Research On Collision Detection Of Deformable Objects Based On Particle Swarm Optimization Algorithm

In recent years, with the distributed interactive simulation and virtual reality technology,the problem of collision detection between graphic again became the research hot spot. It is widely used for aerospace technology in virtual environment of training, the three dimensional software product design, animation design, interactive virtual clothing with the trunk of the human body, virtual campus roaming virtual surgery simulation, the medical field and so on.At the same time, the users’ request is also more and more high. In the virtual world, thequality of deformable objects collision detection, such as cloth, fabric, often determines thevirtual scene simulation effect. Processing of collision detection and collision response are themain technical problems, with the deformation simulation accounts for about forty percent ofthe time. In cloth model collision detection problem, most of the problems are in dealing withthe problems of thousands of points in the random motion, preventing not to penetrate andcross-border. Almost every type of collision detection algorithm is trying to reduce theprocess time and improve the accuracy.In order to solve the collision detection technology of large amount of calculation andthe problem of low efficiency, on the basis of random collision detection, this article choosesthe mass-spring model to shape the objects, starting from the perspective of the improvedparticle swarm algorithm, all possible collision feature points on the map coding into adiscrete solution space, and the solution space merging real-time updates. Then an improvedstochastic particle swarm optimization (PSO) algorithm is proposed, makes up the existingPSO’ ability to deal with the weakness of the discrete function, easily falling into localoptimum, to improve the accuracy of iterative calculation and reduce the time complexity,find collision points highly and effectively, so as to solve the large amount of calculation andtime-consuming long problems of the deformation collision detection technology.