Research On Key Technologies Of Robotic Automatic Polishing System For Miniature Metal Castings

At present,the polishing tasks of miniature metal castings mostly rely on human work.The working environment is extremely bad,the artificial polishing is low,the product uniformity is poor,and the polishing quality is difficut to be ensured.With the increase of labor costs and the protection of workers ' safety,,aritificial polishing is bound to be replaced by machine polishing,and automatic polishing will be realized.Based on the national 863 program subproject “Research and engineering realization on the robotic hybrid force-position control technologies for precision polishing”,this paper researches on the key technologies of robotic miniature metal castings polishing,including the polishing process,robotic system,polishing quality model and robotic force control strategy.Firstly,this paper carries out the polishing processes analysis and off-line simulation with the experience of artificial polishing.According to the requirement of polishing force,this paper proposes a robot system based on the force sensor's robotic active force control and the buffering belt grinder cylinder's passive force control.The detailed structures of belt grinder with passive force control and robotic tool are designed.Then this paper constructs the robotic system,which builds the hardware foundation of this project.Secondly,by analyzing the factors that affect the surface roughness,this paper puts forward a Support Vector Regression(SVR)polishing quality model based on the model inputs of belt speed,robot speed,grinding force and workpiece curvature,and the model output of roughness.What is more,this paper optimizes the parameters of polishing quality model respectively by the method of Grid Search,Genetic Algorithm(GA)and Particle Swam Optimization Algorithm(PSO),to choose a faster method with smaller Mean Square Error(MSE).After that,this paper carries out the polishing quality modeling and its experimental verification with the optimization parameters.Thirdly,this paper puts forward the active/passive force control strategy for the robot automatic polishing system.By comparing three kinds of filtering algorithms,the median filter,average filter and Kalman filter on the effect of disposing original force signal,this paper selects a filtering algorithm and its parameter with stronger anti-interference ability and better predictive performance.Then this paper deduces the gravity compensation formula of the robot end-tool,and eliminates the gravity effect on sensor in different robot postures.Besides,this paper proposes the Cartesian impedance control strategy of robot automatic polishing system,and discusses the impedance arguments effect on control system.Aiming at the problem of changeable polishing environment,this paper proposes a fuzzy adaptive impedance control algorithm for robotic polishing,and the contrast experimental results show that the robustness of fuzzy impedance algorithm is stronger.Finally,according to the designed scheme and algorithm,this paper constructs the calibration experiment of robotic system,and designs a friendly interactive software for the robotic miniature metal castings polishing system,including communicating section,motion controlling section,force monitoring section,and force controlling section.After that,the experiment of calibration with laser tracker,gravity compensation experiment,constant force control experiment,automatic polishing system calibration experiment and robotic automatic polishing experiment are carried out.The experimental results show that the force control precision of the robot automatic polishing system is about ±5N,and the workpiece under constant force polishing acquires better surface finish,lower surface roughness,and more continuous surface.