Collision Detection Optimization in a Multi-Particle System

Collision detection optimization in an event-driven simulation of a multi-particle system is one of the crucial tasks, determining the efficiency of the simulation. We present the event-driven simulation algorithm that employs dynamic computational geometry data structures as a tool for collision detection optimization (CDO). The first successful application of the dynamic generalized Voronoi diagram method for collision detection optimization in a system of moving particles is discussed. A comprehensive comparision of four kinetic data structures in d-dimensional space, performed in a framework of an event-driven simulation of a granular-type materials system, is supported by the experimental results.