Robot joint high-precision control system and method based on full-state feedback

Abstract

The invention discloses a robot joint high-precision control system based on full-state feedback. The control system comprises a state observer, a robot joint friction model and the like, the state observer is constructed based on an equivalent dynamical model of a robot joint, inputs detected joint operation position data and a control signal not giving friction compensation and outputs an estimated value of a transmission system output side variable, and the robot joint friction model inputs a joint speed command signal or a speed signal detected at an output side of the transmission system and output an estimated value of the equivalent friction force, or an estimated value of the equivalent friction torque. The invention further discloses a robot joint high-precision control method based on full-state feedback, equivalent disturbance estimated by the reduced-order expansion state observer is compensated into an input signal of a motor, and disturbance compensation is carried out on a driving joint; and the friction force estimated by the friction model is compensated into an input torque signal of the motor to compensate the friction force.

Description

technical field [0001] The invention relates to a high-precision motion control method for robot joints, in particular to a high-precision control system and method for robot joints based on full state feedback. Background technique [0002] The robot realizes motion transmission by multiple driving joints. Since the stiffness of the driving joints is much lower than that of the connecting parts, the dynamic accuracy of the end effector is mainly determined by the output accuracy of the driving joints. The driving joints of the robot are generally realized by the screw nut pair or the reducer to realize the motion / force transmission, and the elasticity, friction, and clearance of the transmission system seriously affect its output accuracy. The traditional semi-closed-loop control strategy uses the motor encoder signal to construct a closed-loop control, which can only guarantee the output accuracy of the motor, but cannot guarantee the output accuracy outside the closed-loo...

AI Technical Summary

Problems solved by technology

At present, a single data-based or model-based method is often used to compensate for the impact of disturbance. The data-based disturbance compensation method uses sensors or observers to obtain disturbance information. Obtain comprehensive disturbance information, such as the influence of friction during low-speed commutation; the model-based disturbance compensation method needs to establish a mathematical model of the disturbance, and then design a feed-forward compensator to eliminate it. Model Completeness and Accuracy

The nonlinear links of robot-driven joints are complex and the equivalent disturbance varies with the position of the robot. It is difficult to guarantee the effect of disturbance compensation by using a single compensation strategy. It is necessary to explore and research efficient composite disturbance compensation methods

Images