Parallel Robot Trajectory Optimization Based On Machine Vision

With the rapidly improvement of China's scientific and technological level,modern industrial production has an increasing high demand for high-efficiency,high-quality and low-cost industrial production.Many enterprises have launched intelligent industrial upgrading measures for industrial production lines,among which industrial robots are playing an increasingly important role in the intelligent industrial upgrading.Sorting in assembly line production due to the workers in the sorting operation of a large number of mechanical duplication of effort and needs to be intelligent industrial upgrading,and because of its high precision,high speed,as well as good dynamic response performance,parallel robots were applied in assembly line sorting activities.So this paper use parallel robot application in assembly line sorting activities as the research object in related research.Firstly,by analyzing the exist problems of object detection algorithms at home and abroad,a new object detection algorithm is proposed in combination with the sorting operation environment to further improve the real-time performance of object detection.We have completed corresponding machine vision system,the hardware model selection of main components,the machine vision software platform,and the object detection algorithm proposed in this paper.Last,the results of data verification prove that the method can achieve the expected results.Second,as there exist problems of curvature mutation in the corner of the trajectory applied to the door of the sorting operation glyph trajectory planning,this paper proposed the Scott curve to door glyph trajectory planning optimization,for avoiding a straight line with the connection to the easement curve curvature mutation,thus reducing the operation of residual vibration in the process of sorting with parallel robot.We verify the method can achieve expected results through constructing the experimental platform.Finally,for the optimal trajectory planning problems of energy consumption,this paper put forward in the shortest distance as the optimization goal,through the particle swarm optimization(POS)to optimization the door glyph trajectory,getting the shortest trajectory length,thus reducing the coordinate operation and energy consumption.At last,we build sorting parallel robot experiment platform,sorting through visual system and robot motion experiment,prove the method can achieve expected results,and show that the design of the sorting system smooth movement,grasping workpiece accurately to meet the design requirements.