Summary: | 碩士 === 國立清華大學 === 電機工程學系 === 102 === In this research we present a new system and a new algorithm to achieve odor source tracking task, based upon mobile robot with 5 gas sensors.
Research on odor source localization based on mobile robots suggest that although the gas sensor is the most essential part in tracking algorithm, information from wind sensor also take a critical role. Former approaches to this work brought wind direction into algorithms and significantly improved the efficiency and success rate in laminar flow environment.
Inspired by this observation, we build a mobile robot system with a new odor tracking algorithm. The metal oxide (MOX) semiconductor sensors, wind sensor, distance sensors, and microcontroller are used to implement the task. By adding wind speed as a factor, we propose a new algorithm based on odor concentration of 5 MOX sensors and wind information to achieve the odor following task. In order to evaluate the effect of our algorithm, we compare it with several former algorithms using the robot in the controlled environment.
After a complete derivation of our system and algorithm, we present the results of several experiments that compare the success rate, distance overhead and effective capacity of searching and tracking of proposed methods with other leading approaches. The experimental results showed the improvement in tracking efficiency and the direction towards the odor source is given appropriately with our system and algorithm.
|