Research On Products Assembly Planning For Robot Automatic Operation

Product assembly is a time-consuming part of the production process.Compared to the traditional manual assembly,the automatic assembly of products with robots improves the efficiency of product assembly greatly.When assembling a product,it is necessary to consider the impact of equipment resources on the planning process.Based on the analysis of the relevant method at home and abroad,the paper research on the assembly sequence and path planning.Details are summarized as follows:The framework of product assembly planning for robotic automation is established.The paper analyzes the implementation of three key points in assembly planning,including the modeling method,assembly sequence and path planning.The logical relationship between these implementations is expounded.After analyzing the demand of the information in the assembly planning process,a hierarchical model of assembly information is established,and the geometric information model of parts is also established based on the unified expression of NURBS.A hierarchical assembly relation model with device resources information is set up,by improving the traditional interference matrix.The hierarchical assembly relation model can not only express the dynamic constraint relation between each part and their assembly resources,but also reduce the complexity of assembly planning.The paper elaborates the method of assembly sequence planning based on discrete particle swarm optimization(DPSO)algorithm.Considering the influence of device resources on the planning process,a method of judging the ge ometric feasibility of assembly sequence is proposed.To solve the multi-constraint optimization problem in the assembly sequence planning process,a fitness function of the assembly sequence is defined.From this,a method for searching optimal assembly sequence based on DPSO is proposed in the case of considering the equipment resources,which can realize the assembly sequence planning of the parts in the product.The paper elaborates the method of assembly path planning based on rapid-growing random tree-based(RRT)algorithm.According to the assembly sequence planning result,we can't get a dog-leg assembly path.To solve this problem,a method of searching assembly path based on RRT algorithm and a method of pose transformation based on singular val ue decomposition(SVD)algorithm are proposed.In the process of assembly path planning,NURBS control points are used to locate the parts.We can use a method of component interference detection based on hierarchical bounding box,combined with above algo rithm,to compute the collision free paths.Aiming at the robot assembly environment,the movement trajectory of the part is optimized,which improves the efficiency of the assembly.The assembly planning and simulation system is designed to realize the assembly information modeling,assembly sequence and path planning.The assembly planning results are imported into DELMIA,and the 3D visualization simulation of the product assembly process is carried out to verify the feasibility of the assembly planning results.