Robot polishing method based on multiple sensors

Abstract

The invention discloses a robot polishing method based on multiple sensors. The robot polishing method based on the multiple sensors comprises the steps that (1) type selection is conducted on hardware of a robot polishing system, and the composition of hardware of the visual sensor is calibrated; (2) the surfaces of automobile punching parts are subjected to defect collection through the visual sensor, collected image data are manually marked, true value images are obtained, and a defect data set is established; (3) the partition of workpiece surface defects is accomplished by a PixelNet convolutional neural network; (4) signals collected by the torque sensor are preprocessed in three aspects of static calibration, filtering and gravity compensation; and (5) the robot defect polishing system which carries the torque sensor conducts the polishing operation on the defects by utilizing known defect information and a force / position mixed control algorithm and adopting an orthogonal path polishing scheme. The robot polishing method based on the multiple sensors can be widely applied to lots of production and manufacturing fields including the science and technology fields requiring industrial robots such as the aerospace field, the rail transit field, the medical equipment field and the furniture making field.

Description

technical field [0001] The invention belongs to the field of machine vision and sensors, and relates to a multi-sensor-based robot polishing method, which can be used for industrial robot polishing. Background technique [0002] With the gradual maturity of related technologies in hardware and software for industrial robots, the accuracy of the robot body and movement has been greatly improved, and its cost has also been greatly reduced. Aiming at the problem that large-sized and complex-shaped workpieces are difficult to grind, experts and scholars at home and abroad have begun to conduct corresponding research on industrial robot grinding technology on the basis of a large number of research results on industrial robots. At present, the main research direction is based on the factors that affect the grinding of industrial robots, including grinding tool materials, grinding path planning, robot speed control, and grinding contact force. [0003] Aiming at the problems of l...

AI Technical Summary

Problems solved by technology

[0003] Aiming at the problems of low defect grinding efficiency and difficult quality assurance, this paper proposes a robot grinding system based on multi-sensors, which includes two subsystems: a defect detection system based on visual sensors and a defect grinding system based on torque sensors

Images