Massively parallel Modelling & Simulation of large crowd with GPGPU

Author: Chen Dan  

Publisher: Springer Publishing Company

ISSN: 0920-8542

Source: The Journal of Supercomputing, Vol.63, Iss.3, 2013-03, pp. : 675-690

Disclaimer: Any content in publications that violate the sovereignty, the constitution or regulations of the PRC is not accepted or approved by CNPIEC.

Previous Menu Next

Abstract

In peacekeeping, domestic, or combat operations, unanticipated crowd confrontations can occur. As a highly dynamic social group, human crowd in confrontation is a fascinating phenomenon. This paper presents a novel method based on the concept of vector field to formulate the way in which external stimuli may affect the behaviours of individuals in a crowd. Furthermore, Modelling & Simulation (M&S) of large crowds at individual level has long been placed in the highly computation intensive world. This study adopts GPGPU to sustain massively parallel M&S of a confrontation operation involving a large crowd. This approach enables investigation of a crowd consisting of tens of thousands individuals whose size was prohibitively large for conventional M&S technique to support. Experimental results indicate that the approach is efficient in terms of both performance and energy consumption.