Research On Dynamical Behavior Modeling And Applications Of Spherical Objects In Virtual Environment

With the rapid development of the Virtual Reality(VR) technology and the increase of computational power, it is possible to study the objects movement in Virtual Environment(VE). As a kind of object with simple shape, the balls have many utilities. Many fields have gotten the general laws from the research of the balls movement. Especially in sports area, ball games are important sports, such as football, volleyball, golf and so on. Because of the influence of the fluid phenomenon, such as turbulent flow and the transition of boundary layer, the instability of ball movements is very prominent. This has made it a hard work to describe the movement of the ball accurately. The effective way to do that is to design dynamical behavior models and study the ball movement with the models in VE.This thesis presents the methods to model balls in VE. The spherical objects are classified into three categories: the mass-point-ball, the rigid ball and the deformable (or soft) ball. For the first and second kind of balls, this thesis discusses the modeling methods in three levels: the geometry modeling, the physically based modeling and the dynamical behavior modeling(DBM). Combining the mass-spring model with pressure model, this thesis developed a way to modeling the soft ball.The details and innovative points of this research include:1, the research of the mass-point-ball modeling methods. If the scale or the deformation of a ball has nothing to do with the research problems, or its rotation can be omitted, the ball is regarded as a mass point, and its mass-point behavior model is constructed. The model is made up of two sets: the attributes set, which describes the properties and the states of the ball, and the operations set, with which users can interact with the ball, query the ball states, and control the behavior of the ball. If a dynamic system consists of n mass points, it can be treated as one mass point moving in a 6-dimension space. The mass-point-ball model has wide range of applications, from the modeling of the micro-particles to that of the macro-celestial body, from fuzzy objects to some fluid phenomenon, and so forth. 2, the study of the rigid ball modeling methods.The movement styles of the balls on plane or in the air are many and varied. Besides abiding by the general movement laws of the common rigid bodies, the dynamics behavior of balls has its own characteristics. And based on the analysis of the DBM methods of the ordinary rigid bodies, this thesis studies the rigid balls movement on the plane and in the air, and constructs the accurate dynamical behavior model for golf. With this model, problems about the flying distance of golf are studied as an important point, and the research results are very close to the reality.3, the research of the soft ball modeling methodsWith combination of the mass-spring method with pressure model, this thesis designs a dynamic behavior model for the deformable balls. With the application of the model, this thesis presents a method to simulating the deformation of the cornea during the curation of hypermetropia with Laser Thermokeratoplasty. The simulation results have proved that the cornea modeling method is effective. Also, we use the soft ball model for modeling the air spring used in auto damper system. The dynamic model of soft ball is also suitable for the simulation of many things, such as the duration of bumping a balloon, the safety gasbag, the water wave with small swing and so on. Because of the inhere shortcomings of mass-spring model, the soft ball model can only be used for approximate simulation, it can not be fit for precisely studying problems.4, the collision detection algorithms in VE.This thesis also studies the algorithms of collision detection in VE while modeling the dynamical behavior of the balls. To detect the collisions among mass-point-balls, this thesis presents a ray collision detection algorithm. This algorithm can also be used for the collision detection of common rigid body. This thesis has especially studied the collision detection among general objects, and put forth a new algorithm based on k-dop. The new algorithm has enhanced the QuickCD algorithms, and extended the applications of the algorithms based on k-dop. With the extension, the new algorithm can be used for the collision detection in the VE consisting of multi-flying objects and multi-environment ones. The new algorithm is fast and exact. 5, the study of the force models in VE.This thesis has also studied the force models frequently used in VE, including the mutual interaction between the user and the virtual entity.During the research of DBM of spherical objects, we have primarily explored a framework for constructing the virtual physics environment, and designed the software named Virtual Physics Environment System(VPES). VPES consists of four basic functional components: the scene rendering module, the virtual entity constructing module, the force interaction module and the collision detection and response module. With VPES, after the geometry model construted and physics property given, with some initial movement conditions, the virtual objects will change its movement states under different forces naturally.