Research On Virtual Assembly Technology Of Multicomponent With Array Complex Shape

Array multicomponent products with complex shapes are widely used in aerospace,satellite manufacturing and application fields.Due to the transmission of assembly errors and the coupling effect during assembly,the accuracy is uncontrollable and the quality is difficult to be guaranteed in the assembly process.At present,most of the product assembly relies on the manual operation of process personnel and preassembly of products by virtue of practical process experience,that is,repeated adjustment,repeated repair and repeated test.On the one hand,the product friction and bump will lead to the loss of assembly precision on the other hand,and on the other hand,the waste of manpower and material resources and the loss of assembly cost will seriously affect assembly efficiency.In view of the problems mentioned above,this thesis proposes a virtual assembly technology of array complex surface multicomponent products based on measured data,and studies the relevant principles and methods.The specific research contents are as follows:(1)The automatic acquisition technology of array complex shape multi-component products and feature extraction of collected data are studied.An automatic data acquisition system is designed based on the requirements of small size and multi-component automatic measurement.Based on the characteristics of actual data collected in the field,the preprocessing methods of structural data are studied,including point cloud denoising,automatic splicing and registration.The characteristics of the processed structural data were analyzed,and the assembly feature extraction model based on curvature discrimination method and projection comparison method was used to construct bounding boxes according to the feature information,and the joint cloud was segmented by bounding boxes method.(2)Assembly model construction and multi-component virtual assembly sequence optimization design based on measured data were studied.Based on assembly feature information,the assembly model of measured data was constructed.According to the information of assembly,the priority constraint relation for assembly process was analyzed,and the assembly priority was determined by assembly stability factor,and the feasibility sequence of virtual assembly was obtained.An improved genetic algorithm was proposed to optimize the assembly sequence and improve the overall assembly accuracy and speed.The virtual assembly feature parameters were designed,the virtual assembly optimization model was constructed,and the quantum particle swarm optimization algorithm was proposed to solve the assembly feature parameters to complete the assembly pose adjustment.Experiments are designed to verify the content and prove it is feasible and credible.(3)The realization method of virtual assembly function is studied and verified by experimental analysis.Based on the secondary development of Poly Works,a virtual assembly function realization process was designed,and a visual function module was developed to build a virtual assembly quality visual platform.Virtual assembly experiments were carried out on ceramic fiber rigid heat insulation tile of reusable vehicle(typical array complex shape multi-component product),and the experimental results before and after assembly were analyzed.Through verification,it is proved that this technology can provide reliable suggestions for assembly technicians and effectively improve the on-site assembly efficiency.