Research On Automatic Palletizing Technology Of Mixed Production Line

At present, partial production line presents a flexible development trend of variety and customization,therefore,for the robot palletizing technology also put forward higher and higher requirements.However, the current production line of the palletizing robot is generally completing the fixed palletizing action in accordance with the pre-set procedures for a single type, or the target object without fixed posture, which cannot adapt to the delivery of palletizing mixed production, and small scale production line without fixed posture. Therefore, it is necessary to study a new efficient and stable mixed production line of different types and position of object palletizing system and method.Combination with machine vision technology and palletizing robot, automatic palletizing of different objects can be achieved in the mixed production line. First,analyze the whole workflow, the key problems needs to be solved in the whole palletizing system are established. On this basis, the vision system platform and the target tracking method are established, then the target recognition and location methods are analyzed and discussed in detail. Final,the method of obtaining the coordinate data of the target object is determined .The palletizer robot in the process of high-speed operation needs optimize operational tempo and improve efficiency.Arc interpolation methods in a cartesian coordinate systems and joint space three curve interpolation were used to in the optimization of running track, the planning conclusion of simulation was based on Matlab. At last,the software system for the palletizer robot was programmed.The DLL file was used in Matlab and C# hybrid programming,and the DPRAM was used as connection method between PC and PMAC, the processed data is sent to the robot controller in real time,so as to realize the whole palletizing path.Based on the research of robot vision technology and palletizing method, test platform of palletizing robot based on machine vision is built. Experimental tests about identification of positioning, crawling, palletizing and other processes of target box are carried out.Through the analysis of the experimental results, it is proved that the platform can realize the positioning recognition and automatic code release of different types of boxes on the same production line. The feasibility of the method is demonstrated, which is of great significance to the actual process of automatic classification of mixed production.