Numerical Simulation Of Gas/Odor Diffusion In Indoor Ventilated Environments

In natural word, the odor is exploited by many creatures. Inspired by thisphenomenon, at the beginning of1990s, researchers began to use mobile robotequipped with gas/odor sensors to search, track and finally locate the odor sources.The research of gas/odor diffusion is a very important part of mobile robot based odorsource localization (OSL). This thesis focuses on the study of the gas/odor diffusionin indoor ventilated environments, and the dynamic and visual simulation plume of anindoor ventilated environment is established using Visual C++software.To facilitate the research of mobile robot based OSL and evaluate differentstrategies for the OSL, numerical simulation of odor plume in indoor ventilatedenvironments is addressed. On the basis of the FLUENT software package, the3-Dwind field of an indoor ventilated environment is calculated using the Reynoldsaverage approach, where the real wind velocity measured by a3-D anemometer istaken as the inlet boundary condition of the simulation. In a fixed height, the2-Dwind field is obtained by superimposing a normal-distribution fluctuating speed onthe calculated advection velocity, and the2-D odor plume is built by combining theFarrell’s filament concentration model with the2-D simulated wind field.Through comparing the wind and concentration information recorded usingmultiple sensor nodes with the simulated ones, the creditability of the establishedsimulated plume is demonstrated. In addition, in order to build a more completesimulation plume environment for the research of mobile robot based OSL, the gassensor model and motion model of the mobile robot are embedded into the simulationplume environments.