The Optimizaton Of Robotic Cellular Layout By The Improved Partical Swarm Optimizaton Algorithm

In industrial automated manufacturing processes,robots carry out predetermined tasks such as welding,painting and stacking.The similarity of these tasks is that the robots need to perform cycle operations.Some robots and facilities form a robotic celluar manufacturing system.The productivity of the robotic cellular is closly related to its layout.In this paper,we investigate the layout optimization algorithm for improving the productivity of robotic cellular.Firstly,this papper uses DH method,proposed by Denavit and Hartenberg,to calculate the forward and the inverse kinematics of robots.Two half-spherical surfaces and an entire sphere,which simplifies the searching space of optimization algorithm,are used to envelop the workspace of robots.Furthermore,the method of modeling facilities and the interference adjusting between them are discussed.Secondly,the assessment criteria of layout optimization are introduced and the object function is proposed considering the nonlinearity and the dexterity of robots.What's more,the interference in the layout is designed as a punishment function.Thirdly,the advantage and disadvantage of some heuristic algorithms are discussed.Then we focus on the particle swarm optimization(PSO)algorithm and discuss the influence of topological structure and the learning method of the particles.Base on the previous discusstion,a new kind of PSO algorithm,which combines dynamic topological structure and fine search method,is proposed.To vertify the performance of the new PSO algorithm,we choose four benchmark functions and other three different PSO algorithms to test,which runs fifty times on each benchmark fucntions.The results indicate that the improved PSO algorithm performs better than others in three experiments.Finally,we apply the improved PSO algorithm on robotic cellular layout optimization and conduct an experiment using HSR-JR608 rotot.The results indicate that our layout optimization algorithm can create a near optimal layout and reduce the cycle time of robotic cellular.