Automatic Aluminum Alloy Surface Grinding Trajectory Planning of Industrial Robot Based on Weld Seam Recognition and Positioning

Abstract

In this paper, we propose a novel method for planning grinding trajectories on curved surfaces to improve the grinding efficiency of large aluminum alloy surfaces with welds and defect areas. Our method consists of three parts. Firstly, we introduce a deficiency positioning method based on a two-dimensional image and three-dimensional point cloud, which enables us to accurately and quickly locate the three-dimensional defective areas. Secondly, we propose a 2D weld positioning method based on the defect area and obtain the spatial position of the 3D weld by combining the relationship between 2D and 3D images. Additionally, we propose an orthogonal projection method from the point cloud to the aluminum alloy surface to calculate the weld reinforcement. Thirdly, we present a space spiral grinding trajectory planning method for complex curved surfaces based on the characteristics of the weld reinforcement, spatial position, and spatial position information of the defect area. This method shortens the grinding time of the defect area and improves efficiency. Simulation and experimental results show that our grinding trajectory planning method is more efficient than other grinding methods in removing defects from the surface of aluminum alloys. Moreover, the defect area after grinding is smoother than before.

Variable Time-Step Physics Engine with Continuous Compliance Contact Model for Optimal Robotic Grinding Trajectory Planning

“…Figure 12 presents the roughness data for each point both before and after grinding. Notably, the initial roughness at each point appears to be relatively large, averaging 1.6396 [ 39 ]. Following the grinding process, a significant reduction in roughness at each point is observed, with minimal changes, averaging 0.9166.…”

“…We conducted grinding experiments ( Section 4.3 ) on curved surfaces, discovering that optimal roughness during grinding [ 39 ] can be achieved by adjusting the contact force. The experimental findings reveal an average relative error of 2.75%, indicating a discrepancy between the virtual force and the actual contact force.…”

Prediction and Control of Small Deviation in the Time-Delay of the Image Tracker in an Intelligent Electro-Optical Detection System

“…Then, the DSST filter stably outputs the pixel coordinate position of the target in the next frame. The DSST filter is used to extract the image blocks f 1 , f 2 , • • • , f n with gray level feature from the single-sample detection area with resolution M × N [19][20][21]. Then, the filter h t is solved to obtain the gray level response value g 1 , g 2 , • • • , g n corresponding to each image block f 1 , f 2 , • • • , f n .…”

Research on the path extraction method of shoe upper grinding based on 3D vision

“…Read the registered shoe model point cloud model to obtain the coordinates and RGB values of each point in the point cloud model. Set the RGB threshold as shown in equation (14). When the RGB value of a point in the point cloud model falls within the threshold range, the point is saved.…”

“…Yuankai et al [13] proposed a welding seam feature point extraction algorithm based on centroid positioning and the least squares method for fitting the welding seam feature points to obtain the workpiece welding trajectory. Zhao et al [14] proposed a method for planning complex surface spiral grinding trajectories with defects and demonstrated its correctness and efficacy. Zhu et al [15] constructed a local template using the fast point feature histogram (FPFH) descriptor to describe the characteristics of each area of the casting.…”