Intelligent Grinding Feed System Research Based On Machine Vision

Grinding occupies an important position in modern manufacturing, especially inCNC grinding technology, and it got fast development recently. Online measurementtechnology plays a very important role in achieving high-precision, high-quality and high-efficiency of parts processing. Online-contact detection equipment has been widely used.Sonic detection equipments also come into use. But these Detecting devices should beinstalled in the vicinity of the parts grinding. For some special structure grinders whichcannot install these detection devices, it is necessary to use a long-range and non-contactdetection device for machining parts’ real-time detection. So research and application ofimage recognition and detection technology came into being with the development of thetimes. The detected image data need to be integrated into the CNC feed system, and thisrequires a feed controller according to the image change control. Combining the abovepoints, an intelligent grinding feed system based on machine vision is proposed in thispaper.In the detecting portion, first perform spatial transform, component extraction andbinarization on spark images which collected in real-time acquisition of the grindingprocess. Then add up spark values and connect the image values to continuous numericalrecords. Digitize continuous spark images as feedback of control systems for thecompletion of the detection link of the grinding sparks image closed-loop control system.In the controlling portion, fuzzy controller designed in this article is used, feed rate iscontrolled. The controller consists of a main fuzzy controller, auxiliary fuzzy controller aswell as security control rules. Finally grinding tests proves the feasibility and advantagesof this system.In this paper, the authors independently completed the design of the hardwarestructure, image processing programming and fuzzy controller’s design and programming.Also complete the grinding sparks images for real-time acquisition and processing, as wellas the manufacturing of intelligent systems and fuzzy control feed.