Cloth Dynamic Simulation Based On Position Hierarchical Constraints Method

In the study of computer animation,cloth simulation has become a representative of researching computer animation with its characteristics of simple implementation and wide coverage(involving computers,mathematics,and physics).In cloth simulation,the main simulation methods include traditional mechanics simulation methods and position-based dynamic methods.In the position-based dynamic method,constraint correction is added to the traditional mechanics simulation to avoid error accumulation,but the constraint oscillation problem makes the correction effect and the number of iterations appear nonlinear.In the traditional hierarchical method,the use of the hierarchical constraint grid method effectively improves the problem,but as the level of the grid system increases,more and more irregular polygons will exist in the constraint correction grid,so that correction occurs.Offset affects the simulation effect.In the traditional position-based layering method,this paper uses cloth simulation as a carrier to explore the deficiencies of the existing hierarchical methods,and finally completes cloth dynamic simulation based on position hierarchical constraints.The research work of the dissertation is as follows:(1)A position-based sampling hierarchical simulation algorithm is proposed.Aiming at the problem of non-uniform distribution of constrained grids in traditional hierarchical simulation methods,this paper proposes a position-based sampling hierarchical simulation algorithm.In this method,the sampling depth is firstly determined,then the sampling principle is used to simplify the extraction of the particle groups at each level.Finally,the point cloud reconstruction theory is used to perform point reconstruction and the construction of the acceleration model is completed.After the acceleration model is constructed,the position-based dynamic simulation algorithm is used to simulate the initial layer position,and then the accelerated hierarchical structure model is used to continuously accelerate the correction of the particle position from the coarsest layer to the finest layer,and the position correction is finally obtained.Furthermore,the speed is improved.(2)A position-based hierarchical simulation algorithm for decision functions is proposed.Aiming at the problem that the sampled-hierarchical method can not effectively simulate the cloth model with uneven distribution of particles,a position-based hierarchical algorithm for the decision function is proposed.In this method,the convergence of accelerating the cloth simulation is also accelerated by using the hierarchical idea.Firstly,the hierarchy depth isdetermined,the decision function is selected in combination with the distribution characteristics of the original mesh,and the hierarchical particle extraction is completed.Then,the point cloud is used.The reconstruction theory completes the triangular reconstruction.Finally,the initial level simulation is performed using the dynamic simulation method based on position,and the constraint correction is performed using the hierarchical structure model.Compared with the previous position-based sampling hierarchical simulation algorithm,this method is more flexible and adaptable to various requirements,such as unevenly distributed grids,etc.,due to the efficiency of specific scenes,there are further improvement in efficiency.In terms of experimental results,the position-based sampling layering algorithm proposed in this study compared the original position-based hierarchical dynamic simulation method to the performance of the core measurement index(rate of spring change)of cloth simulation under different sampling conditions.Different data,the maximum increase of 18%,the lowest increase of 5%,so the sampling based hierarchical method is more similar to the sampling and the model.The position-based decision function hierarchical algorithm is 23% higher than that based on hierarchical dynamic simulation methods.However,if no suitable decision function is selected,the result may not be ideal.Therefore,the method based on decision function's performance is more the selection of model and decision function.