SEMI-PHYSICAL SIMULATION ENVIRONMENT FOR ROBOTIC ODOR SOURCE LOCALIZATION

With the complexity of odor source plume focused on, the data of plume distribution which is visualized by a sequence of images using OpenDX, a visualization software, is obtained by Fluent, a hydro mechanical dynamic modeling software, computing the odor concentration of the field interested. The sequence of images was projected to the ground to form a virtual dynamic odor distribution environment where the light intensity indicates odor concentration. To research mobile robot olfactory with semi-physical simulation environment consisting of the images mentioned above and a robot equipped with light sensors, the authors make robot to locate the odor source by traveling the plume. To acquire the semi-physical simulation environment performance, adopting moth-algorithm, we conduct an experiment which result show the repeatability, environment friendliness and close to natural environment.