Path Planning Of Spraying Robot For Complex Surface

Most products are in need of the process of spraying to improve the corrosion resistance and exterior appearance prior to their exit from factory.Previously,it mainly relied on manual operations,which gave rise to such problems as high labor intensity,low spraying efficiency and poor stability.Especially,harmful gases would be produced alongside spraying.The emergence of spraying robots has partly accelerated the decline in manual work.The existing spraying robots boast its great advantages in major enterprises with large-scale production and fixed product specifications.However,their application to minor and medium-sized enterprises requires improvement so as to meet the need of operating with small-scale production,and great varieties of products with complex shapes.In this paper,based on the research background of small and medium-sized enterprises' demand for spraying path plan on products with complex shapes,visual recognition technology and spraying technology are combined to complete the task of spraying and perfect its automation as well as quality.The spraying system mainly includes three modules: visual recognition,transmission and spraying path plan.Proper optical equipment is selected in this paper for the special environment of spraying.The image processing algorithms used in visual recognition is analyzed in detail to calculate the placement of workpiece and the identification of workpiece type,such as image grayscale processing,image contrast enhancing,noise reduction processing,image matching,edge detecting and corner detecting.A transfer module with a rotation mechanism is adopted,which consists of a railed trolley for transferring disc-type workpieces and a suspension chain,hoister,a photoelectric sensor and other devices for transferring plate-type workpieces,to finish the transfer of various sprayed workpieces quickly and accurately.After visually identifying the type and placement of a workpiece,combined with the three-dimensional model of the workpiece,the API function in the SolidWorks software is used to complete the triangular meshing of the sprayed workpiece with complex shapes.The triangular planes on the workpiece afterwards are fitted into natural quadric surface and irregular plane.Different spraying models and path evaluation functions are established to work out the corresponding spraying paths for the surface and the plane.After that,the method for the turning and height-setting of the spraying gun are optimized to enhance the uniformity of the paint film and improve the spraying quality.After the optimization for a single surface/plane,the paths of different ones are combined together,in accordance with the ORPP problem(open-loop rural postman problem).Aimed at the shortest spraying path and the fast solution speed the BP neural network algorithm and the ant colony algorithm are combined to work.The simulation results show that compared with the ant colony algorithm alone,the new method has the advantages of shorter spraying path and fewer iterations.Finally,for the convenience of users,a set of control system interfaces based on Visual Studio,Matlab,Robot Studio are designed.The generated spraying path is simulated and verified in Robot Studio,and the uniformity of coating and the stability of robot in operation are simulated and verified in Matlab.