Research On Fast Collision Detection Algorithm Based On Separation Distance In Virtual Assembly

In view of the mis-design of the new product in the actual assembly process,which might be the problem of unassembly of the design size,the unreasonable operation space of the assembly tool and so on,which can be found ahead of real assembly by virtual assembly technology,and revise the design,which can avoid design rework in the real assembly stage.Assembly simulation can plan the assembly sequence and assembly path of parts by man-machine interactive,and to verify the feasibility of the assembly path and the overall assemblability of the product.At present,most collision detection algorithms are restricted by the complexity of parts and the scale of assembly scenes,which doesn't meet the requirements of real-time and accurate simulation of large-scale complex mechanical products,a collision detection method based on image space is studied in this paper,which solves the problem of real-time verification of assembly trajectory for complex products.The component collision type in the virtual assembly platform are analyzed,a collision detection method based on double view and transparent color overlay is studied,which can achieve the effect of avoiding collision between parts,the main research are as follows:In order to improve the detection efficiency of the algorithm by using the "overall strategy step by step",in the broad phrase,choose Axis-Aligned Bounding Box to be resolved the problem of assembly objects embedded in the space position,and then use the translucent color overlay algorithm to remove the non-collision objects quickly,which identify potential collision object.Aiming at the problem of anti-collision between parts,propose a way that calculate the separation distance,and design a potential collision face sequence calculation method based on pixel depth value,which involves twice rendering between faces.Then the potential collision surface is identified,and the projection is drawn to the lateral view.The minimum number of pixels is traversed along the motion axis,and it is regarded as the minimum separation distance,which is the maximum movable distance of the components to be assembled.The concept of double view is put forward,which is the positive view and lateral view of the simulation interface.The normal plane of the moving direction of the assembled object is the positive view,and the lateral view depends on the type of the two potential collision face sequence.In addition,a pixel region subdivision strategy is proposed,which can reduce the accuracy effect that caused by the discrete image sampling.Based on the above research contents,a collision detection algorithm for Virtual Assembly System is proposed.The algorithm transforms the collision detection problem in 3D space into the traversal problem of planar pixels by multi-dimensional display projection,which makes up the shortcoming that most image-space collision detection algorithms can only judge whether the collision occurred,but cannot find the distance,which can avoid collision and interference effectively in assembly process,and can meet the real-time simulation requirements of large complex mechanical products,also verify the assemblability of product design effectively.