An estimation-based plume tracing method in time-variant airflow-field via mobile robot

The odor source localization using a mobile robot in time-variant airflow-field environment is considered. A novel estimation-based plume tracing method is proposed in this paper. Based on the odor-patch path estimated by the flow information when odor clues are found, a searching route is planned for the mobile robot to trace the plume. To have fair and statistic comparisons with both the spiral-surge and zig-zagging plume tracing algorithms, a simulation experiment platform is built and two evaluation criterions, i.e., the approaching index to the source and the mechanical efficiency, are proposed. Experiment results in large-scale time-variant airflow field indicate that estimation-based plume tracing method behaves better.