Analysis of the effect of sampling strategy: flicking for chemical plume tracing by an autonomous robot

摘要

In this paper, we propose an odor sampling strategy with flicking motion for chemical plume tracing(CPT). CPT consists of locating a chemical source according to the chemical particles in the air. The capability of performing CPT by an autonomous robot is highly demanded, yet difficult due to the complex behavior of airflow. From the viewpoint of bio-inspired systems, we focused on the flicking of an insect, by which it swings its antennae (chemical sensors) variably. We expected the flicking provide variable sampling resolutions according to searching phases. We first designed and implemented a variable air sampling mechanism that gave flicking ability to the robot. Then, we introduced the flicking to a common CPT algorithm inspired by the programmed behavior of a silkworm moth, which was composed of three motion phases. For the motion phases, we tested all combinations of the wide or narrow range of the sampling system. Through CPT experiments, results indicated that the flicking could improve CPT performance.