Research On Underwater Core Cutting Wire Arc Cutting Based On Visual Sensing

With the massive development of marine resources,the demand for underwater cutting operations is increasing,and there is an urgent need to realize efficient and automated underwater cutting.The underwater core cutting wire arc cutting technology has the characteristics of simple equipment,safe operation and low overall cost,and has a wide range of application prospects in future underwater cutting operations.The current problem with this technology is the lack of real-time sensing and control of the cutting process.Visual sensing technology is one of the important means to study the automation of welding and cutting process.Therefore,this article attempts to extract the characteristic information of the cutting process through visual sensing,so as to achieve the purpose of effectively predicting and controlling the cutting process and improving cutting efficiency in the future.In this thesis,the underwater wet arc cutting vision sensor system is designed by adopting a closed pipe at the end to drain the water and reducing the thickness of the water between the camera and the arc,which reduces the interference of water on the image quality.The effects of narrow-band filters.Exposure time and other factors on image quality were studied when images were collected.It was determined that 808 nm narrow-band filters and 20 ms exposure time were used as acquisition parameters to obtain clear cut images.In order to extract the characteristic information of the cut image,Matlab software was used to study the image processing method of the acquired image.The processing effects of different filtering and enhancement algorithms on the cut image are compared,and finally median filtering and gamma enhancement are selected to process the image;the improved adaptive threshold adjustment segmentation algorithm is used to accurately segment the cut edge and the cut end;through the edge detection of the image after threshold segmentation,a clear cut edge is identified.It further analyzes the cut feature parameters that affect the cutting efficiency,and realizes the extraction of the cut feature values on the basis of the above-mentioned image processing.In order to establish the relationship between the pixel value of the cut feature parameter and the actual value,the underwater vision sensor system was calibrated,and the detection error was within 5%,which met the calibration requirements.In order to analyze the cutting characteristics based on the cutting state of the front and back of the workpiece,a laser-assisted light source was used to develop a vision system to detect the cutting state of the back of the workpiece.By analyzing the images of the front and back of the cut in the cutting process,three models of arc ignition inside the cut are established,and combined with current and voltage information,it is shown that when the arc is ignited at the bottom of the cut,it is the most unfavorable for cutting stability.In addition,arc bubble is also one of the factors that affect the stability of the cutting process.Through the captured bubble images,it is found when the arc is briefly extinguished,the ends of the cutting wire still exist a certain amount of heat vaporizes the surrounding water to produce bubbles.The bubbles only go through two stages of growth and rupture because there is no gas that is decomposed by the core cutting wire as a support.