Research On Laser Cutting Control System Based On Machine Vision

Laser cutting is one of the most widely used laser processing technologies.With the increasing influence of intelligent manufacturing on the manufacturing industry,the industry urgently needs to develop more efficient and intelligent laser cutting numerical control system.Developing a cutting system based on visual recognition is an important way to realize intelligent laser cutting.Cutting rate and precision are two important indexes to measure numerical control laser cutting system.At present,the overshoot and out-of-step phenomenon is the key to limit the improvement of laser cutting rate.In terms of cutting precision,the processing precision can be further improved by using appropriate image contour extraction method and interpolation algorithm.In this paper,a cutting software and hardware system based on machine vision is developed to improve the cutting speed and precision of the numerical control laser cutting system.The specific work is as follows:(1)Image processing part of the cutting system.The target image is acquired by an industrial camera,and the boundary contour is extracted by image grayscale and filtering.In order to facilitate the subsequent interpolation operation,the redundant pixel information in the image information is reduced by optimizing the contour fitting algorithm,so as to reduce the amount of computation and improve the efficiency.(2)Motion control part of the cutting system: overshoot and out-of-step phenomenon caused by inertial action of physical devices cannot be avoided,but can be compensated by algorithm to achieve smooth start of the control system.Aiming at the out-of-step and overshoot phenomena of the platform device,the optimal acceleration and deceleration control algorithm is studied,and the out-of-step and overshoot phenomena are eliminated successfully.(3)The cutting system test parts: interpolation algorithm based on time division,improve the precision of interpolation,using the system for image acquisition,processing,and realize the motion control for the platform device to simulationcutting test of the object,image algorithm is verified by experiment and the design of motion control algorithm for accuracy with the help of the system work.(4)Linux system environment construction: according to the functional requirements analysis of the control system,select the appropriate embedded hardware platform;Porting system boot programs;Build build environment;Clipping the kernel and setting up NFS server for system development and debugging;The core hardware of the control system is introduced.According to the software structure of the system,the serial communication,man-machine interaction and motion control modules are analyzed.