhibits the positioning error ranging from 5% to 10% due to the influence of the adjacent coordinate simulation. Introducing the adaptation circuit allowed us to reduce the influence of external perturbations during the technological process execution to a minimum (positioning error is <0.5%) Simulation results show that the link with variable transmission coefficient is essential. During the simulation we formulate dependences of the transmission coefficient in the adaptation channel. The results of the research confirm the efficiency of algorithms considered in this paper.